The Cambridge Handbook of Bilingualism 9781316831922, 9781107179219, 2018013801

Table of contents :
Front Matter
Dedication
Contents
Figures and Tables
Contributors
Acknowledgements
Common Abbreviations
Introduction
1 - Becoming Bilingual in Early Childhood
2 - Bilingualism from Childhood through Adolescence
3 - Young Bilingual Adults
4 - Bilingualism in Midlife
5 - Language and Older Bilinguals
6 - Ideologies of Language, Bilingualism, and Monolingualism
7 - Bilingualism and the Law
8 - Language Planning and Policies for Bilingualism
9 - The Economics of Bilingualism
10 - The Nature of Exposure and Input in Early Bilingualism
11 - Becoming Bilingual through Additive Immersive Programs
12 - Foreign Language Learning from Early Childhood to Young Adulthood
13 - Supporting Bilingualism in Adult First Generation Migrants
14 - Unlearning and Relearning of Languages from Childhood
15 - The Measurement of Bilingual Abilities
16 - Highly Proficient and Gifted Bilinguals
17 - Language Choice in Bilingual Interaction
18 - First Language Attrition
19 - Bilingualism in Clinical Linguistics
20 - Doing and Undoing Bilingualism in Education
21 - Second Language Acquisition as a Road to Bilingualism
22 - Bilingualism in Cognitice Science
23 - Bilingualism in Neurolinguistics
24 - Bilingualism and Sign Language Research
25 - Bilingualism and Bidialectalism
26 - Bilingualism and Language Contact
27 - Bilingualism and Multilingualism
References
Language Index
Index of Place Names
Subject Index

Citation preview

The Cambridge Handbook of Bilingualism The ability to speak two or more languages is a common human experience, whether for children born into bilingual families, young people enrolled in foreign language classes, or mature and older adults learning and using more than one language to meet life’s needs and desires. This Handbook offers a developmentally oriented and socially contextualized survey of research into individual bilingualism, comprising the learning, use, and, as the case may be, unlearning of two or more spoken and signed languages and language varieties. A wide range of topics is covered, from ideologies, policy, the law, and economics, to exposure and input, language education, measurement of bilingual abilities, attrition and forgetting, and giftedness in bilinguals. Also explored are cross- and intradisciplinary connections with psychology, clinical linguistics, second language acquisition, education, cognitive science, neurolinguistics, contact linguistics, and sign language research. A N N I C K D E H O U W E R is a professor of language acquisition and multilingualism at the University of Erfurt, Germany. Her book The Acquisition of Two Languages from Birth (Cambridge, 1990) constituted pioneering work in bilingual acquisition, and her 2009 textbooks Bilingual First Language Acquisition and An Introduction to Bilingual Development are used all over the world. L O U R D E S O R T E G A is a professor of second language acquisition at Georgetown University, Washington DC. A widely published scholar, she is best known for her award-winning meta-analysis of L2 instruction in 2000, for her best-selling textbook Understanding Second Language Acquisition (2009, translated into Mandarin in 2016), and for championing a bilingual turn in SLA.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 07 Jan 2020 at 06:17:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

cambridge handbooks in language and linguistics Genuinely broad in scope, each handbook in this series provides a complete state-of -the-field overview of a major sub-discipline within language study and research. Grouped into broad thematic areas, the chapters in each volume encompass the most important issues and topics within each subject, offering a coherent picture of the latest theories and findings. Together, the volumes will build into an integrated overview of the discipline in its entirety.

Published titles The Cambridge Handbook of Phonology, edited by Paul de Lacy The Cambridge Handbook of Linguistic Code-Switching, edited by Barbara E. Bullock and Almeida Jacqueline Toribio The Cambridge Handbook of Child Language, Second Edition, edited by Edith L. Bavin and Letitia Naigles The Cambridge Handbook of Endangered Languages, edited by Peter K. Austin and Julia Sallabank The Cambridge Handbook of Sociolinguistics, edited by Rajend Mesthrie The Cambridge Handbook of Pragmatics, edited by Keith Allan and Kasia M. Jaszczolt The Cambridge Handbook of Language Policy, edited by Bernard Spolsky The Cambridge Handbook of Second Language Acquisition, edited by Julia Herschensohn and Martha Young-Scholten The Cambridge Handbook of Biolinguistics, edited by Cedric Boeckx and Kleanthes K. Grohmann The Cambridge Handbook of Generative Syntax, edited by Marcel den Dikken The Cambridge Handbook of Communication Disorders, edited by Louise Cummings The Cambridge Handbook of Stylistics, edited by Peter Stockwell and Sara Whiteley The Cambridge Handbook of Linguistic Anthropology, edited by N. J. Enfield, Paul Kockelman and Jack Sidnell The Cambridge Handbook of English Corpus Linguistics, edited by Douglas Biber and Randi Reppen The Cambridge Handbook of Bilingual Processing, edited by John W. Schwieter The Cambridge Handbook of Learner Corpus Research, edited by Sylviane Granger, Gae¨tanelle Gilquin and Fanny Meunier The Cambridge Handbook of Linguistic Multicompetence, edited by Li Wei and Vivian Cook The Cambridge Handbook of English Historical Linguistics, edited by Merja Kyto¨ and Pa¨ivi Pahta The Cambridge Handbook of Formal Semantics, edited by Maria Aloni and Paul Dekker The Cambridge Handbook of Morphology, edited by Andrew Hippisley and Greg Stump The Cambridge Handbook of Historical Syntax, edited by Adam Ledgeway and Ian Roberts The Cambridge Handbook of Linguistic Typology, edited by Alexandra Y. Aikhenvald and R. M. W. Dixon The Cambridge Handbook of Areal Linguistics, edited by Raymond Hickey The Cambridge Handbook of Cognitive Linguistics, edited by Barbara Dancygier The Cambridge Handbook of Japanese Linguistics, edited by Yoko Hasegawa The Cambridge Handbook of Spanish Linguistics, edited by Kimberly L. Geeslin The Cambridge Handbook of Bilingualism, edited by Annick De Houwer and Lourdes Ortega

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 07 Jan 2020 at 06:17:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

The Cambridge Handbook of Bilingualism Edited by Annick De Houwer University of Erfurt, Germany

Lourdes Ortega Georgetown University, Washington DC

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 07 Jan 2020 at 06:17:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107179219 DOI: 10.1017/9781316831922 © Cambridge University Press 2019 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2019 Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Ortega, Lourdes, editor. | De Houwer, Annick, editor. Title: The Cambridge handbook of bilingualism / edited by Lourdes Ortega, Annik De Houwer. Description: Cambridge, United Kingdom ; New York, NY : Cambridge University Press, 2018. | Series: Cambridge handbooks in language and linguistics | Includes bibliographical references and indexes. Identifiers: LCCN 2018013801 | ISBN 9781107179219 Subjects: LCSH: Bilingualism – Handbooks, manuals, etc. Classification: LCC P115 .C36 2018 | DDC 404/.2–dc23 LC record available at https://lccn.loc.gov/2018013801 ISBN 978-1-107-17921-9 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 07 Jan 2020 at 06:17:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

We dedicate this handbook to the memory of Jules Ronjat (1864–1925), who was a true pioneer in the field of bilingualism and the first to carry out a book-length study of a bilingual individual (Le de´veloppement du langage observe´ chez un enfant bilingue, 1913), and to Dick Schmidt (1941–2017), whose longitudinal studies of Wes learning English and himself learning Portuguese are an inspiration for the non-deficit study of adult bilingualism.

Downloaded from https://www.cambridge.org/core. University of Groningen, on 13 Jan 2020 at 02:01:29, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Downloaded from https://www.cambridge.org/core. University of Groningen, on 13 Jan 2020 at 02:01:29, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Contents

List of Figures and Tables List of Contributors Acknowledgments List of Common Abbreviations Introduction: Learning, Using, and Unlearning More than One Language Annick De Houwer and Lourdes Ortega Part I Bilingual Learning and Use at Five Stages of Life 1 Becoming Bilingual in Early Childhood Ludovica Serratrice 2 Bilingualism from Childhood through Adolescence

Martha Bigelow and Penelope Collins 3 Young Bilingual Adults Kellie Gonc¸alves 4 Bilingualism in Midlife David Singleton and Simone E. Pfenninger 5 Language and Older Bilinguals Mira Goral Part II The Larger Contexts of Bilingualism 6 Ideologies of Language, Bilingualism, and Monolingualism Janet M. Fuller 7 Bilingualism and the Law Philipp S. Angermeyer 8 Language Planning and Policies for Bilingualism Joseph Lo Bianco 9 The Economics of Bilingualism Franc¸ois Grin Part III Contexts for Bilingual Learning and Unlearning 10 The Nature of Exposure and Input in Early Bilingualism

Sharon Armon-Lotem and Natalia Meir 11 Becoming Bilingual through Additive Immersive Programs

page ix x xiii xiv

1 13 15

36 59 76 101 117

119 135 152 173 191

193 213

Maria Juan-Garau and Roy Lyster

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 15 Jan 2020 at 18:23:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

viii

Contents

12 Foreign Language Learning from Early Childhood to Young Adulthood Carmen Mun˜oz and Nina Spada 13 Supporting Bilingualism in Adult First-Generation Migrants James Simpson 14 Unlearning and Relearning of Languages from Childhood to Later Adulthood Merel Keijzer and Kees de Bot Part IV The Dynamics of Bilingualism across the Lifespan 15 The Measurement of Bilingual Abilities Jeanine Treffers-Daller 16 Highly Proficient and Gifted Bilinguals Adriana Biedron´

and David Birdsong 17 Language Choice in Bilingual Interaction Annick De Houwer 18 First Language Attrition: From Bilingual to Monolingual Proficiency? Barbara Ko¨pke Part V Bilingualism Research across Disciplines 19 Bilingualism in Clinical Linguistics Carol Scheffner Hammer 20 21 22

23

and Lisa A. Edmonds Doing and Undoing Bilingualism in Education Ofelia Garcı´a and Ruanni Tupas Second Language Acquisition as a Road to Bilingualism Lourdes Ortega Bilingualism in Cognitive Science: The Characteristics and Consequences of Bilingual Language Control Kenneth R. Paap Bilingualism in Neurolinguistics: From Static to Dynamic Approaches Arturo E. Hernandez

Part VI Bilingual Connections 24 Bilingualism and Sign Language Research Gladys Tang

and Felix Sze 25 Bilingualism and Bidialectalism

233 250 267 287 289

307 324 349 367

369 390 408

435 466 481

483 Jean-Pierre Chevrot

and Anna Ghimenton 26 Bilingualism and Language Contact

510 Suzanne Aalberse

and Pieter Muysken 27 Bilingualism and Multilingualism Suzanne Quay and Simona Montanari

524 544

References Language Index Index of Place Names Subject Index

561 645 651 654

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 15 Jan 2020 at 18:23:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Figures and Tables

Figures 22.1 Visualization of the processing of cat/gato by Spanish–English bilinguals according to the Revised Hierarchical Model (adaptation of figure 3 in Kroll & Stewart, 1994) page 438 22.2 Visualization of the processing of cat by Spanish–English bilinguals according to the Bilingual Interactive Activation Model (BIA) (adaptation of figure 6.1 in Dijkstra & van Heuven, 1998) 440 22.3 Visualization of the processing of cat/gato by Spanish–English bilinguals according to the BIA+ Model (adaptation of figure 2 in Dijkstra & van Heuven, 2002) 442

Tables 2.1 Predictors of successful language and literacy bilingual outcomes of schooling from a dual contextual and personal perspective 39 2.2 Language as resource heuristic from a dual contextual and personal perspective 40 9.1 Terminology of linguistic diversity adopted in this chapter 175 9.2 Four types of value 181

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 15 Feb 2020 at 14:28:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Contributors

Suzanne Aalberse Universiteit Amsterdam (University of Amsterdam), the Netherlands Philipp S. Angermeyer York University, Canada Sharon Armon-Lotem ‫אילן‬-‫( אוניברסיטת בר‬Bar Ilan University), Israel Adriana Biedron´ Akademia Pomorska w Słupsku (Pomeranian University), Poland Martha Bigelow University of Minnesota, United States of America David Birdsong The University of Texas at Austin, United States of America Jean-Pierre Chevrot Universite´ Grenoble Alpes (University of Grenoble), France Penelope Collins University of California at Irvine, United States of America Kees de Bot Rijksuniversiteit Groningen (University of Groningen), the Netherlands and Pannon Egyetem (University of Pannonia), Hungary Annick De Houwer Universita¨t Erfurt (University of Erfurt), Germany Lisa A. Edmonds Columbia University, United States of America Janet M. Fuller Rijksuniversiteit Groningen (University of Groningen), the Netherlands Ofelia Garcı´a City University of New York, United States of America

Downloaded from https://www.cambridge.org/core. UZH Hauptbibliothek / Zentralbibliothek Zürich, on 08 Jan 2020 at 02:43:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

List of Contributors

xi

Anna Ghimenton Universite´ Lumie`re Lyon 2 (Lumie`re University Lyon 2), France Franc¸ois Grin Universite´ de Gene`ve (University of Geneva), Switzerland Kellie Gonc¸alves Universitet i Oslo (University of Oslo), Norway Mira Goral City University of New York, United States of America and Universitet i Oslo (University of Oslo), Norway Carol Scheffner Hammer Columbia University, United States of America Arturo E. Hernandez University of Houston, United States of America Maria Juan-Garau Universitat de les Illes Balears (University of the Balearic Islands), Spain Merel Keijzer Rijksuniversiteit Groningen (University of Groningen), the Netherlands Barbara Ko¨pke Universite´ de Toulouse 2 – Jean Jaure`s (University of Toulouse 2), France Joseph Lo Bianco The University of Melbourne, Australia Roy Lyster McGill University, Canada Natalia Meir ‫אילן‬-‫( אוניברסיטת בר‬Bar Ilan University), Israel Simona Montanari California State University, United States of America Carmen Mun˜oz Universitat de Barcelona (University of Barcelona), Spain Pieter Muysken Radboud Universiteit (Radboud University), the Netherlands Lourdes Ortega Georgetown University, United States of America Kenneth R. Paap San Francisco State University, United States of America Simone E. Pfenninger Universita¨t Salzburg (University of Salzburg), Austria Suzanne Quay 国際基督教大学 (International Christian University), Japan

Downloaded from https://www.cambridge.org/core. UZH Hauptbibliothek / Zentralbibliothek Zürich, on 08 Jan 2020 at 02:43:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

xii

List of Contributors

Ludovica Serratrice University of Reading, United Kingdom James Simpson University of Leeds, United Kingdom David Singleton Trinity College, Ireland and Pannon Egyetem (University of Pannonia), Hungary Nina Spada University of Toronto, Canada Felix Sze 香港中文大學 (The Chinese University of Hong Kong), China Gladys Tang 香港中文大學 (The Chinese University of Hong Kong), China Jeanine Treffers-Daller University of Reading, United Kingdom Ruanni Tupas National Institute of Education, Singapore

Downloaded from https://www.cambridge.org/core. UZH Hauptbibliothek / Zentralbibliothek Zürich, on 08 Jan 2020 at 02:43:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Acknowledgments

This Handbook has been the center of our lives for about two and a half years. Was it worth it? That is for you, the reader, to ultimately decide. But for us, as editors, it certainly has been. Through our work with each other and our contributors we have learned so much. We have discovered many aspects of the rich and varied research on bilingualism in unexpected corners and in publication venues that we did not know before. Through engaging with each other and with our authors we have come to question some of our own long-held beliefs about bilingualism and have had to keenly argue points of difference, and in the process we have come to new insights. This is perhaps the best gift the making of this Handbook has given us. We thank each and every one of our contributors for the positive and constructive spirit in which we have been able to build this book. We especially thank them for putting up with our often nitpicky fastidiousness throughout the process. We would never have come to this point without the trust that Cambridge University Press and especially Andrew Winnard put in us. Thank you. During the preparation of this Handbook we often did not have the time to spend with our loved ones the way we would have wanted to. Also, we were often too much “elsewhere” in our minds to be able to fully relax when we did spend time with them. We apologize, and thank especially our life partners for putting up with us. We promise, things will be different from now on! Annick De Houwer and Lourdes Ortega

Downloaded from https://www.cambridge.org/core. Columbia University Libraries, on 10 Mar 2020 at 21:12:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Common Abbreviations

AoA AoO BFLA CDI CLI CLIL CODA CPH ERP ESLA FL FLP HL L1 L2

LoE LoR LP MLU PPVT RT SES SL

age of acquisition, implied is: onset of first exposure to a language age of onset of exposure to a language bilingual first language acquisition MacArthur–Bates Communicative Development Inventory (Fenson et al., 1993) crosslinguistic influence content and language integrated learning hearing child of deaf adults critical period hypothesis event-related potential early second language acquisition foreign language family language policy heritage language the language a bilingual person learned first if the person did not hear more than a single language from birth any language a bilingual person learned after they had already been learning an L1; any third, fourth, etc., language a bilingual person learned after they had already learned two languages from birth length of exposure, the time since the onset of exposure to an SL length of residence language policy mean length of utterance Peabody Picture Vocabulary Test (Dunn & Dunn, 2007) response time/reaction time (in experiments) socioeconomic status societal language

Downloaded from https://www.cambridge.org/core. The University of Reading, on 10 Mar 2020 at 19:59:40, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

List of Common Abbreviations

SLA SLPs TE TL

xv

second language acquisition, referring to the process or the field speech-language pathologists translation equivalent target language

Downloaded from https://www.cambridge.org/core. The University of Reading, on 10 Mar 2020 at 19:59:40, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Downloaded from https://www.cambridge.org/core. The University of Reading, on 10 Mar 2020 at 19:59:40, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922

Introduction Learning, Using, and Unlearning More than One Language Annick De Houwer and Lourdes Ortega

Many children are born into families where parents speak different languages to them. Karin was born in Portugal and taken care of at home by her mother, who spoke Portuguese to her. Karin’s father spoke Swedish to her from the start, but was usually not home during the day. As a young infant Karin was very happy to have her father come home after work and play with her. When Karin was around four months old, something changed. As usual, Karin would be in her mother’s arms when her father came home, and Karin’s mother spoke to Karin’s father in Swedish. But for a few weeks, Karin did not immediately show her usual joy at her father’s homecoming. Instead, she “froze up” and looked at her mother’s mouth with a disquieted frown, and refused to cuddle either parent until her attention was led away to something else (Cruz-Ferreira, 2006, p. 62; two other infants described by Cruz-Ferreira showed the same pattern at around the same age). This early disapproving noticing of language difference lays bare a basic fact related to individual bilingualism, viz., the instant attribution of social value based on the perception of people’s use of a particular set of sounds (or gestures, as the case may be). As Chevrot and Ghimenton (Chapter 25, this volume) explain, such social attributions shape and constitute language-related attitudes. The fact that from very early on both bilingually and monolingually raised infants are able to perceive and distinguish accents, regional varieties, and languages (see Serratrice, Chapter 1 this volume, and Chevrot & Ghimenton, Chapter 25, this volume) means that the development of language-related attitudes is an integral part of language development, regardless of the number of language varieties involved. Our very categorization of what constitutes a particular variety or way of speaking is, most likely, intimately tied with the attitudes we have developed from an early age. As De Houwer (Chapter 17, this volume) explains, these attitudes may to a large degree depend on socialization practices, while at the same time examples like

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

2

AN NICK DE HOUWER AND LOU R DES ORTEGA

Karin’s show that social attributions are driven by internal factors as well. Language attitudes may crystallize into wider-held language ideologies (Fuller, Chapter 6, this volume), which find expression in, for instance, specific language policies (Lo Bianco, Chapter 8, this volume), legal systems (Angermeyer, Chapter 7, this volume), and the world of work and employment (Grin, Chapter 9, this volume) – all of which, in turn, impact individual bilingualism. This Handbook seeks to offer a developmentally oriented and socially contextualized, realistic perspective on the learning, use, and, as the case may be, unlearning, of more than a single language variety by individual people at different stages of life. The focus is on bilinguals as people and what we know about them, their bilingual issues, and their bilingual lives. As such, it is language users rather than language-as-structure that take central position. Research findings on individual bilingualism are readily available in North America, Europe, and Australia. We have encouraged our contributors to consider select relevant findings on bilingualism from Latin America, Asia, and/or Africa, as long as these were published in languages that the Handbook readership will likely have access to (that is, English, Spanish, French, and German). Bilingualism is particularly extended and seen as natural in non-Western contexts, whereas much of the research in Western contexts is generated under a societal pressure to conform to monolingual expectations (see further later in this introduction). Therefore, even a brief consideration of more diverse contexts for bilingualism may be particularly informative.

An Inclusive Approach to Bilingualism Learning, using, and unlearning more than a single language variety are very much interrelated. People may learn a language variety through using it, and people may at least partly unlearn a language variety because they are not using it, that is, they may not fully maintain it (see Keijzer & de Bot, Chapter 14, this volume, and Ko¨pke, Chapter 18, this volume). We use the term learning for any kind of language learning, that is, both instructed and uninstructed, both implicit and explicit. For us, language learning equals language acquisition and language development, and these terms are used interchangeably throughout the Handbook. We use the term language as shorthand for any form of linguistic communication that is socially constructed as habitually belonging to a particular way of speaking, signing, or writing. We do not wish to pretend that this is a full definition of what language is (an impossible endeavor, anyhow), but we wish to stress that we are not only interested in bilinguals learning what are known as standard languages. We also are interested in bilinguals learning signed languages (Tang & Sze, Chapter 24, this volume), and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

Introduction

3

other language varieties that are not necessarily standardized, such as ethnic varieties, contact varieties (Aalberse & Muysken, Chapter 26, this volume), and regional and social varieties (Chevrot & Ghimenton, Chapter 25, this volume). Some bilingualism researchers and some of the Handbook authors (Fuller, Chapter 6; Garcı´a & Tupas, Chapter 20) oppose the idea that languages and language varieties can or should be named and enumerated. In this view, the very notion of “language” is a mere ideological invention. We agree that all adult linguistic categories are, ultimately, ideological constructs. However, we see inherent value in studying distinct languages and language varieties in bilingualism because of the fact that already in infancy, prior to any sociopolitical influences, both bilingually and monolingually raised children can perceive, distinguish, and harbor strong attitudes toward different accents, regional varieties, and languages. The learning of more than a single language variety is where bilingualism begins. Learning a language usually starts with learning to understand some of it. This is where bilingualism starts too: it starts when an individual has learned to understand some of at least two languages. Thus, infants who have started to understand some words in each of two languages but cannot yet speak or sign are bilingual. Their language comprehension abilities (Treffers-Daller, Chapter 15, this volume), however, are still very different from young adults who can understand long stretches of oral or signed discourse in several regional and social varieties, or who can read newspaper articles in one language (e.g., standard Dutch) and understand discourse in another (e.g., West Flemish dialect). For literate young adults, being able to read newspaper articles and understand longer stretches of discourse are age-appropriate, expected language skills. Such skills take time to develop in any language, regardless of how many one knows, and regardless of the modality (spoken, signed, written). This insight, however, should not impinge on individuals’ characterization as bilingual or not. Thus we dynamically define a bilingual individual as one who understands at least two languages at age-appropriate levels, regardless of modality (see also De Houwer, to appear). This includes seven-yearolds who can understand Langue des Signes Que´be´coise (sign language of Que´bec) and who can read short French sentences with basic vocabulary. This includes educated English-speaking retirees living in the Provence region of France who can understand the local type of French both in informal conversation and in more formal contexts, for instance at city hall. This includes Japanese engineers who are fluent in Japanese and can read English manuals for complicated machinery, but who may be lost if someone speaks to them in English. This includes teens in the south of Morocco who can understand conversations in the local varieties of Berber and Arabic, in Moroccan French, and in standard French as they hear it on television. Thus, our approach to bilingualism includes what many may call multilingualism, that is, the learning and use of more than two

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

4

AN NICK DE HOUWER AND LOU R DES ORTEGA

languages or language varieties. We asked our authors to adopt an inclusive approach to bilingualism as involving more than one language or language variety, regardless of the precise number. Our inclusive approach to bilingualism as starting with dual language comprehension and as referring to the involvement of more than a single language follows classic views in the field (Baetens Beardsmore, 1982; Elwert, 1959; Grosjean, 1982). It implies that the number of individuals who are counted as bilingual is quite high, and that hence the variability among them in terms of demographic background, geographical location, age, and language abilities will be enormous.

A Lifespan Perspective Indeed, variability is a key concept in bilingualism studies. Also in psycholinguistic processing approaches to bilingualism (Paap, Chapter 22, this volume), this variability is fortunately being taken into account more and more. The variability among bilinguals starts with the youngest of bilinguals: children may hear two languages from birth, usually in the home, or they may start hearing a second language sometime after they have just heard a first, with the first usually heard at home and the second outside the home, for instance, at daycare or preschool (Serratrice, Chapter 1, this volume). This difference in timing and associated circumstances makes for a difference in very young children’s early language learning trajectories (De Houwer, 2018a). Many chapters in this Handbook discuss how the timing of first regular exposure to a new language and the circumstances in which that exposure takes place relate to bilingual development (see in particular the chapters by Gonc¸alves, Chapter 3; Singleton & Pfenninger, Chapter 4; Armon-Lotem & Meir, Chapter 10; Mun˜oz & Spada, Chapter 12; and Ortega, Chapter 21). As Keijzer and de Bot (Chapter 14) as well as Ko¨pke (Chapter 18) show, exposure also plays a role in the extent to which bilinguals maintain abilities in each of their languages across the lifespan. Bilingual development and individual language maintenance are likely differentially affected by different conditions and forces at different lifespan stages. This is because individuals are at different points of psychosocial development as a function of their global stage in life (e.g., Caps, 2008). Thus, any relevant phenomenon under study may call for differential research treatment over different lifetime periods. We consequently have striven to build in a lifespan perspective across all topics and chapters in the Handbook to the extent possible. Part I, comprising five chapters, shows this most clearly, as it traces aspects of bilingual development and use across five distinct life stages: (1) the first six years of life, or early childhood (Serratrice, Chapter 1), (2) middle childhood and adolescence, that is, from about age six until around age

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

Introduction

5

18 (Bigelow & Collins, Chapter 2), (3) young adulthood, which we define here as covering the years from around age 18 until toward the late 30s (Gonc¸alves, Chapter 3), (4) midlife, going somewhere from around age 40 until around age 60 (Singleton & Pfenninger, Chapter 4), and (5) later adulthood, starting at around age 60 (Goral, Chapter 5). Also beyond Part I we have encouraged all Handbook contributors to be as specific as possible about the ages or life stages of any bilingual individuals they discuss, and to include coverage of studies across different life stages where possible and relevant. The life stage divisions in Part I largely follow general practice in developmental psychology, although in that field finer distinctions are often made. For instance, the authoritative text by Steinberg, Bornstein, Vandell, and Rook (2011) makes a distinction between infancy and early childhood, both of which we subsume under early childhood; it treats middle childhood and adolescence as two separate stages as well, rather than our stage (2), where we combine these. By combining these we are not suggesting there are no differences between infancy and early childhood in the Steinberg et al. sense, or between middle childhood and adolescence. Rather, making these distinctions for a general Handbook such as the present one would lead to a level of detail it cannot accommodate. In future research on bilinguals, however, we would encourage scholars to start carrying out comparative work that fully takes into account different life stages as identified by developmental psychologists. Studies of children may lump together data from four- and eight-year-olds, without acknowledging that at those ages, these relatively small age differences represent developmentally quite different life stages. Likewise, research with people who are beyond middle childhood often combines data on individuals with widely different ages and presents results without taking into account life stage variability and how it may impact bilingualism. As we have found out in working with this Handbook, research especially on bilingual adults often omits precise mention of respondents’ ages at time of study, thereby assuming that age or life stage does not matter. And although more and more studies are interested in comparing what are called early and late bilinguals, they tend not to list any information on when exactly study participants started to learn each of their languages. We found this generally cavalier attitude to reporting or taking into account actual ages surprising in light of the fact that, as most pointedly explained in Singleton and Pfenninger’s chapter, scholars have been centrally concerned with the factor of age of first exposure to a language. This concern, alas, has apparently not led to a general developmentally oriented view of bilingualism. We hope that this Handbook can contribute to such a more developmentally oriented view in bilingualism research, which recognizes that each life stage represents a different kind of general life experience for most people, and that this may be reflected in bilingual learning, unlearning, and bilingual use.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

6

AN NICK DE HOUWER AND LOU R DES ORTEGA

Broader Contexts and Opportunities for Bilingualism While different life stages will certainly exert different kinds of effects with regard to individual bilingualism, all bilinguals are directly or more indirectly affected by the global contexts in which their bilingualism develops. These contexts are highlighted in Part II of the Handbook. Fuller (Chapter 6) discusses how language ideologies often lead to linguistic inequality and hegemony, although there are also pluralist ideologies that celebrate bilingualism. Language ideologies are at the basis of specific language policies (Lo Bianco, Chapter 8). Among others, these will affect the ranges and kinds of language choices in education. Language ideologies will also help regulate what language(s) can be used in courtrooms, and will affect what happens to asylum seekers or other bilinguals in legal–lay communication (Angermeyer, Chapter 7). Grin (Chapter 9) explains how economic research can elucidate the complex dynamics of economic (dis)advantages of individual (and sometimes collective) bilingualism, and how policies encouraging linguistic diversity are an economic benefit all around. It is against the backdrop of particular societal language ideologies and contexts that parents may be raising bilingual children. In her interview study of six upper-middle-class highly educated Thai women in Japan, Nakamura (2016) admirably describes how the largely monoglossic ideology in Japan influenced these mothers’ language choice with their children and their own personally held language ideologies. Mothers favored Japanese in talking to their children, and also wanted to practice their own Japanese that way, given their opinion that Japanese was important for both their own and their children’s future earning potential. Similar language ideologies and earnings considerations may also have affected changed home language practices of lower-middle-class caregivers in 173 families in the United States. As part of their longitudinal study on language outcomes of young US children in homes where Spanish (possibly in addition to English) was spoken, Mancilla-Martinez and Lesaux (2011) reported that nearly half (47%) of the children’s parents (mostly mothers) or guardians spoke only or mainly Spanish at home when children were 4.5 years old. That percentage dropped to 22% when parents were asked again when their children were 11 years old. These drastic changes in parental home language choice may have been influenced by generally negative attitudes held against Spanish in the United States. The changes in parental home language choice in the 173 families studied by Mancilla-Martinez and Lesaux (2011) may have affected children’s language choice as well: at age 4.5, 45% of the children spoke mainly or only Spanish at home, but only 17% of the children still used mainly or only Spanish at home by age 11. Many spoke a lot more English as they grew older. In Nakamura’s (2016) study, hardly any of the Thai

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

Introduction

7

mothers’ young children spoke much Thai. Instead, they spoke mainly Japanese. These examples make clear that the contexts in which people become bilingual are of prime importance. Some of the main contexts for bilingual learning are described in Part III of the Handbook. These include naturalistic language exposure and input as relevant to young sequential bilingual children (Armon-Lotem & Meir, Chapter 10), bilingual language immersion programs in primary, secondary, and tertiary education (JuanGarau & Lyster, Chapter 11), foreign language instruction from early childhood to young adulthood (Mun˜oz & Spada, Chapter 12), and host country language-teaching programs for newly arrived adult migrants (Simpson, Chapter 13). Contexts may be conducive to learning a particular language, or to unlearning it, in the sense that it will be used less and at lower levels of proficiency. In the last chapter in Part III, Keijzer and de Bot (Chapter 14) examine some of the contexts in which such unlearning takes place. They focus both on children and on older adults. Keijzer and de Bot argue that total unlearning (or forgetting) is rare or nonexistent, and explain how previously partly unlearned languages can be relearned. The diverse learning contexts across the lifespan described in Part III lead to bilinguals being able to use their languages in dynamic ways. Part IV in the Handbook selectively highlights some of these. First there is the issue of how well bilinguals can use each of their languages, and how we should measure those bilingual abilities at different stages in life (Treffers-Daller, Chapter 15). While there are many bilinguals whose abilities in each language are quite uneven, with language X steadily better developed than language Y, other bilinguals are highly proficient in each language they have learned. Furthermore, while many people develop high abilities in three or four languages, others develop high abilities in several more. Biedron´ and Birdsong (Chapter 16) discuss the circumstances and conditions relating to this high-functioning bilingualism. Once bilinguals can interact in minimally each of two languages, there is the fundamental issue of what determines which language they will use at any given time (De Houwer, Chapter 17). For instance, even though the Thai mothers in Nakamura’s (2016) study did not all speak fluent Japanese, they found it impolite to speak Thai to their young children in the company of Japanese people who did not understand Thai. This resulted in mothers globally not speaking much Thai after migration to Japan, and only speaking more Thai again during rare visits to Thailand. Over a longer period of time, such habitual patterns of language choice may affect bilingual abilities. The ebb and flow of language choice across the lifespan may be such that bilinguals start to show signs of language attrition, that is, their ability in a language appears to diminish. In her discussion of L1 attrition in migrants who after age 12 moved to a place where their L1 was no longer the societal language, Ko¨pke (Chapter 18) explains how the relations between changes in language environments and language proficiency in mature bilinguals are

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

8

AN NICK DE HOUWER AND LOU R DES ORTEGA

far from straightforward. Like Goral (Chapter 5) in Part I, Ko¨pke is careful to distinguish nonpathological attrition from declines in language proficiency that are due to pathological conditions such as a stroke or dementia. These are discussed in chapters in Part V.

Connections among Disciplines and Research Areas As will have become clear from the description so far, individual bilingualism is dynamic and multidimensional. No single discipline can fully capture the complexity and heterogeneity of bilinguals’ language (un)learning and use, which affect many different aspects of life. The chapters in Part I through IV are written from several different disciplinary perspectives. For instance, Serratrice and Armon-Lotem and Meir write from a developmental psycholinguistic perspective; Gonc¸alves and De Houwer from a more sociolinguistic one; Fuller and Angermeyer from a perspective grounded in social anthropology; Juan-Garau, Lyster, Mun˜oz, and Spada from an applied linguistics perspective; and Grin from the perspective of economics. Our primary aim has been to strike a balance between social and psychological aspects of bilingualism. This is because the social cannot be separated from the individual. The learning, use, and unlearning of languages are socially embedded phenomena that rely on and materialize through the application of and reliance on sociocognitive processes. Straddling the social and psychological, there are also important applications to consider. For instance, the relation between learning opportunities and very young children’s early language learning trajectories needs to be taken into account in educational settings and in clinical practice. In policy decisions about provisions for adult later language learning, the role of language ideologies and the symbolic and market values of languages need to be acknowledged. The inter- and crossdisciplinary focus in the Handbook as a whole is highlighted and complemented in Part V. Here we asked authors to write about what they see as major foci of interest in bilingualism in their respective fields. The chapters in Part V are loosely chronologically ordered according to the life stages they discuss. Hammer and Edmonds (Chapter 19) discuss approaches to language impairment in early childhood and in somewhat older children (but they also discuss clinical aspects of bilingual aphasia). Garcı´a and Tupas (Chapter 20) turn to early to middle childhood and adolescence. They problematize how many schools treat children’s bilingualism and, like Bigelow and Collins in Chapter 2 earlier in the Handbook, advocate the use of pedagogies that respect and celebrate that bilingualism by allowing the use of all languages at school (Gonc¸alves and Simpson in respectively Chapter 3 and Chapter 13 similarly advocate such translanguaging approaches for adults). In her wider discussion of how second language acquisition research contributes to insights about bilingualism, Ortega (Chapter 21) focuses mostly on young

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

Introduction

9

adults and people in midlife, complementing the earlier chapters by JuanGarau and Lyster, Mun˜oz and Spada, and Simpson. As Ortega explains, second language acquisition research is often mired by an implicit deficit approach to bilingualism. Such a deficit approach is, sadly, not limited to academic circles, but is also present in clinical practice and education, as the chapters by Hammer and Edmonds and by Garcı´a and Tupas both denounce. Until the mid-2010s, Western public media outlets would generally express such deficit views as well. There has been a positive change in this regard, mainly thanks to work on bilingualism in cognitive science coming out of Ellen Bialystok’s lab in Canada. This work has claimed several cognitive advantages of bilinguals over monolinguals. These claims, however, have met with deep controversy within cognitive science. In his thorough review in Section 5 focusing on research with mainly young adults, Paap (Chapter 22) lays bare some of the arguments that have been made on either side, and stresses that regardless of the validity of claims of a purported bilingual cognitive advantage, bilingualism is highly valuable. Many of the bilingual processing studies that Paap discusses not only have a behavioral component but also examine neurophysiological evidence. Some of this work is reviewed in Hernandez’s Chapter 23, which focuses mainly on adults in midlife and later. Like Hammer and Edmonds, Hernandez discusses bilingual aphasia, but now from a processing perspective. In the final part of the Handbook, we placed what we called “bilingual connections,” that is, chapters that highlight the links between bilingualism and related areas of research. As Tang and Sze (Chapter 24) show in the first chapter in Part VI, research into sign language acquisition and bimodal bilingualism, involving a combination of oral and signed forms of language, has seen a burgeoning of attention and has the potential to shed unique light on traditional questions and preferred answers in the field of bilingualism research as a whole. The use of more than one language code also arises in the case of regional and social varieties that are not necessarily standardized, as individuals may speak both a regional variety and a standard variety broadly seen to belong to the same “language.” After a critical discussion of the problematic nature of satisfactorily distinguishing among varieties, Chevrot and Ghimenton (Chapter 25) explore the boundaries between bidialectalism and bilingualism and point to the fundamental role of social perceptions (attitudes) toward particular ways of speaking. Boundaries between different ways of speaking are also central in Aalberse and Muysken’s Chapter 26, but now from the perspective of contact linguistics. Bilinguals and bidialectals are the ultimate “locus” for language contact (see also De Houwer, to appear), and this individual language contact may, in the long run, lead to the creation of new ways of speaking that are transmitted from generation to generation. The final chapter in Part VI, Chapter 27 by Quay and Montanari, reintroduces the distinction between learning, using, and unlearning two versus more than two languages that the rest of the Handbook has

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

10

AN NICK DE HOUWER AND LOU R DES ORTEGA

generally ignored, given our inclusive approach to bilingualism as involving more than one language, regardless of the precise number. Quay and Montanari describe a number of settings in the world in which multilingualism rather than “strict” bilingualism is the expected reality. They emphasize several similarities between multilingualism and bilingualism in the strict sense, but point out that due to its larger variability, multilingualism is much more complex and dynamic. Importantly, they argue that there are generally no expectations of multilinguals performing like three or more monolinguals in one, whereas often for bilinguals, there is a societal expectation of linguistically performing like a monolingual in each language.

Bilingual–Monolingual Comparisons In our editorial work, we have met up with much research into individual bilingualism that we did not know. Both in that work and in research we were aware of previously, there is generally a large interest in comparing aspects of bilinguals’ language use and sociocognitive abilities to those of monolinguals. One issue here is that many studies do not provide sufficient background information to evaluate the extent of bilingualism or, indeed, monolingualism, of the study participants (see also Surrain & Luk, 2017). Some participants may have been placed in the monolingual participant group but would count as bilingual under our definition of age-appropriate levels of comprehension in two languages. For example, in many parts of today’s world, English as a foreign language is part of compulsory education and access to online media is free. In these contexts, it is difficult to find socalled true monolingual speakers, unless perhaps one samples participants who grew up linguistically insulated from English (but then this would render the age or milieu of such a monolingual baseline noncomparable to the bilinguals under study). Moreover, in many geographies there may be different varieties of what is often considered one language at the nationstate level. Few studies check whether supposedly monolingual participants are perhaps bidialectal. On a much more fundamental level, however, the question can be asked why bilingual–monolingual comparisons are needed to understand bilingualism. By definition, bilingualism is not monolingualism. Yet monolingualism is usually held up as a gold standard, and there is unfortunately even today a monolingual bias present in many studies of bilinguals. In the field of language attrition, for example, Ko¨pke (Chapter 18, this volume) notes that much research relies on bilingual–monolingual comparisons rather than on longitudinal studies, which she sees as more informative and preferable. In studies of early childhood bilingualism, many bilingual–monolingual comparisons are used to speak of so-called delayed or accelerated development, and it is always the monolinguals who are

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

Introduction

11

held up as the standard. Also here, bilinguals could and should be compared to each other. In second language acquisition studies too, the practice of comparing nonnatives to natives is ubiquitous, and language learners are unfairly compared to an idealized notion of native speakers who are monolingual and who show no traces of other languages (Ortega, 2014). Instead, adults learning an additional language should be viewed as bilinguals in the making, worthy of study in their own right. At a very deep level, a monolingual bias shows by the lack of identification of the language that study participants are tested in. Many studies often just assume that the term language (or French langue, or Dutch taal, or German Sprache, or Spanish lengua, etc.) refers to whatever happens to be the societal language of the setting under study. For instance, in Glennen’s (2015) study on the “language abilities” of five- to seven-yearold children from Eastern Europe adopted earlier by families in the United States there is no explicitation of what language children were tested in, although the instruments used were developed for English, and it is described that the adoptees were exposed to English after adoption. Of course, in the case of the internationally adopted children in Glennen (2015), it is a question to what extent children still understood or spoke the language that they had possibly started to acquire prior to adoption (see Keijzer & de Bot, Chapter 14, this volume, for discussion), but this does not relieve studies from the need to at least specify the language they were tested in. The term language abilities all too often refers to just the societal language. Even if studies identify the language tested, as Ortega argues in Chapter 21 (this volume), they show a monolingual bias if they focus only on one single language of the child or adult bilingual, typically the societal language that they are newly learning. As many chapters in this Handbook demonstrate (Treffers-Daller’s Chapter 15 especially), in order to obtain a full picture of a bilingual individual’s language functioning, all the languages that this person has learned must be taken into account. This is especially important in clinical settings (Hammer & Edmonds, Chapter 19). Many researchers thus implicitly and unwittingly adhere to what Fuller (Chapter 6) calls a monoglossic ideology. Finally, rather than seeing bilinguals and monolinguals as fundamentally different, scholars investigating people’s language learning and use should fully acknowledge the important insight that classic scholars such as Baetens Beardsmore (1982), Elwert (1959), and Grosjean (1982) expressed decades earlier, namely that the distinction between bilinguals and monolinguals is fuzzy, and that there is a continuum between being more bilingual and more monolingual (see also Serratrice, Chapter 1, this volume).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

12

AN NICK DE HOUWER AND LOU R DES ORTEGA

Closing Words It has been impossible to cover all the topics and foci relevant to individual bilingualism in a single volume. Yet our selection is meant to cover a broad spectrum, sustaining the lifespan perspective throughout as much as possible. Our particular choices have been guided by a mix of our personal interests and expertise, our assessment of what is generally of interest to scholars working on bilingualism, and our evaluation of whether sufficient research material was available for a particular topic to warrant inclusion in the Handbook. There is, however, sufficient research material to fill several handbooks. Bilingualism research is vibrant and done in many different fields, from migration studies to neurology. We hope this Handbook can inspire scholars from many diverse fields to contribute to this vast and exciting body of work.

Acknowledgment We are grateful to Wolfgang Wo¨lck for his insights and advice throughout our work on this Handbook.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:02:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.001

1 Becoming Bilingual in Early Childhood Ludovica Serratrice

1 Introduction The study of language development in children who acquire two languages, whether from birth (bilingual first language acquisition [BFLA]) (De Houwer, 1990), within the first year of life (Paradis, Genesee, & Crago, 2011), or later in early childhood (early second language acquisition [ESLA]) (De Houwer, 1990), has flourished since the early 1990s. Research on corpus data, and experimental work on a wide range of topics, from speech perception to lexical learning, and from morphological and syntactic development to discourse-pragmatic adequacy, has refined our knowledge and understanding not only of language development in bilingual children, but of the essence of language development itself and of the human mind’s capacity for learning. Although both BFLA and ESLA are ways of becoming bilingual in early childhood, defined here as the period between ages zero and six, there are fundamental differences between these two contexts of acquisition. In BFLA, the general consensus is that children essentially treat their two languages as two first languages. Although young BFLA children usually behave in language-specific ways in each of their languages, they may also show crosslinguistic influence across their two languages. Overall, though, the same developmental mechanisms that apply to monolingual first language acquisition also apply to BFLA. At the same time, acquiring two languages must be different from acquiring only one. In a number of restricted contexts, BFLA children understand and produce language in ways that are significantly different from the ways in which monolinguals do (Serratrice, 2013). In the case of ESLA, children are by definition exposed to a second language after they have started acquiring a first language to some extent,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

16

L U D O V I C A S E R R AT R I C E

and the contexts in which they hear and use their L1 and L2 are likely to be quite different. While BFLA children are typically exposed to two languages in their home environment, ESLA children tend to hear a first language (L1) at home, and only later are they exposed to a second language (L2) when they start attending some regular form of childcare in the community. Recent estimates show BFLA to be three times as frequent as ESLA (De Houwer, 2007; Winsler, Burchinal, Tien et al., 2014). As for an upper cut-off point for ESLA, that is, where second language acquisition (SLA) ceases to be early SLA, different authors have proposed different thresholds between the ages of three (Paradis et al., 2011) and six (De Houwer, 1990), the latter coinciding with the age at which many children start to get formal literacy instruction. Although the linguistic development of bilingual children is still, by and large, measured against the monolingual standard, there is a gradual recognition within the field, and in non-academic circles too, that the bilingual language experience can and, indeed, should be viewed in its own right and not just as an atypical case of language development, as still often is the case. Following classic authors such as Baetens Beardsmore (1982) and Grosjean (1982), scholars have been advocating that bilingualism should be thought of as a gradient measure, rather than as a dichotomous category framed in terms of bilingual versus monolingual (Luk, 2015). Factors such as age of first exposure to each language, length of exposure, current and cumulative amount of input and output feature as critical variables in creating the language profile of bilingual children along a continuum (see Armon-Lotem & Meir, Chapter 10, this volume, for discussion of such factors). Conceptualizing bilingualism as a continuous measure is likely to provide researchers with a more nuanced understanding of the complex picture of language development in bilingual children. This chapter presents a selective overview of the path to language in the years before the start of formal education in children who are born into a bilingual environment, and therefore are simultaneously exposed to two languages (BFLA children), as well as in children who are exposed to a second language later in early childhood (ESLA children). Given the breadth of early childhood bilingualism research, the coverage is necessarily restricted to a limited number of topics. Many important aspects of bilingual development are not dealt with in detail, or at all. These include phonological development in production (Vihman, 2014), bimodal bilingualism where one of the children’s languages is a sign language (Baker & van den Bogaerde, 2008; see Tang and Sze, Chapter 24, this volume), the learning of a new language by young international adoptees who no longer have any access to the language of their country of origin (Genesee & Delcenserie, 2016; see Keijzer & de Bot, Chapter 14, this volume, for discussion of somewhat older international adoptees), and the controversial issue of the purported bilingual executive advantage (Valian, 2015; see

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

17

Paap, Chapter 22, this volume, for extensive discussion as regards adults). The topic of code mixing is partially addressed in connection with the issue of language separation. The chapter covers the following topics in some detail: speech discrimination and the relation between speed of processing and lexical development, the role of gestures accompanying language, determinants of lexical learning, and the relation between vocabulary and grammar within and across languages. More research with preschoolers has been carried out with BFLA than ESLA children. The studies included here reflect this asymmetry.

2 Speech Discrimination First words represent a fundamental linguistic and cognitive developmental milestone, and one that for parents typically signals the start of a child’s journey to language. However, well before children utter their first words they have already learned a great deal about their linguistic environment. There is good evidence that the early perceptual foundations of word learning start prenatally; in the last trimester of gestation, fetuses start processing the prosodic contours of language and music that are well preserved across the abdominal barrier (Shahidullah & Hepper, 1994). The segmental aspects of speech are not yet accessible at this time. Across the more than 7,000 documented languages of the world, three rhythmic classes have been established: stress-timed languages (e.g., English), syllable-timed languages (e.g., Spanish, Tagalog), and moratimed languages (e.g., Japanese). Monolingual newborns can discriminate between the language to which they have been exposed in utero (their “native” language) and an unfamiliar language from another rhythmic class, but they cannot discriminate languages from the same rhythmic class (Mehler, Jusczyk, Lambertz et al., 1988). Studies on newborns who have been exposed to two languages from different rhythmic classes in utero show that such prenatal bilingual exposure affects infants’ preferences soon after birth. Byers-Heinlein, Burns, and Werker (2010) tested the preference of 15 newborns from an English monolingual background and 15 neonates born to English–Tagalog bilingual mothers. Unlike monolingual newborns, who displayed a significant preference for English, the infants of bilingual mothers were equally interested in Tagalog and English. This is consistent with their prenatal experience of both languages. In addition, Byers-Heinlein et al. (2010) showed that the bilingually (as well as the monolingually) exposed newborns could reliably discriminate between Tagalog, a syllable-timed language, and English, a stresstimed language, presumably on the basis of interclass rhythmic differences. Although bilingual infants can discriminate between languages from different rhythmic classes, it is less clear whether they can also

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

18

L U D O V I C A S E R R AT R I C E

discriminate between languages that belong to the same rhythmic class, and, if so, whether they do so on the basis of segmental (i.e., phonemic and/ or phonotactic differences) or rhythmic information (i.e., the distribution of vocalic and consonantal intervals, and the variability of consonantal intervals). Spanish–Catalan four- and five-month-old BFLA infants can reliably distinguish between these two syllable-timed languages (Bosch & Sebastia´n-Galle´s, 2001), but in this language pair infants have access to both segmental and rhythmic cues and therefore it is not clear which type of cue they rely on. At the earlier age of 3.5 months, Basque–Spanish bilinguals can reliably discriminate between their two languages, both also syllable-timed, on the basis of intra-class rhythmic differences alone (Molnar, Gervain, & Carreiras, 2014). Molnar et al. (2014) used low-pass filtered Basque and Spanish speech as stimulus material that prevented infants from accessing segmental cues but that preserved rhythmic characteristics distinguishing Spanish from Basque: the distribution of vocalic intervals, the variability of consonant intervals, and the intensity and duration of stressed vowels carrying phrasal prominence. Thus, by the tender age of 3.5 months, bilingual exposure already determines BFLA infants’ ability to discriminate between their two languages on the basis of suprasegmental acoustic cues. Whether this ability is already present at birth or whether additional language exposure is necessary for intra-class distinction is not yet known. By the middle of their first year, infants cannot only discriminate between languages, but also between different language varieties (Chevrot & Ghimenton, Chapter 25, this volume). While it appears that BFLA children can rely on rhythmic and prosodic information for overall language discrimination, we know considerably less about how the bilingual experience affects phonemic learning. Research on monolingual infants has shown that children are born as “universal learners,” with the ability to acoustically discriminate between phonetic units in any language, not just the language they are hearing (Eimas, Siqueland, Juscyk, & Vigorito, 1971). By the end of the first year of life, this ability is greatly reduced and children’s sensitivity to phonetic contrasts that do not represent phonemic contrasts in their input language is on the wane (these are called nonnative contrasts). At the same time, infants’ ability to discriminate native contrasts, that is, phonetic contrasts in the ambient language that represent phonemic difference, is fine-tuned (Werker & Lalonde, 1988). This narrowing down of phonetic, acoustic discrimination into phonemic categories is instrumental to lexical development as it allows children to focus on the contrasts that are meaningful and relevant for the language they are in the process of learning. The findings for BFLA infants are mixed in this regard, and whether they follow the same developmental trajectory as monolinguals when it comes to phonemic discrimination is not yet clear. Some studies report bilingual–monolingual developmental differences, with BFLA children

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

19

going through a U-shaped learning curve where they are unable to discriminate vowel or consonant contrasts in one of their languages that they successfully identified at a younger age (Bosch & Sebastia´n-Galle´s, 2003). Dealing with two languages entails more complex and variable input and requires specific perceptual reorganization processes. These extra processing demands may account for bilingual children’s temporary inability to perceive some phonemic distinctions. Other behavioral studies have shown that bilingual children can discriminate phonemic contrasts in both of their languages throughout their first year of life and beyond (Sundara, Polka, & Molnar, 2008; Sundara & Scutellaro, 2011), while monolinguals are unable to discriminate nonnative contrasts past the age of six to eight months (Burns, Yoshida, Hill, & Werker, 2007). Methodological differences and variation in language combinations and phonemes tested are at least partly responsible for these discrepant results. The use of techniques that allow the recording of brain activity, as opposed to behavioral responses, is a promising way to tap into the neural mechanisms that underlie speech perception and language processing in young infants. Whole-head magnetoencephalography whereby the magnetic fields produced by an infant’s entire brain are measured showed a neural sensitivity to both languages in 11-month-old Spanish–English bilinguals (Ferjan Ramı´rez, Ramı´rez, Clarke et al., 2017), while agematched English-hearing monolinguals discriminated syllable contrasts only in their native language. Furthermore, unlike monolingual infants, who showed phonemic discrimination in their single language at the neural level, bilinguals showed acoustic but not phonemic discrimination of native and nonnative contrasts (Ferjan Ramı´rez et al., 2017). In early research on early perception, acoustic discrimination is generally seen as a lower-level, earlier, language-universal stage of development that precedes higher-level, phonemic, language-specific discrimination. As Ferjan Ramı´rez et al. (2017) argue, however, “remaining in the universal listening stage indicates a highly adaptive brain response, and a distinct advantage in a situation when two languages are being mapped simultaneously” (p. 12). Another promising neurophysiological method to investigate the transition from universal to language-specific phonemic discrimination is the use of electroencephalography (EEG), which records electrical brain activity, and the analysis of the resulting brain wave patterns. Different types of patterns mean different things. In auditory discrimination studies, scholars investigate the Mismatch Response (MMR), which refers to the brain’s detection of a change in an auditory stimulus after hearing a series of repeated identical stimuli. Of relevance here is the distinction between quick, early brain responses (“components”) and somewhat later responses. The early positive component of the MMR (pMMR), which occurs approximately between 100 ms and 260 ms after a study participant has heard a novel stimulus, indexes a less mature, acoustic level of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

20

L U D O V I C A S E R R AT R I C E

neural encoding and discrimination. The later negative component (nMMR), which occurs between 260 ms and 460 ms post stimulus, signals a more mature phonemic level of analysis. At the age of seven months, English-learning monolingual infants displayed an EEG pattern of neural discriminatory responses for both a native contrast (English) and a nonnative contrast (Spanish), but by the age of 11 months, neural responses showed discrimination only for the native contrast (RiveraGaxiola, Silva-Pereyra, & Kuhl, 2005). Using the same methods and stimuli with a group of Spanish–English bilingual infants, Garcia-Sierra, RiveraGaxiola, Percaccio et al. (2011) found that, unlike monolinguals, the bilinguals did not show any neural discrimination at 6 to 9 months but did so later, at 10 to 12 months. More interestingly, the strength of the nMMR was positively correlated with the amount of input in each language for the bilingual infants. Parental estimates of relative exposure to English showed that by the age of 10 to 12 months, those bilingual infants who had high levels of English input had a stronger nMMR than they did at 6 to 9 months. For the bilinguals with low English input, no significant nMMR increase was observed. The same applied to Spanish: age effects for the increased strength of the more mature nMMR were only found for the children whose parents reported a higher relative exposure to Spanish. Similar findings are reported by Garcia-Sierra, Ramı´rez-Esparza, and Kuhl (2016) in an EEG study in which they measured the pMMR and the nMMR of 19 Spanish–English bilingual and 18 English-hearing monolingual infants at 11 and 14 months of age, all of whom heard native (English and Spanish for the bilinguals; English for the monolinguals) and nonnative contrasts as experimental stimuli (Chinese for the bilinguals; Spanish and Chinese for the monolinguals). Garcia-Sierra et al. also collected direct measures of the amount of input for all children using the Language Environment Analysis System (LENA Foundation, Boulder, Colorado; see www.lena.org) and divided children into a high-input and a low-input group based on counts of the words parents addressed to them. The monolingual children in the high-input group showed a more mature phonemic level of analysis for contrasts in their native language, English, as evidenced by an nMMR. In contrast, in response to native contrasts, the monolinguals in the low-input group and the bilinguals in the high-input group only showed the positive component of the Mismatch Response (pMMR), which is indicative of attentional mechanisms that precede automatic speech perception but that do not yet show neural commitment (see later in this chapter) to the sounds of the native language(s). The Garcia-Sierra et al. (2011) and Garcia-Sierra et al. (2016) studies are particularly important as they provide empirical evidence for the crucial role of language input in the early stages of the developmental timeline, up to 12 or 14 months of age. The transition to a more mature phonemic encoding of the speech stimulus, that is, a stronger nMMR, was attested only for children who had been exposed to a high amount of speech input.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

21

The inclusion of input quantity information as measured by directly observed parental word counts for the monolingual group shed more light on the role of quantity of input in early language development. Even receiving all of language exposure in one language (as a monolingual) is no guarantee that the amount of input will be sufficient to trigger mature phonemic encoding of speech stimuli. Whether children around 12 months of age can move toward mature phonemic encoding has important consequences for the course of later language development (see Armon-Lotem & Meir, Chapter 10, this volume, on the role of quantity of input in childhood bilingualism). Behavioral and brain studies with monolingual infants have repeatedly shown a positive correlation between the discrimination of native contrasts and later language development (Tsao, Liu, & Kuhl, 2004). In their longitudinal study investigating the relation between early speech discrimination and later lexical skills in bilinguals, Garcia-Sierra et al. (2011) confirmed these monolingual patterns with a group of 27 Spanish–English bilingual infants growing up in Texas, where approximately 30% of the population speak Spanish at home. Children’s word production at 15 months in both English and Spanish was positively correlated with earlier measures of phonemic discrimination. Similar results were obtained by Silve´n, Voeten, Kouvo, and Lunde´n (2014), who found that better speech discrimination of Finnish sounds at seven months by Russian–Finnish bilinguals in Finland predicted higher levels of both Finnish and Russian word production in the second year of life. According to Kuhl (2004), early phonemic learning in the native language results in the commitment of neural networks dedicated to encoding language-specific contrasts. The same mechanisms are likely at work for two first languages: this early neural commitment facilitates subsequent learning of patterns that are consistent with those that the infant has already learned (native contrasts) and it filters out those that do not confirm to the familiar patterns (nonnative contrasts). This special relation between early speech perception in the second half of the first year of life and later lexical learning depends on the ability to perceive phonemic distinctions and to successfully associate a sound pattern with a lexical item. An individual infant with advanced speech perception skills in one or both of their languages should therefore show advanced word-learning skills in the same language(s). The ability to isolate discrete, meaningful speech sounds thus is a prerequisite for lexical learning. In addition to building comprehension vocabulary skills in both languages, bilingual children also need to learn to say words. This requires the production of language-specific phonemes that are the building blocks of words in each of their languages (see chapter 8 in Vihman, 2014, for a comprehensive overview of early phonological development in bilingual children). However, word learning is clearly more than just the recognition or articulation of a sequence of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

22

L U D O V I C A S E R R AT R I C E

phonemes in the speech stream, and occurs through interaction and exposure. Fifteen-month word production of the bilingual infants in Garcia et al.’s (2011) study was correlated not only with earlier phonemic discrimination but also with the relative amount of language exposure in each language as measured via parental questionnaires. The next section considers bilingual children’s lexical learning in more depth and further highlights the role of exposure.

3 Lexical Learning 3.1

The Size of Bilingual Children’s Vocabularies and Lexical Growth Vocabulary size is an important measure of young children’s level of language development. Lexical learning, both in word comprehension and in word production, is characterized by a high degree of individual variation in both bilingual and monolingual children (Bates, Dale, & Thal, 1995; De Houwer, Bornstein, & Putnick, 2014). Over and above this interindividual variation, there are both discrepancies and similarities between bilingual and monolingual children with respect to vocabulary size. Researchers have compared just one of the languages bilinguals are acquiring with the single one monolinguals are learning, or they have compared total vocabulary, which combines the vocabulary of both languages for bilinguals. The instrument used most frequently is the MacArthur-Bates Communicative Development Inventory (CDI), a standardized parent report form (Fenson, Dale, Reznick et al., 1993). In a carefully controlled study, De Houwer et al. (2014) used Dutch and French adaptations of the CDI to collect data on comprehension and production vocabulary for 31 Dutch–French BFLA children and 30 Dutch monolinguals in Belgium at the ages of 13 and 20 months. De Houwer et al. matched the bilingual and monolingual children in their study on a number of variables known to affect lexical development in monolinguals (birth order and birth status, age, length of language exposure, family socioeconomic status [SES], and gender), so that they could isolate the effect of hearing one versus two languages from birth (see TreffersDaller, Chapter 15, this volume, on the need to carefully control for relevant sociodemographic variables in any bilingual–monolingual group comparisons). The only differences De Houwer et al. found in this middle-class group of children were a comprehension vocabulary advantage for the bilinguals at 13 months – when both of their languages were combined – and a monolingual advantage for Dutch total word knowledge at 20 months, combining words that were only understood and words that were both produced and understood (no differences emerged for Dutch comprehension or Dutch production separately). For several additional comparisons, there were no bilingual–monolingual differences. De

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

23

Houwer et al. conclude that bilingualism per se does not slow down the course of lexical development; any significant bilingual–monolingual differences in the absence of known independent causes of delay such as hearing problems and developmental disorders are most likely to be attributed to the quantity and quality of children’s input in each of their two languages. Indeed, a strong predictor of variation in early lexical development in both bilingual and monolingual children is the variability in the quantity and quality of child-directed caregiver speech (Hoff & Core, 2013; Rowe, 2012; see also Armon-Lotem & Meir, Chapter 10, this volume). Such quantity and quality may be mediated by parental SES (Hart & Risley, 1995). As reviewed by DeAnda, Arias-Trejo, Poulin-Dubois et al. (2016), some studies find that, in the second year of life, monolingual Englishand Spanish-speaking US middle- to high-SES infants develop language faster and reach higher levels of vocabulary than lower SES peers, while others do not find an SES difference. In their own studies of bilingually raised infants between 15 and 18 months of age investigating links between SES, the proportion of exposure to the tested language, and comprehension and production abilities in the language tested, DeAnda et al. likewise find conflicting results as regards the role of SES. This study is one of the first to systematically investigate the role of SES in very young bilinguals. It is too early to tell whether the early bilingual comprehension advantages found by De Houwer et al. (2014) may be relatable to the fact that children were growing up in middle-SES families, where parents were perhaps more verbal and thus offering children more word-learning opportunities (however, De Houwer, 2014, shows great variability among the middle-class mothers of the children studied in De Houwer et al., 2014, as regards talkativeness). Other studies that, like De Houwer et al. (2014), mostly report a lack of significant differences in the vocabulary skills of bilingual and monolingual preschoolers include Comeau and Genesee (2001); Legacy, Zesiger, Friend, and Poulin-Dubois (2016); Poulin-Dubois, Bialystok, Blaye et al. (2013); and Sundara, Polka, and Genesee (2006). It is likely not a coincidence that all of these studies include French–English bilingual children in Canada, a country where both languages enjoy high prestige and institutional support. The status of each of a bilingual’s two languages likely has implications for lexical learning. Language status has been proposed as an explanation for the lack of bilingual–monolingual differences into the school years for French–English bilinguals in Canada (Thordardottir, 2011; Smithson, Paradis, & Nicoladis, 2014). The favorable sociocultural environment and the educational support that both French and English receive in bilingual regions in Canada create a language learning environment that is highly conducive to bilingualism. This optimal context is not so easily reproducible for other language combinations in other regions of the world where there is a clear asymmetrical difference between a bilingual’s two languages in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

24

L U D O V I C A S E R R AT R I C E

terms of status, and where one language can be seen as a minoritized one (see Fuller, Chapter 6, this volume, and Lo Bianco, Chapter 8, this volume). This has inevitable consequences for the quantity and quality of exposure and the opportunities for uptake and learning. While measurements of children’s lexical knowledge in comprehension and production provide a snapshot of children’s skills at a given point in time, they present only partial information about their evolving language skills. Such measurements do not give any indication of developmental trajectories and of whether they are different between bilingual and monolingual children, or between bilingual children’s two languages. Hoff, Rumiche, Burridge et al. (2014) carried out a longitudinal study of the development of production vocabulary in a high-SES sample of 26 Spanish–English bilingual and 31 English-speaking monolingual children in Florida from 22, 25, and 30–48 months of age. They considered how the role of parental language status affected children’s lexical growth over time. In the bilingual families, children had heard two languages at home from birth, either from parents who were both native Spanish-speaking or from parent pairs where one parent was a native Spanish speaker and the other a native English speaker (classifications were made on the basis of self-report, country of birth, and speaker status of the parents’ parents). In the monolingual families, all children had heard only English at home from birth, and all parents except two fathers were native English speakers. At 22, 25, and 30 months, data collection relied on English and Spanish versions of the CDI. At 48 months, the researchers administered a picture vocabulary test (in both languages for the bilinguals). A comparison to monolingual norms for English on this test revealed that on average, the bilingual children were performing at around the 50th percentile; the English percentile scores of the monolingual sample in Hoff et al. did not differ from those of the bilinguals with one native English-speaking parent, but were significantly higher than those of the bilinguals with two native Spanish-speaking parents (for discussion of using monolingual norms with bilinguals, see Treffers-Daller, Chapter 15, this volume). At all the time points, bilingual children’s total production vocabulary as measured in raw scores and combining Spanish and English was at the same level as that of the monolinguals, but bilingual children from homes with two native Spanish-speaking parents made more gains across time compared to monolinguals and bilinguals with just one native Spanishspeaking parent. This suggests that the children who heard Spanish at home from two native-speaking parents maintained their English skills and at the same time learned more Spanish words than children in families with only one native Spanish-speaking parent. The relation between parental language status and skills in the minority language has been reported elsewhere (Gathercole & Thomas, 2009). When English vocabulary was considered only in terms of relative change using z-scores calculated over the whole sample, it was the children with one native

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

25

Spanish-speaking parent who had a significant and consistent longitudinal advantage over the children with two native Spanish-speaking parents. Both groups of bilinguals, however, increased more slowly in English compared to English-speaking monolingual peers. An important conclusion from the Hoff et al. (2014) study was that in bilingual households with one native Spanish speaker and one native English speaker, there was a positive correlation between English use at home – presumably predominantly from the native-speaking parent – and children’s English vocabulary, and a weak and nonsignificant negative correlation with Spanish vocabulary. In contrast, in families where both parents were native speakers of Spanish, the use of nonnative English at home did not correlate positively with growth in English vocabulary, but was strongly and negatively correlated with Spanish vocabulary. In essence, hearing nonnative input in English did not appear to benefit children in English; the only effect was that input in native Spanish appeared to be reduced and therefore vocabulary skills in Spanish were adversely affected. This conclusion is in line with previous findings that nonnative input is not as conducive to language development as native input (Hammer, Davison, Lawrence, & Miccio, 2009; Paradis, 2011). However, as argued by Armon-Lotem and Meir in Chapter 10 of this volume, there are many kinds of nonnative input (the quality of input is likely not dichotomous but must be a gradient, just as bilingualism is), and without documenting the proficiency of the so-called nonnative parents it remains a question what it is exactly in the input that affects children’s language development. In another longitudinal study, this time of younger Russian–Finnish bilinguals in Finland between the ages of 14 and 36 months, Silve´n et al. (2014) also showed smaller Finnish vocabulary size in production and slower growth rate for the bilinguals than for Finnish monolingual peers. However, when total vocabulary was considered, the Russian–Finnish bilinguals produced more words at 18 months and were learning new words faster than the monolinguals. The bilinguals’ rate of vocabulary growth was slower in Russian, the minority home language, compared to Finnish, the language children encountered not only in the home, but also in daycare and preschool. Longitudinal studies of early bilingual vocabulary development thus show that growth in the minority language is particularly sensitive to input, a conclusion also put forth by Armon-Lotem and Meir (Chapter 10, this volume). Children in households where they hear native input in the minority language from both parents – which may have been the case for some of the children in the Hoff et al. (2014) study – may follow steeper growth trajectories in the minority language than those who are exposed to native input in the minority language by only one parent – the case for all of the children in the Silve´n et al. (2014) study. Confirmation of this would have important implications for home language maintenance and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

26

L U D O V I C A S E R R AT R I C E

the opportunities that children need to sustain growth in both of their languages. As there will be individual differences across both native and nonnative speakers, what would really advance the field would be more nuanced studies of native and nonnative child-directed input alike. Word learning, both in comprehension and in production, is a lifelong undertaking. Differences that may be relatively small or nonexistent between bilingual and monolingual children at a younger age, when the range of experiences and potential topics of conversation are still rather limited, tend to become more pronounced as children grow older and their social circle expands. This is particularly true in the school years when learning academic vocabulary is likely restricted to the language of schooling, and when opportunities for rich and meaningful interactions with peers tend to be in the main societal language (see further Bigelow & Collins, Chapter 2, this volume). Research on lexical comprehension skills in primary school-aged children consistently reports a gap between sameaged bilingual and monolingual children, with the former showing smaller vocabularies in the school language, which is a second language for many bilingual children (Bialystok, Luk, Peets, & Young, 2010). When bilingual children do not grow up in contexts in which the two languages are spoken in the wider community (e.g., in Catalonia in Spain where both Catalan and Spanish are societal languages), they tend to have fewer opportunities in the (minority) home language and as a consequence show slower growth in that language, resulting in fewer minoritylanguage than majority-language words in their lexicon (Silve´n et al., 2014). Deliberate parental strategies fostering a rich array of language and literacy activities in the minority language can counter vocabulary impoverishment: early literacy activities are known to make a significant difference in promoting vocabulary growth in monolinguals (Raikes, Alexander Pan, Luze et al., 2006). Nurturing the development of the home language in minority contexts, however, requires time and a concerted effort that may not be within reach for all bilingual families. The following section further highlights the role of language exposure in lexical learning.

3.2

The Role of Processing Speed and Relative Exposure for Lexical Learning For monolingual children, it has been found that their ability to identify familiar words in the speech stream and the speed with which they do this (processing speed) are positively correlated with concurrent vocabulary size and with lexical growth in the second year of life (Fernald, Perfors, & Marchman, 2006). Input from caregivers influences both vocabulary knowledge and lexical processing skills; in turn, the bidirectional relationship between vocabulary and processing speed mediates the efficient uptake of information in the speech signal (Hurtado, Marchman, & Fernald, 2008).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

27

Monolingual children are by definition exposed to a particular language for 100% of all language exposure. Bilingual children’s exposure to a particular language by definition amounts to less than 100% of all language exposure. In order to fully take into account that the language experience, knowledge, and processing of young bilinguals is holistically distributed across two languages, scholars typically measure exposure in relative terms, that is, in terms of the amount of time that, for instance, children spend in an English-speaking environment hearing English, as opposed to the amount of time spent in a Spanish-speaking environment where the other language is primarily heard (see Armon-Lotem & Meir, Chapter 10, this volume, for discussion of advantages and disadvantages of using relative measures of exposure). In the first study to investigate real-time language comprehension in BFLA toddlers, Marchman, Fernald, and Hurtado (2010) showed that – similarly to monolingual children – lexical processing speed and vocabulary size were positively correlated within each of the two languages of 26 30-month-old Spanish–English bilinguals growing up in California. Hurtado, Gru¨ter, Marchman, and Fernald (2014) additionally investigated the role of language exposure and tested another group of 37 Spanish–English BFLA children in California at the ages of 30 and 36 months. An important aspect here is the absence of monolingual controls, and the use of relative measures not only for language exposure but also for aspects of children’s language. For vocabulary size Hurtado et al. (2014) used a relative measure to derive an index of vocabulary knowledge reflecting how many Spanish words children knew relative to the number of English words. The same logic applied to processing speed, where the measure indexed how fast children were in the lexical processing task in Spanish relative to how fast they were in the same task in English. Positive correlations were found between vocabulary knowledge and processing speed, that is, those children who knew relatively more words in Spanish were also relatively more efficient at processing Spanish words; this replicates earlier findings on the relation between lexical knowledge and lexical processing in monolinguals. Relative amount of exposure in each language was also positively correlated with vocabulary size: children with relatively more exposure in Spanish knew relatively more words in Spanish, and children with relatively more exposure to English produced more English words relative to Spanish words. This result held constant for the parental questionnaire measures at 30 months and the standardized language measures at 36 months. Relative speed of lexical processing was also positively related to relative amount of exposure, thus showing that it is not just knowledge – as measured by relative vocabulary size – that is significantly affected, but also the ease with which children process words in the input. There was also a longitudinal relation between relative exposure, relative speed of processing, and comprehension and production vocabulary: relative exposure and speed jointly predicted production

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

28

L U D O V I C A S E R R AT R I C E

vocabulary, while speed of processing was the most significant predictor for comprehension vocabulary. Whether the predictive power of language exposure and speed of processing for vocabulary lasts beyond bilingual children’s early childhood years, as has been shown for monolinguals (Marchman & Fernald, 2008), is not yet clear.

3.3 Language Differentiation An issue that is likely related to children’s processing speed in each of their languages, the size of bilingual children’s vocabularies, and the rate of lexical growth is the relative independence of the two languages that children are learning. Whether BFLA children start with an undifferentiated lexicon or whether they can discriminate between their two languages right from the start holds important cues to the organization of language in the mind of a bilingual learner. As described in Section 2, in early speech perception bilinguals do appear to make a distinction among their languages from the first year of life. For children’s language production, two criteria have typically been adopted to investigate the language differentiation issue: language choice, and the presence of translation equivalents (TEs). Language choice refers to children’s ability to choose language-specific words in context-sensitive ways, that is, for instance, using Italian words when speaking to an Italian-speaking interlocutor, and German words with a German-speaking interlocutor. Like bilingual adults, bilingual children have been shown to mix their languages, that is, to not stick to just one language in speaking to a particular person (see also De Houwer, Chapter 17, this volume). Such mixing was originally interpreted as a sign of an undifferentiated system where children would indistinctly draw from their two languages regardless of the appropriateness of doing so (Volterra & Taeschner, 1978). In a seminal paper, Genesee (1989) argued against this interpretation of language mixing and proposed that evidence of differentiation of languages in bilingual development was predicated upon the functional separation in different contexts of language use. In his reanalysis of previous evidence Genesee (1989) showed that although children mix, they do so much less frequently than if they were just randomly drawing lexical items from their two languages. Rather, there is robust evidence that even very young bilingual children make language choices similarly to adults, that they choose their languages as a function of topic and addressee, and that they resort to code-mixing to fill lexical gaps (Lanvers, 2001). For instance, bilingual children are sensitive to the pragmatic need to use Spanish with a Spanish speaker and English with an English speaker, and they make language choices accordingly (Deuchar & Quay, 2000). Already at the one-word and two-word stages, BFLA children are capable of making appropriate language choices not only with familiar adults but also with strangers (Genesee, Boivin, & Nicoladis, 1996), and on the basis of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

29

varying rates of mixing by the interlocutors themselves (Comeau, Genesee, & Lapaquette, 2003). Thus, BFLA children can pragmatically differentiate between their languages from the earliest stages of development. Appropriate language choice rests, at least in part, on the availability of translation equivalents in bilingual children’s lexicons, that is, on words for the same meaning that children acquire in each language (e.g., English horse and Italian cavallo). TEs are germane to the debate over whether bilingual children start out with one undifferentiated lexical system, or whether language differentiation starts early on. The presence of two labels for the same meaning lends support to the latter. It also shows that a constraint like mutual exclusivity, which refers to preschoolers’ preference for mapping words and meanings one-to-one and their general resistance to learning synonyms, and which has repeatedly been shown to apply in monolingual lexical development, is superseded by bilingual children’s pragmatic need to have two language-specific labels for the same concept (Kalashnikova, Mattock, & Monaghan, 2015). Although there is considerable inter-individual variation in that children with larger vocabularies understand more TEs, TEs are already present in BFLA children’s comprehension vocabularies by the age of 13 months (De Houwer, Bornstein, & De Coster, 2006). In production, even before BFLA children can say 50 words, TEs typically account for between 20% and 30% of their total lexicon, and form–meaning mappings in both languages are facilitated by phonological similarity. In a study of 24 Catalan–Spanish BFLA children aged 18 months, Bosch and Ramon-Casas (2014) showed that TEs that were identical in phonological form (e.g., [si], yes) accounted for 28% of the words in the bilinguals’ total lexicon (the sum of their Catalan and Spanish words). In contrast, TEs that were similar (e.g., [pan] versus [pa], bread) or dissimilar (e.g., [pero] versus [gos], dog) made up less than 2% of children’s total vocabulary, and they were only present in children with larger vocabularies. The TE findings show that BFLA children do not have an undifferentiated lexicon; at the same time, children learning phonologically closer languages with large numbers of cognates benefit from double the exposure to a given phonological form–meaning mapping, and hence are bound to have a larger proportion of TEs.

4 The Relation between Gesture and Language Development In the early stages of language development, words are not the only means through which children communicate. Between the ages of 8 and 12 months, monolingual children start to use holdout gestures, that is, gestures where an object is held and extended as if to offer it (CameronFaulkner, Theakston, Lieven, & Tomasello, 2015) and use index finger

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

30

L U D O V I C A S E R R AT R I C E

pointing with a referential intent. In the second year of life, they also use iconic gestures. These are gestures that convey actions or attributes associated with objects. Examples are moving the index and middle fingers to indicate walking, or placing two hands with extended fingers to the side of each ear to represent a rabbit. The use of gesture is related to vocabulary learning. One of the ways in which both are explicitly connected in the input occurs when parents translate children’s gestures into their lexical equivalents. In a study of 10 parent–child dyads of English–Spanish BFLA children between 2.5 and ¨ zc¸alıs¸kan, and Hoff (2016) showed that, just like 3.5 years of age, Mateo, O for monolinguals, bilingual children used deictic and “give” gestures to request something before they used particular words, and that parents translated those gestures into words, thus offering children lexical learning opportunities. Crucially, those referents for which parents translated from gesture to word were more likely to appear in children’s vocabulary than gestures that had not been translated. One of the questions that has preoccupied researchers investigating the relation between gesture and language is whether the onset and development of gestures in young children are determined by cognitive or rather by linguistic development. In the case of monolingual children, this is a difficult question to answer as the two go hand in hand. As with many other elements of language, studying the emergence and development of gestures in young bilinguals offers the opportunity to tease apart this confound. This is what Nicoladis, Mayberry, and Genesee (1999) set out to do in a longitudinal study of five French–English BFLA learners between the ages of 2 and 3.5. By the age of 3.5, as many as 91% of the children’s gestures accompanied speech – in line with adult usage – and most of these gestures conveyed the same meaning as their speech. Beats (i.e., by-phasic upward/downward movements) and iconic gestures were related to mean length of utterance (MLU) growth in each of the children’s languages, thus showing a language-specific pattern of development. Iconic gestures conveying visual information on actions, locations, shapes, and size appeared alongside predicates in the children’s language, that is, together with verbs and adjectives, which contributed to longer utterances and thus to the increase of MLU. Although children did not produce many beat gestures, such gestures were related to MLU growth in each language, thus lending support to the hypothesis that gesture development is paced by language-specific development rather than by overall cognitive development. In a subsequent study with eight French–English bilingual children, now between the older ages of 3.5 and nearly 5, Nicoladis (2002) confirmed the same language-specific pattern. While there was no relation between conventional and deictic gestures on the one hand, and language proficiency as measured by MLU or comprehension vocabulary on the other, children used more iconic gestures in their

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

31

dominant than in their nondominant language, and there was a positive association between MLU and iconic gestures. In this study, dominance was defined in terms of the MLU measured in words in each language, the number of utterances in each language, and the score on an often used test of word comprehension in both French and English (the Peabody Picture Vocabulary Test, Dunn & Dunn, 2007); the dominant language was the language that had higher scores in two of the three measures. Iconic gestures were not used in the absence of speech, thus suggesting that children did not resort to them as a language substitute. Rather, these findings indicate that complex language is a prerequisite for the use of iconic gestures. As bilingual children are their own maturational controls, these findings have more general implications for the relation of gesture to language and cognitive development. The former, and not the latter, sets the pace for iconic gestures in young children.

5 The Relation between Vocabulary and Grammar The research on the relation between gesture and language in the early stages of BFLA suggests a language-specific connection rather than an overall effect of cognitive development and maturation, and, so far, little in the way of transfer across languages. Another important relation that has long been the subject of inquiry in both bilingual and monolingual language development is that between vocabulary and grammar. According to emergentist and usage-based lexicalist approaches to language development (e.g., Bates & Goodman, 1999), there is no sharp distinction between vocabulary and grammar: learning a word entails learning not only its semantics, but also its morphosyntactic and collocational properties (Behrens, 2000). Along these lines, in what is known as the critical mass hypothesis, scholars have argued that early grammatical development is kick-started by the accumulation of a sufficiently large number of lexical exemplars that trigger the abstraction of grammatical regularities (Marchman & Bates, 1994). The lexically specific nature of early word combinations appearing in pivot frames comprised of one constant and one variable item (e.g., mommy there, mommy sock, mommy gone) has equally been argued to be key to children’s ability to generalize over instances and form abstractions (Lieven, Pine, & Baldwin, 1997). At the same time, it is also possible that cognitive and environmental influences might be responsible for the observed link between lexical and grammatical learning. Whatever cognitive skills a child relies on for word learning (such as the ability to phonemically discriminate as discussed earlier) may also be implicated in the learning of syntactic relations. And, similarly, whatever is conducive in the child’s language environment to word learning, such as the quantity and the quality of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

32

L U D O V I C A S E R R AT R I C E

the input, may have a facilitative effect on the development of word combinations for bilingual and monolingual children alike. In the case of bilingual children who are learning two languages with their associated sets of lexical items and syntactic constructions, the question of interest revolves around the degree to which vocabulary and grammar are connected within and across languages. Marchman, Martı´nez-Sussman, and Dale (2004) assessed the early Spanish and the English vocabulary and grammatical skills of 113 Spanish–English BFLA two-year-olds. They used the English and the Spanish versions of the CDI Words and Sentences to measure vocabulary and grammar, and a parental background interview to estimate the relative amount of exposure to the two languages. The within-language association between lexical and grammatical skills for these bilingual toddlers was strong for both Spanish and English; in contrast, the cross-language relation between vocabulary and grammar was only weak. This suggests that the mechanisms that connect early lexical and grammatical development are language specific rather than language universal. In a similar study, Conboy and Thal (2006) collected measures of lexical and grammatical development from 64 Spanish–English BFLA children between 19 and 31 months of age, also using the English and Spanish version of the CDI Words and Sentences; for 34 of the children, longitudinal data were collected up until the age of 31 months. Besides testing the maturational view that would predict a similar pattern of development in both languages, that is, the emergence of grammatical forms as a function of increasing age, Conboy and Thal (2006) addressed the issue of whether grammatical development is driven by the number of word meanings children know or by the number of different words that they know (these tend to be identical in monolingual children). In the former case, grammatical development in each language would be predicted by the number of meanings for which children had a word in either language, that is, pooled across both languages. In the latter case, grammatical development would be independently associated with the size of the lexicon in each language. Similarly to what Marchman et al. (2004) found, grammatical development in Spanish was associated with Spanish vocabulary size, and grammatical development in English was associated with English vocabulary size, thus favoring a languagespecific form-driven tie, rather than a less language-specific meaning- or concept-driven tie between the lexicon and grammar. In a study that included 196 Spanish–English BFLA children of different language abilities between the older ages of nearly five and just over seven, Simon-Cereijido and Gutie´rrez-Clellen (2009) replicated these findings of significant and positive within-language associations between vocabulary and grammar and the lack of a similar link across languages. Notably, this result held for both the typically developing children in their sample and children who had been diagnosed with a language delay, thus pointing to the importance of clinical assessment in both languages of a bilingual

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

33

child when there is a suspicion of a developmental language disorder (see further Hammer & Edmonds, Chapter 19, this volume, for assessments of a possible language delay; on the need more generally to take into account both languages for the assessment of bilingual abilities, see TreffersDaller, Chapter 15, this volume). One might ask whether the timing of exposure to an L2 in the preschool years affects the developmental link between the lexicon and grammar. Do the within-language lexicon–grammar relations found for BFLA also hold for ESLA children? Given that ESLA children may be exposed to an L2 after lexical and grammatical learning has already taken place to a lesser or greater extent in their L1, they could, at least in principle, avail themselves of extant lexical and grammatical knowledge in their L1 when learning their L2. ESLA children also come to an L2 with higher levels of cognitive abilities than they had when they first started to learn their L1, and this may affect the course of subsequent language learning. Finally, it is also possible that the link between vocabulary and grammar may be more important in earlier versus later stages of development. Kohnert, Kan, and Conboy (2010) studied a group of 19 ESLA children who were L1 speakers of Hmong born in the United States and first exposed to their L2, English, between the ages of 1.5 and just over 5 years of age. Between the ages of nearly 3 and 5 years and 2 months, lexical measures in both languages were collected through the administration of comprehension vocabulary tests and a story-telling task. The narrative task provided a measure of lexical diversity (the number of different words [NDW]) and a measure of grammatical complexity (MLU). In L2 English, there were robust within-language associations between lexical and grammatical skills, even after controlling for age and L1 vocabulary. This suggests that neither general cognitive maturation nor L1 transfer could account for the observed link between word knowledge and grammatical complexity in English. In L1 Hmong, the relation between vocabulary and grammar was weaker than in L2 English and only significant when the NDW rather than the comprehension measure was considered in relation to MLU. This asymmetry between the L2 and the L1 findings may be due to the fact that for English, MLU was calculated in morphemes, but in Hmong (a language with no inflectional morphology), it was calculated in words. Another possibility is that, given that children were overall more advanced in Hmong, they may have reached a stage of linguistic development where the lexicon–grammar link had become somewhat weaker and was undetectable by the relatively crude measures of lexical and grammatical skills. Similarly to previous studies of BFLA children, there were no significant crosslinguistic associations between lexicon and grammar in this group of ESLA children, although there was a moderate positive correlation between L1 and L2 vocabulary in terms of NDW.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

34

L U D O V I C A S E R R AT R I C E

The ESLA children in the Kohnert et al. (2010) study were on average older than the BFLA children in studies exploring lexicon–grammar links (Conboy & Thal, 2006; Marchman et al., 2004), but they were still of preschool age. What happens when children enter school clearly has important implications for later academic success (see Bigelow & Collins, Chapter 2, this volume), and understanding more about the relation between vocabulary and grammar in these later stages is of crucial importance. The cross-sectional and longitudinal studies that have investigated the relation between the lexicon and grammar in simultaneous and sequential bilingual children have largely found evidence for within-language relations across domains rather than for crosslinguistic relations. These results paint a picture of bilingual development proceeding along parallel and independent lines. The language-specific nature of early grammatical development in BFLA children also speaks to the issue of language differentiation in the morphosyntactic domain. BFLA children typically follow qualitatively similar developmental paths to those of monolingual peers (De Houwer, 1990; Meisel, 1989) and show evidence of early morphosyntactic separation. And yet another line of inquiry has repeatedly reported the existence of crosslinguistic influence (CLI) at the level of grammar (see Serratrice, 2013, for a review). CLI has been observed in a wide range of language combinations for a range of morphosyntactic phenomena, from referential expressions (Serratrice & Herve´, 2015) to word order (Nicoladis & Gavrila, 2015) and to argument structure (Chan, 2010). These findings are not necessarily incompatible with the lexicon–grammar results. The first likely explanation is that in the lexicon–grammar studies referenced earlier, grammar is narrowly defined in terms of MLU and children’s production of closed class words that relate to grammatical categories, while studies investigating grammatical CLI tend to focus on specific morphosyntactic constructions. Processing studies of adult bilinguals have now provided substantial evidence for the cross-language coactivation not only of the lexicon but of syntax too (Hartsuiker, Beerts, Loncke et al., 2016; see Ortega, Chapter 21, this volume), and some evidence to this effect is emerging for bilingual children as well (Herve´, Serratrice, & Corley, 2016; Kidd, Chan, & Chiu, 2015). Despite recent inroads into the field, we still know relatively little about the determinants of CLI in young bilingual children. The study of how the lexicon and syntax are coactivated in child bilinguals will make a contribution not only to the understanding of language acquisition and processing in this group of children but also more generally to the relations between language, processing, and cognition.

6 Conclusion Becoming bilingual, whether from birth or soon after, or subsequently in early childhood prior to schooling, entails a complex interaction between

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

Becoming Bilingual in Early Childhood

35

what children bring to the learning task, that is, among others, speech segmentation skills, speed of processing, and the linguistic and cultural environments in which they grow up. Quantity and quality of input are strong predictors of children’s early lexical skills, which in turn are closely related with emerging grammatical skills. Cultural practices like book reading, story-telling, and singing songs that are associated with larger vocabularies in monolingual children have also been found to be of importance in bilingual children (Song, Tamis-LeMonda, Yoshikawa et al., 2012). The relative status of the languages involved, whether they are both societal languages enjoying high prestige – as is the case for French and English in Canada – or whether there is a more asymmetrical minority–majority relationship – as for Spanish and English in the United States – has important consequences for the opportunities that children have to develop and maintain skills in both of their languages. Additional environmental factors such as socioeconomic status have barely been explored, but they are likely also inextricably tied to the course of bilingual language development. Studying language development in bilingual children is both theoretically important and practically necessary. From a theoretical point of view, turning to bilingual acquisition can shed light on some unresolved issues as regards the language-specific or language-universal nature of language acquisition. Bilingual children are their own cognitive and maturational control. This fact, combined with the fact that the quantity and quality of input may vary in their two languages, allows the investigation of language-general and language-specific predictors of phonological, lexical, and grammatical development. Alongside the study of these linguistic behaviors, technological advances are now making it possible to investigate the neural correlates of bilingualism, and we are beginning to see how the bilingual experience affects changes in the brains of bilingual speakers. From a practical point of view, more and more children are becoming bilingual in childhood in an ever more interconnected world. We need to understand the course of language acquisition in this large and expanding proportion of the world’s population. The findings of research on early bilingualism have important consequences for children’s harmonious bilingual development, for their later life chances, and for understanding difference from delay in atypical development. Further research on childhood bilingualism is therefore not just theoretically important but a pragmatic necessity.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 14:55:50, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.002

2 Bilingualism from Childhood through Adolescence Martha Bigelow and Penelope Collins

1 Introduction Bilingualism from childhood through adolescence is shaped by opportunities or missed opportunities for language learning, most prominently arising from the life-central experience of schooling, as well as the powerful force society plays in legitimizing the languages children and teens speak or want to speak. In large part, language learning opportunities occur through direct interactions with adults and peers within the proximal systems of children’s homes and schools. However, children’s experiences within their schools and communities are also influenced by more distal systems, such as their culture, their nationality, and how their local community fits within the broader society. In this chapter we tap into multiple perspectives to understand the desires and needs for bilingualism – defined as learning and using more than one single language or language variety – among school-age children and youth, their families, and their communities. We argue that it is often the work of educators, with the support of policy makers and families, to turn these desires into reality. We recognize the complexities of being or becoming bilingual in school, home, and community settings, and we hope to illuminate some of these complexities. In order to do so, we establish a frame of reference for the close connection that exists between young bilinguals feeling like legitimate users of their languages and their access to language learning opportunities. We leverage research insights vis-a`-vis young bilinguals’ lived experiences to help us think of what it means to speak and learn minority versus majority languages in a given context, such as learning Korean versus English in the United States, or learning Turkish versus Dutch in the Netherlands. We also consider world languages such as

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

37

English in many non-English-speaking geographies in the world or French, German, and other so-called foreign or modern languages in school contexts. We also explore the desires for language learning of bilinguals of a heritage, ancestral, or Indigenous language. While we recognize the fluidity of these categories, we feel it is helpful to understand hegemonic or dominant practices and systems related to learning a minority, community, or home language versus a majority language of the mainstream society or the nation-state, and by comparison to world or foreign languages. Questions such as who has the opportunity to be educated in their mother tongue and who is required to learn someone else’s language permeate our discussion of bilingualism during childhood and adolescence, because the majority of influences and time for these life stages comes through schooling, from kindergarten through to secondary (K–12) graduation. Essentially, we discuss how context – histories, relationships, dreams and desires, programs, and politics – all matter for school-age bilingualism, as young people navigate their worlds using their languages, dialects, and dreams of who to become.

2 A Dual Contextual and Personal Perspective on Being and Becoming Bilingual Multiple ways of being and knowing are constantly at work in the world of bilingual children and youth. Imagine, for example, a child feeling affirmed by the adults in her life for knowing how to speak the language of her ancestors, or for taking on a language that uses a script that is different from what most people around her know. Imagine what it feels like for a child to be a secondary student and have a teacher who speaks her home language(s) and who can even talk about her successes at school with her parents without an interpreter. Conversely, what would it mean to a child if her bilingualism were completely unknown to everyone at school? For school-age children, the emotions that accompany different ways of being and becoming with language are as integral to their identities as their potential to be or become bilinguals and multilinguals. Across grade levels, ages, and social contexts, children and youth experience language learning and language proficiency differently. There are also dramatic shifts in terms of access to language learning opportunities. The normal ups and downs of language learning may have to do with the mode they use (e.g., speaking, texting), their comfort level in terms of their interlocutors (e.g., peers, grandparents), or their motivation for wanting to persist with a previous or new language (e.g., staying connected with old friends, going to college). Youth are agentic in their bilingualism, that is, they make choices and are proactive and reflective in self-regulating their

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

38

MARTHA BIGELOW AND PENELOPE COLLINS

language learning. But they also benefit from feeling like legitimate users of their languages. Legitimacy is about exploring who a child can be in a given place and time. A good illustration comes from Ennser-Kananen (2014), who documents the legitimacy discourses experienced by Jana and Karina, two 15-year-old trilingual Latvian twin sisters enrolled in a German as a foreign language class in a secondary school in the United States. These two students, who identified as white, struggled to see themselves as legitimate users of Latvian, their first language, in the general school context because of the societal racialization of monolingualism, associating whiteness with speaking only English, their second language (see Fuller, Chapter 6, this volume, for discussion of language ideologies). At the same time, despite being exited from English language classes, their legitimacy as English users was unstable because their use of English seemed to submerge their Latvian identity markers. Furthermore, they expressed constant struggles with distancing themselves from other English learners, and feeling incomplete legitimacy as members of the group of mainstream, white, monolingual English speakers. On the other hand, in their German class, they garnered legitimacy from the teacher by striving to complete assignments and be accurate. Other research has documented how immigrant youth may choose nonmainstream language styles because of how they are racialized in their new home (e.g., Bigelow, 2010), further illustrating how language and racial identities are closely connected. While students negotiate being bilingual on a moment-by-moment basis, research in language policy and education has pointed to some larger contextual predictors of successful bilingualism, including the likelihood that an individual will not only be orally proficient but also literate in multiple languages as a result of schooling that supports what De Houwer (2015a) has called harmonious bilingualism. We show this dual contextual and personal perspective in Table 2.1, where in the first column we list positive circumstances originally outlined by Durgunog˘lu and Verhoeven (1998) that may predict a child’s likelihood for successful biliteracy – in the first rows in their minority language(s) and in the later rows in their majority language(s) – and in the second column we reframe these factors from the hypothetical perspective of a school-age child thinking of her or his lived experience in terms that signal harmonious bilingualism during the schooling years. Decisions adults make about which languages children should be encouraged to use and learn are likely to be grounded in certain attitudes or ideologies about the legitimacy of the languages and, by default, the speakers of those languages (see Fuller, Chapter 6, this volume). Ruiz’s (1984) wellknown orientations to language planning – language as problem, right, and resource – provide a helpful framework for the purpose of considering what is possible in a given linguistic context with regard to majority, minority, and world languages, in the best-case scenario when language is viewed as

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

39

Table 2.1 Predictors of successful language and literacy bilingual outcomes of schooling from a dual contextual and personal perspective Larger contextual predictors of success

Harmonious bilingualism from a child‘s/ youth’s lived experience perspective

When learning a minority/societal language Historically and politically, the home My teachers and classmates recognize the language(s) are valued and supported by name of my language(s) and show that the majority language speakers. they are impressed. There is perceived economic utility to My family always tells me not to forget my maintaining the home language, e.g., back language because we might go home migration. someday. The home language is important for coethnic I can talk to my grandma and cousins in ties and religious identity. Kenya. Pastor Tomas speaks Spanish at church. A writing system and reading materials exist I can read many kinds of books and in the home language. newspapers in my language. Literacy activities at home offer a strong Everyone at my house is always reading foundation for further literacy something. development. When learning a majority/societal language Majority language instruction considers the cultural and linguistic backgrounds of the language-minority students, as evidenced by teacher preparation, curriculum, and materials. The home environment values general literacy development, including the majority language. There is a match between values and practices of homes and schools. The individuals have access to majority language resources through school, media, and interaction with individuals.

My teachers like it when I speak and read Spanish. Some of them try to speak Spanish.

I get to read some of my own comic books in French to a second grader who is learning French.

I get to take library books home with me every week. I have friends who speak English better than me. The majority language is viewed as essential I want to keep learning English so I can be for social advancement, economic a teacher someday. prosperity, and empowerment. Note. The first column captures our synthesis of the factors discussed by Durgunog˘lu and Verhoeven (1998, pp. x–xv).

resource. We personalize this issue in Table 2.2, where in the first column we list questions to evaluate whether a language-as-resource orientation has been achieved in a given school context, selected from the heuristic list offered by Hult and Hornberger (2016, p. 41) for the purposes of language planning and policy, and in the second column we reformulate them from the viewpoint of a school-age child or teen, interrogating her or his lived experience of language as resource. Our purpose with Tables 2.1 and 2.2 is to offer a translational framework that encourages the exploration of research insights about legitimacy and access through not only the contextual plane but also the personal

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

40

MARTHA BIGELOW AND PENELOPE COLLINS

Table 2.2 Language as resource heuristic from a dual contextual and personal perspective Language as resource heuristic

From the viewpoint of a school-age student

What languages are represented as resources? Who decides which languages are what kinds of resources?

Do teachers see my languages as beneficial to my life and my future? What languages are celebrated by my peers and parents? Do I speak any of those languages? Who benefits from which linguistic Can I imagine using all of my languages for resources? many different things? Who benefits from my languages? Do I benefit? What resources (human, financial, symbolic, Who or what helps me develop my etc.) are provided for supporting which languages? languages? Who is the target population for language Can I imagine participating in the language education (i.e., linguistic minorities, learning opportunities around me? linguistic majorities, or both)? What “ideological and implementational Can I imagine being or becoming bilingual spaces” are present in policies that allow without any cost to my most valuable for the development of educational relationships? programs that expand students’ bilingual repertoires? How do educational programs facilitate the Does school help me imagine learning development of lifelong bi/ languages my whole life? multilingualism? How is the development of intercultural Do I have any opportunities to share about understanding included in educational my life and to learn about other people’s programs and curricula? lives in school? Note. The first column captures our selection of 8 of 15 questions proposed by Hult and Hornberger (2016, p. 41) as a heuristic to identify language-as-resource orientation for language planning and policy purposes.

dimension of the experiences of children and youth. We believe this dual perspective can be useful for educators and policy makers when trying to gain clarity about which languages should be taught in elementary, secondary, and high school programs. We should not neglect either perspective in working out how to connect instruction with the lived experience, and the possibilities for who students can be as users of multiple languages. To sum up, parents and educators make decisions for youth that have profound consequences on their being or becoming bilinguals. Considerations may include whether to require the study of a language and whether and how to honor languages that the students already speak, promote languages that students are at risk of losing, or offer opportunities for students to learn languages they hope to one day speak. All of these issues about imagining bilingualism hinge on the inner workings of home, school, and community and the degree to which individuals, systems, and synergy among them can offer pathways to language learning for school-age students.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

41

3 Becoming Literate in a New Language During childhood, an important watershed event for many children will be learning to read and write in one or more languages. Indeed, socialization into literacy is central during these ages, and schools are the main formal setting where many children will begin to develop literacy skills that must serve them well for the rest of their lives. It is therefore important to consider children’s social and cognitive experiences as they learn to engage with and make sense of text. Simply put, this typically involves learning to decode and encode print, and constructing meaning from texts (Verhoeven & van Leeuwe, 2012). All learners acquiring literacy for the first time, regardless of age (child, adolescent, adult), must determine the relations between oral language and the printed word. When learning to decode, they must figure out how written words are pronounced; when learning to encode, they must figure out how to write spoken words. To do both, students must learn to identify or manipulate the sounds of the given language, or become phonologically aware; they must learn to visually recognize the units of the writing system of the language (e.g., the letters of the English alphabet); and for many languages they must learn how phonology maps onto orthography, or spelling patterns. Finally, and critically, these processes must lead to creating meaning from text, which is possible to do only when the decoding process links to words the learner knows. Moreover, creating meaning from text often involves interactions with peers or teachers. We turn now to research about the experiences of bilingual children in classrooms, encompassing their interactions with text to construct meaning, the influence of identity on their learning, and the influence of peers and teachers on their learning experiences. Typically, language-minority children develop the skills involved in decoding and encoding as well, or better, than their monolingual peers (August & Shanahan, 2006), but with some differences. For example, in our own work (specifically by the second author of this chapter, Collins, who published under the name Chiappe prior to 2007), we have shown that when attempting new and unfamiliar words, language-minority children may use and rely upon the letter-sound rules to a greater extent than monolingual children (Chiappe & Siegel, 1999). In addition to decoding and encoding skills, however, children must construct meaning from the texts they read, and in order to build these written text comprehension skills they rely heavily on their listening comprehension skills (Verhoeven & van Leeuwe, 2012). Whereas some language-minority students may still be developing their oral and written comprehension of the majority language, the same component processes as for monolingual, majority-languagespeaking peers are implicated in their reading comprehension (August & Shanahan, 2006). These components include breadth and depth of vocabulary knowledge, syntactic knowledge, relevant background knowledge,

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

42

MARTHA BIGELOW AND PENELOPE COLLINS

understanding of genres, and text and discourse structures, as well as strategies to unpack texts (August & Shanahan, 2006; Lesaux & Harris, 2017; Melby-Lervaº g & Lervaº g, 2014). In sum, reading development for bilingual language-minority children and monolingual children who are speakers of the majority language has the same basic components. Although language-minority children who are still developing oral proficiency in the majority language tend to show overall lower literacy achievement and mastery of the academic register than their monolingual peers (e.g., August & Shanahan, 2006), the differences in achievement may reflect factors beyond the mastery of syntax and vocabulary. These differences may also be related to the degree to which languageminority children have been immersed in the majority culture, since more immersion usually means more access to the cultural knowledge needed to unpack the texts encountered in school. For example, for a group of 94 Chinese-Canadian immigrant adolescents enrolled in grades 7–12 in a school in Ontario, Canada, Jia, Gottardo, Koh et al. (2014) found that the degree of acculturation in the majority culture made unique contributions to reading comprehension, above and beyond measures of English decoding and vocabulary. Importantly, the degree of identification with Chinese culture did not make a difference, for good or for bad, in these adolescents’ English reading comprehension scores. Thus, greater immersion in the target language supports proficiency and literacy development in that language, without any necessity for immigrant youth to sacrifice affiliation to the home language and literacy heritage. We return to this point in Section 6. Many high schools across the world have begun to respond to the educational requirements of a new type of emergent bilingual student in need of literacy development: adolescents who, because of conflict and displacement, have had only limited and interrupted formal experiences of schooling in their home languages, and thus have limited literacy in any language when they begin their schooling. They suddenly must learn to read and write and do schoolwork in a majority language in which they are not fluent, and they must do so as teens, which makes for a markedly different initial literacy experience at the typical age of four to six. They are known in the field of educational linguistics as SLIFE: students with limited or interrupted formal education. (If they are adults over 21 years old, then the acronym LESLLA is often used: literacy education and second language learning for adults; see Simpson, Chapter 13, this volume.) SLIFE are often seen as not fitting into typical secondary schools. This evaluation of unfit stems from assumptions about what mainstream adolescents should be able to do and calls into question what secondary school educators know or must know about teaching initial print literacy skills in the context of subject areas in which SLIFE may (also) lack background knowledge. Rather than educators and researchers understanding the ways in which secondary schools are rigid, SLIFE are often characterized as having

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

43

innumerable deficits that, sadly, can even lead teachers to give up on them or find them “ineducable” (Bigelow, 2010). There are varying levels of literacy among SLIFE and some educational policies include students with no print literacy together with those with two or more years of missed schooling. Clearly, vast differences exist among adolescents labeled SLIFE and while educators are striving to identify those with the most need for instruction in basic literacy skills, students may start their schooling in a class with peers who are literate in their shared native language. The classroom dynamics that this produces were documented in Bigelow and King’s (2015) study of a newcomer secondary classroom in which Somali teens had mixed levels of literacy in Somali. Despite the teacher’s deliberate efforts to be inclusive and respectful of all students’ languages and cultures, print literacy was still a skill that garnered power in the classroom. This phenomenon had its origins in the migration realities of the role of gender in refugee camp life, where the boys were often able to attend school more regularly than the girls. Higher print literacy among Somali boys, in turn, resulted in an easier transition to “doing school” in the US school, and higher-status contributions to class discussions where Somali words for key concepts in class were negotiated. In this study, any amount of literacy or schooling became a way in which the boys’ input was able to override the girls’ Somali contributions to the classroom learning community. This is a good reminder, once again, that the experience of being an immigrant in schools is mediated by macrolevel established systems as well as micro-level everyday interactions among youth, including those interactions among coethnic peers. This is not to say that immigrant- and refugee-background youth are not doing literacy work through multimodal communication and using electronic devices. Vanek, King, and Bigelow (2018) found that East African youth are adept at the multiplex skill of establishing social presence through social media using images, employing the affordances of the tool (i.e., “like” on Facebook), and using all of their languages to share and comment. We found that youth were also able to front ideas linked to their religious, gendered, sexual, family, refugee, political, and other intersecting positions. Integrating social media into classroom language learning thus offers innumerable contexts for authentic, bilingual, multimodal communication among peers and for literacy development.

4 Opportunities for Bilingualism for Language-Majority Children and Youth While (im)migrant and refugee background youth are adapting to school and learning a new language, language-majority children and youth are also seeking opportunities to become bilingual. Obstacles to bilingualism include lack of opportunities that are intertwined with ideologies of

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

44

MARTHA BIGELOW AND PENELOPE COLLINS

monolingualism, particularly when it comes to English-speaking countries like the United States or the United Kingdom. Traditional language programs where students study the language as a separate class, disconnected from content learning, certainly still exist, but other options for children may also exist. The most important example of innovation in foreign language instruction is foreign language immersion education, initially implemented in the mid-1960s in Canada and currently widespread all around the world (see Juan-Garau & Lyster, Chapter 11, this volume). While there are many types of immersion programs, a basic tenet is that children learn grade-level content through a new language. There are one-way immersion programs, whereby schooling is delivered in a minority language to students from a majority-language background, and two-way immersion programs, in which majority- and minoritylanguage students join the same program in order to learn academic content through each other’s language. It is difficult to know exactly how many immersion programs there are in the United States or other countries, but we know that they are on the rise all around the world, with more and more programs targeting different languages for different reasons: languages that are minoritized but important in a given context, such as French in Anglophone Canada or Spanish in the United States; Indigenous or regional languages of increasing local importance, such as Catalan, Basque, Gaelic, Navajo, or Maori; powerful languages that were traditionally less commonly taught, such as Mandarin; and English as a foreign language in many countries. More and more parents of languagemajority students are trusting in the potential of immersion programs, particularly at the elementary level, to build language proficiency in an additional language at no cost to content learning or home language development. What the research has shown is that language-majority students in one-way and two-way immersion programs do as well as or better than students schooled only in English on standardized achievement tests administered in English. The benefits of immersion for language-minority students have been less clear by comparison (cf. Steele, Slater, Zamarro et al., 2017). A contested topic in immersion education is whether or how to use the majority language in immersion (and other types of) classrooms through crosslinguistic or translanguaging pedagogies (for crosslinguistic pedagogies, see Juan-Garau & Lyster, Chapter 11, this volume; for translanguaging pedagogies, see Garcı´a & Tupas, Chapter 20, this volume). Immersion programs attribute their success to strict language policies that designate at least 50% of the instruction as a time to use only the target or minority language, accompanied by strong pedagogies for making content comprehensible to young children. Research has shown that while it is not easy to stay in the target language, seepage from the students’ stronger or more familiar language can result in worsened outcomes in the new language of immersion (Fortune & Tedick, 2015). Thus, Juan-Garau and Lyster

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

45

(Chapter 11, this volume) are justifiably concerned that opening up the classroom to other languages could easily result in less use of the target language and thus weakened outcomes for immersion. On the other hand, translanguaging proponents like Garcı´a and Tupas (Chapter 20, this volume) are also right when they say that it is impossible to quell multilinguals’ constant connections among their languages, because students will use all their languages covertly, even in a classroom that suppresses their stronger language(s) (see Paap, Chapter 22, this volume, on coactivation). This reality then suggests that explicit target-language-only policies impact the productive modes most. In addition, it must be acknowledged that proponents of translanguaging are usually thinking of the needs of language-minority students and the linguistic insecurity that monolingually biased schooling instills in them. Be that as it may, support for and encouragement of using both languages in instruction is more ecological. In addition, experimental evidence is emerging that learning might be comparable under single-language or mixed-language conditions (Anto´n, Thierry, Goborov et al., 2016). This debate will likely need to be resolved empirically as well as ideologically, with deep recognition of the differences between language-majority children learning in immersion programs, with vast instructional support, and language-minority children learning through the majority context, with little instructional support. At the secondary level, immersive language programs are on the increase, but traditional foreign language programs still prevail (for a review, see Mun˜oz & Spada, Chapter 12, this volume). In the United States, one effort that has gained traction across many states is the awarding of Seals of Biliteracy (http://sealofbiliteracy.org/). This is a credential of language proficiency that school districts can award to learners who by high school graduation have achieved a high level of proficiency, as measured in multiple ways, and depending on the district. Some states negotiate with colleges and universities to make it possible for students who have obtained the Seal to receive college credits or to fulfill modern language requirements. While anyone can receive a Seal of Biliteracy (a language-majority or language-minority learner), those who can enroll in many years of language classes are more likely to receive the Seal because of their opportunities to become literate. The implication is that youth who at home speak a so-called small language, a language not taught in school, or an endangered language are less likely to obtain the Seal, and thus they benefit less from the opportunities it affords, including legitimation of the language and financial benefits in postsecondary schooling. As parents all over the world choose language learning paths with and for their children, we see the influence of neoliberal ideologies treating language as a commodity that can be leveraged for profit-making purposes. Language learning choices are often seen as paths to higher education, better job prospects, and presumably more wealth, if a person learns

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

46

MARTHA BIGELOW AND PENELOPE COLLINS

a language or languages of (global) power, often English and more recently Mandarin (for economics-based discussion of earnings related to bilingualism, see Grin, Chapter 9, this volume). While these motivations often appeal to adults, benefits that are abstract and in the distant future may not sustain the sort of day-to-day motivation needed to learn a language, particularly one perceived as optional by language-majority youth. In Markwardt’s (1948) classic study, one key reason for language learning was to be a cultivated person. While this is likely still applicable to many learners, today we think of motivation more in terms of self-concept construction (Csize´r & Do¨rnyei, 2005), international posture (Yashima, 2009), and identity and investment in the communities in which the learner hopes to use the new language (Peirce, 1995). Youth, including majority-language-speaking youth, may be interested in learning languages to tap into popular culture, for altruistic reasons such as service learning, or to communicate with peers across the globe. Maybe they choose a language to study because they like the teacher. Thus, it is impractical to limit world language options to hypothetical future financial gains, and foreign language educators and students should imagine bilingualism for many purposes, including but also reaching beyond monetary justifications.

5 Language Rights and Additive Language Learning for Language-Minority Children and Youth Transnational mobility and migration have spawned pressing linguistic and educational needs for language-minority children and youth. All over the world, initiatives respond to these needs by seeking to support the development and maintenance of students’ home or heritage language(s) while nurturing the academic and social needs in the majority languages of the given society where youth live. In this section, we first discuss the affordances and benefits of dual-language programs for language-minority children, which are mostly at the primary grades. We then explore recent work done on heritage and Indigenous language programs, often at the secondary level. We view Indigenous language recovery, reclamation, or revitalization as a distinct part of heritage language education.

5.1 Dual-Language Programs and Legitimizing Two Languages During the primary school years, in schools where the choice is available, parents of language-minority children can choose to send their children to dual-language programs. These are programs where the children will learn grade-level content through a combination of the home language and the majority language side by side in the same classrooms with majoritylanguage-speaking peers who are being schooled to become bilinguals in

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

47

the majority language they already speak at home and what for them is a new (foreign) language. While one-way immersion programs are designed for language-majority youth, two-way or dual-language programs cater to both language-majority and language-minority youth. These programs tend to be very successful for all learners. What we know about language-minority students who participate in high-quality two-way programs, in particular – for instance, about so-called English language learners (ELLs) in the United States through studies by Collier and Thomas (2017) and Lindholm-Leary and Block (2010) – is that they do as well as or better than ELLs schooled only in English on standardized measures in, for example, reading and mathematics tests administered in English. We also know that if they remain in two-way programs, typically by grade 6 (when students typically are 11 or 12 years old) they perform as well as average-performing native or ELLs on standardized measures. They can even surpass the performance of average native English speakers once they are classified as proficient in English. These two-way programs often meet the linguistic and educational needs of diverse communities through grade-level content teaching in contexts where it is possible to strike a balance between speakers of two languages, not in small part because of the availability of certified elementary teachers who can teach content in the minority language. The research suggests generally positive views not only about the academic outcomes but also about these types of programs being spaces that mitigate the subordination and discrimination that minority-language children experience throughout their education. Nevertheless, we want to problematize a few things, with the help of Palmer’s (2008) study of discourse patterns and dialogic identity construction of language-minority students in a second-grade two-way immersion program in the United States. First, Palmer found that children are aware of power issues in these programs, namely, that Spanish speakers must learn English very well, whereas English speakers have the option to learn Spanish, and even small gains in the latter but not the former are applauded. Second, while these programs strive to be empowering to Latinx children, the Englishspeaking children often dominate classroom discourse. Palmer found that the Spanish produced by the English speakers was rarely corrected by Spanish-speaking students, while English speakers often corrected their classmates’ English. Third, despite incubating strong academic identities in dual-language classrooms, students are often confronted with oppressive instruction elsewhere in the school, and thus “leave their academic selves behind and enter into other classrooms silent and timid” (p. 659). Palmer argues that it is critical to attend to these imbalances in order for equity pedagogy to prevail for the minority-language-speaking children in dual-language programs such as the ones researched in her study. These findings give us pause – how is it possible to meet the academic, cultural, and linguistic needs of both minority and majority children? Clearly, we

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

48

MARTHA BIGELOW AND PENELOPE COLLINS

need to return to the notion of who is and who is not a legitimate speaker or learner of a language, which is always contextual.

5.2 Learners of Heritage and Indigenous Languages Youth from minority-language backgrounds who are losing or have lost proficiency in their home, ancestral, or community language are often eager to reclaim that language. One might say they have become languagemajority speakers because they may see the language of the nation-state in their context as their primary and strongest language. Yet the shift to the majority language has not resulted in experiencing greater success in school (McCarty & Lee, 2014). Whether these children and adolescents belong to immigrant or to Indigenous communities, we see alignment among them because many efforts of critical heritage language education and Indigenous language education are or can be synergistic. Heritage and Indigenous language students share the need for language learning pedagogies that are culturally sustaining (see Paris, 2012). These are pedagogies that both support self-determination and acknowledge deep and intergenerational trauma, often with the goal of reclaiming or revitalizing a language that may have been lost for more than a generation (Hermes, Bang, & Marin, 2012). Both types of youth have very particular strengths because of their proximity to speakers of the heritage or Indigenous language, cultural ways of being that can inform curriculum and instruction, and potentially unique ways of seeing the world due to being minoritized. Learners of heritage or Indigenous languages know how to “read the world” if not “the word,” as critical pedagogues Freire and Macedo (1987) famously put it, and how to leverage their culturally grounded lived experiences using family and community ways of knowing. In this sense, literacy is far more than reading and writing – it is a way of being in the world. Heritage and Indigenous language learners may be elementary or secondary students, but for this discussion, we focus on the secondary school experience. In the United States, heritage language students are narrowly defined as first-, second-, or third-generation immigrant youth who have had prior exposure to their home language through their family or community but have had limited to no opportunity to develop formal knowledge and school registers in the minority language. Later, as adolescents or college students, they seek formal instruction in it (Leeman & King, 2015). There has been an explosion of linguistic and educational attention to heritage language learners, with curriculums being developed specifically for them at the university level and to a lesser extent at the high school level. Heritage language learners want language learning opportunities tailored to them – experiences that allow them to be themselves linguistically, dialectically, culturally, and transnationally. Cushing-Leubner and Eik (2018) offer a visionary curriculum for secondary Latinx youth in the

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

49

United States that welcomes students with vastly different levels of proficiency in Spanish and considers the political and racio-linguistic climate Latinx youth navigate (Flores & Rosa, 2015). In collaboration with Latinx youth, they developed Jo´venes con Derechos (Youth with Rights), a collective originating from the Spanish as a heritage language class in which youth use their varieties of Spanish and English to study critical race theories, justice movements, and Indigenous communities across the Americas, including borderlands identities. This curriculum activates youth because it draws on their dissatisfaction with schooling, their experiences of discrimination, and the deprivation of their linguistic rights in and out of school. It leads to youth carrying out their own research using the tradition of Youth Participatory Action Research to explore issues, improve their own lives, and inform others. For example, they explored questions such as “How do language, accent, and dialect affect how people are treated?” and “How can we ensure that our school is a place where all languages are respected?” This approach to heritage language instruction is at the cutting edge of pedagogies that legitimize Latinx linguistic identities and give them ways to express themselves, amplify their voices, challenge dominant narratives, and invite others to join their movement (see https://voice.adobe.com/a/kY40r/). Educational programs designed to support the learning of an ancestral, Indigenous language are spread worldwide, with well-known cases being Maori in New Zealand, Basque in the Basque Country, and Gaelic in Scotland. Simons and Lewis (2013) estimate that 19% of the world’s more than 7,000 living languages are no longer being learned by children. But they also note that the endangerment of languages that were in use only less than 100 years ago is not equally distributed geographically: the incidence of moribund languages can be as small as 10% in a region like Sub-Saharan Africa and as alarming as 75% in Australia, Canada, and the United States. In the United States, McCarty and Lee (2014) report, there are approximately 170 Native American languages spoken. There are more than 700,000 American Indian, Alaska Native, and Native Hawaiian students enrolled in the public school system, yet in the 2010 US census only 1 in 10 of these young people reported speaking a Native American language. The level of endangerment of Indigenous languages and the minimal access to their ancestral languages that so many children around the world have are the consequence of attempted ethnicide and linguicide by governments whose concerns are mostly with the languages of the nationstate and the rights and privileges of language majorities within their borders (see Lo Bianco, Chapter 8, this volume). In response, however, there are many grassroots efforts to promote bilingualism involving an Indigenous language. Programs and even entire schools are devoted to this effort. These programs center on strong and positive relationships among all members of the classrooms, and with communities, often with a strong spiritual thread in the curriculum and instruction. But these school efforts

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

50

MARTHA BIGELOW AND PENELOPE COLLINS

are particularly vulnerable to local budget cuts, national changes in educational policies, and broad migration patterns, including inter-village migration and moving to big urban centers. Under the push and pull of these multiple influences, many Indigenous language youth are left with great linguistic insecurity as they do language learning, unlearning, and relearning over the critical years of schooling, as Wyman (2013) documented with Yup’ik youth in Alaska. Communities also encounter roadblocks to bringing Indigenous language learning into settler schools because, for example, there is often a requirement to be a licensed teacher in the language and to demonstrate a specific level of proficiency in the language, and foreign language standards may not reflect Indigenous language aims or realities. Additionally, standardized testing often assumes monolingual norms. Furthermore, doing Indigenous revitalization within the very schools that were instrumental in causing those same Indigenous languages to become extinct or nearly become extinct is fraught with distrust. It is understandable, then, that revitalization efforts also take place outside schools, through communities and informal or autonomous learning, as Hermes and King (2013) have shown with Indigenous parents and their children using an innovative movie-based software platform to learn Ojibwe together at home. Solutions to Indigenous language revitalization include policies that open up possibilities for inclusion in education across all school grades, pedagogies that are sustaining and grounded in local communities and their ways of knowing and cultures, and efforts to build the notion among youth that it is possible and desirable to learn an ancestral language – that they can in fact become legitimate users of an Indigenous language.

6 Academic Achievement Gaps in Bilingual School Students: Setting the Record Straight Despite vibrant grassroots efforts and best (neoliberal) intentions in many parts of the world to promote bilingualism during the schooling of children and youth, most bilingual children and youth are in programs that push the majority language to such an extreme that home or native languages are lost. That is, much of schooling is subtractive (see Garcı´a & Tupas, Chapter 20, this volume). In most contexts, resources go toward majority-language instruction or to programs that transition students from more to less home/community language support and toward the societal language. Students’ home languages are a casualty society is unfortunately willing to live with, especially in defining academic success as performing at or above the mean on standardized tests, graduating from high school, and getting admitted to college. In the United States, a large-scale survey of schools conducted by Lewis and Gray (2016) found, among other things, that most high schools enroll

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

51

students who are officially classified as English language learners (that is, not yet fluent in English) and receive supplemental English language instruction, but that they do not offer any native language instruction in the vast majority of cases, even in the most common native language of the English learners in the school. They also found that only 16% of high schools had a newcomer program, which means that teens with very little proficiency in English, the language of instruction, and possibly new to school, are in classes with content and language demands that exceed their preparation. While these conditions of low proficiency and insufficient preparation are typically framed as challenges for the students, we would like to frame them as challenges for schools. Most educational systems all over the world are driven by test scores and worry about a so-called achievement gap between mainstream students, whose schooling is in the language they have spoken all their lives with their families and communities, and language-minority students, who must conform to be schooled in a language that is unfamiliar to them. All too often the blame is placed with the linguistic shortcomings of the minority students in the majority language, and any investment in maintaining and using their home languages is viewed as a lack of national loyalty, at worst, or as an act of educational self-sabotage, at best. It is therefore important to set the record straight. The findings reported by Jia et al. (2014), discussed earlier in Section 3, suggest that increased acculturation into the majority culture provides language-minority students with greater access to the very linguistic, rhetorical, cultural, and background knowledges that are tapped as majority-language measures of academic achievement, particularly in literacy. At the same time, however, this study also shows that engagement in the majority community does not work in opposition to identification with the heritage culture. In other words, it is important not to imagine a bilingual’s two languages and two worlds as subtracting from each other or competing against each other. Much to the contrary, the research is conclusive in showing supportive interactions in the two languages (e.g., Collier & Thomas, 2017; Lindholm-Leary & Block, 2010). Specifically, academic achievement, proficiency, and literacy are boosted when the home language of students is not dislodged by the language of schooling and instead opportunities are nurtured to use both frequently and meaningfully. We review here four studies that are particularly telling in turning up the same findings with rigorous designs and in varied contexts. In an unusually large reanalysis of the Programme for International Student Assessment (PISA) data for 2012, collected from 120,000 students across 18 countries, Ag˘ırdag˘ and Vanlaar (2018) found that minoritylanguage students across all but three of the countries examined showed lower mathematics and reading achievement than majority-language students. However, when minority students who reported speaking the minority language often with their parents were examined separately, in most

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

52

MARTHA BIGELOW AND PENELOPE COLLINS

countries those minority students did not achieve less. Moreover, in three countries (Canada, Finland, and Singapore), frequent minority language use actually predicted higher mathematics and reading achievement compared to majority-language peers. At the same time, a positive relation was also found, for the entire sample and all countries, between frequently speaking the language of the school context and mathematics and reading achievement. Thus, this large-scale, international study suggests speaking both the home and the majority language more leads to better school achievement. Another two large-scale studies by Winsler and colleagues demonstrate a similar supportive relation between proficiency in the minority language and language and literacy gains in the majority language, this time in the context of the United States. In a very large sample of more than 2,000 low-income children enrolled in preschools in Miami, Florida, Winsler, Kim, and Richard (2014) found that the supportive synergy can be seen at work already at the very young age of four. Winsler, Burchinal, Tien et al. (2014) found larger gains in early English literacy over time for immigrant children who spoke only the heritage language in the home versus those who spoke only English in the home. Their data came from a nationally representative sample of more than 10,000 children across the United States, collected longitudinally from nine months until age five or six in kindergarten. A fourth study, in Sweden, by Bylund, Abrahamsson, and Hyltenstam (2012), suggests that the positive relation between the two languages of a bilingual is sustained in the long term. These researchers gave a very difficult battery of tests in Spanish and Swedish to 30 college-educated adults who had emigrated from Latin America to Sweden between the ages of 1 and 11. They found that those who performed best in their home language (Spanish) also performed best in the majority language (Swedish). Although this compelling body of research is unfortunately less well known than it should be, there is growing recognition not just among researchers but also among educators and policy makers of the importance of both supporting and promoting minority language development for the sake of better, longterm language and literacy development in the majority language, which in turn also supports better academic performance. We know that other broad contextual factors beyond language, such as poverty, traumatic experiences occurring before, during, or after immigration, and cultural differences between the heritage community and school, may influence children and youth’s development of majority language literacy skills (NASEM, 2017). Further, the academic achievement of immigrant children who are schooled in the majority language varies by country. Whereas in Canada and Australia, English language learners tend to show comparable, or stronger, academic achievement than monolingual English speakers, in the United States, language-minority students who are classified by districts as developing proficiency in English tend to show lower levels of literacy attainments (NASEM, 2017). Students

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

53

classified by districts as developing proficiency in English in the United States are more likely to repeat a grade or drop out of high school than their English-proficient monolingual peers, even those from low socioeconomic backgrounds (Prevoo, Malda, Mesman, & van IJzendoorn, 2016). For this reason, it is useful to consider the psychological processes involved in second language literacy acquisition to include motivation through the lens of the expectancy-value theory (Eccles, 2009). The expectancy-value model of achievement choice focuses on the cultural, social, and other influences on the development of children’s achievement motivation and its impact on choice, performance, and persistence. When applying this model to language-minority children and youth, we find that students are most likely to be invested in learning English when they are confident they can master it and when it has high value to them. A variety of factors shape minority language students’ confidence in their ability to master the majority language in oral and written modes. These include their prior successes, the fits between their current knowledge and the new demands in the tasks to be mastered, their focus on mastery rather than relative performance, and their belief that mastery is primarily a function of their effort (e.g., Proctor, Daley, Louick et al., 2014). However, there is growing evidence that minority children’s perceptions of their school efficacy is not just informed by their achievement, but it is also shaped by their classroom learning environment and school climate (LeClair, Doll, Osborn, & Jones, 2009). For one, language-minority children often face a less-challenging curriculum – in the sense of less rich in learning opportunities – in which instruction is teacher centered and students have few opportunities to use language (Schleppegrell, 2013; Wedin, 2010). To illustrate, language-minority students have few opportunities to have more than short turns at talking, and these are often interrupted and with corrections. Such experiences not only limit their opportunities to build proficiency but also undermine their sense of efficacy and diminish their engagement in language learning as well as content learning. Similarly, school climate and classroom experiences can shape the value language-minority students attach to developing oral and written proficiency in the majority language. Their proficiency is influenced by the relevance of the material to their goals, values, identities, and life space, by the enjoyment they experience while doing the learning tasks, and by the extent to which they experience a sense of autonomy and choice as they enter and engage the learning tasks (Eccles, 2009). Teachers are also known to exert an important influence. Teachers who harbor strong monolingual attitudes also hold low expectations of their minority-language students, as Pulinx, Van Avermaet, and Ag˘ırdag˘ (2017) discovered in a survey of 775 teachers in 48 secondary schools in Flanders, Belgium. More than 75% of the teachers strongly agreed with statements that “non-Dutch speaking pupils should not be allowed to

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

54

MARTHA BIGELOW AND PENELOPE COLLINS

speak their home language at school” and that “the most important cause of academic failure of non-Dutch speaking pupils is their insufficient proficiency in Dutch.” These monolingual beliefs were strongly associated with a low level of trust in students’ self-efficacy, or “the level of confidence teachers have in their pupils to meet individual obligations and expectations regarding school work and effort” (p. 546). Berthele (2012) also found an empirical link between teachers’ negative attitudes against code-switching and ethnic stereotyping and lower evaluations of foreign language proficiency. Using an experimental methodology, he found that 157 future primary education teachers in French-speaking Switzerland systematically evaluated Swiss German primary school pupils learning French (a) considerably lower if they heard a sixth grader’s voice mixing three German words in their French oral description and they had been told the pupil’s name was Dragan or Goran (Serbian names), (b) higher if they heard the same recordings without any code-switching and were told the pupil’s name was Luca or Tim (common names in German-speaking Switzerland), and (c) even higher (suggesting a halo effect) if they heard the recordings without code-switching and attributed to Dragan or Goran, the ethnically marked names. Conversely, however, teachers can create classroom contexts that promote achievement by affirming students’ sense of belonging and individual worth (Lopez, 2012). More specifically, language-minority students, particularly those who are still learning the majority language, show greater investment and achievement in language arts classes (i.e., classes that focus on reading, writing, and communicating in the language of instruction) when teachers incorporate student perspectives and experiences in instruction, promote autonomy, and provide a safe classroom environment with high levels of emotional warmth (Lopez, 2012). Similarly, in classrooms where teachers adopt additive approaches to second language acquisition, that is, where language-minority students’ cultural knowledge and heritage languages are viewed as complementary to educational goals, students are more likely to value and be engaged in learning. In contrast, classrooms in which teachers adopt subtractive approaches to second language acquisition and emphasize assimilation into the majority culture – much as the teachers who Pulinx et al. (2017) surveyed – are more likely to yield students who are resistant to language and literacy development in the majority language (Lopez, 2012). Schleppegrell (2013) illustrates how teachers can build language-minority students’ sense of efficacy in learning the different registers of school, both written and oral, through the use of metalanguage, which she takes from systemic functional linguistics. She notes that metalanguage can support language-minority students, particularly those who are still learning the majority language, in accomplishing challenging, language-driven tasks, thereby enhancing their understanding of how academic language works and boosting children’s confidence that they can master language and the curriculum.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

55

7 The Lives of Bilingual Children and Adolescents beyond School In this chapter thus far we have highlighted how the context and system of schooling impacts the desires and needs of children and youth in their paths toward staying or becoming bilingual. We have done so because we believe that the formal experiences of literacy and schooling are formative during the life stages of childhood and adolescence. We would like to acknowledge, however, that the lives of children and teens are rich with out-of-school influences that can also contribute to their development of a harmonious bilingual identity. Indeed, our purpose with Tables 2.1 and 2.2 (Section 2) was to offer a translational framework that encourages the exploration of research insights about legitimacy and access through the experiences of children and youth in and out of school. We believe this dual perspective can be useful for educators and policy makers when trying to gain clarity about which languages should be taught in elementary, secondary, and high school programs. It is also critical to recognize the fluidity and variability that exists within contexts, not only over time but also at any given point in time. Language users are never the same across all their contexts of language use, and neither are their contexts. Consider, for example, how dramatically youth need to change the way they use language for learning school content, interacting in friendship groups, including social media and online gaming spaces, or interacting at home in the family. Youth may also need to navigate de facto language policies at school and at home that may impose the use of a single language. In other spaces, youth may need to be talented code-switchers to belong to a multilingual, possibly transnational community. A social domain of increasing importance for children and youth is through digitally mediated spaces. There is tremendous variability among these spaces, including social media such as Snapchat, massively multiplayer online games such as World of Warcraft or Minecraft, storysharing sites such as Wattpad, and online forums for participants to share fanfiction and media, such as Otakutalk. However, what these digital environments have in common is a global membership enabling youth to connect socially with others who share similar interests, enabling them to form transcultural identities. Black (2006) described the journey of Nanako, who as a young adolescent immigrated from Shanghai to Canada. Initially, she spoke no English and found it difficult to connect socially at school. After joining Fanfiction.com, she began writing and publicly posting her fanfiction in English. Over the years, her identity evolved. While she initially established herself as an anime fan, she later developed an identity of a successful author in this space who was noted

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

56

MARTHA BIGELOW AND PENELOPE COLLINS

and celebrated for her skills in translanguaging and mixing Mandarin with English. The digital world may provide youth such as Nanako with a space for harmonious bilingualism, where they have the freedom and community support to build proficiency in their adopted language, in Nanako’s case, English, rather than have their nascent proficiency be treated as a deficit. One of the affordances these sites provide is the multimodal way of participating, which enables youth to augment their communication using video, audio, pictures, and emojis, thereby allowing them to focus on self-expression and response to members of the virtual community. In these spaces, youth have the opportunity to practice their language and written skills by communicating with others, including native speakers, without being forced to identify whether they themselves are language learners. Those who choose to self-identify as language learners often do so to solicit feedback, and in return, often receive supportive, encouraging, and constructive feedback. Indeed, by interweaving Mandarin with English in her fiction, Nanako was recognized as an accomplished user of multiple languages and community members reported that they were learning Mandarin by reading her fanfiction (Black, 2006). In short, through the creation of transnational affinity spaces that are socially supportive and multimodal, the digital world may diminish many of the barriers to promoting harmonious bilingualism that exist in the real world.

8 Language Brokering by Bilingual Children and Youth Contexts for bilingual language use outside of school can be affirming for bilingual youth not only in the virtual but also in the so-called real (face-toface) world. We choose here to examine language brokering as one such context that many immigrant and first-generation children and youth experience, as they interpret and mediate communication between their parents, extended families, and members of their community with the host community. Language brokers may be self-selected or chosen by their families due to their possessing a natural gift for language, and Biedron´ and Birdsong (Chapter 16, this volume) discuss this possibility. Here we focus on the personal and social dimensions of language brokering. Young bilinguals’ assistance extends well beyond providing their families and communities with linguistic support in a range of situations, such as communicating at parent–teacher meetings or the doctor’s office. It also includes serving as a mediator or bridge between their heritage and host cultures. Serving as language brokers for their immigrant families provides youth with the opportunity to develop deep understandings of both their heritage culture and the culture of the dominant society, as well as

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

Bilingualism from Childhood through Adolescence

57

a meta-understanding of how their heritage and host cultures view one another (Guan, Nash, & Orellana, 2016). Whereas being a language broker for one’s family may foster a deep sense of self and connection with the family background, children’s role as language brokers may also give rise to difficulties. For example, children and adolescents may still be developing the language skills or knowledge needed to mediate between their family and members of the majority society, which can be particularly stressful as they feel responsible for helping their families. Sherman and Homola´cˇ (2017) relate the story of Tam, a university student in the Czech Republic who had emigrated from Vietnam with her family. As a young child, she had to interpret instructions from Czech government officials to her Vietnamese parents. While she was fully aware of the fact that her parents could face consequences for not following directions, Tam was equally aware that she did not understand everything in this adult situation. Despite her limited understanding of Czech administrative terms, their translations to Vietnamese, and the context itself, Tam focused on managing the language problems and later gained confidence as a broker. Alternately, immigrant youth face the stressful choice of what part of a message not to convey to their parents (Guan et al., 2016). As immigrant families face racism and discrimination, youth often protect their families by choosing not to translate derogatory comments or by selectively translating a negative message if there were necessary implications for the task at hand. Yet when describing these situations, immigrant youth often discuss how they overcame these challenges and their devotion to helping their families (Guan et al., 2016; Sherman & Homola´cˇ, 2017). As youth reflect on their different roles in society and across cultures, they may develop a range of identities. For example, Khalid, a 19-yearold Iraqi American, adopted a multilayered identity based on social context, considering himself an outsider or Middle Eastern person when reading and translating the news to his father, but noting that he was “one of them” with his friends (Guan et al., 2016). Alternately, youth may adopt a hyphenated identity as did 18-year-old Andy, who identified as Asian-American, enabling him to embrace both cultures without being seen as either too ethnic or too assimilated (Guan et al., 2016). And still other language brokers, such as 20-year-old Ana, embrace their heritage identities. Whereas Spanish may be a marker of immigrant and minority status for the Latinx community in the United States, Ana advocates for Spanish: So a lot of people would come in and speak English to me and I would be speaking Spanish to them because I felt like messing around with them. (Laughs). I’d be like, “It’s ten dollars” and they are like, “Oh, okay.” (Laughs). They only speak to me in English, but I speak to them in Spanish to make them speak Spanish. (Guan et al., 2016, pp. 160–161)

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

58

MARTHA BIGELOW AND PENELOPE COLLINS

In addition to working through their cultural identities, language brokers also develop attitudes about what it means to have limited fluency in the societal language, whether it may be considered a disadvantage, a disability, or a form of maintaining the heritage culture and community (Guan et al., 2016). These identities are strongly influenced by the linguistic, cultural, and sociopolitical contexts of their families and dominant culture’s policies and attitudes about the legitimacy of the languages involved.

9 Conclusion This chapter began with the idea that the desires and needs of children and youth matter in their paths toward staying or becoming bilingual. We have explored opportunities for becoming bilingual through the recognition of social, emotional, and cognitive aspects of the phenomenon. We hope our discussion and examples illustrate that young bilinguals traversing the years of childhood and adolescence have powerful contextual opportunities for developing bilingualism, in school settings and out of school, as well as restrictions. Both are grounded in sociopolitical and ideological forces that are always circulating among learners, their families, and educators. We believe that educators and researchers must never lose sight of the complexities of creating and sustaining bilingualism as well as challenging language ideologies that threaten the feeling children and youth have of being legitimate speakers of multiple languages. Imagining bilingualism among children and youth during the school years may come easy for some – families who already live transnationally and bilingually, or children who participate in strong, multiyear, articulated, content-based language programs with excellent teachers. But for most, bilingualism is hard won. It takes imagination and creativity to conceptualize the curricular and instructional spaces to make bilingualism possible, often starting with developing awareness of powerful and monolingual educational traditions and systems. We see language programming as at risk for all children and youth because of the monolingual mindset that pervades human consciousness as well as national discourses. Our hope is with the children and adolescents themselves, both in recognition of the undeniable potential they have for language learning and confidence in their ability to show adults new pathways to language learning that include their voices and experiences.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:58:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.003

3 Young Bilingual Adults Kellie Gonçalves

1 Introduction This chapter is about language learning and language use in young adulthood. It is primarily focused on young adults between 20 and around 40 years of age who are learning and/or using more than a single language and are mobile, that is, on the move. I introduce the term global hybrids in this chapter to denote young mobile adults in a range of contexts, regardless of whether they are backpackers, young professionals, migrant workers, refugees, or part of a newly established transnational young family. A plethora of other terms besides global hybrids exists within the fields of sociology, human geography, cultural studies, and migration studies referring to different kinds of individuals based on the types of mobility they are involved in (Urry, 2007). Elliott and Urry (2010, p. ix) estimated that at the time of their writing individuals across the globe were traveling a total of as many as 23 billion kilometers each year. They predicted that by 2050 this number is to increase fourfold, to 106 billion kilometers each year. The rate of mobility in our contemporary world has led Elliott and Urry to claim that we are currently experiencing a “golden age of mobility” (p. ix) in which the ever-rising movement of individuals, financial resources, objects, information, and ideas across the globe leads to major social change. Migration is one form of human mobility. According to the Global Migration Data Analysis Centre (2017), by 2015 international migration had reached an all-time high, with 244 million people on the move for different reasons. Migration will be a salient phenomenon in the lives of young mobile adults, understood in this chapter as the changing of one’s place of residence to another country or a very different region in the same country, whether it be for a well-specified period of time (other than vacation or business trips) or more indefinitely.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

60

K E L L I E G O N Ç A LV E S

Young mobile adults often need to learn one or more new languages in the host country in order to communicate in various social contexts and make sense of their lives on a daily basis. Indeed, mobility and globalization are closely linked with language contact, language learning, multilingual societies, and transnational families (Lanza & Wei, 2016). Language as a resource – whether a symbolic, interactional, material, or ideological one – flows, changes, and is used, learned, and drawn on by individuals for different communicative purposes in order to navigate their daily lives and get things done, however successful or unsuccessful they may be in doing so. The aim of this chapter is to provide an overview of studies that address issues of bi- and multilingual (henceforth: bilingual) language learning and the different sociocultural linguistic practices associated with the migration of young adults. It also identifies main topics for what remains to be done in the future. Introducing the term global hybrids is not meant to simplify the reality of individuals’ intricate historical trajectories, nor to disregard the complexity of issues surrounding race, class, citizenship, and gender within a second language (L2) context. In fact, by talking about global hybrids, I precisely aim to capture the complexity of individuals’ lives and the messiness involved in any type of categorization with regard to language learning, bilingual language practices, and adults’ changing identities as they traverse different social contexts around the world in both languageminority and language-majority settings, while acknowledging that in the 21st century, diversity and mobility are “normal” (Heller, 2011, p. 7). The “global” in global hybrids points to the commonality that individuals engaged in mobility, regardless of what kind, are part and parcel of a new social order and global economy. The “hybrid” in global hybrids makes manifest that the theoretical framework adopted draws on poststructuralist perspectives of bilingualism, while acknowledging that the postmodern mobile subject is equipped with fragmented and multiple identities that emerge and are negotiated in social interaction (see De Houwer, Chapter 17, this volume, on the negotiation of identities in bilingual interaction). In essence, use of the term global hybrid is meant to capture the desires, sentiments, and complexities experienced by a myriad of young migrants, regardless of social status. Anyone who has moved to a different country would probably agree that it is not an easy endeavor. In fact, being uprooted from one’s home, family, and social networks is inevitably a challenging experience for most young mobile adults. There are numerous reasons. Moving involves, among other things, encountering different sociocultural and linguistic ideologies, values, and practices, coupled with perhaps socioeconomic and political changes. These kinds of experiences vary according to the different types of young mobile adults that exist, the cultural capital they are imbued with, and the types of networks they have access to. For the purposes of this chapter, I focus on the following kinds of global hybrids: (1) young

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

61

adults who move abroad for self-development (including targeted language learning, study, or volunteering) for a specific amount of time; (2) young adults seeking a more prosperous or adventurous life abroad, without any specific job prospects necessarily established; (3) young adults who are caught up in persecution or war and move to a different country as a refugee or asylum seeker in order to simply survive; (4) young adults who move or have moved to a new country because of marriage (whether arranged or not); and, related to that, (5) young migrated adults who start a bilingual family. This list is by no means exhaustive (e.g., young professionals who get a well-paying job in another country either temporarily or permanently are not discussed here; there appears to be very little research on such individuals). Rather, it is meant to provide an indication of the types or categories of young global hybrids currently on the move. Of course, young global hybrids do not necessarily neatly fit into any single category. In fact, some categories overlap, shift overtime, and thus, like many other categorical distinctions in linguistics and the social sciences more generally, they are fuzzy. In Section 2 I focus on global hybrids as individuals, that is, not as part of a couple or family. They include the categories (1), (2), and (3) referenced earlier, that is, young adults choosing to spend some longer time away from home for self-development, young adults who have voluntarily moved to find a better life, and young adults who as a result of circumstances beyond their control find themselves in a new country. My discussion here is partly determined by the extent to which these global hybrids have been covered in the research literature so far. I show how the people studied are affected by issues of bilingual language learning and social practices within their respective social contexts. Section 3 looks specifically at global hybrids as young adult partners within the marital unit. Here I provide an overview of studies in the field of sociolinguistics that mainly consider bilingual language practices when one foreign partner or global hybrid is involved. Section 4 looks at approaches to global hybrids as young parents, including the recently established subfield of family language policy. Finally, Section 5 addresses methodological challenges encountered in the investigation of bilingualism among young adult global hybrids and identifies some main research needs.

2 Global Hybrids on the Move and Their Language Learning and Use A first category of global hybrids consists of young adults who move abroad for a set period of time in order to study, see the world, or do volunteer work. Such young adults have been the subject of many studies that have focused on their second language learning in an international study abroad context (Brooks, Water, & Pimlott-Wilson, 2012; Juan-Garau & Lyster,

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

62

K E L L I E G O N Ç A LV E S

Chapter 11, this volume; Ortega, Chapter 21, this volume). Most of the research in this line of work concerns the learning of English as a foreign language. Recent work taking a more ethnographic approach has turned its attention to other contexts. For example, based on a multi-sited ethnography, Garrido’s (2018) work explores how voluntary work, geographical mobility, and language learning intersect in a transnational social movement called Emmaus (www.emmaus-international.org). Garrido analyzed the role of language through narratives of young (wanna-be) mobile volunteers for solidarity work in the Emmaus network. These volunteers need to learn and use a lingua franca (mainly English or Spanish) for voluntary work that is justified by lifestyle mobility, intercultural exchanges, so-called linguistic enrichment, and, for some volunteers, even reasons of solidarity. Garrido showed how young volunteers have to navigate nation-state laws constraining people’s mobility and the linguistic differences that exist among international volunteers and locals. She found that volunteers’ construction of Emmaus as a safe zone and even adventurous engagement became a site where language learning was often regarded as a motivation for mobility, but also often contrasted to the actual sociolinguistic realities, where linguistic differences were seen as a barrier against mobility. A small but growing number of studies investigates the language learning of migrants within their workplace contexts. Many worksites employing such global hybrids offer on-site courses to help workers learn the societal language. In her critical ethnography of bilingual life in a manufacturing factory in Toronto, Canada, Goldstein (1997) focused on female Portuguese immigrants who were line workers at the lower end of the socioeconomic scale and who were taking English language courses at work, during work hours (English is the societal language in Toronto). However, these courses did not lead to more English use on the production floor. Most of these women instead continued to use Portuguese, their first language (L1). English was therefore not necessarily endowed with authority among the line workers of Goldstein’s study. In fact, due to their strong and dense ties within their Portuguese-speaking networks, one of which was the workplace, their L1 was maintained for emotional reasons of solidarity and sisterhood. In Gonc¸alves (2015), I looked at young Portuguese-speaking migrant domestic workers’ use, nonuse, learning, or not learning of English in their multilingual workplace in the United States. I examined the language practices in these women’s various social networks by focusing on their linguistic practices within their homes, local Portuguese-speaking community, and the cleaning company they work for, where Portuguese rather than English is associated with economic gain and symbolic capital. In particular, the study investigated emic perceptions about learning English and language use among domestic workers, their coworkers, and their English-speaking clients. The findings revealed that despite having a

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

63

proficient knowledge of English, a few domestic workers were more concerned about how they pronounced English rather than about the actual content of their talk. Individuals were often so worried about their accents in English and the potential risk of unintelligibility with their clients that it often left them silent. In a study of the same setting focusing on Portuguese, Gonc¸alves and Schluter (2017) found that in spite of Portuguese having minority status on both national and regional levels in the United States, on the local level within the company hierarchy in the same cleaning company, it was endowed with linguistic authority. Portuguese served as the company’s lingua franca for internal communicative purposes. This meant that many of the company employees did not necessarily need to use and thus learn English for their daily social practices within their workplace setting, nor for upward mobility within the company. Because many of the female adult migrants studied in Gonc¸alves (2015) and Gonc¸alves and Schluter (2017) did not speak English well enough to converse with their English-speaking clients, they relied on their multilingual employer and company owner, Magda, to serve as their main language broker (Gonc¸alves & Schluter, 2017). In fact, Magda engaged in what Gonc¸alves and Schluter called inter-employee brokering as well as in employee–client brokering in order for successful communication to be achieved for both company-internal and company-external purposes. Thus, due to the high value of Portuguese in the women’s families, the local community, and even their own workplace setting, many of them did not require English for daily communicative interactions. Portuguese fulfilled these women’s communicative needs, so that many of them did not feel the need or have the motivation or energy (after a physically demanding day of cleaning) to learn English. Several women also discussed their diasporic reconstructions of the homeland and plans to return, which became another significant factor why they did not learn English despite its regional, national, and global status. Stro¨mmer’s (2016) ethnographic case study of a former student of hers examined opportunities (or the lack thereof) for language learning in a typical entry-level job for immigrants in Finland. Her work focused on examining the factors that supported or prevented the learning of the societal language, Finnish, by Kifibin, a man in his 30s from Uganda. Kifibin’s L1 is Luganda, but he was educated in English. After completing his BA in Uganda, Kifibin moved to Finland to study in an international English-speaking MA program at a Finnish university. In order to finance his studies he took on part-time employment in a cleaning company. Subsequent to successfully obtaining his master’s degree and moving to a larger urban area in Finland, Kifibin hoped to continue his education and eventually pursue a doctoral degree in Finland. Unable to find employment in his area of expertise, however, Kifibin continued to work in cleaning companies, but he now did so full-time, which left him few

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

64

K E L L I E G O N Ç A LV E S

opportunities for language socialization at work since he often worked alone. This led to isolation, silence, and Kifibin’s invisibility in the workplace and thus little opportunity to interact with others. Kifibin’s main linguistic resources initially were radio programs, which enabled him to expand his linguistic and thus multilingual repertoire. Thanks to Stro¨mmer’s ethnographic longitudinal approach coupled with her participants’ general interest in the project, who consisted of Kifibin, his manager, and two supervisors, Kifibin managed to improve his situation and changed his language practices in a positive sense. These ranged from him becoming more aware of his learning opportunities by talking with Finnish wait staff in the kitchen he cleaned and by drawing on more Finnish resources such as photographs of Finnish signs found throughout his workplace. Another factor was his supervisors’ willingness to use more Finnish with him. Stro¨mmer’s study provides another example of how the workplace is considered a crucial site for language socialization for young adult migrants, where language learning, group membership, and individuals’ legitimate status come to the fore (e.g., Roberts, 2010). Block (2007) raises salient points about the social contexts for language learning and what he calls the critical experiences that adult migrant language learners find themselves in, which correlate with an individual’s changed sense of self as a result of uprooting one’s entire life. Navigating the challenges and alterations of one’s past, and one’s culture and language, within any new setting is a necessary undertaking (however difficult it may be) and done by individuals to protect what Giddens (1991) has called their “ontological security,” but also because identities are ascribed to individuals by others. Block further states that: these processes of reconstruction and repositioning do not take place in predictable manners and it is certainly not the case that naturalistic context guarantees sustained contact with longer-term inhabitants of the second language context. (Block, 2007, p. 75)

In other words, there is no single point in which such processes of adjustment converge, nor is there any guarantee that the rate of adult migrants’ language learning within naturalistic settings will correlate with the amount of time spent in the new host country. This is illustrated in Cammarota and Porquier (1986), who investigated the complexity of language acquisition among political refugees coming to France from Argentina and Chile in the 1970s to flee the military dictatorship there at the time. Drawing on a psycho-sociolinguistic framework, they looked at processes of acquisition, modalities of learning, and the specificities of what was at the time called interlanguage. Among other factors, Cammarota and Porquier found the following to be of relevance in order to account for language acquisition among the political refugees they studied: ambivalent feeling toward the home country and the host country, turbulence within familial relations, feelings of guilt, and the

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

65

representation the refugees made of French and of the acquisition of French. Factors like these go well beyond length of input and quantity of exposure as possible explanatory factors for later language learning (Ortega, Chapter 21, this volume) and also need to be addressed in explanations of young migrant adults’ bilingual language development. Indeed, studies of adult L2 language learning by adult migrant learners that takes place outside of the classroom have to consider the complex social contexts and the settings in which they find themselves so as to gain a better understanding of how learners’ daily, real lives shape and affect their L2 learning, L2 investment, and overall development (Gonc¸alves, 2013, 2015). Such holistic approaches need to account for language use within migrants’ social networks encompassing familial, social, transnational, and professional ties. Crucially, they also need to take into consideration the time or lack of time that would-be learners are offered, or can afford to take, in order to be able to learn and use the target language of the host country, which may not overlap with the language(s) used in the workplace (Gonc¸alves, 2015). Assuming that global hybrids need to learn a language for communicative, instrumental, and integration purposes, and assuming that they have the time and resources to take language classes, the question arises as to what kinds of pedagogical approaches would be most suitable for adult migrant learners (see further Simpson, Chapter 13, this volume). In Gonc¸alves (2015), I argued for new directions in English as a Lingua Franca and English as an International Language pedagogy that take into account factors affecting and shaping migrants’ language learning experiences and their second language development. Among other approaches, I advocated for the use of translanguaging practices (Garcı´a & Tupas, Chapter 20, this volume). However, translanguaging practices in classrooms for young adults and global hybrids remain to be explored. Beyond translanguaging practices, contexts aiming to help young hybrids develop their repertoire in the new language, whichever it is, should develop programs that take into account various learner-related factors for this particular population. At a minimum, they would need to include the amount of time spent in the host country, prior language learning experiences in classrooms and informal settings, their home and family status (e.g., married, divorced, single, children), the kind of community learners reside in, the language(s) used at the workplace, and the specific needs for repertoire building for diverse functions, social network ties at the local and transnational levels, diasporic reconstructions of the homeland in terms of whether the learner in question plans to return or stay in the host country, deportation vulnerabilities (for undocumented migrants), and, in particular for asylum seekers and refugees, traumatic life experiences. This is highlighted in Søndergaard’s (2017) study of cognitive factors involved in classroom language learning among refugees in Sweden. Søndergaard compared male refugees who had suffered from

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

66

K E L L I E G O N Ç A LV E S

post-traumatic stress disorder (PTSD) to male refugees who had not. PTSD can be triggered by severe life-threatening experiences, including war, interpersonal violence, and torture. Informants suffering from PTSD showed slower rates of language learning. Individuals’ pain was reported to be caused by disturbing memories of past events, personality changes, irritability, difficulties in concentrating, or anxiety, depression, and disassociation. Søndergaard suggested that when planning language classes for refugees and especially those who are suffering from PTSD, teachers should be sympathetic and aware of memory impairments in their learners. He advocated stress-free and predictable learning environments. By acknowledging the many complex social factors that likely play a role in young global hybrids’ new language learning, educators, practitioners, and students alike may be in a better position to assess individuals’ goals, needs, and motivation (or lack thereof) regarding the sociolinguistic realities of language learning and the many challenges they entail.

3 Young Adults as Members of a Couple in Bilingual Settings This section discusses language choices in couples that have the option of using more than a single language. It does not solely focus on global hybrid couples, that is, on couples consisting of at least one spouse who has moved to a different region or country, but also refers to work on young adults’ language choices in bilingual communities. Based on his observations of language choice in global hybrid couples many decades ago (then called mixed couples), Siguan (1980) proposed that the geographical place where a globally hybrid couple resides is the most prominent factor determining their language choice in communication with each other. Secondary factors include the social status of the two languages, feelings of solidarity associated with the minority language, and issues of gender. With respect to this last secondary factor, Siguan suggested that the language of the male would ultimately prevail (he was only referring to heterosexual partnerships). As reviewed in Piller (2002), studies of language choice by bilingual young adults in bilingual communities (not involving any migration) have viewed young adults’ marital choices as fundamental in explaining language shift toward the majority language. Especially groundbreaking here was Gal’s (1978) ethnographic study investigating the language choice patterns of young women in the small Hungarian–German bilingual village of Oberwart, Austria. At the time of Gal’s research, women’s life possibilities in Oberwart depended mainly on who they married (p. 12). As the title (“Peasant men don’t get wives”) of her seminal article reveals, Gal describes how rather than marry a local peasant, young women in Oberwart preferred to marry local men working outside of agriculture who

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

67

could provide them with a more comfortable and independent lifestyle. The former tended to be mainly speakers of Hungarian, while the latter were mostly German-speaking. Thus, the two languages in this bilingual community represented different levels of social status, with German symbolizing upward mobility. Different from men or older women in the community, young women spoke German far more frequently and with more interlocutors than Hungarian, leading Gal to argue that their language choices correlated with their aspired-to social positions. In her anthropological study of language use among women in the Breton– French bilingual region of Brittany in France, McDonald (1994) found a similar pattern. Starting in the 1920s (earlier than in Oberwart), language shift has been taking place from the minority language of Breton to French, with French indexing socioeconomic mobility or, as McDonald (1994) describes, a move “from cow-shit to finery” (p. 91). McDonald’s findings were similar to Gal’s (1978) in that local minority-language-speaking men could not find local Breton-speaking wives. There are alternatives to one language gaining the upper hand in couples’ communication. On the Solomon Islands, in what Lincoln (1979/1980) termed dual-linguality, husbands and wives addressed each other in diverse languages, one of which belonged to the Papuan language family, while the second belonged to the Austronesian language family. Both spouses understood the other one’s L1. Thus, in a situation such as this, each bilingual spouse always simply speaks their respective L1 with the other, in what is increasingly called dilingual conversations (De Houwer, 2009). In the rural area of Misio´n La Paz in northern Argentina, people typically speak only a single Indigenous language but normally understand two more (Campbell & Grodona, 2010), resulting in dilingual conversations all the time, a fact that Campbell and Grodona evaluate as highly unusual (see also De Houwer, Chapter 17, this volume). Spouses in this community typically each speak different languages to each other as well, and no status differences are attributed to any particular language. Dual-linguality in couples was also found to take place in the northwest Amazon Basin spanning Brazil and Colombia, where people are not allowed to marry someone from their own language group and must always use their own language in a marriage (Grimes, 1985). Global hybrid couples might also practice dilingual conversations (e.g., Piller, 2002), but so far there have been few reports of couples doing only that. In an interview study of 51 global hybrid young couples residing in Japan, only a single one consistently used dilingual conversations (Yamamoto, 1995). The couples consisted of a Japanese and an immigrant non-Japanese spouse with English as an L1. A third of the couples spoke both English and Japanese with each other. About two-thirds of the other couples spoke just a single language with each other. Within this group, the majority (again two-thirds) used just English. The main reason for choosing English

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

68

K E L L I E G O N Ç A LV E S

was that Japanese spouses were more proficient in English than vice versa. This result is not surprising considering the status of English globally (see Fuller, Chapter 6, this volume). Other research has drawn on various interdisciplinary frameworks to explore the connections among language attitudes, language choice, language ideologies, and identity construction and performance within global hybrid couples. In a rich study drawing on the frameworks of the political economy of language and language use as an act of identity, Walters (1996) investigated linguistic practices and beliefs related to language of highly educated foreign wives married to equally highly educated Tunisian men in Tunisia. In this simultaneously diglossic and bilingual community, educated people routinely code-switch between the vernacular Tunisian Arabic (the only variety uneducated speakers in Tunisia have access to), Modern Standard Arabic, and French. English is highly valued. The wives in this study, who had English as their L1, mostly communicated with their husbands in both English and French; their husbands’ L1 was Tunisian Arabic. Walters describes the many ambiguities and challenges that the women faced with regard to their learning and using of Arabic, especially its local vernacular variety (see Chevrot & Ghimenton, Chapter 25, this volume, on bidialectalism). The women Walters (1996) interviewed generally wished to learn Tunisian Arabic but often did not attain a high level of proficiency in this local, in-group variety. They usually found their husbands to be unsupportive in helping them to learn Tunisian Arabic, but often did manage to learn to speak it through interactions with their husbands’ family members, usually their mothers-in-law. Walters shows how the wives’ experiences regarding language learning and linguistic access to Tunisian Arabic were linked to Tunisians’ notions of identity and these women’s gendered roles as mothers and wives as well as to their personal identities and notions of self. Most wives made conscious decisions regarding the learning and use of specific linguistic codes in order to negotiate power relations within their marriages and extended families. Thus, the foreign wives’ ability or inability to speak Tunisian Arabic had a symbolic meaning for all involved (Walters, 1996, p. 516). Piller (2002) draws on feminist and postmodern approaches to second language learning, multilingualism and cross-cultural communication in her study of 36 couples where one spouse had German as an L1 and the other English. Based on conversations among these couples recorded in the absence of third parties, Piller’s analysis focuses on understanding the couples’ discursive construction of hybridity, that is, performing identities based on one’s bilingual language repertoires, by exploring individuals’ sense of self and how this is influenced by learning a second language and culture as a mature adult. Like Walters (1996) did, Piller (2002) highlights the fact that language choices on the micro level are always shaped

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

69

by larger macro-social structures and dominant discourses within given societies. In Gonc¸alves (2013), I report on a three-year ethnographic study of nine global hybrid couples in the tourist area of Interlaken in German-speaking Switzerland, where a diglossic situation prevails. I analyzed how young adults with English as an L1 who are married to German-speaking Swiss come to terms with their culturally hybrid identities as a result of learning (or not learning, as the case might be) the local central Swiss dialect and/or Standard Swiss German as well as through engaging in local sociocultural practices of daily life. Results indicated that the foreign spouse was often left frustrated and disillusioned by both regional and local language classes. Following national language policies, these classes were offered exclusively in Standard German (see Lo Bianco, Chapter 8, this volume, on language policies). Swiss German courses did exist; these were mainly offered in the capital city of Bern. However, these required participants to have a basic working knowledge of Standard German, the time to travel to Bern, and the economic means to finance these classes. Informants were thus presented with an expensive double-edged sword, which required them to pay for a class and learn a language they would only need for literacy purposes and that is regarded as “useless” within the context of daily social interactions with local Swiss. Furthermore, informants who did manage to learn Standard German and attempted to speak it with their Swiss interlocutors were met with convergent accommodation (see De Houwer, Chapter 17, this volume) toward English. Successful learners of the local dialect learned it either through their children, elderly neighbors, and family members, or through taking on what informants considered to be low-skilled jobs that, however, provided individuals with oral practice, such as waitressing. Some of the foreign spouses, most of whom were women, decided to give up trying to learn German altogether, partly because they were not willing to take on what they considered a menial job. Instead, they continued with their daily lives in English. While life in English is possible in the tourist town of Interlaken, the spouses who used solely English never felt completely integrated in their local community, due to their lack of local language skills. Of the nine couples in this study, only one couple spoke the local Bernese dialect together. Five couples claimed to use English only and the remaining three stated it was generally a mix between English and the local dialect depending on the context, topic, and interlocutors. These and similar studies indicate that specific language choices within global hybrid couples are never neutral, but complex processes that depend on a number of factors, including personal, gendered, social, geographic, political, cultural, linguistic, and psychological ones. Specifically, spouses’ proficiency levels in the languages in question, the status of the languages involved, the attitudes and ideologies surrounding language use and practices, and the language(s) of the community at large

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

70

K E L L I E G O N Ç A LV E S

are important considerations. In addition, spouses making up a global hybrid couple are from different cultures and different L1 backgrounds, and may not be very familiar with the other one’s culture and language. Furthermore, language choice plays a pivotal role in how we depict and define not only ourselves, but also others (De Houwer, Chapter 17, this volume). Therefore, decisions regarding one’s language choice within the intimate relationship of a global hybrid couple are anything but simple.

4 Young Adults as Parents in a Bilingual Family When a globally hybrid couple has a child, the inevitable question arises as to what language or languages to use within the family once the new member has arrived. The question may even be relevant during the pregnancy period, as studies of newborn babies show some language learning in utero (see Serratrice, Chapter 1, this volume), and parents may want to go into the parenting phase well prepared. Popular literature concerning the hardships and complexities surrounding parental language choices have resulted in guidebooks, Internet blogs, and websites for parents interested in successfully raising bi- and multilingual children. Harding-Esch and Riley’s (2003) book is a case in point. In their preface, these authors state that: This handbook is for all parents who might be considering bringing up their children as bilinguals [. . .] it is not about bilingual societies [. . .] The linguistic problems of immigrant groups are not discussed either: they involve social and political issues that go well beyond the scope of this book. (p. xiii, italics added)

This extract highlights some key issues associated with decisions regarding language practices in a potentially bilingual family. The authors make several disclaimers that accentuate the intricacies involved in such decisions: these are subject to each family’s individual context and thus specific practices. As such, decisions regarding language practices in families in bilingual societies will differ from those of immigrant families who for socioeconomic or political reasons have had to migrate to a largely monolingual society. Similarly, family language practices will look very different for the so-called expat family moving to a different country on a temporary basis as opposed to a family who has moved to another country for an indefinite amount of time. The question of what factors help determine bilingual parents’ language practices with their children has sparked the interests of scholars across interdisciplinary boundaries and how these practices achieve success in raising bilingual children. In addition to the more psycholinguistically oriented literature focusing on the role of input in bilingual acquisition (see Armon-Lotem & Meir, Chapter 10, this volume), there is now an

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

71

extensive body of sociolinguistic and applied linguistic literature that studies what has been termed family language policy (FLP). King, Fogle, and Logan-Terry (2008) have defined FLP as “explicit and overt planning in relation to language use within the home among family members” (p. 907). Such explicit planning does not necessarily result in a strict adherence to the plan, though (De Houwer, 1999). Furthermore, as Okita (2002), De Houwer (2009), and Schwartz (2010) have noted, parents in potentially bilingual families do not always consciously or clearly plan their initial family language use. In the context of this chapter, I consider FLP to refer to the broader notion of family language use and practice as Palviainen and Boyd (2013) understand it. In their study of three bilingual Swedish–Finnish families, FLP is considered a semi-planned and jointly constructed enterprise among different family members in which children are instrumental participants who take on an agentive role (see also Curdt-Christiansen, 2014). In her review focusing on sociolinguistic and sociopsychological aspects of intergenerational language transmission in an environment where the societal language may not be spoken at home, Schwartz (2010) discusses the complexities that parents encounter when attempting to transmit a minority language to their children. Among others, Schwartz highlights the role of family structure and family language ideology in shaping family language practice and management. Schwartz’s summary and her ensuing methodological recommendations make clear that interdisciplinary efforts drawing on linguistics, sociology, and education are needed to fully understand the complexities involved. Several studies within FLP have been conducted that address Schwartz’s call (see, e.g., many contributions in Lanza & Li Wei, 2016). Most certainly, FLP needs to look well beyond just parental use (or not) of the so-called one-person one-language (OPOL) strategy often hailed as the best and only method for raising bilingual children (see also Quay & Montanari, Chapter 27, this volume). Under the OPOL strategy, parents each address their children in just a single language, and exhort others to do the same. As Marley (2013) describes based on her experiences as part of a bilingual family in the United Kingdom, using the OPOL strategy does not necessarily mean that children will actually speak two languages. Marley addressed her two children in her L1, English, while their father addressed them in his L1, Moroccan Arabic. However, the children always spoke just English at home (see also De Houwer, 2009). After gaining Internet access through which they could communicate online (in the written mode) with their Arabic-speaking cousins in Morocco, though, the children started to communicate with them in Arabic using the Roman alphabet (a practice shared by the cousins). Marley (2013) concludes from this that thanks to computer-mediated communication, her children’s attitudes toward Moroccan Arabic became more positive, and

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

72

K E L L I E G O N Ç A LV E S

she regards online communication as a powerful tool that can help binational families globally to support their minority language. Closely related to language choices within the family are choices that financially privileged parents are able to make in connection with the selection of schools for their offspring. In a rare study of nine such families in Tokyo, Japan, who had the choice among Japanese-medium and Englishmedium schools, Velliars and Willis (2014) explored the reasons for which both the global hybrid and local Japanese parent couples interviewed mostly chose English-medium schools. As with FLP, these reasons were of a social, cultural, and linguistic nature. In addition, both educational and practical concerns informed parental choices. There is no doubt that within multilingual contexts, language choice and language attitudes that may underlie parents’ decisions to speak a particular language to their offspring are inevitably linked to various social, economic, and political arrangements and ensuing changes that such processes entail. Parents’ and families’ language ideologies often reflect powerful macro-social structures and attitudes and ideologies about the status of majority and minority languages (King, Fogle, & Logan-Terry, 2008), child–caregiver interactions, and child language development (De Houwer, 1999), which inevitably affect how languages are learned, maintained, unlearned, and employed within various familial contexts around the world (see Fuller, Chapter 6, this volume, on language ideologies more broadly).

5 Research Needs to Better Understand Young Adult Bilinguals The methodological hurdles facing researchers interested in young global hybrids are many. As I have argued elsewhere (Gonc¸alves, 2013; Gonc¸alves & Schluter, 2017), in addition to concerns of personal trust, language barriers among researchers and informants are a major obstacle. This is especially relevant in relatively hidden domains such as workplaces and the home. Researchers must be proficient in the participants’ L1 (or in one of their L1s) in order to understand the nuances afforded by speaking the same language or dialect. High levels of proficiency may facilitate access to further participants and their diverse communities of practice. Additional methodological challenges are associated with gaining access to people as part of a marital unit, which in many cultures constitutes a highly intimate and private relational unit. This may be why research on the linguistic practices of and within global hybrid couples is scarce and still not considered “a standard site of academic inquiry” (Piller, 2002, p. 7). Piller (2002) documents some of the reasons potential participants in her study gave for not being willing to actually participate. These included time constraints, conflicts with spouses over bilingualism, and

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

73

migration, as well as the gendered notion of “uncooperative husbands,” leading Piller to claim that for some of her would-be participants, admitting to some of these issues “was too threatening to their performance of happy couplehood for them to participate” (p. 45). In Gonc¸alves (2013) two additional factors emerged, namely issues of privacy and personal acquaintances. Residing in the same place as my informants meant that my status as an “insider” and “outsider” had both advantages and disadvantages as regards participant recruitment. This in turn affected individuals’ willingness to talk to me about their relationships and the foreign spouse’s language learning process and development. However difficult the work with global hybrid couples is, a major shortcoming of it is that few studies trace the socialization processes of language learning among these young global hybrids longitudinally. If and when language learning is addressed, it is usually done through a snapshot at a given point in time rather than through tracing a global hybrid spouse’s learning trajectories over an extended period of time. This effectively renders the documentation of the often convoluted process of second language development among young adults impossible. In fact, many studies largely disregard a foreign spouse’s language learning journey altogether. The majority of studies done on global hybrid couples tend to focus primarily on the women’s experiences within such marriages or shift their attention to the bilingual development of their children, intergenerational transmission, and private language planning. While research on multilingual and transcultural families is growing (e.g., Lanza & Li Wei, 2016) and is addressing new and different questions that are vital to our understanding of how such families are viewed and constructed through their multilingual practices both online and offline, much work remains to be done. Before we can begin to understand how couples and eventually families consciously or less consciously plan their language choices, we need to know much more about the language learning process and development of the global hybrid spouse in the relationship (see also De Houwer, 2009, pp. 86–90). This spouse may be at a disadvantage in terms of language, but also because of other sociocultural and even political and economic factors. Investigating global hybrid spouses’ second language learning will allow us to better understand the complex journey of individuals’ language learning development as young and mature adults and how such processes shape language choices and future private language planning, policies, and practices, not only within couples, but eventually also within their multilingual families. Related to needing more of a focus on language learning in global hybrid couples, more research is needed that addresses individuals’ dynamically changing identities, existing power asymmetries within the couple, and the sociolinguistic and emotional challenges presented by multilingualism as a result of moving countries (e.g., Dewaele & Salomidou, 2017). All

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

74

K E L L I E G O N Ç A LV E S

these factors may influence what languages are used and transmitted to children in the future. These factors need to be seen against the backdrop of macro-level social processes and the contemporary changing and challenging sociopolitical and economic conditions of each context. These affect how language(s) are viewed, valued, and utilized on the micro level of the family. In addition to the methodological challenges involved in this type of research, we must also be willing to reconceptualize particular notions within contemporary global contexts. The notion of family is one of them. Regardless of whether reference here is to nuclear families or more extended families, sociologists agree that the family has been and continues to be a dynamic system prone to change. The significance of the family within bilingual development cannot be underestimated given its historical foundations (cf. De Houwer, 2009, pp. 10–11), but has required some reconceptualization that is more in line with the rapidly changing world of the 21st century. Lanza (2007) introduced the element of social practice to FLP in order to understand the ways in which family members communicate, transmit languages, and engage in bilingual practices. Lanza (2007) stressed that each family is unique in that it has its own norms for language use, resulting in family members having their “own ways of speaking, acting, and believing” (p. 47). Within the context of FLP, the family plays a key role in terms of maintenance, preservation, and transmission. Future research needs to push the boundaries of studying bilingualism in global hybrid families by going beyond the traditional nuclear family and typical two-parent middle-class household. This can be done by investigating bilingual practices in single-parent families, patchwork families, LGBT households, and transnational families that include caregivers such as grandparents, nannies, and au pairs, while keeping in mind that many families today are busy managing more than two and even three languages (Quay & Montanari, Chapter 27, this volume). Many individuals have the financial means to invest in their futures and in the learning of different languages if necessary. Much of the research on global hybrids has focused on those that are privileged and elite. Those who are neither also need our attention. We need to learn more about young adult language learners who are located at the other end of the socioeconomic scale, and who have suffered and continue to suffer from sociopolitical, economic, and linguistic inequalities as a result. The underprivileged are often invisible individuals working several jobs or double shifts in order to make ends meet or to send remittances back home. These are individuals who due to time restrictions or financial limitations often cannot afford to take language classes. Their language learning experiences are often limited to workplace settings in which language does not play a large role or in other often hidden naturalistic contexts. For language learners experiencing mobility, researchers who study bilingualism

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

Young Bilingual Adults

75

should turn their attention to the social, economic, political, and linguistic implications brought on by the new social order resulting from globalization processes and neoliberal policies (see Lo Bianco, Chapter 8, this volume, on language policies in particular). Mobility and migration (regardless of what kind) affect how languages are learned and used by millions of individuals in different societies in various geographical settings around the world today. In our attempt to better understand language learning processes, we will need to make room for much needed methodological improvements and theoretical advancements as regards the bilingualism of young adults on the move.

6 Conclusion Much work remains to be done on how young adult global hybrids cope with the many linguistic challenges they undoubtedly meet up with. Moving around in a globalized world where national boundaries are becoming quite permeable means that young global hybrids may be faced with changes in the economic, political, social, and cultural systems they were previously familiar with. As the research reviewed in this chapter shows, such changes inevitably affect sociocultural practices such as language choice and the extent to which a new language may or may not be learned. In any discussion of young global hybrids as adult language learners, we cannot veer away from the cultural politics surrounding language and the powerful language ideologies that circulate on the macro level. These are often lived out on the micro level. In essence, we cannot dispense with the reality that language is a resource but also a commodity. Furthermore, whether one has particular language skills or not plays out as a form of gatekeeping. Such gatekeeping is found within various geographical, social, political, cultural, racial, gendered, and institutional as well as private contexts. A more profound understanding of such contexts will help explain how young global hybrids handle their language learning and language use as they navigate the world.

Acknowledgment This work was partly supported by the Research Council of Norway through its Centres of Excellence funding scheme, project number 223265. I would like to thank Elizabeth Lanza, Kristin Vold Lexander, Maria Antonia Obojska, and Rafael Lomeu Gomes at MultiLing for their comments on earlier drafts of this chapter.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.004

4 Bilingualism in Midlife David Singleton and Simone E. Pfenninger

1 Introduction Research into adult bilinguals raises fundamental theoretical and indeed practical questions. Is there a point beyond which it is in some sense too late to begin an additional language? Can a native language be acquired and consolidated to such an extent that it cannot be affected by fluctuations in input and experience in later life? What kind and degree of language exposure and use is necessary for bilingual individuals to maintain both their languages into maturity, and to continue to perform in both at a similarly functional level? What are the roles of socioaffective and environmental factors in these developments? These are questions of fundamental importance, considering current efforts to promote bilingualism in Europe and to boost the teaching of foreign languages (FLs) around the world (see Mun˜oz & Spada, Chapter 12, this volume), as well as the fact that global migration has become a natural feature – as is evidenced by more than 3% of the world’s population, or almost 244 million, being international migrants (DELMI, 2014, quoted in Forsberg Lundell & Bartning, 2015b; see also Migration Policy Institute, no date). In this chapter, we reflect on the capacity for bilingualism in the adult years of life, both in the case of adults who were once bilingual children, and in the case of adults who become bilingual or add new languages beyond their first during midlife. We do so by addressing two widespread myths that have to do with age and the timing of bilingualism. One is that when the onset of bilingualism is early, that is, from birth or during the years prior to formal schooling, the natural and expected outcome is an adult bilingual who possesses a so-called balanced proficiency in both languages. The other myth is that when the onset of bilingualism is relatively late, that is, when an additional language or languages are acquired in adulthood, success is impossible because, simply, people are assumed to be too old to learn. Our goal is to debunk both myths. In

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

77

the first half of the chapter, we argue that a near-perfect command of a given language in mature years is in no way guaranteed by the fact of an early start (even of a very early start) of acquisition in an input-rich naturalistic setting; nor is it the case that those naturalistic learners who begin the second language (L2) later in life inevitably fail to attain the levels reached by younger beginners. Bilingual users of any age never perform exactly like monolingual native speakers, and this should not come as a surprise: They are not monolingual native speakers (see also Treffers-Daller, Chapter 15, this volume). And to quote an Irish expression, even the dogs in the street know that when there are two languages in the same head they will talk to each other, quarrel with each other, and generally influence each other. The longer the contact, the greater, other things being equal, that influence is likely to be. In the second half of the chapter, we inspect some neurological research that addresses the question of differences between adults’ and children’s brains related to their capacity for second language attainment. Finally, we discuss methodological challenges that will need to be tackled in the future if researchers are to do justice to the complex question that we take as central in the study of mature adult bilinguals, namely: how do specific individual characteristics and environmental variables interact with the age at which bilingualism sets in life?

2 Is an Early Start a Guarantee for Highly Proficient Bilingualism Later On? The linguistic environments in which bilingualism is encountered vary enormously. It is often very difficult to determine precisely why some young people start to speak both languages and actively continue to use both of them, maintaining them into adulthood, whereas in other cases productive use is associated with just one of their languages, the other becoming merely receptively proficient in adulthood. Sometimes all use of one of the languages may cease, so that there is effectively a transition from bilingualism to monolingualism (De Houwer, 1999). De Houwer (2007) found that despite a bilingual environment, 25% of the 1–20-yearold children studied in her survey did not use the language that either one or both of their parents spoke. This phenomenon has been attributed to a variety of factors: frequency of input, parents’ linguistic choices and their discourse and interaction strategies, socioeconomic status, and unequal status of the languages in question. In this section, we highlight factors that mediate access to language use and therefore language maintenance in adulthood. We examine select social factors that are important for adults in the case of simultaneous or very early bilinguals, and we then explore the two specific cases of heritage speakers and bimodal bilinguals.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

78

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

2.1

Social Factors Determining the Maintenance of Two Languages into Adulthood For simultaneous or very early bilinguals, a powerful influence moderating maintenance of bilingualism into adulthood is degree of cultural and ethnic identification within the local community or, conversely, stigmatization outside that community. Identification or stigmatization are important because to varying degrees they either support or hinder bilingual language use and, indeed, bicultural participation more generally. An interesting study showing the importance of social factors is Albirini’s (2014). Albirini compared two groups of bilinguals who had grown up in the United States, one Palestinian Arabic–English and the other Egyptian Arabic–English. They comprised a mixture of US-born and foreign-born individuals with immigration ages between one and seven. They were all in their 20s at the time of testing. It turned out that across a number of measures, the Palestinian participants had a significantly higher level of proficiency in Palestinian Arabic than the Egyptian participants had in Egyptian Arabic. The critical factor in this regard was found to be the higher level of use of Arabic among the Palestinians. Qualitative data from interviews indicated that the advantages accruing to Palestinian speakers from their greater use of (and, relatedly, exposure to) Arabic seemed to be attributable to their commitment to Arabic as a main marker of their heritage and identity, to the encouragement they received from their families to maintain their command of Arabic, and to the wider Arabic-speaking social networks they had access to. Immigrants and their offspring from particular cultural and religious affiliations are the targets of stigmatization more often than others, which can have consequences for the access to their languages that they end up experiencing. Eijberts and Roggeband (2016) explored how first- and second-generation migrant women of Turkish and Moroccan descent in the Netherlands reacted – individually and collectively – to negative stereotyping. Only some of the women may have been early bilinguals, as theirs was a large and heterogeneous sample regarding age of onset of bilingualism, as well as with respect to age at the time of interviewing, educational level, and proficiency in Dutch. The responses and coping mechanisms their study uncovered also ranged greatly. Some women concealed their affiliation while others accepted the immutable reality of stigma; some withdrew from situations in which the stigma might be forced to the fore, while others attempted to compensate for their stigmatized status by being more likeable; yet others chose to directly confront and challenge sources of stigmatization while some increased their social identification with the stigmatized group. One can readily see how some of these strategies (for example, concealment of affiliation) might lead to a lessening of use of the language specifically associated with the relevant group, while others (for example, increased identification with the group in question)

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

79

might lead to an increase in the use of the relevant language. The implications for variation in bilingual learning (and non-learning) are clear. One of the critical questions concerning bilingual language development and maintenance into adulthood is how much input is enough for language acquisition in two languages to be successful. The issue has begun to attract keen empirical attention in the case of simultaneous or early child bilinguals who are tested during childhood (see Armon-Lotem & Meir, Chapter 10, this volume). As yet, however, very few studies have addressed the issue with regard to adult participants. One of them is by Bylund, Abrahamsson, and Hyltenstam (2012). Proceeding from the assumption that those bilingual speakers who reach nativelike L2 proficiency as adults do so because their first language (L1) may undergo some attrition, these researchers examined L1 and L2 knowledge in adult L1 Spanish–L2 Swedish bilinguals whose ages of first exposure to the L2 ranged from 1 to 11 years (the average length of residence in the L2 environment was 23.7 years). The participants’ reported frequency of L1 use was less than 30%. Contrary to the authors’ expectations, they found a positive correlation between L1 and L2 proficiency, such that those who performed like native speakers in the one language were likely to do so in the other language as well, and vice versa. While the findings by Bylund et al. (2012) are encouraging and clearly show supportive relations between the L1 and the L2 (mirroring those found for four- and five-year-olds by Winsler, Kim, & Richard, 2014), the concern remains that insufficient input in the minority language will thwart the likelihood of maintaining bilingualism during the adult years. Even in contexts where there is consistent use in the home of a given language by both parents, societal conditions may militate against the attainment and maintenance of a fully-fledged proficiency in it. This has been shown for many studies of child bilinguals who are learning two languages in a minority–majority relationship (see review by ArmonLotem & Meir, Chapter 10, this volume). But this is true even where the home language is supposed to enjoy an official majority-language status. One can cite in this connection the instance of the situation in the Gaeltachtaı´, the officially designated Irish-speaking areas in Ireland, where increasing inward migration of people who use English has drastically reduced the amount of interaction in Irish available in everyday life. This, together with the higher status of English, is having a major impact, to the ´ Giollaga´in, and O ´ Curna´in (2014), extent that according to Pe´terva´ry, O reaching and traversing adulthood with what would once have been considered a native command of the Irish language is very difficult to achieve.

2.2 Heritage Speakers and the Role of Literacy and Schooling Our discussion thus far shows that sociolinguistic factors influence language learning, including the characteristics of language status and

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

80

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

prestige, as well as the community the bilingual individual is raised in, for example, in those cases where one of the two languages that the child is acquiring is a minority language (with possibly lower status), and the other one is the majority language of the wider society. In this regard there has been increasing research on understanding adult heritage speakers (Polinsky & Kagan, 2007). Speakers of heritage languages (alternatively referred to as ethnic minority languages or community languages) are the children of immigrants born in the host country, where the majority language(s) is/are spoken, or immigrant children who arrived in the host country some time in childhood; they are thus either simultaneous or sequential bilinguals but are often bilinguals with more familiarity and comfort in the majority language (even though this may be their L2) and a weaker or more limited functionality in the minority language (even though this may be their L1, or one of their L1s in the case of simultaneous bilinguals). According to Montrul (2012), what heritage speakers have in common is that by the time they reach adulthood, the heritage language is their weaker language: “[T]ypical outcomes of the heritage language acquisition process by the time these children reach early adulthood are nonnative like competence and use of the language, better ability with receptive than productive language, non-uniform levels of proficiency, and linguistic gaps that resemble the patterns attested in second language acquisition” (p. 183). Heritage speakers and adult L2 learners are similar insofar as they both show signs of transfer from the majority language and patterns of development of the heritage language that differ from those of monolingual speakers in the country of origin, although for certain structures, heritage speakers may behave closer to native speakers than to L2 learners. Many of the findings from heritage speakers come from studies of young adults in their early 20s, rather than mature adults. Nevertheless, considering this research is instructive, since patterns of language use in midlife adults with similar language biographies may well turn out to be very ¨ zsoy, and van Hout (2007) analyzed noun, adversimilar. Treffers-Daller, O bial, and relative clauses in the speech of three different groups of Turkish–German bilinguals and one monolingual Turkish-speaking control group. A first group of 20-year-old Turkish–German bilingual students had been born and raised in Germany and were thus second-generation heritage speakers of Turkish. A second group had also been born and raised in Germany but went back to Turkey with their parents at approximately the age of 13; therefore, they had been residing back in Turkey for about seven years at the time of testing. A third, younger group of 16-yearold heritage speakers had just recently returned to Turkey. The monolingual control group consisted of 20-year-old Turkish students who had been born and raised in Turkey. The study’s most conspicuous findings were two. First, the sociolinguistic background of the speakers (i.e., their linguistic biography) was clearly linked to the bilinguals’ use of the structures under investigation. Specifically, the first and third groups, who had had

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

81

the least contact with Turkish as spoken in Turkey, had a more limited stylistic repertoire and made use of simpler types of embeddings, whereas informants in the second group, who had had the most contact with Turkish, chose from a variety of constructions. Second, even in adulthood, several aspects of Turkish grammar were not completely acquired by Turkish migrants of the second generation in Germany (i.e., the first group), who had never been to a Turkish-medium school. By contrast, qualitative analyses from additional informants of an intermediate generation, who had spent a few years in the Turkish educational system in childhood or adolescence prior to emigrating to Germany, suggested they may have acquired some of the more complex constructions – yet still did not use them in the same way as monolingual (control group) and bilingual (second group) speakers who lived in Turkey. The lack of development of the L1 of migrants and the relatively poor performance of many immigrants in Western Europe can thus in part be ascribed to the lack of availability of school education in the minority language; as TreffersDaller et al. conclude, “[s]chools could play an important role in furthering L1 knowledge among [. . .] Turkish students [in Germany]” (p. 270). The importance of continuing instruction in the L1 in the school years as a factor to explain later adult language use is also a main theme in Montrul’s (2002, 2008) studies of adult heritage speakers who had been exposed to two or more languages early in life, yet never acquired high levels of proficiency in their L1 or later lost aspects of their L1. She has found that for these simultaneous or early bilinguals, an earlier starting age of the L2 means possibly greater loss of the L1. In other words, language attrition will be more severe in individuals who have been exposed to the two languages very early than in late-timed sequential bilinguals (for discussion of the dynamics of language unlearning, see Keijzer & de Bot, Chapter 14, this volume, and Ko¨pke, Chapter 18, this volume). However, parallel education in the L1 (the heritage language) and the L2 (the majority language), along with socioaffective factors (motivation, language preference, prestige, socioeconomic standing), as well as input and overall levels of education, interact with, and mitigate, age effects. What is more, Montrul (2012) notes that, just like L2 learners, heritage language learners seem to benefit from form-focused language instruction of the kinds discussed by Mun˜oz and Spada (Chapter 12, this volume). As evidenced in a large body of bilingualism research (see Murphy & Evangelou, 2016), young learners’ heritage language need not be avoided or discouraged in school but, on the contrary, should be encouraged, as the L1 has positive consequences on the L2 (see Garcı´a & Tupas, Chapter 20, this volume, for discussion of other reasons too). Montrul’s data elucidate the importance of age-related social, psychological, and contextual factors as opposed to strictly maturational factors that invoke effects of aging in the brain. It becomes clear that owing to the complex status of age as a “macro-variable” (Montrul, 2008, p. 1), which covaries with environmental factors, the age factor has to be considered (and researched) in light

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

82

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

of macro-cultural and micro-cultural phenomena that can have a bearing on interpersonal relations that influence, shape, increase, or decrease variables such as motivation, literacy, and schooling.

2.3 Bimodal Bilingualism within Deaf Communities Bilingualism involving signed and spoken language has much in common with bilingualism as we encounter it in other contexts. What is special about it, of course, is that it is bimodal. This means, inter alia, that signing– speaking bilinguals can communicate in both their languages simultaneously (see Tang & Sze, Chapter 24, this volume, for extensive discussion). Another distinguishing feature is that Deaf communities (which includes hearing signers who associate and align themselves with such communities) have the status of a minority group in every single national society and ethnicity in the world. Kanto, Huttunen, and Laakso (2013, p. 244) characterize Deaf communities as exhibiting “strong social bonding” and transmitting “cultural knowledge (values, customs, and information)” to their members. As Grosjean (2010) and Tang and Sze (Chapter 24, this volume) discuss, there is even greater diversity in language input and use reported among bimodal bilinguals than by individuals who are bilingual in spoken languages. This is due to wide variability in speech perception and recognition in persons with similar hearing impairments, differing language backgrounds, varying linguistic environments (e.g., sign language contacts), and varying degrees of involvement in the Deaf community among its members. Worth noting, however, is that profoundly deaf children who are exposed only to oral language in their early years are deprived from initial language access and thus cannot acquire an L1. If they later encounter a signed language as older children, adolescents, or adults, the proficiency they attain in it (in terms of, for example, syntactic processing) differs from the proficiency of signers who had access to relevant language input from birth (Lieberman & Mayberry, 2015; Mayberry & Lock, 2003). Mayberry and Lock (2003) infer that “a lack of language experience in early life seriously compromises development of the ability to learn any language throughout life” (p. 382), which they attribute to the effects on relevant neural development of a critically impoverished environment (cf. Mun˜oz & Singleton, 2011). One kind of bilingualism in Deaf communities can result from the use of assistive technology (Lieberman & Mayberry, 2015), specifically from the use of cochlear implants, which are electrode arrays surgically implanted in the cochlea of the inner ear, providing a sense of sound in the context of profound hearing loss. Within Deaf communities, however, the ethical and social implications of cochlear implants are hotly debated, with some arguing they spell ethnocide for the community (e.g., Hladek, 2002). These implants can be offered to signing children, who then become bilingual in

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

83

sign and spoken language. Research into such bilingualism is only just beginning, but some early findings indicate that in these children, the acquisition of spoken language benefits from the prior acquisition of sign (Hassanzadeh, 2012). On the other hand, given the technological advances in cochlear implantation, more and more children receive an implant in the first year of life, before they have been able to learn to sign. Lieberman and Mayberry (2015) argue that, given what we know about the importance of early language exposure, it is desirable that scientific research continue to investigate the potential benefits of early sign language exposure for all deaf children, irrespective of the amount and type of assistive technology from which they may subsequently benefit. Cochlear implants are very much less available to adults; for example, in the United Kingdom, they are offered only to profoundly deaf adults who are also blind (Davis, 2014). Consequently, data on cochlear implant-facilitated spoken language–signed language bilingualism in adults are very thin on the ground and the phenomenon awaits systematic research. Another kind of bimodal bilinguals within Deaf communities is hearing children of deaf adults (CODAs). They grow up to be adults who are typically bilingual in a signed and a spoken language, but there is wide variation in the degree of proficiency and balance between their languages. This variation is usually not something that develops in adulthood, but is largely a continuation of a pattern discernible from earlier ages (Pizer, Walters, & Meier, 2013). Sign language seems to be weaker when parents’ deafness is not such as to oblige family members to rely absolutely on sign for communication with them. There appears to be an ideology common to CODAs, which is that effort in sign language is appropriate to the degree that it overcomes potential communication barriers (Pizer et al., 2013, p. 84). Another factor appears to be the extent to which there have been hearing siblings in the family or other signers available to interact with. Of 13 American CODAs interviewed by Pizer et al., only 3 emerged for whom English never surpassed their use of American Sign Language (ASL) and who used it with high levels of proficiency. Pizer et al. (2013) explain this high proficiency in ASL in terms of many opportunities to use it in the family. On the other hand, however, language experiences as adults, including social networking with other CODAs or professional engagement with interpretation, shape patterns of adult language use beyond the early years of primary socialization in the family and give rise to an adult CODA identity that is distinct from other identities seen in Deaf communities (p. 81). Evidence from different types of bilinguals within Deaf communities – and indeed all other evidence across different populations of simultaneous or very early bilinguals reviewed in this section – reflects Grosjean’s (1997) “complementarity principle,” according to which the level of fluency attained in a language (more precisely, in a language skill) will depend on the need for that language and will be domain specific: “Bilinguals

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

84

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

usually acquire and use their languages for different purposes, in different domains of life, with different people. Different aspects of life require different languages” (p. 165). Grosjean’s insight has had far-reaching linguistic and methodological implications. It explains not only why bilinguals’ language abilities rarely are identical across languages, and why the fact that they have traditionally been evaluated against native speakers (or monolinguals) must be considered a fatal flaw, but also why bilinguals’ language repertoires change over time: “as the environment changes and the needs for particular language skills also change, so will their competence in their language” (Grosjean, 1998, p. 132). New situations, new interlocutors, and new language functions will involve new linguistic needs and will therefore change the language configuration of the person involved. The corollaries are that adult bilingual interactions must be studied and linked to attained degrees of bilingualism (see De Houwer, Chapter 17, this volume) and their linguistic competencies measured bilingually, that is, in both languages (see Treffers-Daller, Chapter 15, this volume).

3 Is It True That Mature Adults Cannot Learn a New Language Well? We now turn to adult L2 learners who, unlike adult heritage speakers or many bimodal bilinguals, acquire additional languages in adulthood. Investigations into proficiency in L2 and ultimate attainment – typically defined as the highest level of proficiency that can be achieved – are usually undertaken in relation to age of acquisition, as well as linguistic accuracy, language use, immersion, and acculturation (e.g., Hammer & Dewaele, 2015).

3.1 The Critical Period Hypothesis for Bilingual Learning Age of onset of language learning has very often been linked to the critical period hypothesis (CPH) (also discussed by Biedron´ & Birdsong, Chapter 16, this volume). The CPH posits that there are biological constraints on what language acquirers can attain beyond a certain maturational point. The location of this maturational point is much disputed. The CPH is actually an array of hypotheses with widely differing predictions: “the end of the critical period for language in humans has proven [. . .] difficult to find, with estimates ranging from 1 year of age to adolescence” (Aram, Bates, Eisele et al., 1997, p. 85). The very existence of such a critical point is called into question by such variability. There is also much discussion about what kinds of linguistic capacities are supposed to be affected by any putative critical periods at different ages and why (e.g., Granena & Long, 2013b).

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

85

Some L1 findings seen as favoring the CPH position relate to children deprived of the experience of language in childhood. When such individuals experience post-childhood restoration to a language-rich environment, they tend to progress in language development in a limited or unusual manner. The value of such evidence in regard to the CPH has, however, been questioned (including by CPH advocates, e.g., Lenneberg, 1967), because it is interpretable in terms of the overall physiological and cognitive damage done to an individual by isolation and deprivation of interaction (Singleton & Mun˜oz, 2011, p. 407). Other L1 evidence comes from profoundly deaf subjects who did not have access to sign language input in their early years and who then acquired a sign language as their L1 later (e.g., Lieberman, Borovsky, Hatrak, & Mayberry, 2015; Mayberry & Lock, 2003; see Section 2.3). Research into such cases has not found that language completely fails to develop, but it has revealed vulnerable areas where development in the language of later signers differs from that of early signers. Again, though, a very plausible interpretation is that such differences reflect not the termination of a critical period, but the general effects resulting from being cut off from language-mediated social interaction during the period when cognitive development is most intense. Findings from research on the learning of additional languages that are taken to support the CPH largely stem from studies of immigrants with varying ages at the time of immigration. They typically show that the younger immigrants are when they relocate to a place where the societal language is not their L1, the more likely it is that they end up with nativelike command of the new language (e.g., Granena & Long, 2013b). It is not the case, however, as is clear from our review in Section 2, that all early arrivals in a new language community end up with nativelike proficiency in the new language. Nor is it the case, as we discuss in what follows, that those who arrive later always fail to attain such a level. Myriad studies that have investigated age relations as they emerge for sequential bilinguals (first- or second-generation immigrants) in naturalistic settings have shown that age of onset tends to have an effect on (selfreported) proficiency, on self-perceived competence, and on language choice for emotional expression. However, numerous other factors, such as context of acquisition, frequency of use, L1, gender, and acculturation, also play into the equation (see Armon-Lotem & Meir, Chapter 10, this volume, for discussion of some of these factors for sequential bilinguals in childhood). Thus, the idea that late L2 learners cannot score within the native speaker range, or early learner range, has been dismantled into finer components. In fact, late starters can also be good learners of an L2, and they can compensate for a late start by means of other factors such as aptitude, identity, motivation, and attitudes that can possibly override the impact of the purported age factor and explain variability. As Rastelli (2014, p. 65) puts it, it is not so much about whether adults can attain an L2 to a nativelike level, but how they do it.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

86

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

Social-psychological factors are found to be of utmost importance to explain the degrees of success experienced when adults undertake the learning of an additional language. In particular, motivation, attitude, and a deliberately chosen and cultivated sense of a new self in the new language appear to shape adult learning outcomes. As we see in the remainder of Section 3, these factors have been very widely found to relate to acquisitional effects in adult learners of minoritized Indigenous and regional languages, and they are equally relevant for the learning of foreign or world languages, as well as majority languages, in contexts of migration.

3.2

Successful Adult L2 Learners of Indigenous and Regional Languages The learning of Indigenous (e.g., Ma¯ori in New Zealand) and regional languages (e.g., Scottish Gaelic in Scotland) is a particularly challenging task for adult members of communities where languages need revitalization, given the fact that exposure to such languages often requires conscious choice and deliberate effort, and because minoritization is a common experience for speakers and learners of Indigenous as well as regional languages, vis-a`-vis more widely used majority languages in the same geography. Gorter (2017, p. 87) illustrates the problem by noting that if someone asks for something in a shop in Frisian (in Friesland in the Netherlands, where Dutch is dominant) or Basque (in the Basque Country in Spain, where Spanish is dominant), they are likely to receive as a response from the salesperson “I am sorry, but I don’t understand you,” whereas if the client asked in a third language, say, English, the salesperson would probably make an effort and try to understand. This kind of situation may lead to the “hiding” of the Indigenous or regional language and/or to “enhanced feelings of anxiety” when people do use it (p. 87) (see De Houwer, Chapter 17, this volume, for discussion of the symbolic value of particular language choices). The key elements for the successful acquisition of both Indigenous and regional languages among adults seem to be very high motivation, plentiful interaction, and friendly relationships involving the language in question. This is shown in three studies we review next. Ra¯tima and Papesch (2014) tell the story of the second author, Te Rita Papesch, an English-speaking Ma¯ori woman living in New Zealand who in adulthood set about the project of learning the Ma¯ori language, embarking on this first in her late 20s. Te Rita was primarily motivated to learn the language as a means to better understanding her own Ma¯ori identity. She was clearly very open to change, and her commitment to learning the language went hand in hand with the taking on of important cultural responsibilities. She worked hard on building good relationships with her mentors and fellow speakers; among the latter were her own children,

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

87

who learned the Ma¯ori language at university so that the family were gradually able to establish a home immersion situation. Te Rita also drew on elements of her life such as songs to aid her learning. The results of Te Rita Papesch’s language project were outstanding. In an environment where the very survival of the Ma¯ori language is now in jeopardy, she reached the highest level of mastery of the language (“complete proficiency”) on the New Zealand Ma¯ori Language Commission’s fivelevel scale (Ra¯tima & Papesch, 2014). Strikingly similar favorable factors are uncovered by King and Hermes (2014) in the case of successful adult learning of a Native American language, Ojibwe. King and Hermes first note that the teaching and learning of Ojibwe typically make use of methodologies that tend to be rather passive or academic in nature and to feature little in the way of interaction in the language. Such methodologies tend, therefore, “to impede communicative use of the language by learners” (p. 276). The authors then go on to relate the successful experience of two very determined adult learners of Ojibwe who, because of their strong desire to make progress, stepped outside the accepted pattern and engaged with Ojibwe in interactive and friendly ways. Nance, McLeod, O’Rourke, and Dunmoreb (2016) offer interesting insights into adult L2 motivation to learn and use Scottish Gaelic as new speakers, a term that denotes “those who did not acquire the language through traditional intergenerational transmission in the home, but instead acquired it through immersion education, adult education or other formal or informal means” (p. 167). New speakers may or may not have an ethnocultural affiliation with the language in question. Nance et al. (2016) specifically investigated Scottish Gaelic rhotic production among a group of 19 generally successful adult new speakers living in Scotland, who enjoyed some status in the community by holding “professions such as immersion school teaching, translation, language officers, and other positions requiring extensive expertise in Gaelic” (p. 168). Nance et al. found that some of these new speakers made the traditional velarized/ palatalized rhotic distinction and others did not. When the distinction was not made, this did not always reflect L1 influence or what is often portrayed as falling short in relation to nativelike phonology, but sometimes appears to have been the result of an ideal L2 self (Do¨rnyei & Ushioda, 2009) shaping the process of motivation according to perceptions of what the individual speakers found appropriate for themselves. For example, some speakers “preferred an ideal self that was more oriented toward a new-speaker model and considered a native-speaker target as inauthentic” (Nance et al., 2016, p. 185).

3.3 Successful Adult Foreign Language Learning Social-psychological factors are also relevant in the context of adult foreign language (FL) learning. This is illustrated by Moyer (2013), who

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

88

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

reviews a number of studies suggesting a link between strong L2 motivation in adult learners and nativelikeness in their FLs. Among these studies is an early investigation of her own in which Moyer found that, of 24 highly motivated graduate students of German as an FL who also taught German, 14 overlapped with native controls in terms of the results on one or more tests of German phonology (Moyer, 1999). One of the participants was deemed to pass as a native across the board. He had begun to study German at age 22 as a monolingual English speaker and was “largely selftaught” (p. 98); his motivation to achieve nativelike status as well as his fascination with German language and culture were exceptionally high. Focusing on English rather than German, Nikolov’s (2000) study of 13 adult Hungarian learners of English as an FL in Hungary (with experience of just one academic term of study abroad in the United States) is also informative and converges with Moyer’s (1999) findings. Of the 13, one was deemed to be a native speaker of English by a large majority (89%) of the English native speaker judges, and a further four were rated as native speakers by a sizeable percentage of the judges (42–56%). What these successful learners shared was a professional identification with English, an integrative orientation toward Anglophone culture, a determination to find opportunities for increasing and maintaining their FL proficiency, and an outgoing, sociable approach to English-speaking contacts. The part played by motivation and attitude is once again striking. The findings of Samimy’s (2008) study of a 39-year-old American adult acquirer and user of Arabic has some features in common with those of Moyer (1999) and Nikolov (2000) but also exhibits some differences. Her subject, Mark, was a white English-speaking American who did not migrate to an Arabic-speaking country, although he spent quite considerable periods in such countries studying and actively seeking opportunities to use Arabic. In the United States, he had a professional connection to Arabic in his role as a doctoral student of Arabic as an FL who also taught the language, but he experienced certain obstacles with respect to integrating with the expatriate Arabic-speaking community he knew in the United States. As Samimy puts it, “[b]eing white, blonde, and blue eyed, and married to a non-Arab woman has limited his access to the target language community” (p. 412). This failed to discourage Mark, however. What helped him overcome the difficulties was “his perseverance and commitment to bringing the Arabic world to him by creating ‘Arabic bubbles’ [in which he thought in Arabic] in every possible situation” (p. 412). (Biedron´ and Birdsong, Chapter 16, this volume, offer examples of equally determined language learners who learn multiple FLs.) In the end, the social dimension of Mark’s motivation was rewarded. He made of his Arabic professors his “community allies” (p. 412), who helped him attain access to the social networks of Arabic teaching circles. The outcome of this learning and use project was highly successful: Mark’s proficiency was assessed as Superior on the American Council on the Teaching

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

89

of Foreign Languages (ACTFL) proficiency scale (equivalent to level C2 of the Common European Framework of Reference for Languages; see American Council on the Teaching of Foreign Languages, no date), and the nativelike quality of his performance in Arabic was often commented on by Arabs. An interesting illustration in the context of adult English FL learning comes from Andrew (2012), whose study of English as an FL in Mexico helps explain how social-psychological factors interact with age, which itself can be considered a social-psychological rather than a biological variable alone. Through classroom observation and a series of periodic interviews with seven Mexican adults during an English as an FL course, Andrew collected stories to identify critical moments and events that enabled her “to understand the meaning the participants gave to their language-learning experiences and how age factored into their lives” (p. 67). Her data showed that these adult learners’ perceptions of age and age identity contributed to their language acquisition process. Analyzing the experiences of four female adults between ages 34 and 59, she found that their attainment was constrained not necessarily by cognitive ability per se, but by their negative construction of later adulthood identity as evinced, for example, in occasional ageist comments. However, unlike the older men in Andrew’s study (who were over 60 years of age), these mature female adults viewed language learning, and the learning of English, as part of their ongoing narratives of professional development, and this narrative led to positive learning experiences (see Goral, Chapter 5, this volume, on adult bilinguals over 60). For the four middle-aged women, learning English seemed to constitute a means for them to acquire cultural capital, given the prestige of English in Mexico, rather than a means for communication in daily life. Their learning had to be carefully juggled with their primary responsibilities to their families and jobs. Even though the learners firmly believed that younger people are at an advantage when it comes to the task of studying an FL, they also pointed out that adults are often more focused, more motivated, and more serious than adolescent or young adults. Thus, Andrew’s findings strongly suggest that the experience of learning a language varies according to each individual’s position in the lifespan and involves both linguistic and nonlinguistic dimensions.

3.4

Successful Adult Language Learning by Cultural and Economic Migrants Turning to adult learners who, unlike learners of Indigenous or regional languages, did migrate to an environment where contact with their target language was an inevitable part of their everyday experience, similar social-psychological factors are relevant as in the contexts discussed thus far, including motivation, attitude, and an emerging sense of a new self.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

90

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

For example, Diskin and Regan (2015) studied Polish and Chinese immigrants to Ireland who wanted to learn English to fulfill economic and higher educational aspirations (see Grin, Chapter 9, this volume, on the links between language and issues pertaining to economics). The researchers focused on the migrants’ acquisition of English discourse-pragmatic markers (e.g., like, you know, I mean) and found such markers to be fairly rapidly and copiously acquired, approximating, indeed, to native speaker levels of frequency. These markers were increasingly used as length of residence increased and when the migrants had an integrative attitude to Irish culture. The exception, interestingly, was the clause-final use of like, which participants apparently felt to be too marked for Irishness and not representative of more general English native speaker usage. Let us consider more closely adult language learning evidence from advantaged migrants, or what Forsberg Lundell and Bartning (2015b) call cultural migrants, that is, “people who choose, out of their free will, to move permanently to another country and to learn the language” (p. 4). This population differs from what Benson and O’Reilly (2009) call lifestyle migrants in that the latter have as their main motivation the search for a better life, not seldom preceded by a traumatic or stressful event (for discussion of the language learning needs of this migrant population, see Simpson, Chapter 13, this volume). Cultural migrants, on the other hand, have a more focused motivation, possibly also linked to L2 learning motivation; due to their curiosity and interest in the target language context, they are typically highly motivated to immerse themselves in the majority culture and language, they are well educated and well integrated, and thus have optimal prerequisites for nativelike attainment. Kinsella and Singleton (2014) studied 20 native English speakers originally from the United Kingdom, Ireland, and the United States, all residents in France, whose average age of moving to France and thus first significant exposure to French was 28.6 years (ranging from 22 to 60). The participants were all successful learners in the sense that they all, by their own account, regularly passed for native speakers of French. The majority had emigrated to France for emotional rather than economic reasons and all but one reported feelings of attachment to the country: “For the majority, France had a romantic appeal, and some of the participants reported experiencing strong emotion on their first visit to the country. With the exception of only one participant, a considerable degree of Francophilia prevailed across the sample” (p. 445). None reported feeling any sense of loss in relation to their L1 linguistic identity. In the context of the study they were all given the tasks of identifying some regional French accents and completing a lexico-grammatical test. Six of the 20 participants had met their future French partners before coming to live in France. These six found making contact with native French speakers unproblematic. The same was not necessarily true of the sample at large. For some participants, it was only after several years’ residence in France that they felt they had the

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

91

amount of daily contact with native French speakers that they wished for. Three of the 20 participants scored within native speaker ranges on all tasks. All three (i) had spent a very long time in France (on average, nearly 30 years), (ii) conducted their social life mostly through French, (iii) identified themselves closely with the Francophone community (so closely that they wished to live the rest of their lives in France and be buried there), and (iv) had French partners. More classic cases of copious and intense interaction in the L2 and of integrative motivation (and behavior) would be hard to find, and indeed these adult learners fit the classic definition of integrativeness or “a genuine interest in learning the second language in order to come closer to the other language community” (Gardner, 2001, p. 5). The seven studies included in Forsberg Lundell and Bartning’s (2015a) collection illustrate that many cultural migrants who begin learning the new language in adulthood reach very high levels of L2 attainment and are highly functional in their L2 in everyday life. Some of the authors present their findings against the backdrop of models that explain how L2 acquisition is affected by sociopsychological factors, such as Schumann’s (1978) Acculturation Model, which links adult L2 learning with sociocultural and psychological processes of acculturation. According to this model, adult L2 learners may develop ego boundaries and attitudes that can constrain acquisition. Although after the 1980s not many bilingualism studies explicitly analyzed adult L2 acquisition themes in light of Schumann’s model, there is ample evidence that those learners who are well integrated, who have positive affect toward their host community, and who perceive themselves as close in relation to the target culture reach high L2 levels. Thus, several researchers have redirected attention to the notion of acculturation, because, as Hammer and Dewaele (2015) note, a main strength of Schumann’s model is not only the inclusion of affective (psychological) and integrative (social) variables but also its potential robustness to uncover the reasons behind high levels of L2 attainment. Nevertheless, there are studies that have failed to find a relationship between acculturation and degree of L2 attainment. Granena (2015), who like Hammer and Dewaele (2015) also chose Schumann’s Acculturation Model as a theoretical starting point, found that neither age of onset of L2 learning nor socioaffect were able to predict morphosyntactic attainment. She analyzed timed and untimed grammaticality judgments of Chinese L1–Spanish L2 sequential bilinguals (between 18 and 50 years old at the time of testing) as a function of self-assessment of four nonbiological factors: perceived L2 learning success, identification with the L2 culture, satisfaction with one’s own pronunciation, and desire to pass as a native speaker. None of these factors had a significant impact on attainment by either early bilinguals with age of Spanish learning onset ranging from three to six years or late bilinguals with L2 learning onsets of 16 years and older.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

92

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

Despite the obvious strengths of Schumann’s (1978) Acculturation Model, it has also been questioned whether the concept needs to be reformed and actualized. For instance, as Forsberg Lundell and Bartning (2015c) rightly note, Schumann’s model runs into conceptual problems when regarded in the context of today’s globalization. The questions that arise, among others, are whether we can acculturate into a global culture, and whether the power of globalization will lead to a less pronounced relationship between culture and language: “Will ‘cultures’ situate themselves at other levels than those of nation and language, such as subfields of interest, for example, skateboarding, environmental issues or architecture?” (p. 225). Another interesting aspect raised in Forsberg Lundell and Bartning’s (2015a) volume is the suggestion that bilingualism and second language acquisition researchers should perhaps be more interested in adult learners’ possibilities to function in society, that is, in their communicative proficiency in various migratory contexts, rather than their grammatical intuition or voice onset time.

4 How Different Are Child and Adult L2 Acquisition? Evidence from the Mature Adult Brain We are in essence arguing that a difference between child and adult learners is not so apparent because social-psychological factors interact with chronological and biological age and create great diversity of outcomes for early as well as late bilinguals, such that bilingual outcomes are neither deterministic of success for children (Section 2) nor deterministic of failure for adults (Section 3). This position does not necessarily lead us to claim that a child–adult difference in language learning does not exist. We review here some relevant evidence, with special attention to emerging neurolinguistic evidence about the mature adult brain. One striking age-related difference is a reverse pattern of relative advantages seen in child versus adult sequential bilinguals when they are compared on rate of learning versus level of attainment. Ever since Krashen, Long, and Scarcella (1979), it has been well known that under naturalistic learning circumstances, the rate of L2 acquisition of adult immigrants is faster than that of child immigrants at least for a while; eventually, however – typically after three years of naturalistic immersion in the target language environment – young starters generally catch up with and surpass adults. In other words, the L2 acquiring advantage that children generally have over adults in such circumstances seems to require a certain period of time to manifest itself. Thus, a much cited study by Snow and Hoefnagel-Ho¨hle (1978) of English-speaking migrants to the Netherlands of various ages found that after the first four or five months of residence in the Netherlands, adults’ proficiency in their newly acquired L2, Dutch, was markedly superior to that of children in all areas except pronunciation.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

93

The older participants’ advantage began noticeably to be eroded in the following months, but the initial rate advantage for the adults was clear. We can plausibly attribute this to the more advanced cognitive stage of the adults. It is not the case that young starters eventually surpass late-starting adults because adult migrant L2 learning necessarily simply ceases. Adamuti-Trache (2013) found that adult immigrants to Canada continued to improve their new language skills during at least their first two years after arrival – especially those who started from little prior knowledge. However, learning may be slow, as Adamuti-Trache mentions that the least proficient of the individuals surveyed were performing in the L2 only “fairly well” after four years and that some, the least integrated socially, actually reported a decline in proficiency. The slowness of rate of learning for these least proficient participants, as well as the reverse pattern some reported, may in fact be attributable to low levels of social integration – something younger migrants tend to suffer from less. Early and (at least some) late learners may appear to be capable of achieving the same proficiency in the long term as native speakers, but some researchers hypothesize that either (i) younger learners may process L2 information differently from older learners, which may be rooted in different brain mechanisms (e.g., Clahsen & Felser, 2006; Morgan-Short, Sanz, Steinhauer, & Ullman, 2010; Wartenburger, Heekeren, Abutalebi et al., 2003), or (ii) they may process it similarly but under difficult circumstances (Herschensohn, 2007; McDonald, 2006). Siding with the notion of different underlying resources being available to children and adults, Rastelli (2014), among several authors, claims that children and adults acquire an L2 in fundamentally different ways. He suggests that although an adult’s brain is adaptive, and also capable of responding with anatomical and functional rapid changes to environmental challenges, it cannot recreate the exact conditions in which early language acquisition took place: “what late language learners cannot catch up with anymore is a subset of L2 grammar and L2 syntax which lies beyond the scope of [statistical learning]” (p. 69). Rastelli claims that early language acquisition takes advantage of a child’s brain’s neuroanatomical and function maturation (or the brain’s maturation in infancy, to use Rastelli’s words), and that adults rely on lifelong (rather than maximal) plasticity and the capacity of the adapting (rather than maturing) brain. The question then is, as Rastelli (p. 72) points out, whether the acquisition of an L2 in adulthood – or the acquisition of specific aspects of the L2, at least – must take place within a period of maximal plasticity or whether it can rely on lifelong plasticity instead. His answer to this question is an optimistic one. An adult’s brain, Rastelli says, is a plastic and adaptive mechanism; when something is lost, adults recruit additional neural, cognitive, and, importantly, environmental resources to cope with the learning task. Thus, the later the age of onset of bilingual learning, the more crucial environmental factors (quantity and quality of input, intensive practice, instruction, etc.) become;

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

94

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

although adults “must do it their own way” (p. 95), learning can still take place, and adults are “destined to catch up” (p. 84) in at least most areas of the target language (see Hernandez, Chapter 23, this volume, for further discussion of similarities and differences in the neurological correlates of language learning in children and adults; see Goral, Chapter 5, this volume, on language learning after 60). Critical or sensitive periods for second language learning in the human cortex have so far been mainly investigated using structural magnetic resonance imaging (for a summary, see Kliesch, Giroud, Pfenninger, & Meyer, 2018). In particular, studies have demonstrated an increase of grey matter volume prior to adolescence, followed by a decrease during and after adolescence (Giedd, Clasen, Lenroot et al., 2006). These macro brain changes are accompanied by cellular plasticity such as a vast production of axons and synapses before puberty, followed by pruning in later adolescence (Andersen, Thompson, Rutstein et al., 2000). It has been argued that the brain remodels during this developmental period and neural circuits are shaped from an unrestricted possibility to learn toward effective functioning in order to fulfill specific environmental needs (Crews, He, & Hodge, 2007). Although the duration and the end of the period is still a matter of ongoing discussion (see Section 3.1), the idea is that this peak in cortical development during adolescence (as measured by cortical volume) is correlated with a high capacity to learn an L2 (Uylings, 2006). However, it has been shown that a phase of peak in cortical development is not equally important for different components of language. The functional Magnetic Resonance Imaging (fMRI) study by Wartenburger et al. (2003) demonstrated that the neural representation of grammatical processing was depending on the age of acquisition, while the neural activation during semantic judgment was solely changing as a function of proficiency level. This finding was supported by Perani, Paulesu, Galle´s et al. (1998), who showed decreasing differences in fMRI activations between L1 and L2 learners with increasing proficiency level during semantic processing, independently of age of learning onset. This evidence therefore suggests that a high proficiency level in L2 semantics is achievable even when individuals start learning after adolescence. What is more, it has been reported for many years now that learning requires dendritic and synaptic plasticity and that such plasticity does not stop after adolescence but that it continues into adulthood (e.g., Lo¨vde´n, Wenger, Maº rtensson et al., 2013). In fact, there is conclusive evidence for substantial structural and functional neural plasticity in adulthood as a function of inter-individual trajectories of learning during L2 learning (e.g., Stein, Federspiel, Koenig et al., 2012). In other words, it is predominantly inter-individual differences in proficiency increases that are reflected in brain plasticity rather than initial or resulting proficiency levels. For instance, in Maº rtensson, Eriksson, Bodammer et al. (2012),

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

95

struggling learners showed stronger increase of grey matter in middle frontal areas compared to more successful learners. Finally, while aging may be characterized by a decrease in cognitive functions such as working memory, attention, and processing speed (see Goral, Chapter 5, this volume), the Scaffolding Theory of Aging and Cognition (Park & Reuter-Lorenz, 2009) suggests that there may also be brain compensatory mechanisms supporting the scaffolding for a higher level of cognitive functioning, such as frontal recruitment during difficult tasks, dedifferentiation, neurogenesis, and lateralization of functioning. Furthermore, experiences such as new learning, engagement, exercise, and cognitive training may positively influence the level of cognitive functioning (Park & Reuter-Lorenz, 2009).

5 Methodological Challenges We would like to close this chapter on bilingualism during mature adulthood with a discussion of methodological issues in research on age effects in adult language learning. Some of the methodological difficulties we discuss in this section are concerned with the design of studies and the sampling of participants, others pertain to the collection of data, while yet others relate to the processing and analysis of the results. All these challenges are interrelated and compounded. One of the main challenges with research on the factor of age in language learning is systematically and inextricably intertwined with other, co-occurring variables that are contextual, affective, and personal in nature. This means that the significance of starting age and biological age is rather difficult to model, precisely because we cannot disentangle effects of maturation (i.e., effects of aging in the brain) from effects of age-related circumstances (see also Goral, Chapter 5, this volume, and Biedron´ & Birdsong, Chapter 16, this volume). According to Forsberg Lundell and Bartning (2015c), given that age is confounded with other factors, research ought to be carried out on populations that seem socially and psychologically enabled for the task of acquiring an additional language as adults, such as migrants with class, race, and gender privileges. In a similar vein, DeKeyser (2013) points to the importance of gathering data from immigrants who are not beyond middle age, to avoid effects of cognitive aging on testing results, and who have had a fair chance to learn the new language. The direction of this kind of design is toward obtaining data from groups whose privileged attributes are known and can be taken into account when the results are assessed. A further issue is that of generalizability, most often linked to randomized research designs. In actual fact, in classroom research, random assignment to experimental conditions has frequently not been implemented – because of the very frequent use of intact classes, that is,

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

96

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

cluster-randomized interventions, for convenience or to preserve ecological validity. In naturalistic settings too, age researchers often use convenience samples that are not representative of the immigrant population, as they are strongly biased toward the more educated and hence tend to minimize any age effects (DeKeyser, 2013). There may be good reasons for these practices (see earlier in this chapter), but they place severe limits on the statistical generalizability of the relevant findings – when a convenience sample, not obtained randomly, does not represent, for example, a variety of socioeconomic levels. On this point, it should be kept in mind that an increase in sample size leads to an increase in the generalizability of one sample to other samples that the same sampling procedure would produce, but that it does not lead to an increase in the generalizability of any sample estimate to its corresponding population characteristic (e.g., Lee & Baskerville, 2003). When age research is carried out in classrooms, failure to use randomized designs is partly due to practical constraints, and we argue that disregarding these practical constraints and insisting on randomization as the gold standard would be misguided in these contexts (see Pfenninger & Singleton, 2017, pp. 47–48). Often schools do not allow for the random sampling of participants or the random assignment into groups, simply because individuals are already in intact classes. Moreover, even if we could ensure random assignment at the individual level, the downside of such carefully controlled – and artificial – research would be that it would lack ecological validity because the contexts and activities in question would not represent those ordinarily encountered by language learners (see also Duff, 2010). One alternative would be to use cluster randomization, where the unit of randomization is the class or school rather than the student, but this approach would need to be assessed when analyzing the data. With respect to sample size, DeKeyser (2013) also suggests that the sample sizes often need to be in the hundreds, if not bigger, in order to produce meaningful and interpretable results in naturalistic settings (he stipulates at least 20 participants per subgroup). We, of course, agree that small samples may have a debilitating impact on statistical power. Many advanced statistical procedures require larger sample sizes than the more traditional and simpler univariate analyses. However, here too, systematic sampling plays a crucial role in hindering the ideal of large sample sizes. According to DeKeyser, it is vital to obtain large samples of individuals who are largely isolated from their L1 community and thus engage in minimal L1 use, so as to test the limits of ultimate attainment in the L2. If he is right, then this would mean collecting data over very wide geographical areas, making large sample sizes more challenging to attain. When it comes to language learning of mature adults with their diverse language backgrounds, past experiences, and learning conditions, the use of longitudinal research designs is strongly recommended (see Ortega & Iberri-Shea, 2005). Most of the questions investigated in age-related studies

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

97

are fundamentally questions of time and timing that can be meaningfully interpreted only within a full longitudinal perspective. For instance, what do we know about the pace and pattern of FL development throughout mandatory school time (in an instructional setting) or throughout the lifetime of L2 learners (in a naturalistic setting)? What critical transition points in L2/FL development need to be taken into account when planning educational policy for more mature learners? Ultimately it is through cumulative longitudinal findings that age researchers are able to contribute meaningful characterizations of the gradual process of attaining L2 competence at different ages and with different ages of learning onset. Researchers also need to document more precisely than in the past what (kinds of) structures are increasingly problematic as a function of age. They also need to incorporate more explicit considerations of how type of structure interacts with co-occurring variables such as particular types of motivation and particular kinds of learners. Illustrations of this come from the studies by Hammer and Dewaele (2015) and Granena (2015) described earlier. Despite similar research questions and theoretical frameworks, the two studies obtained fundamentally different results – in favor of and against the impact of sociopsychological factors on attainment of adult learners, respectively. Granena (2015) speculated that her choice of morphosyntax as the language-dependent variable to investigate may have been infelicitous in that motivation may have a greater impact on L2 phonology rather than on morphosyntax. We would further argue that the existence of conflicting findings such as in these two studies also points to the difficulty of identifying certain types or dimensions of motivation and to how these may relate to the choices and progress adults make in their language learning in a given setting. As we have seen, people’s motivations to learn and speak particular languages are inescapably bound up with the macro-level and micro-level sociohistorical, cultural, and interactional contexts in which they live and with which they engage. It is rather difficult to capture dynamic and evolving motivations if, as Granena (2015) or Hammer and Dewaele (2015) did, we use solely psychometric approaches, which, by necessity, restrict themselves to a predetermined number of motivational dimensions and thus do not allow for richly detailed analyses. Qualitative analyses seem indispensable in the face of such complexity and variability. Mixed-methods research is highly recommended, as a qualitative component can inject an exploratory and feedback strand into an otherwise quantitative study, or it can be used to corroborate quantitatively arrived at results. As Moyer (2013, p. 1) says, “a host of interrelated variables is at play, having to do with learner orientation and experience,” which cries out for “a more nuanced look at what underlies age effects” in adult language learning. Often we simply cannot afford to ignore life beyond the numbers. It is clear, for instance, that measuring the age variable merely in terms of the number of years spent in an L2 environment is an oversimplistic way of proceeding (cf.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

98

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

Mun˜oz & Singleton, 2011, pp. 10–19). The qualitative dimension can enable research to get right down to the individual level, to take note of very personal circumstances, attitudes, and quirks that would not otherwise figure in the reckoning, and to arrive at a “flavor” of learners’ perceptions and reactions that is very often indispensable when it comes to constructing a true-to-life interpretation of the quantitative data collected from bilinguals. This does not mean, of course, that there is no place for more quantitative research approaches to investigating age effects. However, all too often, age researchers have resorted to the use of the general linear model, a family of statistical models (e.g., t-tests, ANOVA, or multiple regression models) that assumes a normal distribution among other features as a default. This unquestioned reliance has led to a particularly vexing problem in the age research, namely, the discretization (or binning) of participants into random age or age-of-onset groups (e.g., grouping 20–29-year-olds in one category, the 30–39-year-olds in another, and the 40–49-year-olds in yet another) and, consequently, a loss of information. Such grouping creates the impression that 20- and 29-year-olds tend to be more alike than 29- and 30-year-olds and 40–49-year-olds, which is entirely due to arbitrary cut-off choices and, according to influential statistician Jacob Cohen (1983), comes close to throwing away a third of the data (see also Vanhove, 2013). Rather than drawing categorical boundaries where none exist and squeezing continuous data so as to perform comparisons using traditional t-tests, ANOVAs, etc., age researchers are better advised to employ generalized additive models (GAMs) for nonlinear data or linear mixed-effects regressing models (LMERs) for linear data, which require no prior aggregation and are run on the unaveraged data (see later in this chapter). The overreliance on t-tests, ANOVA, or multiple regression in age research leads to the neglect of intra-learner, inter-learner, and item variability, forces researchers to discard all results on any subject with even a single missing measurement, and is based on the independence assumption, which means stipulating that multiple responses from the same subject have to be regarded as independent from each other. Instead, we urge researchers to use LMER models because they offer distinct advantages (Pfenninger & Singleton, 2017). They greatly improve generalizability by permitting multiple random factors to be modeled for a random sample and by accounting of both participant and item variability, thus allowing for the simultaneous generalization of the results to new participants and new items. They help researchers assess the impact of context-varying factors in relation to individual differences variables by integrating individual-level and contextual-level data in the same statistical analysis. They enable measurements within and between participants that are crossed at the same level of sampling (e.g., the children from different families in different communities) or nested in a hierarchical

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

Bilingualism in Midlife

99

fashion (e.g., learners within classes within cohorts within schools). Last but not least, they are robust against missing data and imbalanced designs, obviating the need to replace missing values using debatable imputation techniques, which is advantageous in longitudinal studies as well as experimental studies. In sum, LMER models encourage researchers to shift from a myopic focus on a single factor, such as age, to examining multiple relations among a number of variables, including contextual variables.

6 Conclusion In this chapter we first explored some factors that have been offered as explanation for variation in the extent to which childhood bilingualism is maintained into adulthood. We saw that – from the situation of people growing up with Irish and English to that of people growing up with American Sign Language and English – the reasons for the less welldeveloped nature of one or the other language in adulthood often go back, in fact, to the experience and attitudes lived through in childhood. Also of relevance in this regard is the fact that the bilingual’s languages are typically complementary, as Grosjean (1997) proposed, in the sense of being used in different domains and for different purposes, and such that in many cases one language accounts for more domains and purposes than the other. Scientific evidence has mounted steadily against the contention that bilinguals who have been brought up using two languages automatically converge on the grammars of native speakers and automatically experience enduring success with bilingualism. We then went on to address the issue of the success of late bilinguals. We began our treatment with a foray into the assertions of the critical period hypothesis, which pour very cold water on the idea that late bilinguals can achieve extremely high proficiency in the later learned language, and we cited evidence of adult learners whose attainments would seem to at least put a question mark over such a blanket view. We further took some wary steps into neurological research, raising the question of the differences between adults’ and children’s brains, whatever the degree of similarity in their capacity for second language attainment. We reported substantial differences of perspective between researchers on this matter, with the one consensual point appearing to be that some level of cerebral plasticity remains possible throughout life. Finally, in our discussion of methodological challenges, we suggested that not only do researchers need to investigate how specific individual characteristics and environmental variables (e.g., the impact of the learning context or compositional effects within the sample) mitigate the age factor, and what impact these interactions between age and individual characteristics have under different conditions of performance, but that also context needs to be granted appropriate weight in the analyses, and

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

100

DAV I D SI N GL ETO N A N D SI M O N E E. P F EN N I NG E R

that there are appropriate quantitative, qualitative, and mixed methods tools for achieving this goal. In contrast to the myriad studies focusing on bilingualism in early life, investigation of the levels of attainment achieved and maintained by bilinguals in adulthood is limited, and much yet remains to be explored with respect both to the limits of continuing adult bilingualism and to late L2 acquisition. What we do know at this point is that long-term achievement in L2 acquisition is characterized by high inter- and intra-individual variability. These can be explained by the fact that mature adults’ motivations to learn and speak particular languages are inescapably bound up with the macro-level and micro-level sociohistorical, cultural, and interactional contexts in which they live and with which they engage.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 14:59:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.005

5 Language and Older Bilinguals Mira Goral

1 Introduction This chapter addresses aspects of language competence and processing in bilingual older adults. Although there is considerable controversy regarding the point at which processes of aging might begin, the operational definition of older here refers to adults who are 60 years old and older. Bilingual older adults references a variety of speakers, including (1) individuals who are bilingual (including childhood, early, and late bilinguals) and who are becoming older, and (2) older adults who are becoming bilinguals (due to life changes or by choice). Questions concerning language learning and unlearning pertain to both subgroups, but the literature has focused on differing aspects of language change in each of these two types of older bilinguals. Accordingly, the main question addressed here for the former type is how age-related changes might affect one or both of an older bilingual’s languages. For the latter type, the main questions pertain to processes of second language (L2) learning and first language (L1) loss. The notations L1 and L2 are used here when the bilingual speakers under study do not have two first languages but rather have acquired their first language before their second language; when relevant, a distinction between order of learning and hierarchy of proficiency levels is made. Following this brief introduction, this chapter addresses age-related language changes in healthy older bilingual adults, and then focuses on specific bilingual phenomena as manifested in older adulthood, including first language maintenance and loss, language mixing, and language learning, before concluding with a summary and avenues for future directions. Bilingual proficiency is dynamic. As they change their environment, interlocutors, and habits along the lifespan, bilingual speakers rarely maintain the same levels of language proficiency and frequency of use in each of their languages across time. This is most obviously the case for people who speak multiple rather than just two languages, as they increase and decrease their levels of use and proficiency (in this chapter, speakers of multiple languages

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

102

MIRA GORAL

are included in the term bilinguals). The study of changes in bilingual language abilities has focused mainly on bilingualism-related factors, and to a lesser degree on lifespan-related factors. Thus, amid a large body of literature on language skills in healthy as well as in neurologically impaired older monolingual adults, fewer studies have addressed questions concerning age-related changes in language skills in healthy bilingual older adults. Greater attention has been devoted to aspects of changes in language usage, especially in bilinguals who are not childhood bilinguals, as is evident from studies of individuals’ maintenance of proficiency levels in their first language (L1 maintenance), dominance shifts from one language to another, and changes in L1 and L2 proficiency levels across individuals’ lifespans and across generations. Among those studies, a relatively large set has examined the phenomenon of L1 attrition (see Keijzer & de Bot, Chapter 14, this volume, and Ko¨pke, Chapter 18, this volume). Another growing body of studies has focused on age-related cognitive changes, examining these changes in bilingual speakers as compared to monolingual speakers, and exploring the possible cognitive advantages that bilingualism may afford (see Paap, Chapter 22, this volume, for a general discussion). Another subset of studies addresses language maintenance and loss in atypical aging, that is, in bilingual individuals who demonstrate language impairment due to acute-onset brain damage or neurodegenerative brain disease (see Hernandez, Chapter 23, this volume). The present chapter focuses on changes in language competences and processing that have been reported in studies of healthy older bilingual adults. Following Grosjean (1989), I take the view that bilingual speakers may be better understood if studied within the context of bilingualism itself, rather than through comparisons with monolingual speakers; however, studies that compared bilingual and monolingual groups and are relevant to the current discussion are included here. The focus in this chapter is on changes associated with older age, while fully recognizing that many additional changes may result from changes in environment, social context, and personal motivation, among other variables. The majority of the studies reviewed here have focused on English and other European languages, including Spanish, Dutch, and French. Additionally, studies of Chinese– English, Hebrew–English, and Kashmiri–Urdu speakers have been included. Bimodal bilinguals have received little attention in this literature.

2 Age-Related Language Changes in Healthy Older Bilingual Adults Many adults experience a number of changes as they grow old that may affect language and communication. These include decreased hearing and vision acuity, and, at times, decreased opportunities for communication.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

103

The substantial literature on language changes associated with older adulthood has demonstrated, however, that many language abilities remain preserved or even improve across the adult lifespan (e.g., Thornton & Little, 2006). In particular, older adults tend to have larger lexical repertoires and richer semantic knowledge than younger adults. Their storytelling skills tend to be superior, as do aspects of their spoken and written language comprehension. The body of literature addressing such issues does not usually focus on the number of languages that individuals know. Studies focusing more specifically on older bilingual individuals have also found positive agerelated changes, particularly for lexical knowledge. For example, Bialystok and Luk (2012) reported higher comprehension vocabulary scores for bilingual (and monolingual) older participants (mean age = 68.9 for the bilingual older group) than for younger participants (mean age = 23.5 for the bilingual younger group). The authors aggregated data from 20 studies conducted with bilingual speakers of English and another language (including a number of European and Asian languages) and monolingual English speakers, and reported slightly but statistically significant higher scores for the older than the younger participants. Furthermore, comparable performance on word recognition tasks has been reported for older and younger bilingual adults (e.g., Johns, Sheppard, Jones, & Taler, 2016). Johns et al. (2016) used a lexical decision task with 300 words and compared the performance of younger (age range 18–30) and older (65 and older) bilingual English–French and monolingual English-speaking participants living in Canada. They reported no significant differences between the response times of the older and younger bilingual participant groups. However, Johns et al. (2016) did find greater variability in performance among the older group than among the younger participants. This variability is consistent with previous results with older (presumably monolingual) participants (e.g., Hultsch, MacDonald, & Dixon, 2002). Several language skills, however, have been shown to decline with age. This is most notably the case for lexical retrieval, especially in tasks requiring the retrieval of single lexical items (Kave´, Knafo, & Gilboa, 2010; Schmid & Keijzer, 2009). Two explanations have been put forward to account for greater word retrieval difficulty in older age. One associates the difficulty with increased lexical connections and lexical competition, the other with impaired lexical–phonological connections (see later in this chapter). If we assume that lexical production difficulty in older age may be due to competition among lexical items, greater interference may be expected for older than younger adults. This is because the lexical repertoires of older adults, associated with lifelong experiences and learning, tend to be larger than those of younger adults (e.g., Ramscar, Hendrix, Shaoul et al., 2014). Therefore, assuming that bilinguals generally have larger lexicons than monolinguals, greater retrieval deficits might be expected for bilinguals than for monolinguals. Yet studies have found little

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

104

MIRA GORAL

difference between bilingual and monolingual older adults in their performance on language tests, including tests of lexical retrieval, suggesting that knowing many words does not, in fact, make it harder to find them. For example, in their large-scale analysis of performance on a Hebrew picture-naming test across the lifespan, Kave´ et al. (2010) found comparable age-related decline in naming success among immigrant (who were bilinguals and learned Hebrew as an L2) and nonimmigrant (who acquired Hebrew as their L1) participants (although the immigrant group named fewer items than the nonimmigrant one). Similarly, Ashaie and Obler (2014) reported comparable age-related decreased accuracy on a picturenaming test for bilingual Kashmiri–Urdu and monolingual Kashmiri older speakers. Rosselli, Ardila, Araujo et al. (2000) found no differences in sentence repetition and verbal fluency performance among bilingual Spanish–English older adults and age-matched monolingual speakers of Spanish and English, save for lower productivity on semantic fluency tests in both languages for the bilinguals. Congruent with these findings, a study of 708 monolinguals between the ages of 18 and 88 (Shafto, James, Abrams et al., 2017) suggests that age-related increase in word retrieval difficulties is not associated with age-related increase in verbal knowledge. In contrast to the prediction of greater age-related decline in lexical retrieval among bilinguals than monolinguals, if we consider the interplay between language and general cognition, smaller magnitudes of agerelated language change could be expected for bilinguals than for monolingual older adults in certain language domains. It has been suggested that age-related cognitive decline in domains such as processing speed and reduced working memory capacity (note, however, that such changes may begin in early rather than late adulthood) may underlie certain aspects of age-related language decline. It may be the case, however, that changes in general cognitive abilities do not underlie language changes in aging but rather that better general cognitive abilities help compensate for agerelated language difficulties. Compensation could account for individual differences found among older adults. Compensation is also consistent with the fact that age-related lexical retrieval problems reported on single word naming tests are typically not found during connected speech production (Kave´ & Goral, 2017). This suggests that speakers compensate for word retrieval difficulty by selecting alternative words or using the context to facilitate retrieval. The notion that better cognitive skills could ameliorate language difficulties among bilinguals is consistent with assertions that bilinguals may have superior cognitive skills (there is controversy on this point, however; see Paap, Chapter 22, this volume). Additional domains of interest in the study of language in older adults include discourse production and discourse comprehension, as well as changes in the language used by younger adults when addressing older adults (elderspeak). Scholars have conducted little research with bilingual speakers so far to address these aspects.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

105

3 Selective versus Comparable Age-Related Language Changes If we expect certain age-related language changes such as increased difficulty with word retrieval, we might ask whether aging affects the two languages of older bilinguals to the same extent. Reasons to expect different consequences of aging for a bilingual’s two languages include the effects of frequency of language use. Older bilingual adults may maintain frequent use of both their languages, or may find themselves in life situations that call for selective use of just a single language. Furthermore, for bilingual individuals who did not acquire their two languages simultaneously from earliest childhood onward, it is possible that as they grow old they increase or decrease their use of the earlier-acquired L1. Relatively few studies provide an answer to this issue of selective versus comparable age-related changes across the languages of bilingual older adults. One domain where the two languages of older bilinguals have been compared is that of sentence processing. Whitford and Titone (2016) found comparable age-related differences across the two languages of older bilingual participants living in Canada. In a within-subject design, the authors reported on an eye-tracking study comparing French–English sequential bilingual adults (age range 61–87) reading in both their L1 (French) and their L2 (English). Whitford and Titone found an effect of age (reduced reading fluency was associated with older age), an effect of language (there was lower reading fluency in L2), but no age by language interaction, suggesting that whatever changes were associated with age affected both languages. An additional domain that would be of interest in the study of agerelated changes in the languages of bilingual older adults is the lexical domain. As suggested by the relatively intact semantic performance of older adults amid their relatively impaired lexical retrieval, one reason for the word-retrieval difficulties generally characterizing the performance of older adults on single word production tests may be weakened connections between semantic and phonological networks within the mental lexicon. It is likely that lexical frequency plays a role in the strength of connections within the lexical network, and if so, greater retrieval difficulties may be expected in the language used less often. Whereas lifelong experience with a language strengthens connections within the lexical system, reduced experience in one of the languages might result in weakened connections within the lexical system of that language. Future studies would be needed to examine the effect of language use on age-related lexical retrieval changes in older bilinguals. Another reason to expect selective age-related change is rooted in the hypothesis that differing neuropsychological systems are associated with the two languages of bilinguals who learn their L2 in adulthood. If so, and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

106

MIRA GORAL

if one language system is more vulnerable to processes of aging, differential patterns of change may be observed. At least one study that followed bilingual individuals over time has reported relatively stable performance in English (L2) among Dutch (L1) participants living in Australia (Clyne, 2011). Future studies that directly compare performance in the two languages of bilingual older adults will be able to support or refute this prediction.

4 First Language Maintenance and Loss in Older Age One subgroup of bilingual individuals consists of those who live in a largely monolingual environment and therefore have limited opportunities to use one of their two or more languages. The extensive use of one language along with limited exposure to and use of another may result in the progressive decrease of proficiency in the lesser-used language. This is known as language attrition (see Keijzer & de Bot, Chapter 14, this volume, and Ko¨pke, Chapter 18, this volume). A large body of studies has examined language attrition versus language maintenance among bilingual individuals. Evidence suggests that younger individuals who undergo a change in linguistic environment while they are acquiring their L1 experience greater L1 attrition than adults who had acquired most of their L1 prior to a change in linguistic environment (Flores, 2010). Older individuals who speak a language other than the language spoken by the majority in the region where they live, particularly those who immigrated at an older age, have often been found to have greater L1 proficiency and lower L2 proficiency as compared to younger members of the same minority-language community. Greater L1 maintenance has been associated with older age, likely due to L1 use in individuals who are less acculturated to the new environment. For a sample of L1 Dutch immigrants to English-speaking Australia who spoke Dutch proficiently at the first point of data collection and still used it 16 years later, de Bot and Clyne (1994) found little evidence of L1 attrition, although lexical richness in informal conversation had diminished somewhat (some transfer from English was also noticeable). In addition, de Bot and Clyne interviewed a subsample of five nonfluent Dutch speakers in Dutch. These individuals had been studied earlier, in 1971, when four of them were in their mid-40s and one of them was in their mid-60s. Twenty years later, by the time the four were in their mid-60s and the one in their mid-80s, three had improved in English and all had increased their use of two-way switching between languages, as well as their use of transfer from English into Dutch, suggesting change in L1 practices even in older age. A critical question in the study of L1 attrition in older adults is whether processes of L1 attrition can be distinguished from processes of age-related changes (see also Keijzer & de Bot, Chapter 14, this volume). Goral, Libben,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

107

Obler, Jarema et al. (2008) highlighted this challenge and attempted to answer the question. They compared accuracy and response time (RT) on lexical decision tasks in both Hebrew and English by three groups of Hebrew–English bilinguals. All bilinguals had Hebrew as their L1 and English as their L2, which they had learned in an instructional setting. In their L1, both older (age range 55–64) and younger (age range 19–38) Hebrew–English bilinguals who were immersed in their L2 environment (in the United States) showed lower RTs compared to bilinguals (age range 18–40) who were immersed in their L1 environment (in Israel). At the same time, the older US group showed slower RTs in Hebrew than the younger US group. However, the older group was not slower to react to English than either of the younger groups, thus indicating there was no general slowing due to age. These combined findings are consistent with an attrition rather than an aging effect in the word processing by older bilingual immigrants. It is often suggested that L1 versus L2 usage among sequential bilinguals living in their L2 environment corresponds to private versus public domains, such that immigrants tend to speak the L1 at home with their family members and local communities, and the L2 in the wider society, at the workplace, and so forth. This view, however, may be a simplification of the complex patterns of language use among immigrant communities. Schrauf (2009) examined L2 (English) proficiency and use in older Spanish–English bilinguals (mean age = 69 years) who had emigrated from Puerto Rico to the United States in their 20s and lived there in a Puerto Rican neighborhood. Schrauf demonstrated that one main group of interlocutors with whom older immigrants spoke their L2 consisted of their children, thus showing that the L2 was used in both the private (e.g., home) and public (e.g., work) domains. It may be the case that focusing on the interlocutors with whom each language is spoken rather than on the domains where the languages are spoken will offer a better understanding of language choice in immigrant communities generally and in older adults, particularly (see De Houwer, Chapter 17, this volume, on language choice in bilinguals in general). In studies of patterns of language choice in immigrant communities, age is typically found to be a powerful predictor of language use, with older participants and participants who were older at the time of immigration using more L1 and less L2 than people who are younger and typically more acculturated (see Gonc¸alves, Chapter 3, this volume, for a different perspective on language choice and use by young adult immigrants in both public and private domains). However, it has been hypothesized that whereas immigrants typically increase their use of L2 and decrease their use of L1 in early adulthood, as they grow older, they may reverse this trend and may increase their use of the L1. This is known as the language reversion hypothesis (Clyne, 1977). Schmid and Keijzer (2009) studied German L1 and Dutch L1 speakers living in English-speaking parts of Canada as well as German L1 speakers living in the Netherlands, where

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

108

MIRA GORAL

the main language is Dutch; these were compared to German L1 speakers in Germany and Dutch L1 speakers in the Netherlands. The 249 participants were divided into five age groups, with the youngest group being under 57 years of age and the oldest one aged 72. Among others, Schmid and Keijzer (2009) examined whether there was an effect of age in L1 attrition and whether processes of L1 attrition slow down (or disappear) in older age. Groups were compared on measures of lexical fluency and lexical retrieval in free speech. Participants living in an L2 environment performed less well in their L1 than participants living in an L1 environment. This difference is consistent with language attrition. However, the difference was smallest among the oldest participants, leading Schmid and Keijzer (2009) to suggest that language reversion might have been at play. However, they emphasize that many other factors have to be considered to fully explain the patterns they found. One such factor, language usage, may play an important role in determining degree of L1 maintenance not only in immigrant but also in nonimmigrant older adults. For example, Mu¨ller (2009) examined reported and observed language use by French–English bilingual older adults in Louisiana, United States. The usual language in public life is English, but several varieties of French have been used in mostly the private domain ever since Louisiana was a French colony in the 18th century. In her yearlong participant observation study, Mu¨ller held natural conversations with 13 older adults (age range mid-60s to late 90s) with French as an L1 and English as an L2 learned through school. Mu¨ller spoke English to them, but showed an interest in the participants’ knowledge and use of French. Most participants did not initiate switching to French during those conversations. Mu¨ller did observe them speaking French with selected other residents and in certain ritual-based settings such as prayer meetings. All staff communication with residents was in English. This pattern of language use was consistent with that of the larger community in Louisiana that has witnessed a decline in the use of any French variety, representing language shift, with fewer individuals acquiring French as their L1. Future longitudinal studies assessing the two languages of older bilinguals living in primarily monolingual versus primarily bilingual communities should shed light on the interaction among age, environment, and language maintenance (de Bot & Makoni, 2005; Keijzer & Schmid, 2016).

5 Language Mixing in Older Bilinguals A hallmark phenomenon in bilingual speakers is language mixing. That is, bilingual speakers who communicate with bilingual interlocutors often use both of their languages, mixing them within conversations and at times within utterances (see further De Houwer, Chapter 17, this volume).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

109

Language mixing may occur for a number of psycholinguistic and sociolinguistic reasons, can be related or unrelated to language proficiency, and has been a topic of much research and debate (see further Aalberse & Muysken, Chapter 26, this volume). Given changes in language and general cognition (including monitoring and control) associated with older age, it may be reasonable to expect age-related changes in language-mixing behavior among older bilingual adults. Yet there is little evidence to suggest that healthy older adults are less able than younger bilinguals to voluntarily control language mixing (but see Ardila & Ramos, 2008). Gollan, Sandoval, and Salmon (2011) compared the performance of older (mean age: 77 years) and younger (mean age: 18 years) bilingual speakers of Spanish (L1) and English (L2) living in the United States on a verbal fluency task requiring participants to produce as many words in a given semantic category as they could in 60 seconds. Performance was elicited in each language separately. There were no differences between the groups in the frequency of intrusion errors (i.e., the production of a word in the nontarget language). Later, Gollan and Goldrick (2016) also found comparable frequencies of intrusion errors by older and younger Spanish–English participants who were matched on the ability to name pictures in both languages. Participants were asked to read aloud texts containing language switches. Although older participants read more slowly than the younger ones, all produced more intrusion errors when they read paragraphs that included less typical language switches (e.g., a noun phrase containing a noun in language A and a preposition and a noun in language B) than those containing typical switches (e.g., a noun phrase containing a noun and a preposition in language A and a second noun in language B). This was the case especially when paragraphs contained more, rather than fewer, switches. To date, inappropriate language choice, including the occurrence of intrusion errors in conversation, seems not to be associated with normal aging in bilinguals; rather, it has been associated primarily with atypical aging, especially with neurodegenerative disease (Friedland & Miller, 1999). Another line of research that has investigated language mixing concerns the processing costs associated with mixed versus unilingual conditions. It has been suggested that bilinguals may be slower when they have to switch between their languages as compared to when they produce, or are presented with, just a single language. Previous findings have reported slowed processing (i.e., a switch cost) in nonvoluntary switching tasks, that is, when participants are instructed to switch language according to a specific cue. The switch cost is typically reflected in longer RTs for “switch” items (items cued to be named in language A following an item named in language B) than for “stay” items (items cued to be named in the same language as the previous item). The switch cost has been a robust finding in the literature on bilinguals, with some studies reporting an asymmetric cost. Switch cost asymmetry has been associated with language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

110

MIRA GORAL

dominance, as, typically, switching to the stronger L1 is associated with longer RTs than switching to the less-proficient L2. This asymmetry has been interpreted as reflective of the greater inhibition of a stronger language required for naming in a weaker language. Research comparing younger to older bilinguals has found a larger switch cost for the older group (e.g., Hernandez & Kohnert, 2013). However, Weissberger, Wierenga, Bondi, and Gollan (2012) – who likewise demonstrated an age-related increase in switch cost as manifested in both increased RTs and decreased accuracy for older Spanish–English bilinguals compared to younger Spanish–English bilinguals – also reported that the switch cost among the older group appeared greater for nonlinguistic (visual) as compared to linguistic tasks. The authors took their results to suggest partially preserved language-switching abilities in older age. Furthermore, when participants are instructed to switch if and when they wish, they do, and older adults do not show the same increased switch cost (in terms of RT or accuracy) that they show when the switching is cued (Gollan & Ferreira, 2009). Such results are consistent with findings that bilinguals – younger as well as older – are more successful in resolving tipof-the-tongue states (the phenomenon that you know something that you cannot momentarily recall) and in other naming tasks when they are allowed to use either language, as compared to when they are instructed to use only one of their languages. Another form of language mixing has been examined by eliciting autobiographical memories from older bilingual immigrants in each of their languages when memories are cued based on a word in the L1 or L2. Such studies track the age of the memories to see whether they occurred prior to or after immigration, and to discover the likelihood of a specific pattern of language choice and of language mixing while participants recall the memory. Generally, older individuals often recall many memories from between the ages of 10 and 30. This is known as the reminiscence bump. In addition, immigrants tend to recall more memories from the five years before and after the time of immigration (immigration bump). Bilinguals tend to mix their languages more when the language of recall differs from the language in which the events recalled took place (e.g., Altman, Schrauf, & Walters, 2013). In their study of 12 English–Hebrew bilingual adults aged 64 to 79 who immigrated to Israel, Altman et al. (2013) demonstrated the existence of a clear reminiscent bump. For bilinguals who had emigrated at a later age (age 65 and up), there was also evidence of an immigration bump. There may also be an interaction between the age of immigration and the age of the reminiscence bump. Such results point to the role of experience and context in language choice. Whether the contribution of such variables changes along the adult lifespan is a question for future research. Bilinguals may employ language mixing as a strategy when they cannot find the word in the target language. Some have interpreted this kind of language mixing as evidence for language attrition. If they are related,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

111

then one would expect other clear signs of attrition in the same population. In his study of English L1 teachers of English as a second language living in Spain, Porte (2003) found increased lexical mixing but little evidence for retrieval difficulty. Porte’s findings are consistent with Clyne (2011), who has argued that language-mixing behavior among individuals who are immersed in their L2 environment does not necessarily indicate L1 attrition.

6 Language Learning in Older Immigrants Learning a new language in older age can take place due to a change in language environment, such as is often the case with immigration. People who emigrate from their home country to another country where a language other than their L1 is spoken experience varying degrees of success in learning the language of their new environment. In those instances, the extent to which the L2 is learned depends on a number of variables, including the individual’s age and sociolinguistic context. It has been demonstrated that people who live in communities that consist mostly of speakers of the majority language tend to achieve higher levels of proficiency in their L2 (the majority language) than individuals who live in communities that are densely populated with other speakers of the same L1. Learning of an L2 can be additive, that is, the learning of the new language can occur in addition to the use and maintenance of the L1. In many contexts, however, increasing L2 proficiency co-occurs with increasing acculturation into the mainstream of the new society and decreasing L1 proficiency. Whether L2 learning will be additive is often dependent on opportunities to speak both languages as well as additional variables, such as attitudes by the newcomers toward the L1 and the culture associated with it, and attitudes by the hosting society toward immigration and individual differences. Such attitudes may, for instance, help determine the language learning opportunities offered to newcomers (see, e.g., Simpson, Chapter 13, this volume). Limited L2 acquisition among older immigrant groups can have vast implications. For example, when only limited L2 proficiency has been achieved, a number of barriers for access to services have been identified. This is intuitive, and has been documented in a few research studies. In one study, Chinese speakers with limited English proficiency living in Toronto, an English-speaking city in Canada, were recruited to examine how older people (age range not specified) make decisions about health care and medicine use. The data revealed several challenges associated with using interpreters, such as the interpreters reframing or modifying the message, conveying only part of a complex message (e.g., conveying only one of several questions asked at once), and summarizing answers provided by the participant (Ballantyne, Yang, & Boon, 2013). In addition to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

112

MIRA GORAL

language proficiency, additional barriers to health access have been identified, including cultural differences, trust in the medical profession, and discrimination (e.g., Krieger, 2014).

7 Language Learning in Older (Nonimmigrant) Adults New language learning in older age can also occur without a change in the language environment. The literature on healthy older adults suggests that people continue to learn words and expand their vocabularies as they age. Moreover, studies that have examined processes of learning in older age have found that learning is certainly possible in older participants (sometimes as well as younger comparison groups), suggesting that neuroplasticity continues into older adulthood. In recent years, researchers have examined language learning in older age, both in group studies (e.g., Cox & Sanz, 2015) and in case reports (e.g., Cohen & Li, 2013), demonstrating that older adults are capable of effectively learning a new language. In several studies that compared older to younger learners, no group differences were found. For example, bilingual and monolingual older adults successfully learned Latin both with and without explicit instruction (Cox, 2017). Cox (2017) examined 45 adults who were 60 years old or older (23 Spanish–English bilingual and 22 monolingual English speakers living in the United States). Both participant groups learned new morphosyntactic rules (thematic roles in Latin) and appeared to maintain their knowledge when tested two weeks later. Similarly, it has been shown that older adults do not differ from younger adults in single word learning (Whiting, Chenery, & Copland, 2011). Other studies have confirmed successful L2 learning in older age but have pointed to subtle differences between younger and older learners. For example, Schwab, Schuler, Stillman et al. (2016) set out to examine statistical learning of grammatical categories among older individuals. In two experiments comparing younger (age range 19–24) and older (> 65) English speakers living in the United States, the authors demonstrated that the older participants engaged in similar statistical learning and performed similarly to the younger participants following the training period in which an artificial grammar was introduced. However, Schwab et al. (2016) found that the older adults demonstrated smaller differences between their ratings of grammatical and ungrammatical strings, a finding that the authors suggested indicated a less efficient learning mechanism. Marcotte and Ansaldo (2014) studied younger and older French-speaking adults living in bilingual French–English-speaking Montreal in Canada, who were learning Spanish. The authors compared 10 younger (mean age: 22 years) and 10 older (mean age: 70 years) participants who engaged in learning two types of Spanish words: cognates and non-cognates of French. They found that older participants were less accurate than younger ones in the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

113

early training phases, but had learned just as well by the end of the learning phase, reaching similar levels of accuracy and RTs post training. Yet the older participants took longer to reach criterion in the training phases. Furthermore, neuroimaging findings for naming in pre- and post-training showed significant differences in neural activation levels found for the younger versus the older adults, with no interaction between age and word type. The authors also reported distinct neural patterns for the two age groups. These may reflect different strategies used for learning. It may be the case that older adults benefit less than younger adults from explicit instructions (e.g., Lenet, Sanz, Lado et al., 2011). They may also face particular difficulty with learning the phonology of a new language, at least when instructed with methods that have been found useful for younger learners (Ingvalson, Nowicki, Zong, & Wong, 2017). Age-related differences in language learning, like age-related changes in other language abilities, may be due to decline in general cognitive abilities. Supporting evidence points to an interaction between language learning and general cognitive domains, such as working memory, as suggested by Mackey and Sachs (2012). They found that among nine older speakers of Spanish who were learning English in the United States, participants with better working memory capacity learned to form English questions better than those with lower working memory spans. Following the suggestion that lifelong bilingualism may contribute to cognitive reserve (although the evidence has been highly controversial, see Paap, Chapter 22, this volume), Antoniou, Gunasekera, and Wong (2013) proposed that perhaps not only lifelong bilingualism but also language learning in older age could contribute to preserved cognitive processing. Antoniou et al. (2013) argue that given the evidence for positive effects of various kinds of cognitive training on cognitive performance of older adults, and given that the neuronal substrates associated with language learning largely overlap with those that have been shown to decline with aging, there is merit in predicting that teaching older individuals a second language could have positive effects on cognitive functioning in older age. The literature on purported enhanced cognitive abilities in bilingual versus monolingual individuals has yet to demonstrate whether it is the lifelong use of more than one language or the learning of another language that may lead to superior cognitive abilities. Data are still needed, but recent results suggest no cognitive advantages for older adults who have been learning a new L2 (Ramos, Ferna´ndez Garcı´a, Anto´n et al., 2017).

8 Conclusion This chapter reviewed changes in language skills in typically aging individuals, focusing on two broad populations: bilingual adults who become older, and older individuals who become bilinguals.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

114

MIRA GORAL

The research literature has so far given relatively little attention to the comparison of age-related changes in each of the languages of aging bilinguals who maintain both their languages into older adulthood. The evidence available points to largely comparable, subtle difficulties in both languages, although there is preliminary evidence that the L1 may be better preserved than the L2, at least in a subset of bilinguals. Among immigrants living in their L2 environment, processes of L1 attrition may be reversed in older age. Whether both languages show evidence of the typical subtle changes that are seen in old age in monolinguals or whether one language shows greater resilience may depend on the subtype of bilingual populations examined and on factors such as language use. Studies of L1 maintenance, L1 attrition, and L2 learning in older age suggest that L1 language abilities are largely stable in older adulthood, maybe more so than among younger generations of immigrant communities. In contrast, L2 learning may be less efficient at older than younger ages, although studies have provided evidence that older people do continue to learn into old age and are capable of learning a second language. The studies of older adults who are bilingual or are becoming bilingual highlight the importance of those variables that have been found relevant in the study of bilingual speakers more generally, such as the extent of language usage and language exposure. In addition, the role of general cognitive abilities, which have been documented to change in older age, merits attention. With regards to all of these variables, it is critical to address both internal (e.g., attitudes toward each language and the motivation to use it, reading habits in both languages) and external (e.g., location of residence, social networks) contributors to language proficiency in order to better understand its dynamic nature. One challenge that is shared by both the study of aging and the study of bilingualism is great inter-individual variability. Inter-individual variability has been demonstrated to be greater among older as compared to younger participant groups, and for bilinguals as compared to monolinguals. One avenue to reduce the effects of this variability is the approach of studying within-group rather than between-group processes. That is, studying language behaviors within older adults, including longitudinal studies, may be preferable to comparing older adults to younger ones. Similarly, studying bilingual individuals and the variables that contribute to their performance may be preferable to comparing bilingual speakers to monolingual groups. The variability challenge is compounded by the large set of factors that affect the dynamic processes of learning, using, and maintaining more than one language. For example, among immigrant groups, adults who are older are also those who have lived longer in their L1 environment, may be less acculturated in their L2 environment, and may have fewer opportunities to use their L2. Thus, older adults’ lower L2 proficiency and higher L1 proficiency, compared to younger

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Language and Older Bilinguals

115

individuals in their community, can be related to any and all of these variables. Furthermore, it may be difficult to disentangle the contribution of age-related and attrition-related changes in older bilinguals. It is thus of interest to compare language performance in both of a bilingual’s languages directly, a comparison, however, that has received little attention in the literature to date. Such comparisons introduce their own challenges, including the need for comparable assessment measures and the need to control the myriad influencing variables. The phenomena associated with language mixing in bilingual individuals have posed a great challenge to the study of bilinguals. There is relatively little consensus on what might constitute atypical languagemixing behavior. It is unclear that increased mixing can be associated with decreased proficiency, as many highly proficient bilinguals mix their languages regularly. Thus, it is difficult to ascertain whether language-mixing behaviors among older adults reflect age-related decline. Future studies can further mine the wealth of mixing patterns during spontaneous conversations among bilingual speakers to achieve greater understanding of these phenomena. In recent years, an increasing number of investigations have focused on the potential of older adults to learn generally, and to learn a second language specifically. The hypothesized neuroplasticity into older age is consistent with those approaches that question the biological base for the so-called critical period in second language learning (see, e.g., Biedron´ & Birdsong, Chapter 16, and Singleton & Pfenninger, Chapter 4, this volume). The older brain’s capacity for learning suggests, for example, that the reason why older immigrants do not achieve the same levels of L2 proficiency as do younger generations may be associated less with age-related compromised competences and more with sociolinguistic factors. The potential benefits of learning a new language in older age may represent a valuable avenue for maintaining stable linguistic competences along the adult lifespan. Additional studies examining the efficiency of foreign and second language learning in older adults can add to our understanding of language abilities in older age. Overall, it can be concluded that the language abilities of bilingual speakers remain largely stable into old age, although stable does not mean static, as dynamic changes in competence are inherent to bilingualism, and language learning can continue into old age.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:46:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.006

6 Ideologies of Language, Bilingualism, and Monolingualism Janet M. Fuller

1 Introduction Language ideologies are ideas about language structure and use that index political and economic interests of individuals and the social groups and nations to which they belong. Because ideologies are relevant in political realms, they may be reflected in explicit policies and laws regarding language status in a society (see Lo Bianco, Chapter 8, this volume, for further discussion of bilingualism and language policy). The study of these ideologies necessarily involves looking at linguistic inequality and the concept of hegemony, which can be defined as a form of dominance in which power relations are depicted as irrefutable truths, leading to what is called “internalization” (Woolard, 1998). That is, even those disadvantaged by hegemonic ideologies accept them as immutable facts, not social realities that can be changed or discontinued. Hegemony and internalization may make explicit policies unnecessary; for instance, in the United States, there is no official language, but the ideological link between English and national identity is strong enough to make policy unnecessary for ensuring that English is used in government and educational contexts. There is, of course, debate and conflict over the use of other languages in addition to English in such contexts, and this illustrates that even when there is widespread internalization of an ideology, it is never complete. However, competing ideologies must make reference to hegemonic ideologies, so even in resistance to hegemony, the competing ideologies also serve to reify the power asymmetry. Despite the inescapable presence of dominant ideologies, different ideologies may be evoked in different social contexts. For example, in an analysis of the Zakarapattia region in southwest Ukraine, Dickinson (2010) discusses how ideologies about language are often about stereotypes of social groups and the languages they speak, but also about the practical

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

120

JANET M. FULLER

aspects of life in a multilingual community. In some cases, there is a onenation-one-language ideology, which positions the national language (Ukrainian) as a marker of civic participation, and speaking other languages as unwillingness to integrate. In other contexts, however, the right of people to speak “their own” language (for example, an immigrant language) is paramount, and languages other than Ukrainian are accommodated. In yet other contexts, what Dickinson (2010) calls a “market language ideology” prevails, in which the goal is communication and speakers use whatever linguistic resources they have – sometimes both speaking a lingua franca that is the native language of neither speaker, or each speaking the language in which they are most proficient, if it is understood by the other. Thus, there is situational variation in how ideologies are made manifest. Contemporary social theory holds that the social world, including ideologies as well as identities of individuals, are not fixed but are produced and reproduced through Discourse. The term Discourse (with a capital D) is used to describe social behaviors, including but not limited to language, that construct social categories and their values (see further discussion and sources in Fuller, 2012). Throughout this chapter, I use the term Discourse in this way, to refer to language and cultural performances that are part of the reproduction of ideologies. The focus in this chapter is research on particular ideologies and how they affect bi/multilingualism. After a discussion of key terms in Section 2, I address these ideologies in three broad categories: monoglossic ideologies (Section 3), standard language ideologies (Section 4), and pluralist ideologies (Sections 5 and 6). While the first two categories reflect ideas about the dominance of a single language variety, the very necessity of such claims illustrates that bilingualism and bidialectalism abound in these contexts – if there were no challenges to the single standard code, there would be no need to champion monolingualism. As the discussion of the research shows, monoglossic ideologies are evoked to erase the reality of bilingualism. They are, however, usually challenged with pluralist ideologies that celebrate bi/multilingualism. All of these different types of ideologies are then revisited in Section 7, which focuses on English as a global language. Section 8 is a discussion of methodologies used to study language ideologies and a forecast of future directions for the study of language ideologies. A conclusion that summarizes the research themes and future directions completes this chapter.

2 Key Concepts in the Study of Language Ideology Underlying the study of ideologies are terms for and ideas about different ways of speaking. The term variety is used to refer to a way of speaking that has identifiable linguistic features, but could be considered a language,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

121

dialect, style, or register. All of these categories, but most saliently the categories of “language” and “dialect,” are ideological constructs. This specifically refers to the widespread belief, also held by many linguists, that languages are internally homogenous, bounded systems (see also Garcı´a & Tupas, Chapter 20, this volume). The conceptualization of distinct, “pure” languages is widespread and firmly rooted, but many sociolinguists argue that speakers do not speak languages per se but instead use semiotic resources that are associated with particular languages, which in turn are associated with particular social groups (often, but not exclusively, national groups). Nonetheless, most speakers have strong feelings about languages as immutable and dialects as subordinate, and such ideologies are addressed in this chapter. In discussing monoglossic and pluralist ideologies, I refer to varieties that are considered separate languages; in the discussion of standard language ideology and Global English, I discuss varieties that are considered dialects of the same language. (See also Chevrot & Ghimenton, Chapter 25, this volume, for a discussion of contact between two varieties that are considered dialects of the same language.) In this chapter on ideology, the constructs of language varieties and their social statuses are central. I use the terms dominant language and majority language to refer to the language that has the highest status in a community, usually because it is recognized as the national or regional language, whether it is deemed the official language in terms of policy or not. This language is used in government services and education, perhaps not as the only language in those contexts, but as the one that is accepted as the norm. A minority language is a language that does not have this status, and is spoken in more limited domains and takes its value as part of ethnolinguistic background. A final term that requires definition is essentialism. In the study of language ideologies, this term refers to the idea that people tend to see social groups as made up of people who share certain traits that are immutable and essential to their nature. Thus social categories (such as national or racial groups) are viewed as static and homogenous. Language ideologies may include essentialist connections between language and social categories, that is, ideas that speakers of certain languages must belong to certain social groups and thus share characteristics with other speakers of that language, characteristics that make them fundamentally different from other linguistic groups. As a specialized domain of research, the study of language ideologies draws primarily from linguistic anthropology but also critical studies more broadly. Research on this topic did not begin to proliferate until the 1980s and 1990s, leading up to the seminal article by Gal and Irvine (1995), who introduced a framework for the study of language ideologies. There is some overlap between language ideology and language attitude research in sociolinguistics. Language attitude research looks at ideas about language held by speakers; language ideology research focuses on

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

122

JANET M. FULLER

the closely related societal Discourses and how they are circulated. Language attitudes and ideologies clearly interact and influence each other, and the lines between them may become blurred. However, there are also methodological differences, with language ideology research focusing more exclusively on discourse analytical methods and traditional language attitude research employing methods that seek to elicit speakers’ views, often via surveys. An important central concept in the study of ideologies is the sociolinguistic hierarchy. Linguistic varieties in a society are not all considered socially equal, but are hierarchically ranked in terms of their social value. Distinct languages as well as varieties considered dialects of the same language are hierarchically ordered; these parallels are addressed further in the discussion of monoglossic and pluralist ideologies. Gal and Irvine (1995) outline the concepts of iconicity, erasure, and recursivity. The relationship between language and social groups is seen as iconic: language does not just index a social group but is perceived as sharing features with it. For example, in Dickinson’s (2010) aforementioned research in the Zakarpattia region, she notes that although this region was recognized as part of the Ukraine, it was also viewed as a wild and foreign “other.” Descriptions of the dialect of this region by outsiders always included reference to indecipherable words for everyday objects, and in this way the “otherness” of the language depicted the “otherness” of the people. Erasure occurs when certain information about languages is ignored in order to support a particular ideology. Essentialist ideas about the connection between languages and social groups require erasure of the many exceptions to such simplistic correspondences – for instance, essentialist ideas about “sounding black” erase the reality of the many African Americans who do not speak what is considered African American English, and people other than African Americans who do. Finally, relationships between languages are recursive – so, for example, the hierarchical relationship between English and Spanish (and other indigenous and immigrant languages in the United States) is replicated between dialects, with the Standard – that is, some idealized standard variety of English and/or Spanish – seen as superior to regional, ethnic, or social class varieties of these two languages. In addition to these three semiotic processes, the perception of languages as straightforward indices of social identity categories is a recurring theme in the study of language ideologies. Such essentialist, generalizing viewpoints are present when people who are assumed to form part of a particular group (for example, “women,” “Mexicans”) are automatically assumed to share a particular “essence” that makes them different from people who are considered to belong to other groups. This “essence” can refer to any kind of character or behavior. The study of language ideologies includes the study of the fact that often, particular ways of speaking are assumed to indicate a particular group membership.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

123

As discussed in the following sections, scholars have investigated specific ideologies through various methods and in a wide variety of cultural contexts, reflecting great diversity in patterns of language use and social norms that govern them, but also uncovering recurrent themes in ideologies about language that are discussed in the following section as monoglossic, standard, and pluralist ideologies.

3 Monoglossic Ideologies A number of related ideologies value monolingualism over multilingualism for both individuals and societies. These ideologies overvalue the dominant language and devalue other languages and their speakers. As such they have deleterious consequences for societies and people: they shrink the odds that a minority or an indigenous language will be maintained intergenerationally, and they contribute to the oppression and subordination of multilinguals as less legitimate language users and less worthy citizens in modern nation-states. In my research (Fuller, 2012, 2013), I have discussed what I refer to as “normative monolingualism,” which contains two parts: first, the “natural” state for an individual or a political state is assumed to be monolingual; much erasure is involved in this position, as countries with multiple official languages abound. Second, if more than one language is spoken, the languages must remain strictly separate. Normative monolingualism thus incorporates the one-nation-one-language ideology as well as language purity ideologies (Blommaert, Leppa¨nen, Pahta, & Ra¨isa¨nen, 2012). It is also related to the “mother tongue ideology,” which is the idea that individuals have just one mother tongue, and ideally this is the language that is associated with their nationality (Weber & Horner, 2012, pp. 18–20). This ideology may be reflected in policies and regulations about what language is used for government, business, and education. But whether supported in policy or not, the mother tongue ideology privileges the experiences of monolinguals and erases or degrades the experiences of bilinguals and those whose primary language is not the national language of their home country. The United States is a classic example of a country showing the hegemonic ideology of normative monolingualism. Speaking languages other than English is sometimes seen as “un-American” (Fuller, 2013). Subtirelu (2013) analyzes transcripts from US congressional debates and proceedings and finds that those both for and against making multilingual voting materials available to linguistic minorities drew on the ideology of English proficiency as an eligibility criterion for membership in US society. This ideology is hardly unique to the United States, however. Zentz (2014) discusses how Indonesian language policies encourage citizens to love their local languages but use the national language. This erases the reality

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

124

JANET M. FULLER

of language acquisition, as the effect of this policy is ultimately language shift, as no matter how much they are loved, languages that are not used are not passed on to the next generation. King (2000) discusses ideologies about Spanish and Quichua as the majority and minority languages in the southern Ecuadorian Andes (with parallels in neighboring communities in Peru). Her research shows that alongside the value of Quichua for in-group contexts is the ideology that bilingualism is onerous, and that there is survival of the fittest in language contact, the implicit message being that language shift is inevitable. Again, despite love for the minority language, she found that the language used most in the home was the majority language, Spanish, and that the ideologies surrounding Quichua did not bode well for its maintenance. Often, the value of the dominant language means not only the devaluing of other languages but, more directly, their speakers. While such views of minority-spoken languages and nonstandard dialects are commonplace (see, for example, Rubdy, 2001, for Singapore English), the status of signed languages is viewed as even lower (see Tang & Sze, Chapter 24, this volume, on sign languages). There is an even stronger tendency for signed languages to be seen as not linguistic systems in their own right but as compensation for “real” languages, as Cooper and Nguye˜ˆ n (2015) note in their study of Vietnamese Deaf education. Krausneker (2015) also discusses ideologies about Austrian Sign Language that include the idea that it is not really a language, and also that signed languages ghettoize their users. This latter issue of the role of language in integration into mainstream society, although it has different aspects for signed and spoken languages, is present in discussions of both. A common view is that mastering the dominant language is necessary for participation in society, and that low proficiency in this language is indicative of low motivation to integrate and belong to the mainstream. Krausneker’s (2015) study shows the attitude that it is a hearing world and thus spoken languages are superior, and that cochlear implants enable integration into this superior world for the Deaf, who should want to belong to the hearing world and not isolate themselves by using sign language. Although I go on to discuss parallels between majority languages and the hearing world in terms of attitudes about superiority, note that discrimination against the Deaf is much more complicated, and involves decisions made by hearing parents of Deaf children who are faced with having to learn to sign to communicate with their children. The ideology of superiority of certain languages is, however, ubiquitous. Horner (2015) discusses European Discourses about language and citizenship, with a focus on policies regarding language requirements and testing. Practices of requiring immigrants to achieve a particular level of proficiency in the language rely on the idea of the authority of the national language as well as on views of language as the key to integration (see Simpson, Chapter 13, this volume). Critical perspectives on these

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

125

Discourses counter that learning the national language is not a guarantee for inclusion. In many cases, how people speak the language may be used to perpetuate inequalities (see Lippi-Green, 2012, for a discussion of discrimination of those with ethnic, regional, and foreign accents in English). Further, ideologies about proficiency as something that can be objectively tested abound in these Discourses, but scholars have conducted research that questions the validity of language testing for citizenship (Horner, 2015). While much focus within monoglossic ideologies lies on the societal level, there are of course also implications for language use by individuals and in smaller communities. This has been documented, for example, at the level of education, thus directly affecting children in a key domain of life. Cummins (2007) discusses the practice in all arenas of multilingual education – foreign-language classrooms, bilingual classrooms, and dominant-language schools with multilingual students – of dictating that languages should be kept strictly separate, prohibiting mixing of the languages in utterances or conversation, and in many cases even the use of translation (see also Garcı´a & Tupas, Chapter 20, this volume). All instruction by the teacher and participation by the students should be in the target language, although this may be a language that the students are just beginning to learn. Cummins (2007) notes there is no research support for the position that this is a more effective means to teach a language. Certainly, the message that this restriction on bilingual discourse sends is very clear: languages should remain strictly separate, regardless of the consequences for communication and learning. Further, it positions languages as objects of study as opposed to means of communication, and insists on what many bilinguals and multilinguals would maintain is an unnatural and inauthentic way of speaking. Monoglossic ideologies in educational contexts may restrict the acquisition of minority languages, in particular for younger speakers. But what of the influence of language ideologies on the language use of adults? Research on language policies in the workplace indicates that this is an additional arena in which ideas about the separation of language – as well as the nature of bilingualism – play a role. Zentella discusses the phenomenon of being “hired for speaking Spanish, and then fired for speaking Spanish” (2014, p. 623), and notes that some monolinguals see the use of languages other than English (and by the process of iconicity, bilinguals themselves) as a threat to workplace unity and efficiency, as opposed to a communicative resource or part of individual rights. As also discussed in Fuller (2013), in many court cases addressing language use in the workplace, decisions requiring the use of English clearly rely on an understanding of bilingualism that sees two languages as interchangeable; there is no reason to speak Spanish, Tagalog, or Haitian Creole if you can speak English. (See also Angermeyer, Chapter 7, this volume, for discussion of ideologies of bilingualism that are relevant in legal contexts.)

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

126

JANET M. FULLER

4 Standard Language Ideologies The standard language ideology holds that a particular variety of a given language is superior to other related varieties, and that there are contexts in which this variety is deemed to be the most, or even only, appropriate way to speak – or, more importantly, write. Inherent in this ideology is the belief that languages are internally homogenous, bounded systems. There are clear connections between monoglossic ideologies and standard language ideologies; as mentioned in Section 2, the concept of recursion refers to the fact that ideologies are repeated on different levels of language, so that ideologies about bilingualism may recur in the Discourses about multiple dialects. The idea that nonstandard dialects are linguistically inferior to the standard is, however, even stronger than the idea that dominant languages are somehow more linguistically legitimate. Minority languages may be recognized as legitimate languages for formal, public spheres in other contexts (for example, immigrant languages may be the national language of another country), while nonstandard dialects rarely have such a source of authority. In some contexts, the nonstandard variety is not a regional dialect but a local variety of a former colonial language. An example of this is so-called Singlish (Singapore English). Work on the Speak Good English movement in Singapore (Rubdy, 2001) shows how this campaign seeks to replace Singlish with a global variety of English, linking “good” English with greater economic productivity and efficiency. Overtly, the argument for standard English is about intelligibility and pragmatic considerations of participation in the global economy. Wee (2014) shows resistance to this ideology with the Speak Good Singlish movement, in which speakers claim both social solidarity with and linguistic legitimacy of Singlish. Underlying such campaigns to use forms that represent global norms is the theme of language purity and the age-old ideology that there is some “original” form of a language that is inherently better. This is a widespread ideology, noted across languages and continents; for example, in addition to Singapore we see the same Discourse about standard Spanish in the United States (Leeman & Martinez, 2007) and Tamil varieties in Sri Lanka (Davis, 2012). In some cases, the criticism of nonstandard varieties includes the idea that it is not just the language that is deficient but the speakers themselves are lower class, uneducated, of limited intelligence, and so on, in a clear case of iconicity. Lippi-Green (2012) also discusses another aspect of the standard language ideology with regard to Black English in the United States: that speaking Black English (often called African American Vernacular English, or AAVE, by sociolinguists) is a rejection of mainstream culture. Like speaking a minority language, speaking a nonstandard dialect may be seen as a choice (as stated earlier, an act of defiance), and a choice that

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

127

distances the speaker from the cultural capital of the standard and restricts upward mobility. This choice may be seen as insulting to speakers who only speak the dominant language or standard dialect, because it questions the objective, universal desirability of membership in the social group associated with the standard; it also makes it easier for them to construct language minority speakers as irrational and backward, as they have chosen their inferior position in society. While this is especially true of ethnic varieties, such as African American English, Discourses about regional varieties, such as Appalachian English in the United States, may also include this ideology. The importance of standard language ideologies in education is a critical aspect of research on this topic. Davis (2012) discusses how the standard language ideology works for varieties of Tamil at a Tamil-medium school in Sri Lanka. Although Jaffna Tamil has been at the top of the hierarchy due to its status as the literary language, other varieties were acknowledged as appropriate for the home or community (in speech as well as some written contexts such as plays, magazine articles, or radio or movie scripts). These non-Jaffna varieties of Tamil were often linked to regional or social groups, and were associated with Up-country or Muslim speakers. Thus, the evaluations of their linguistic features were inextricably intertwined with the value of these identity categories. Davis describes this as a diglossia ideology. In diglossia, the “high” and “low” varieties of a language are said to be used in separate social contexts that do not overlap, the “high” variety used in formal and institutional settings, and the “low” variety in private and in-group domains (see discussion of Ferguson and Fishman’s classic work on diglossia in Davis, 2012, and Fuller, 2013). A diglossia ideology makes the sociolinguistic arrangement of the varieties seem like a natural order, as opposed to a cultural construction. Davis (2012) found in her study that, although there are some challenges to the standard language ideology, the hegemony of Jaffna Tamil remains strong. Both monoglossic ideologies and standard language ideologies rely on the idea of a sociolinguistic hierarchy, with the language or dialect at the top often ascribed linguistic as well as social superiority. Further, there is an iconic relationship between languages and their speakers, and erasure of any experiences or practices that might conflict with the hegemony of the dominant variety.

5 Pluralist Ideologies Problematized Although there are many examples of contexts and countries where there is explicit preference for monolingualism in the dominant variety, there are also many spaces in which pluralist ideologies reign and linguistic diversity is valued. There are many different pluralistic Discourses. Some embrace part of normative monolingualism (i.e., the strict separation of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

128

JANET M. FULLER

languages), and most also assume essentialist relationships between languages and their speakers. Thus, as this section shows, seemingly contradictory ideologies are created that value multilingual repertoires while preserving underlying assumptions about languages and their relationships to each other and to their speakers. A classic arena for pluralist ideologies is bilingual education, where there is explicit value and promotion of speaking two (or more) languages. However, educational settings often dictate strict separation of languages. In some cases monolingual patterns of language use are even policed more than in other settings because of the socialization mission of schooling and prescriptivist ideas about language that are paramount in institutions of learning (Fuller, 2012). Indeed, it appears that educational contexts provide some of the strongest ideologies about language separation. (See further in this volume discussion by Juan-Garau & Lyster, Chapter 11, and Garcı´a & Tupas, Chapter 20.) Thus, various classroom studies have found that the ways in which languages are valued even within a pluralist education context may differ, with the societally dominant language positioned as the language of authority or academic success, while the minority language may play a lesser role in the classroom, which mirrors its marginalized role in the community. Jaffe (2003) writes about this dynamic in a Corsican bilingual classroom, noting that literacy practices sought to fulfill both the goal of giving children a sense of ownership of Corsican (despite, or perhaps because of, the fact that most were French dominant) and establishing Corsican as on par with French in terms of literacy tradition and social power. However, most of the work in Corsican was done in the form of group projects, and the evaluation of Corsican literacy skills was not as rigorous as for French; thus the classroom ultimately reinforced the ideology in the wider community that French was more important than Corsican. Similarly, Freeman (2000) reports that in a Spanish–English bilingual middle school program in the United States, Spanish instruction was aimed at beginning-level students, although some of the students in the classroom had learned Spanish in the home. Further, proficiency in Spanish was not required for advancement to the next grade. Although certainly English hegemony in wider society also played a role in the reported lack of interest on the part of many students in learning and speaking Spanish, the institutional role of Spanish in this program positioned it as less important than English. However, there are examples of some different ways of valuing languages that do not adhere to such clearly hierarchical patterns. In arenas that speakers encounter once they finish school, it appears more likely that there is at least tacit acceptance of bilingual discourse or hybrid codes. For example, Swigart (2000) portrays Urban Wolof (a code that includes linguistic features from both Wolof and French) as having what she calls an alternative legitimacy alongside French in Dakar, Senegal. While

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

129

French enjoys official status, and is used for all official and written business, Urban Wolof is also a widely used lingua franca in Dakar. In the thriving informal economic sector, in mass media and advertising, in interactions on the street and in the home, the cultural capital of Urban Wolof is high. Jourdan and Angeli (2014) discuss four different pluralist ideologies: reciprocal multilingualism, hierarchical multilingualism, linguistic pragmatism, and linguistic nationalism, all of which were present to some extent on the Solomon Islands. Reciprocal multilingualism, which was a dominant ideology in precolonial times, values multilingual competence, without strong normative expectations of how different languages should be spoken (see also Quay & Montanari, Chapter 27, this volume). Hierarchical multilingualism, which developed under colonial rule, is essentially multilingualism in which a sociolinguistic hierarchy is present. Speakers of lower-status varieties are seen as racially, socially, and technologically inferior. While these two competing ideologies from previous eras continue to influence the Discourses about language on the Solomon Islands, Jourdan and Angeli (2014) also found ideologies of linguistic pragmatism or instrumentalism, in which English is not seen as intrinsically superior, but as possessing instrumental value for participation in the national and global markets. Finally, the ideology of linguistic nationalism is in this case not a one-nation-one-language ideology but the idea that indigenous languages should be celebrated and ascribed national language status. In this view, national identities can be indexed through many languages that are indigenous to this geographical territory, and all of these should be officially recognized (see Webb, 2002, for a discussion of this in South Africa). Although pluralist because many languages are involved, this ideology is derived from the same underlying concept as the one-nation-one-language ideology, that is, the essentialization of the relationship between language, people, and territory.

6 Language Ideologies in Support of Plurilingualism Perhaps the general theme that ties all of the ideologies discussed thus far together (monoglossic, standard, and pluralistic) is the idea that particular languages are always associated with particular social categories, most notably national and ethnic groups. This can be a motivation for language maintenance and revitalization (e.g., see King, 2000, for Quichua; Dickinson, 2010, for minority languages in the Ukraine; and Coupland, 2010, for Welsh). However, all of these studies make reference to the multiple ideologies involved, and it is clear that speakers’ identification with a local language does not mean that other languages are not only spoken in their communities, but are also part of their social identities.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

130

JANET M. FULLER

Spotti (2011) argues that anti-essentialist views on multilingualism are needed, given that there is no one-to-one correlation between language and social identity. Further, he cautions against a blind celebration of linguistic pluralism by sociolinguists, saying that scholars must recognize that bilingualism may be stigmatized and that ignoring this stigma is denying that there is linguistic discrimination. I would also note that pluralist ideologies can lead to the assumption that minority group members necessarily identify with their minority group language, as opposed to the majority language or both languages. It is furthermore important to acknowledge the existence of pluralist ideologies that embrace the mixing of languages. Researchers have discussed this in terms of translanguaging (Garcı´a, 2009b), polylanguaging (Ag & Jørgensen, 2013), and metrolingualism (Otsuji & Pennycook, 2010). Garcı´a’s notion of translanguaging (2009b; see also Garcı´a & Tupas, Chapter 20, this volume) focuses on the educational benefits of letting learners use all of their linguistic resources, under the premise, discussed in Section 2, that speakers do not use languages but linguistic and semiotic features, and in some cases the linguistic features may be assigned to two or more ideologically distinct codes. In contrast to the monoglossic ideology often found in language classrooms, a translanguaging approach advocates that students use all of their linguistic resources. Because multilingual discourse does not support the dichotomies between languages, it fosters positive identity formation for bilinguals. Work by Ag and Jørgensen (2013) looks at the linguistic performances and underlying ideologies among bilingual adolescents in Denmark, noting that many come from households in which bilingualism extends back generations and many have not necessarily been socialized with a monolingual norm. While they are quite capable of using monolingual standard Danish, and can assess the situations in which this is expected, the teens do not extend this norm to other contexts. Instead, they use various features from different languages, registers, and styles in their casual interactions. The underlying idea here is that this fluidity is a natural state for “[poly]languagers” – which refers to all speakers, as we all have multiple registers and styles, if not dialects and languages (see Chevrot & Ghimenton, Chapter 25, this volume, on the fuzziness of boundaries). The imposition of a monolingual norm and restrictions on what features can be used together is the result of particular patterns of socialization (see De Houwer, Chapter 17, this volume). Such norms are often an explicit aspect of primary and secondary education (see Section 3), but are an ideology that is also present, albeit tacitly, in other contexts and at other times of the lifespan. Otsuji and Pennycook (2010) introduce another term for polylingual practices, namely metrolingualism, which focuses not on the aspect of “mixing” or “hybridity” in linguistic practices, but on the loosening of the association of particular codes with particular identities or social

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

131

categories. They present metrolingualism as a challenge to retrolingual mono/multilingual dichotomies, where certain codes “belong” to members of particular social groups. Metrolingual practices and ideologies are thus potentially a creative and productive ideology for social inclusion. Nevertheless, the use of metrolingual practices also opens the door for cultural appropriation, which is most frequently discussed as the exploitation of artifacts and symbols of a minority culture by members of the dominant culture, often in ways that mock or trivialize them. (An oftcited example of cultural appropriation is the use of pictures of Native Americans in feathered headdresses as sports team mascots in the United States.) The use of linguistic forms that index particular identities can also be seen as cultural appropriation. While ostensibly anyone can use any language without pejorative intent or interpretation, how linguistic forms are used makes a difference. For example, Fuller (2013) addresses the blurry distinction between so-called Spanglish and Mock Spanish, with the former being a creative mixture of Spanish and English but the latter consisting of the misuse of simplified features of Spanish, usually by non-Spanish speakers. There is no clear linguistic dividing line between the two, however, nor is there even a clear distinction in terms of intent. That is, if a speaker does not know that no problemo is mock Spanish, and produces it with the intent of indexing a Spanish-positive identity, it is still a reproduction of racist ideologies in the same way that wearing blackface is a reproduction of racist practices mocking African Americans, even if your intent was simply to better resemble a famous black public figure. Intent matters, but it does not negate historical and cultural connotations. The concept of metrolingualism is thus important to the study of language ideologies because it presents scenarios in which the presence and use of multiple codes is the norm as opposed to the exception, and also allows us the opportunity to address ideologies of language ownership. This topic is addressed again in the next section on Global English.

7 Global English and Native Speaker Ideologies Part of the ideologies about English globally are, of course, the perpetuation of standard language ideologies beyond the borders of countries where English is the dominant language in government, education, and daily life for the majority of speakers (for example, Great Britain, the United States, Australia). Further, ideologies about linguistic pragmatism and the usefulness of English discussed earlier also are central in the Discourses about English worldwide. In addition, however, within the field of Teaching English to Speakers of Other Languages (TESOL) there are Discourses about English that reflect ideas about what a “native speaker” is and about the nature of language learning.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

132

JANET M. FULLER

Proctor (2014) writes about English in India and how both Englishdominant and English-excluded ideologies focus on the benefits of English in terms of social mobility. However, the English-dominant Discourse frames English as a commodity that anyone can acquire, while the English-excluded Discourse recognizes that there are barriers to the acquisition of English and its cultural capital. The English-dominant Discourse negates the effect of privilege and frames language learning as the result of personal effort as opposed to the outcome of a lifestyle that is not attainable for all. Inherent to both of these perspectives is the ideology of the instrumental value of a language. While Proctor’s (2014) research does not use the term native speaker to describe English-dominant speakers in India, this term is ubiquitous in discussions of bilingualism and language teaching, and there is an underlying Discourse about “native speakers” as “owning” a language and being inherently better teachers, an ideology of nativespeakerism. Particularly in the TESOL literature, this position has been amply challenged (see, for example, Llurda, 2016). Kohn (2011) discusses how English as a Lingua Franca (ELF) is increasingly the norm in global communication, and viewing one’s proficiency as developing does not preclude feeling ownership over the language. However, as Llurda (2016) and many others have discussed, even when the difficulty in drawing the line between native and nonnative speakers is recognized, and the absolute authority of native speakers as teachers is rejected, there is much unconscious and even blatant reproduction of this hierarchy by so-called native and nonnative teachers and employers alike. There is further a correlation between nativespeakerism and racism noted in the research on ideologies about native speakers. For example, research in Japan (Rivers & Ross, 2013) shows that there is a preference for English language teachers who are not just “native speakers” but also white, since “whiteness” is considered the norm in English-speaking countries. This also privileges white teachers who are not “native speakers” of English, as they quite literally look the part. The ideologies surrounding English as a global language are linked to a history of colonialism and imperialism. It is no wonder, then, that another Discourse about English is about English as a threat to other languages (Phillipson, 2017). House (2003) discusses a “widespread assumption” of English as a “killer language” and attempts to refute this claim, showing how English is used productively as a lingua franca in Europe. Pandey (2016) provides another perspective on Global English in her discussion of linguistic exhibitionism in prize-winning literature – in particular, in postcolonial literature written in English that contains lexical items from other languages that create an illusion of multilingualism. She argues that this pattern of language use can reinforce linguistic hierarchies that marginalize local languages and privilege English. The inclusion of words from languages other than English, while on the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

Ideologies of Language, Bilingualism, and Monolingualism

133

surface exhibiting diversity, at a deeper level presents multilingualism as simply a few terms that have direct equivalents in English. Thus these literary works, although indexing bilingualism, at the same time demote it to something trivial and put all languages other than English in subordinate roles.

8 Methodologies and Future Advances Language ideologies can be investigated in a variety of ways, although, as can be surmised from the research reviewed in this chapter, this almost always involves some sort of discourse analytical method, often focusing on public discourse and texts. However, there are a few other defining characteristics of language ideology research methods. Both quantitative and qualitative approaches may be employed, and the data used for empirical study vary widely; ideologies can be found in private conversations, public discussions, mass media, and educational practices (among other contexts). Two research methods, however, have gained particular prominence in language ideology research in the past decade. The first approach is that of corpus linguistics, which involves the quantitative analysis of large corpora that may come from a variety of sources – transcripts of spoken conversations, newspaper articles, or other texts. Corpus linguistic methodology is applicable to many lines of research, of which language ideology is only one. Language ideology researchers using corpus linguistic approaches typically identify particular terms that are linked to ideologies, and analyze their relative frequency, distribution, and collocations to make interpretations about ideological stances in the corpus. For example, Fitzsimmons-Doolan (2014) focuses on the 2010 Arizona Department of Education website to identify ideologies about language. The analysis examines collocations of the words language, literacy and English to ascertain what stances are taken with regard to language in schooling. The ideologies unearthed reveal both standard language ideologies and monoglossic ideologies. Another approach that has recently increased in prominence in the study of language ideologies is the analysis of linguistic landscapes, broadly defined as the use of language in the public sphere, including signs, business names, window displays, advertisements, and graffiti. These public visual displays are more than reflections of how languages are used in a community: They are a manifestation of language ideologies. For example, Coupland (2010) argues that the linguistic landscape in Wales shows an aspirational political ideology of bilingualism rather than the objective reality of patterns of language use. Indeed, language use in broader society serves to construct identities and power relations (Cenoz & Gorter, 2006).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

134

JANET M. FULLER

Although corpora of spoken and written discourse and linguistic landscapes have emerged as sites for the identification of language ideologies, scholars continue to conduct research on many other types of data: popular media (Lippi-Green, 2012), ethnographic evidence and interviews (Proctor, 2014), literary works (Pandey, 2016), language policies (Ricento, 2000), and educational practices (Cummins, 2007; Garcı´a, 2009b). While language ideology research is primarily qualitative in nature, the combination with quantitative methods such as in corpus linguistics and some linguistic landscape methods provides further empirical grounding in the field. Further, methodologies that provide links between Discourses and policies (such as Ricento, 2000, or Wee, 2014) should be further explored to move beyond identifying ideologies toward assessing their consequences.

9 Conclusion This overview of themes in the study of language ideologies has mainly focused on hegemonic ideologies that reproduce inequalities (i.e., monoglossic and standard language ideologies) and has shown that some aspects of these ideologies also persist in Discourses that overtly stress the value of linguistic diversity. There are pervasive ideologies about language that hold that all languages are not equal and that there is a sociolinguistic hierarchy. Such ideologies fuel essentialist attitudes toward speakers of particular languages or dialects as having certain traits, which can lead to discrimination. The consequences of ideological stances about language can be seen in many different arenas. Language policies are clearly governed by ideas about language(s) and language rights, and the values of particular languages in specific contexts. Policies and their underlying ideologies also influence how we all use language, including the individual choices we make and the resulting maintenance or loss of language varieties on the societal level. I have also stressed that ideologies are multiple and often conflicting, even within a single interaction or individual. Researchers studying bilingualism must continue to address language ideologies in all their complexity, and should not limit themselves to describing conflicting ideologies, but they should also explain how individual speakers may position themselves with regard to hegemonic ideologies. That is, hegemony is neither complete nor static. To address ideologies and their influence, scholars must continue to look at both public and private spheres of activity. Future directions in language ideology research should reflect the increasingly important new media arenas of language use, where there is potential for new practices to provide a context for both reproduction of and shifts in language ideologies.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:04:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.007

7 Bilingualism and the Law Philipp S. Angermeyer

1 Introduction Language is in many ways central to law and the legal process. The law is codified through language and many legal disputes are primarily about the meaning of specific words or syntactic structures in specific instances of language use. For example, this is true of disputes about the interpretation of legal statutes or contracts, or disputes about speech acts, such as giving consent or making a threat. Language is also central to how such disputes are negotiated and resolved in various arenas of the legal system, such as police interrogations or courtroom examinations. All such manifestations of law through language involve specific standardized linguistic forms that are sanctioned within a given jurisdiction, typically a single standard language variety and its legal register. Access to justice is thus mediated through this variety, and, as a consequence, legal systems privilege those with proficiency in it, and they disadvantage those who are more proficient in other language varieties. The latter often struggle to make their voices heard and to understand the language use of legal professionals, and they may find their own language evaluated negatively. Bilingualism (or multilingualism) thus constitutes an inherent challenge for legal systems to live up to the promise of equality before the law, as the languages spoken by individuals are decidedly unequal, often explicitly so. In response to linguistic diversity, many jurisdictions have seen a growing recognition of the need for interpreting in legal settings, accompanied by an increasing professionalization of interpreting practices. This is particularly true in courts, where it is widely acknowledged that an accused who does not speak the language of the court is entitled to court interpreting in order to understand the charges and engage in a defense against them (see, e.g., European Convention on Human Rights, Article 6.3.; US Court Interpreters Act of 1978; Canadian Charter of Rights and Freedoms, Section 14). At the same time, such rights to interpretation may not be applied consistently across legal settings, or for all languages and language varieties. Some jurisdictions recognize two or more languages as

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

136

PHILIPP S. ANGERMEYER

official, thus making the law itself bilingual and permitting a choice of language within its institutions. However, as is shown in this chapter, even in such instances, there may be unequal treatment of languages with regard to practices of legal written record keeping (i.e., entextualization), and language choice may be indexical of unequal social identities. This chapter provides an overview of ways in which bilinguals and their bilingualism are managed and evaluated in the legal sphere. The investigation of such issues falls broadly within the interdisciplinary field of language and law that brings together scholars from linguistics, law, and social sciences, such as legal anthropology or law and society. This encompasses also the growing field of forensic linguistics, which aims to apply linguistic analysis to address language-related questions that arise in the judicial process. Researchers in this field are organized in the International Association of Forensic Linguists (IAFL), and have found specialized outlets for publication in journals such as the International Journal of Speech Language and the Law or Language and Law = Linguagem e Direito, and in book series with Oxford University Press and University of Chicago Press. Linguistic research on multilingualism in the legal sphere often focuses on discourse analysis of genres of institutional interaction in particular jurisdictions, such as police interrogations, courtroom talk (especially cross-examination in adversarial courts), or asylum interviews. Often scholars are drawn to such data to address larger macro-sociolinguistic issues of language ideologies and inequality in multilingual societies. At the same time, many scholars have used their findings to advocate for changes in institutional practice in law and law enforcement. Other scholars report on research that is conducted in the service of institutions, such as language analysis for the determination of origin in asylum cases, or giving expertise on questions such as language proficiency or the meaning of forms in other languages. As the discussion of research findings in this chapter makes clear, in the legal sphere injustice results not merely from the fact that different languages represent barriers to communication, but also from language ideologies.

2 Bilinguals in Monolingual Jurisdictions: Key Options for Communication Much linguistic research on linguistic diversity in the legal sphere is concerned with so-called legal–lay communication, that is, communication in institutional settings between lay members of the public and legal professionals, such as police officers, attorneys, or judges. Like all kinds of institutional interaction, these are marked by a considerable power asymmetry between the participants, which manifests itself discursively in differential rights in speaking and turn-taking. In addition, as noted by

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

137

Conley and O’Barr (1990), among others, such communication is generally marked by differences between the linguistic and discursive resources that are available to participants. Lay participants may face comprehension difficulties when faced with spoken or written legal discourse, and they may struggle with communicating in ways that are effective and acceptable to legal professionals. Linguistic diversity thus does not introduce communication problems into the legal sphere, but rather amplifies existing ones, making them more apparent to all parties (and harder to ignore for the institution). Where legal–lay communication involves speakers of different languages, communication can be achieved in several ways, with participants either speaking the same language, or speaking different languages and making use of an interpreter. A first option for legal–lay communication is same-language interaction, which occurs when one or more participants speak in a second (or third) language in which they have limited proficiency. Typically this involves lay participants using the institution’s official language as a second language. This, as we see later, has potentially negative consequences for multilinguals’ ability to make themselves heard and especially their ability to comprehend legal arguments. Alternatively, legal professionals may speak the lay participant’s nonofficial language. This is especially common in attorney–client interaction, if attorneys share their client’s linguistic background or have acquired second language proficiency in the language. In the United States, the use of Spanish by legal professionals (both as first and second language, or L1 and L2) has been documented in a variety of contexts, including protective order interviews (Trinch, 2001), police interrogations (Berk-Seligson, 2009), and arbitration hearings in small claims court (Angermeyer, 2015, p. 49). In other cases, legal–lay communication may involve a lingua franca that is neither the language of the institution, nor the lay participant’s first language. This often occurs with the use of English in refugee contexts in continental Europe (see, e.g., Maryns, 2006). However, whenever nonofficial languages are used, this is confined to oral domains that are relatively informal, and it generally requires entextualization in the official language in the form of depositions or witness statements that are produced by the institutional agent, significantly increasing the potential for miscommunication or distortion (Berk-Seligson, 2009; Jacquemet, 2009). In contrast to the options for legal–lay communication between speakers of different languages just described, which are conducted in the same language, interactions may be interpreter-mediated if the primary participants speak different languages. This process is constrained by many factors such as the training, pay, and availability of qualified interpreters, which may differ greatly by language, jurisdiction, and legal setting. Interpreting is used most consistently in courts and in asylum interviews, but less reliably in police interrogations or during

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

138

PHILIPP S. ANGERMEYER

the pretrial phase of the judicial process (Morris, 2008; Berk-Seligson, 2009). When interpreting is available, institutions may come to view it as the most reliable method of communication, and they may discourage lay participants from using the language of the institution. Consequently, bilinguals (other than the interpreter) generally have to act as monolinguals and refrain from bilingual language use (Angermeyer, 2015), as bilingual litigants are not allowed to codeswitch, bilingual attorneys are not allowed to question clients in their shared nonofficial language, and bilingual jurors are not allowed to consider original-language testimony, but have to listen to the interpreted version instead. As Rock (2017) noted, such policies do not take into account that many bilinguals might be used to drawing on their entire linguistic repertoires to communicate, i.e., through what has become known as translanguaging (e.g., Garcı´a & Tupas, Chapter 20, this volume).

3 Legal Interpreting In all legal settings, the provision of interpreting depends on the correct identification of lay participants’ language varieties. Interpreting may be undermined by incorrect assumptions about proficiency and intelligibility. Institutions frequently assume that individuals are speakers of the official language of their country of origin, following a “monoglot standard” language ideology (Blommaert, 2009). For example, Haviland (2003) discusses a murder trial in Oregon in which a Spanish interpreter was assigned to interpret for the accused, a speaker of the indigenous Mexican language Mixtec, who only had minimal proficiency in Spanish. Similarly, Marszalenko (2014) notes the use of interpreters for Tagalog with Filipinos who are native speakers of other languages of the Philippines such as Ilocano, Cebuano, or others. In many such instances, lay participants are speakers of languages that are marginalized and not viewed as fully legitimate languages in the country of origin. This in itself is likely to limit the availability of interpreters, even if institutions were willing to employ them. A comparable situation is found with languages that are regarded as dialectal, nonstandard, or “incorrect” variants of the official language. For example, in Anglophone jurisdictions, well-studied cases include Jamaican Patois (Brown-Blake & Chambers, 2007), Australian Aboriginal English (Eades, 2008), and African American Vernacular English (AAVE) (Rickford & King, 2016). Interpreting is generally not available for speakers of such varieties, despite the fact that they may experience considerable communication problems in court. For example, Rickford and King (2016) demonstrate that jurors and legal professionals in the George Zimmerman trial had difficulty understanding

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

139

the key witness, Rachel Jeantel, when she spoke in African American Vernacular English. While these comprehension problems were apparent throughout the testimony, legal professionals and observers blamed them on the witness’s alleged failure to speak “clearly,” not on the other participants’ lack of familiarity with AAVE. Such an attribution of blame to other speakers of another language or variety can also be found in situations where language boundaries are not in dispute, but where failure or inability to speak the language of the court is treated as a moral shortcoming on the part of the individual (Haviland, 2003). This shows that speakers’ linguistic usages are evaluated socially, based on language ideologies, a point further discussed in later sections.

4 Bilingual Courtroom Interaction and the Limits of Interpreting Compared to other legal settings, language use in court is more regulated and less flexible, as participants’ speaking rights are strongly constrained based on their role in a given trial phase. This is also true of language choice, as courts operate in a formal, legal register of the official language of record (see De Houwer, Chapter 17, this volume, on language choice in other settings). As noted earlier, many jurisdictions explicitly recognize that the right of due process requires the provision of interpreting to any accused who does not speak the language of the court. Some jurisdictions extend the provision of court interpreting to all parties in civil trials (e.g., civil court in New York State; see Angermeyer, 2015). Combined with increased migration worldwide, such policies have led to a growing use of court interpreting, along with an increased professionalization and standardization of court interpreting services, as evidenced by the development of specialized training offered in interpreting and translation programs, for certain language dyads (e.g., see Hale, 2004), or the creation of professional organizations such as the National Association of Judicial Interpreters and Translators (NAJIT) in the United States. As a consequence, court interpreting has become a frequent topic of research in language and law and translation studies, in a wide variety of jurisdictions, including, among others, Australia (Hale, 2004; Lee, 2010), Austria (Kadric´, 2001), France (Licoppe & Vernier, 2013), Hong Kong (Ng, 2009), Japan (Marszalenko, 2014), South Africa (Moeketsi, 1999), and the United States (Angermeyer, 2015; Berk-Seligson, 1990). This research has uncovered the many subtle ways in which institutional policies and the practices of interpreters and legal professionals, ironically, may compromise the delivery of justice for individuals who rely on interpreting. The pragmatics of interpreter-mediated interaction and the interpreter’s translation choices have been found to be powerful forces that

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

140

PHILIPP S. ANGERMEYER

may influence the interaction between other participants. In her pioneering study of Spanish–English court interpreting in the United States, Berk-Seligson (2009) compared interpreters’ target renditions to their corresponding source utterances. She found that interpreters often altered the pragmatic force of a question or piece of testimony, for example by downplaying speaker agency and blame attribution through the use of agentless passives or impersonal constructions. To test the impact of the interpreters’ performance, Berk-Seligson also conducted experiments with mock jurors that showed that their evaluation of the credibility of a witness was influenced by the speech style of the interpreter (e.g., whether the interpreter’s speech was marked by hesitation and hedging, or not). Similarly, Hale (2004), working on Spanish–English court interpreting in Australia, argued that interpreters influenced the perception of testimony. For example, they made testimony appear less coherent by leaving discourse markers untranslated, or made it appear more convincing to legal professionals by using a formal register to translate a nonstandard variety. At the same time, Hale found that the interpreters in her study had a tendency to translate attorneys’ questions in ways that made them less controlling (putting fewer constraints on the range of acceptable answers). Investigating Korean–English interpreting in Australia, Lee (2010) found that when Korean-speaking witnesses used indirect reported speech in their testimony, interpreters often rendered this as direct reported speech. When doing so, Lee argued, interpreters inaccurately create the impression that witnesses are giving verbatim quotes, but they also conform to a general discursive preference in the legal system for direct reported speech. Researchers’ interest in comparing legal interpreters’ renditions to the corresponding source talk has revealed how important the choice between direct and indirect translation is in shaping the pragmatics of interpretermediated court interaction. This choice constitutes a resource for stancetaking (Angermeyer, 2015; Wadensjo¨, 1998). As such, it is telling of the interpreters’ perceptions of their own role in the interactions, their stances toward the institution and the other participants. In direct translation, interpreters represent “the whole appearance of another person’s utterance” (Wadensjo¨, 1998, p. 19), including all reference to person deixis. This means that interpreters speak in the voice of the translated speaker, rather than in their own, using first person singular to refer to the source speaker, and using second person singular to refer to the addressee of the source talk, even if that person is not a recipient of the target utterance (Angermeyer, 2015). Indirect translation, by contrast, involves a shift in person deixis to the participation framework of the target speech, with first person referring to the interpreter, second person to the target recipient, and third person to all other participants, including the source speaker. Indirect translation thus often takes on the form of indirect reported speech with interpreters “presenting the other’s words and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

141

simultaneously emphasizing personal non-involvement in what they voice” (Wadensjo¨, 1998, p. 19). In many jurisdictions, court interpreting is subject to specialized translation norms that require interpreters to produce target renditions exactly matching the source in propositional content and form, and that explicitly prohibit the use of indirect translation. This is in stark contrast to other forms of community interpreting, such as in medical or educational contexts, where reported speech is commonplace. The explicit legal norms of interpreting, and especially the insistence on direct translation, distinguish court interpreting from other forms of community interpreting. As a result, they may conflict with the expectations of minority language speakers, who are likely to already be familiar with community interpreting in other contexts, and lead to pragmatic misunderstandings (see, e.g., Angermeyer, 2015, p. 86). Such explicit regulations of interpreting can be understood as resulting from the production of written court records (so-called legal entextualization practices), in which utterances by the interpreter enter the court record as utterances of the source speaker. As Haviland (2003) notes, such regulations are also based on a language ideology of referential transparency that assumes that translation can be done verbatim, without a change of meaning. In referential transparency, communication is viewed solely as a matter of putting preexisting thoughts into words, so that it is deemed possible to “substitut[e] one language’s word for another’s, as though the word, or code, is merely an exotic costume for a shared meaning” (Haviland, 2003, p. 772). A characteristic feature of court interpreting is the asymmetrical distribution of consecutive and simultaneous interpreting modes that is common in many jurisdictions (Berk-Seligson, 1990, p. 38; Kadric´, 2001). When court interpreters translate for the court, from a nonofficial language into the language of the court, they use short consecutive interpreting mode, with the source speaker forced to pause to let the interpreter translate, leading to fragmentation of their testimony and prompting interruptions from other participants (Angermeyer, 2015). By contrast, in the opposite direction, when the interpreter translates into a nonofficial language so that lay participants can understand the proceedings, this is done in simultaneous interpreting mode as source speakers speak continuously without pausing. However, as simultaneous interpreting places a greater cognitive demand on interpreters than short consecutive interpreting, translations may be less accurate and important information may be omitted (Angermeyer, 2015, p. 133). The asymmetrical distribution of interpreting modes thus doubly disadvantages participants who rely on interpreting, in that they are more likely to be interrupted and less likely to receive the same amount of information than other participants. While this distribution of interpreting modes is found in many jurisdictions, it is generally not explicitly codified, but rather appears to result from the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

142

PHILIPP S. ANGERMEYER

pragmatics of the interaction. Consecutive interpreting arises automatically when the questioning attorney and responding witness rely on the translation to understand one another. By contrast, if they speak the same language, they are not likely to wait for the interpreter before responding to one another’s turns, unless they are compelled to do so, for example, by the judge (Licoppe & Vernier, 2013). Accurate simultaneous interpreting such as so-called conference interpreting at multinational institutions requires specific working conditions in which teams of highly trained interpreters work out of soundproof booths, frequently alternating to avoid fatigue. While such conditions do exist in high-profile international courts or tribunals, such as the International Criminal Tribunal for the former Yugoslavia in The Hague (Elias-Bursac´, 2012), they are very costly and have not generally been adopted in national jurisdictions. Moreover, some jurisdictions have sought to reduce the financial cost of any form of interpreting. In 2011, the government of the United Kingdom outsourced the provision of court interpreting services to a private company, a decision that has been widely criticized by legal professionals, forensic linguists, and interpreters (organized in the group Professional Interpreters for Justice). Opponents argue that privatization has led to degradation in the quality and availability of court interpreting (with trials delayed by interpreter absences), ultimately compromising the delivery of justice for individuals who rely on interpreting. Where court interpreting needs to be defended against budget cuts, advocates emphasize its importance for safeguarding the constitutional rights of defendants, while interpreting for people other than defendants receives less attention. Consequently it is not surprising that court interpreting is generally not available for jurors with limited proficiency in the language of the court, with the occasional exception of interpreting for signed languages (Napier & Spencer, 2008). Speakers of other languages are thus typically excluded from the jury pool, with consequences for the justice system as a whole. Moreover, courts in the United States have sometimes excluded Spanish–English bilinguals from juries for trials with Spanish-speaking defendants on the grounds that they understand the original testimony and may choose to ignore its interpretation into English.

5 Police Interrogations and Misunderstandings Language choice in interactions with police officers is generally more variable and dynamic than it is in court, and less likely to involve a professional interpreter. Linguistic research on such interactions has been spurred by the fact that police interrogations are increasingly recorded on audio or video, resulting in recordings becoming available as evidence for forensic linguistic analysis. Studies have been conducted in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

143

several countries, including Israel (Morris, 2008), Japan (Marszalenko, 2014), Sweden (Wadensjo¨, 1998), and the United States (Pavlenko, 2008; see also Berk-Seligson, 2009). Despite the varied national contexts, they find parallels in the issues facing speakers of nonofficial languages in dealings with police. In some jurisdictions, bilingual police officers have been observed to conduct interviews in nonofficial languages, or to act as interpreters for other police officers. In a detailed analysis of transcripts of police interrogations with Spanish-speaking suspects in the United States, BerkSeligson (2009) argues that interpreting by police officers amounts to “linguistic police misconduct” (p. 171), leading to coerced confessions. Her analysis shows that police officers pursue the goal of obtaining evidence in support of their investigation, preferably, a confession. Police officers who act as interpreters do not stop pursuing these goals, and as a result they interpret selectively, favoring information that matches their theory of the crime. They also try to gain the detainee’s trust by claiming common ground as speakers of the same language. This may lead them to downplay the severity of accusations, or discourage the invocation of the suspect’s rights (see later in this chapter). A further concern arises from the entextualization of police interviews through written statements that are produced by police officers in the official language and that are to be signed by the lay participant. For otherlanguage speakers, this typically involves an ad hoc sight translation of the document (see, e.g., Marszalenko, 2014, p. 180). When this back translation is made by police officers rather than by neutral interpreters, this can be viewed as compromising the integrity of the statement, and it may mask miscommunication during the interrogation itself. For example, Berk-Seligson (2009) discusses a case in which a police officer with very limited L2 proficiency in Spanish acted as an interpreter. The interaction was marked by persistent communication problems and was rife with misunderstandings, but these did not prevent the production of a statement containing a confession. This demonstrates that linguistic difference does not merely represent a communication barrier, but can actually be exploited to exercise coercion and control. In light of such issues, forensic linguists as well as legal scholars have argued that recordings of original other-language testimony be made and preserved as evidence (see also Morris, 2008). Suspects in police custody have important rights that are intended to guarantee due process of law and to protect the integrity of the investigation. Depending on the jurisdiction, these define the suspect’s rights and obligations toward the police and explain access to legal representation. Importantly, police officers are obligated to inform suspects about these rights. This represents a challenge when suspects lack proficiency in the official language. A group of linguists and legal scholars convened by Diana Eades and Aneta Pavlenko has recently developed a set of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

144

PHILIPP S. ANGERMEYER

Guidelines for communicating rights to non-native speakers of English in Australia, England and Wales, and the USA. These have been endorsed by several national and international associations of linguists and applied linguists (Communication of Rights Group, 2016). Based on research in forensic linguistics and psycholinguistics on second language acquisition and comprehension, the guidelines include recommendations for the wording of the police caution itself, as well as for the process of communication, including with regard to language choice and interpreter availability. In particular, they recommend the use of standardized versions of the police caution in plain English as well as in other languages. The guidelines also advise legal professionals to determine the understanding of rights by asking suspects to restate them in their own words, rather than to rely on the common strategy of asking yes/no questions such as Do you understand? As shown by Eades (2008) and Berk-Seligson (2009), the power asymmetry inherent in police interrogations or courtroom examination may compel speakers of nonnative or marginalized varieties to respond affirmatively to yes/no questions, regardless of whether they have understood or even whether they agree with the proposition, a phenomenon termed gratuitous concurrence. The issue of a suspect’s understanding of police warnings such as the Miranda rights in the United States is often contested at trial. When suspects have made self-incriminating statements to police, these are admissible as evidence only if the statements were made voluntarily, after the suspects were informed of their rights. Reporting on a case in which she acted as linguistic expert, Pavlenko (2008) analyzes the recording of a police interrogation of a native speaker of Russian in the United States. Pavlenko demonstrates that the suspect’s L2 English was marked by numerous errors representing transfers of Russian morphosyntactic structures into English, as well as by a reliance on highfrequency lexical items. She further presents evidence of miscommunication, showing that the suspect’s comprehension of English was constrained by morphosyntactic and lexical difficulties, including lack of familiarity with legal terms such as detained or waiver. Pavlenko (2008) concludes that, in her opinion, the suspect did not understand the Miranda warning, and therefore did not waive rights knowingly and voluntarily. Consequently, Pavlenko argues for the need to distinguish between, on the one hand, basic interactional competence in an L2 that enables a learner to “answer basic biographical questions” (p. 9) and, on the other hand, the cognitive academic language proficiency required to process complex legal texts. This recommendation is counter to legal practice, however. In a survey of judgments in US American appeals courts, Solan and Tiersma (2005, p. 82) note that judges are generally not willing to overturn the validity of Miranda waivers if a suspect has used any English at all during the interrogation or at trial (see also Berk-Seligson, 2009).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

145

6 Bilingual Law and the Problem of Functional Linguistic Relativity As discussed thus far, much of the research on bilingualism in legal settings is concerned with situations of individuals who speak a language variety that differs from that of the institution. A different situation exists when jurisdictions have designated more than one language as an official language, whether at the regional or national level in states such as Canada or South Africa, or in supranational entities like the European Union. Multilingual jurisdictions are typically based on the assumption that equal language rights entail that co-official languages are functionally equivalent. However, research in multilingual jurisdictions has challenged such assumptions and has shown that the choice between official languages can have unexpected and unintended legal consequences. Unlike rights to interpreting that derive from an individual’s right to due process, multilingual jurisdictions derive from the language rights of population groups, such as linguistic minority groups that are protected by the state. The provision of such rights falls largely under the field of language policy and planning (Lo Bianco, Chapter 8, this volume). However, official bilingualism has important implications for the legal system. For one, institutional multilingualism introduces a language choice for individuals that is generally independent of their proficiency in another language. For example, in Canada, where English and French have equal status, individuals in federal court have the right to speak either language (even if their language choice may run counter to the preferences and expectations of other participants). Moreover, their testimony will enter the record in the original language, never in translation (Morris, 2008). This differs markedly from the situation of speakers of nonofficial languages whose participation is always mediated by translation, as noted in Section 4. Furthermore, beyond lay–legal interaction, official multilingualism has wide consequences throughout all aspects of the legal system, as both the codification of the law and its interpretation by the courts are multilingual. In Canada, official bilingualism means that the law is bilingual in English and French. At the federal level, legislation is drafted in two parallel language versions, which are debated and enacted together by the bilingual parliament. As a consequence, the English and the French versions of the law are both “original and equally authoritative expressions of the law” (Sullivan, 2004, p. 1006), and neither version counts as a mere translation of the other. When there is a discrepancy between the two language versions, Canadian courts have traditionally applied a “shared meaning rule,” which involves disregarding any elements that are found in only one version, on the assumption that both versions

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

146

PHILIPP S. ANGERMEYER

express a single, shared legislative intent. However, Sullivan (2004) points out that differences in meaning are also introduced by the fact that the English and French versions of the same law draw their terminology from distinct legal traditions, namely the French civil law that applies in Quebec, and the British common law that applies elsewhere in Canada. As a result, common law courts may interpret the same legal statute differently than civil law courts. In addition to English and French, several indigenous languages have coofficial status in Canadian territories, such as Inuktitut in Nunavut. However, as the territory’s parliament enacts legislation in English and French only, but not in Inuktitut, any translation into Inuktitut exists for information purposes only, and is not legally enforceable (Sullivan, 2004). At the same time, legislation does refer to traditional indigenous concepts, for example with regard to wildlife preservation, and Sullivan argues that courts will therefore have to investigate the meaning of such concepts in Inuktitut when they interpret statutes in which they are used. Comparable issues, albeit on a larger scale, arise in supranational entities such as the European Union (EU). As of August 2018, the European Union has 24 official languages. EU citizens have the right to access all documents EU institutions put out in any of these official languages. Moreover, as EU treaties and legislation are enacted by national parliaments of member states, each language version becomes legally binding. Consequently, the European Court of Justice has to consider multiple language versions of such laws, in an effort to resolve any discrepancies between them (McAuliffe, 2012). The large number of official languages leads to a strong reliance on interpreting and translation, and to considerable financial costs. At the same time, different institutions in the EU have designated working languages in order to save costs and facilitate communication with the European Commission using English, French, and German. The European Court of Justice uses French as its sole working language. Yet having a working language can lead to the paradoxical outcome of using court interpreters without a need, when all participants are native speakers of the same language. A comparable situation is found in South Africa, where nine African languages were made co-official with Afrikaans and English after the fall of the apartheid regime, yet English and Afrikaans continue to be used as the main working languages of the court system. As Moeketsi (1999) describes, this can lead to situations where court interpreters are used even though all participants are native speakers of the same African language. Practices of language choice in legal institutions are closely tied to legal traditions that may resist change even where it is politically desired. Ng (2009) shows how the British common law system is maintained in Hong Kong following the transition of sovereignty from Britain to China. Court users have the choice of English or Chinese, but this choice of language has profound consequences for the way that trials are conducted.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

147

Ng observes that adherence to the juridical formalism of the British common law system is tied to the use of English and is much reduced when trials are conducted in Chinese. Trials typically involve disputes between Cantonese speakers, and their ability to participate in the proceedings is much more direct and effective in Chinese-language trials. In Englishlanguage trials, lawyers may elicit testimony in Cantonese, but for the most part, participation by lay participants is mediated and constrained by court interpreting. More formal and monologic trial components such as the judgment or opening and closing statements are spoken in English, resulting in a linguistic distancing of the trial from the original context of the dispute (Ng, 2009, p. 238). In his ethnographic study of a bilingual Hopi tribal court, Richland (2008) similarly finds an ideological link tying languages to particular forms of argumentation and jurisprudence. As in many indigenous societies in North America and elsewhere, proficiency in the ancestral language is unevenly distributed; yet participants who choose Hopi over English are more successful in their appeals to traditional Hopi notions of morality and responsibility that are endorsed by the tribal court. Richland (2008) observes negotiations over language choice that reveal in some participants a language ideology in which Hopi “bears a unique and nontranslatable capacity to contextualize controversies of ‘land and tradition’” (p. 102). In sum, studies on bilingual jurisdictions converge to suggest a functional linguistic relativity, as language choice is shown to have an impact on how participants interact and how judicial institutions operate. Consequently, they challenge the belief in “referential transparency” (Haviland, 2003), which is widespread in legal approaches to multilingualism, namely the idea that languages are functionally equivalent and interchangeable systems for the communication of fixed, languageindependent meanings.

7 The Language of Bilinguals as Evidence In legal settings, the languages of bilinguals are not merely tools for communication, but they can become part of the evidence used to make legal decisions. Where bilinguals’ language choice factors into the assessment of their credibility and trustworthiness, it becomes part of so-called demeanor evidence. For example, converging to an interlocutor’s dominant language may later be deemed a way to seek their approval, or may be understood as such (see also De Houwer, Chapter 17, this volume, on convergence). As noted earlier, Berk-Seligson (2009) argues that the bilingual police officers in her study use Spanish to gain the trust of Spanishspeaking suspects. Similarly, Angermeyer (2015) shows that small claims litigants who use a court interpreter switch to English to respond and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

148

PHILIPP S. ANGERMEYER

engage more directly with English-speaking legal professionals, even as these regularly discourage such code-switching and insist on interpreter use instead. More commonly, divergence of language choice is seen as indexing confrontational stances (De Houwer, Chapter 17, this volume). For example, Haviland (2003) discusses the case of two bilingual health care workers in the United States who were fired for violating their employer’s “Englishonly” policy. Deciding in favor of the employer, the court found that this policy was not discriminatory against bilinguals who could easily speak English instead. In this view, the workers’ choices to speak Spanish are interpreted as “willful acts of disobedience” (Haviland, 2003, p. 772). Similarly, it is not uncommon for bilinguals who have requested an interpreter to be accused by other participants of hiding their “true” proficiency in the standard language (Angermeyer, 2015, p. 9). Such assessments derive in part from a failure to recognize the uneven distribution of skills, as a person who may be able to have a conversation in an L2 may nonetheless lack the academic proficiency required to understand the complex sentences and specialized vocabulary that are characteristic of legal language (see also Pavlenko, 2008, as discussed earlier). These examples illustrate the fact that bilingual or bidialectal speakers are evaluated on the basis of their language choices, no matter which language they choose. Moreover, the examples show that such evaluations are often based on stereotypes and on unscientific assumptions about bilingualism and second language acquisition. Importantly, such stereotyping is not limited to nonofficial language varieties, as speakers of official varieties in multilingual jurisdictions may face similar attitudes (see Section 3). In other instances, the evidentiary status of linguistic choices becomes even more explicit, as linguists are called upon to analyze linguistic forms to help in law enforcement or in the judicial process (Solan & Tiersma, 2005). For example, forensic linguists frequently analyze anonymous phone calls or written texts such as threats or ransom notes in order to create a profile of the speaker or author with regard to characteristics such as region of origin, gender, age, or social class. One area of forensic linguistics of particular relevance to bilingualism is the practice of language analysis for the determination of origin (LADO). LADO is used in many countries to assess the validity of refugee claims by claimants whose country of origin is in doubt. Following the assumption that there is a connection between the way a person speaks and his or her place of origin, an asylum seeker’s speech is evaluated in comparison to a given vernacular variety that is known to be spoken in the alleged country of origin. In recent years, such assessments have increasingly been performed by private companies, which often base their analysis on impressionistic judgments by native speakers. Some linguists have criticized this practice and have called upon governments to follow a set

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

149

of guidelines that outline recommended practices and constrain the applicability of LADO (Language and National Origin Group, 2004). The signatories to the guidelines emphasize that language analysis should only be conducted by trained linguists who have expertise in the relevant language variety and knowledge about how it differs systematically from neighboring varieties. Moreover, they call upon governments to recognize the limitations inherent in LADO, for example, that it provides information about a person’s place of socialization, rather than citizenship. The authors emphasize that LADO “should be used with considerable caution when addressing questions of national origin, nationality or citizenship” (Language and National Origin Group, 2004, p. 265). It is particularly unreliable in cases where speakers of a single vernacular variety may live on either side of a political boundary as is the case, for instance, with ethnic Hazaras. They speak Hazargi Dari but may live either in Afghanistan or in Pakistan (Language and National Origin Group, 2004, p. 263). LADO often involves an interview with the asylum seeker that is conducted solely for the purpose of language analysis. However, applicants’ linguistic repertoires may also be evaluated during the regular asylum interview that otherwise focuses on the content of their speech, namely on their reasons for seeking asylum (Maryns, 2006). Blommaert (2009) describes the case of “Joseph,” an asylum seeker whose claim to be Rwandan was denied by the British Home Office on the grounds that he was not a fluent speaker of Kinyarwanda, a fact that became apparent during an interview in which an interpreter for Kinyarwanda was present. In the rejection letter to the applicant, the Home Office contends that every Rwandan “will normally be able to speak” Kinyarwanda as it is “the medium of instruction in schools at primary level” (Blommaert, 2009, p. 420). Blommaert shows that such a “monoglot ideology” (Fuller, Chapter 6, this volume) of national homogeneity is deeply misguided in contexts like Rwanda, where the genocide and its related war caused soldiers, refugees, and their languages to cross borders, and school-based language socialization was nonexistent. The Home Office suspected that Joseph is Ugandan, based on his familiarity with Runyankole, a language that is mainly spoken in Uganda, but has spread into Rwanda through networks of migration, refugees, and Hutu rebels. However, Blommaert (2009) shows that the asylum officers failed to pick up on an important sociolinguistic clue: Joseph consistently used the term Kinyankole to refer to this language, which is not used in Uganda, but would be used in Rwanda, where the prefix ki- is often used with language names (p. 424). Blommaert (2009) thus shows that language ideologies and a lack of sociolinguistic expertise may lead refugee agencies to misinterpret language evidence, even if they claim that their decisions are based on objective assessments of language data. His analysis also shows how institutional lack of understanding of sociolinguistic and cultural conditions becomes a liability for the applicant. Along similar lines, Jacquemet

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

150

PHILIPP S. ANGERMEYER

(2009) notes in his discussion of asylum hearings in Italy that misspellings of proper names by officials paired with an inconsistent Romanization of names written in Arabic script can lead to accusations that an applicant is using a “false name” (p. 540). The linguistic research literature on asylum hearings abounds with examples in which cases are decided on the basis of apparent inconsistencies between different tellings of the same refugee story, and following a language ideology in which such inconsistency is viewed as evidence of lack of truthfulness (Eades, 2012). These apparent inconsistencies are identified on the basis of entextualizations of prior tellings, and scholars have documented many ways in which multilingualism and linguistic diversity interfere in entextualization processes, whether participants communicate through an interpreter or in an L2. Maryns’s (2006) study of asylum interviews in Belgium provides a particularly striking example, where an applicant’s testimony that “one man [. . .] carry me, help me” is entexualized in the agent’s report as “A man named Karimi helped me” (p. 1). As noted also in the foregoing discussion of multilingual police interviews, the persistent issues around the reliability of entextualization have led linguists to advocate for more widespread recording of testimony given in nonofficial languages.

8 Present Advances and Future Needs in Research Methodologies As this chapter shows, research on bilingualism and the law tends to focus on specific genres of bilingual legal–lay communication. At the same time, scholars increasingly seek to approach such genres as elements in an intertextual chain of interaction and entextualization, exploring the transformation of disputes across time and the ways in which different genres relate to one another (see, e.g., contributions in Heffer, Rock, & Conley, 2013), or how processes like legal interpreting differ across settings such as police interrogations and courtroom interaction (Marszalenko, 2014; Morris, 2008). Such holistic approaches to bilingualism in the legal process face methodological challenges, however, as access to necessary data is possible only if researchers collaborate with the institutions. Moreover, research on courtroom interaction is constrained by the fact that many jurisdictions do not permit researchers to make recordings in the courtroom. Researcher access to other types of lay–legal interaction is similarly restricted, be it police interrogations, attorney–client interaction, or asylum interviews. As a result, research in language and law has often been a fortuitous by-product of consultant work, when defense attorneys or prosecutors solicit forensic expertise from linguistic experts (Berk-Seligson, 2009; Blommaert, 2009; Haviland, 2003; Pavlenko, 2008). Similarly, when scholars have been able to gain direct access to legal settings, this has generally occurred in collaboration

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

Bilingualism and the Law

151

with institutions and, at least in part, in pursuit of institutional objectives. One such area of research is the impact of digital communication technology on legal–lay interaction, where technological advances produce new challenges and opportunities. For example, digital teleconferencing is increasingly used to avoid the transportation of prisoners for brief courtroom appearances in the pretrial phase. Licoppe and Vernier (2013) explore how such teleconferencing alters the provision of court interpreting in French courts, as the interpreter is physically separated from the other language speaker. They find that simultaneous interpreting is impossible when teleconferencing relies on a single audio channel for all interaction, and consequently, consecutive interpreting is used throughout hearings, whether the defendant is addressed or not. Other innovative uses of digital media are also likely to generate new avenues for research in the future. For example, the frequent filming of police arrests by bystanders, as well as the use of body cameras by police officers, is likely to spur new research into police–lay encounters, as they produce new types of data for forensic linguistic analysis, while at the same time prompting questions about the impact of the camera on the encounter itself.

9 Conclusion Studies of bilingualism in the legal sphere shine a spotlight on a central challenge in our globalizing world: how to achieve justice and equality in the face of linguistic and cultural diversity. Equal access to justice depends on access to information and on the ability to make one’s voice heard, yet this is clearly constrained by language, that is, by the match or mismatch between participants’ linguistic repertoires. However, as work by Haviland (2003), Blommaert (2009), Eades (2012), and others demonstrates, injustice results not merely from the fact that different languages represent barriers to communication, but also from language ideologies. Implicit assumptions about language, about bilingualism, second language acquisition, or translation deeply impact the ways in which linguistic difference is managed and evaluated in the legal sphere. It is incumbent upon scholars in language and law to identify and problematize such ideologies whenever possible.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:51:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.008

8 Language Planning and Policies for Bilingualism Joseph Lo Bianco

1 Introduction In 1992, an eminent US public policy analyst, Aaron Wildavsky, published an article in the Journal of Policy Analysis and Management entitled “Finding Universalistic Solutions to Particularistic Problems: Bilingualism Resolved through a Second Language Requirement for Elementary Public Schools.” The collocation “bilingualism resolved” in Wildavsky’s title reveals a great deal of what I argue in this chapter: essentially, that what bilingualism is taken to be presents itself as a significant part of the public policy challenge for professionals concerned with promoting bilingualism, be they academic researchers, speech-language pathologists (SLPs), interpreters and translators, teachers, or communities committed to promoting bilingualism, for whom it should be taken not as a problem to solve, but as a public good, a community asset, and an intellectual resource. In a nutshell, in this chapter I take the position that language policies should be comprehensive and rights based. I also strive to show how the practices of language policy (LP) making are connected to different kinds of knowledge about bilingualism, framed by different constituencies or protagonist voices. Furthermore, I draw from poststructuralist and knowledge-utilization theorists in the wider field of public policy, who have problematized policy as technique, in order to argue that it is important to understand the different knowledges, ideologies, interests, and discourses these various protagonist voices contribute in framing what the problem to tackle in policy is, and what bilingualism should mean in a given policy debate against the constellation of legal, political, and sociolinguistic features of a given context. All too often LP is imagined to unfold along a binary opposition between policy makers and administrators as unknowing powerholders, on the one hand, and practitioners as knowledgeable impotents, on the other. Among the practitioners, LP researchers are counted as trying to speak truth to power. But I insist that this is a crude and inaccurate understanding of the complex dynamics of public policy

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

153

making and of LP. In LP, knowledge and power – and what counts as evidence – are always enmeshed in the subjectivity, interests, and emotions of the communities involved. Throughout the chapter, I develop an argument, in part by gradually unpacking Wildavsky’s (1992) proposal and logic, that language is often a proxy for a range of nonlinguistic social issues, and that LP is often an attempt at managing social diversity in the world. In 1992 as now, and in the United States as in Australia (the two contexts I use as central illustrations in this chapter), discussions of public policy pertaining to language and language in education have been framed as revolving around four serious social cohesion-eroding problems: student underperformance, student low self-esteem, political fragmentation, and minority oppression. As felt already in Wildavsky (1992), language policy and planning arguments are often characterized by a tension between minority population language needs and universal language study entitlements. In opposition to this, in this chapter I assume that a fundamental expectation in all justice-aspiring societies is for all children being schooled everywhere to experience serious, sustained, and supported language education that fosters the achievement of proficiency bilingualism. Universal language study entitlements, by comparison, tend to represent a limited-ambition bilingualism for elites. I also submit as a fundamental insight for LP that language teaching itself is a policy-making activity, because teachers influence the linguistic repertoires and attitudes of learners. Finally, it must be clearly understood that language policy processes are available to those who aim to restrict and contain bilingualism as much as to those who work to advance pluralistic multilingual agendas. Participating in LP debates and LP making is akin to protagonist voices of insiders and outsiders struggling to speak. How much influence different constituencies can have in shaping convincing arguments in support of rights-based bilingualism and pluralistic multilingual worldviews depends on how well they prepare themselves to advocate effectively for bilingualism, prior to organized participating in policy-changing debates. At the end of the chapter, therefore, I suggest some guiding questions that can aid this advocacy capacity building.

2 Policy and Interests: Australian Arguments The Journal of Policy Analysis and Management where Wildavsky (1992) published his piece is the oldest journal in the field of public policy, and since its establishment in 1981 has been a prestigious academic publishing outlet. As of the time of this writing, it is ranked second in the Thomson and Reuters Journal Citation Reports list of 46 journals under the category of Public Administration. Then as now (early 2018), it contains material on how public policies are made, how public resources are managed and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

154

JOSEPH LO BIANCO

evaluated, and how government can be more efficient and effective. These are important topics and it should not be surprising that a controversial issue in language education, how to educate Spanish-speaking immigrant children in US schools, would appear in such a journal, nor that a major figure in policy analytical studies would author it. Yet it was and remains rare for academic policy analysts to discuss any language topic, or for this or any other policy journals to carry such analysis. As a result, policy arguments concerning languages have overwhelmingly been left to insiders: scholars and teachers of languages, members of language academies, SLPs, interpreters, translators, and others among the professions focused on detaching language from its embeddness in communication and on analyzing it; and, of course, the specialized community of scholars in LP with its dedicated outlets, including the journal Language Problems and Language Planning, founded in 1977, and the newer but now wellestablished journals Current Issues in Language Planning, since 2000, and Language Policy, since 2002. Communities of speakers of languages are also often considered insiders rightly concerned with policy arguments pertaining to languages. In an analysis of the same dynamic in Australia as that signaled by Wildavsky (1992) in the United States, Lo Bianco and Aliani (2013) identified the voices in public debate about languages there, and they tracked the typical arguments these voices make and the interests they represent. It is important to appreciate that, in public policy circles, the term interests defines the core around which political constituencies coalesce. Politics is often understood as little more than an arena in which competing interests (Benditt, 1975) meet each other, contest, compromise, prevail, or lose, only to return for another battle. In other words, interests are the “main component of political motivation” (Rebenstorf, 2004, p. 89). In the Australian example, we have called these interests protagonist voices (Lo Bianco & Aliani, 2013) and identified a group of relatively stable ones that have contested the past 35 years of language policy discussion. Of course language debates do not actually appear to be very stable, and in some ways they vary greatly, reflecting different geographic and demographic patterns in different parts of Australia, changing events, and the relative openness or closure of different educational jurisdictions to innovation in language education. The many practical changes in policy priorities, programs, enrollment targets, and funding schemes also suggest fragmentation rather than stability. And yet, underlying all this surface dynamism, at the rhetorical-argumentative level, the landscape has been far less changing. Our research has uncovered a stable configuration of four broad protagonist voices that have influenced the focus of language education policy, sometimes in concert, at other times in conflict with each other (Lo Bianco & Aliani, 2013). These tend to advocate four broadly predictable emphases

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

155

for language policy, often in contradiction to each other, and occasionally in coordinated alliance together, as follows: (1) Language professionals, comprising language teacher associations, members of learned academies (both the humanities and sciences), organizations of interpreters and translators, and language academics, but also medical and legal personnel concerned with the delivery of specialist services to language minority populations, or even more widely, such as SLPs (see Hammer & Edmonds, Chapter 19, this volume). This language interest has mostly promoted universal second language study and wider language awareness in schools and expanded provision in post-school settings. (2) Immigrant community organizations, who have promoted what in North America are usually termed heritage languages and in Australia are called community languages. They have advanced the case for mother tongue maintenance, though often within a wider second language policy for all; parents who enact family language policies of bilingual parenting also fit into this category (see Gonc¸alves, Chapter 3, this volume). (3) Indigenous community organizations, which have focused attention on language revival, transitional and maintenance bilingual education, and nonstandard Englishes, and occasionally have advocated constitutional recognition for the original Australian languages. (4) Diplomatic, trade, and security elites, the most powerful group, who have made the case for languages in what is framed as the national interest. They aim to lend their claim priority over competing claims on policy and have often adopted the air of non-sectional advocacy, and have tended to focus on law and governance areas. These four protagonist voices have dominated public agitation about how much language education there should be, what language learning should be for, in whose interests it should operate, what should be funded, and how and to whom programs should be delivered. This has lent stability to language debates in Australia despite some 67 official reports at the federal level alone over 35 years, among other shifts and turns in programing. As a result, discussion about bilingualism in Australia essentially constitutes one continuous policy conversation around levels and beneficiaries of public resourcing, curriculum priority, teacher education, and a struggle to represent so-called ordinary citizens. In their relationships, these groups have traversed similar rhetorical territory to Wildavsky (1992), especially the opposition between particularism (i.e., minority groups) and universalism (i.e., the nation-state). Australian diplomatic, trade, and security elite voices who enter language debates frequently make reference to a similar dichotomy, invoking trade, commerce, and national security problems in preference to the wishes of minority groups. Variously represented by “captains of industry,” military

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

156

JOSEPH LO BIANCO

personnel and former diplomats, this voice is united in advocating for “Asia,” or what Lo Bianco and Slaughter (2016) have called The Australian Asia Project. “Asia” (or “the region”) in Australian politics is a wide-reaching claim for a reorientation of the entire society toward the countries of Australia’s geography, in preference to Europe or the United States. In this context, language education and curriculum reform are linked to a kind of national cultural and economic reconstruction. In language teaching, this means giving priority to key regional languages of trade, commerce, and security – especially Chinese, Japanese, Korean, and Indonesian, and occasionally also Hindi. The Australian Asia Project also makes a claim on the entire curriculum, insisting on inclusion of Asian perspectives across all learning areas, such as geography, history, or social studies. The shorthand for this collection of demands is “Asia literacy” and sometimes “Asia capability” and is advocated for all students at all levels of education and training. Engagement in language debates by such prominent nonspecialists has ensured that public discourse about language teaching in Australia has remained a regular feature of media coverage about general reforms to education, linked closely to a wider set of demands for national cultural transformation. This public discourse has also grown to become a staple of Australian public policy reform well beyond education.

3 The “Running Sore” of Bilingualism, and Other Inflammations However, as with the name and connotation of the term bilingualism in US debates, Australian policy interests differ radically from each other in how they name the focus of what is to be done in language education, and what bilingualism means. This struggle of names is part of a tussle between insiders and outsiders about who is to prevail in setting the course of language education policy. It is important to recognize that there is more to disagreements about what to call the problem that policy should focus on than mere differences of perspective or the different social positions of participants. In fact, the names, categories, and aims for language education that insiders use to talk about languages, with their specialist conceptual distinctions and research evidence, compared to those of outsiders to the education system, involve a struggle for different kinds of language planning. Policy specialists and elites, when they do focus on language questions, tend to position themselves as being concerned with so-called national problems, to meet the alleged national interest, and to achieve purported universal goals. This kind of discourse is meant to contrast with what they consider to be narrow, sectional, or sectarian interests pursued by ethnic or indigenous communities, or with the academic interests of language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

157

professionals pleading for special support for their specific areas. In effect, these outsiders consider that their status as non-insiders is an advantage in policy discussions, lending them objectivity and impartiality, exactly what Wildavsky (1992) describes as “universalistic” solutions to “particularistic” problems. An important aspect of Wildavsky’s (1992) argument is based on tackling what he considers the serious political problem of bilingualism, which he describes as a “running sore” (p. 311). His piece begins ominously: “Among the perennial problems of our time, sure to cause consternation and heartburn no matter what is done, is bilingualism” (p. 310). Nothing about how these “perennial problems” are expressed relates to language learning as such; instead, language here is a proxy for a range of nonlinguistic social issues. In this representation, opponents of bilingualism believe that what advocates of bilingual education really want is publicly funded “reverse discrimination” (p. 311), so the question of providing (mostly transitional) Spanish-medium instruction for minority students in US public schools is invoked as something much larger in society. It is evident that there is a lack of congruence between the essential concepts related to “two-language education” operationalized by insiders and outsiders, suggesting no agreement about the basic entity to tackle in policy. The significance of this dissonance for LP-making processes and debates cannot be sufficiently underscored. To understand this dissonance, it is relevant to look at the collocations insiders and outsiders link with the phenomenon of bilingualism. In Wildavsky’s piece, bilingualism is part of the flow of civil activist causes from past decades: It is [. . .] symptomatic of a number of elemental conflicts based on different conceptions of equality that constitute running sores in American public life. We pick at them but we do not resolve them. The particularism (treating people on the basis of who they are rather than what they do) that results from adherence to greater equality of condition is at war with the universalism that its critics say maintains an inegalitarian status quo. Everyone is angry but no one is persuaded. (Wildavsky, 1992, p. 311)

As his argument develops, Wildavsky contrasts two contending rhetorical forces, unilingualism and bilingualism, whereby essentially these are names for two policy positions that are contesting the management of the public good of education. He claims that “each side perceives the other’s policy preferences to have twin adverse effects” (p. 311). For unilinguists, bilingualism “detracts from the unity of the American nation” and risks “fragmenting the culture” (p. 311). For bilingualists, by contrast, it is unilingualism that has adverse effects, those of “oppressing the Spanish-speaking community” and lowering “student performance, including generating lower self-esteem” (p. 311).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

158

JOSEPH LO BIANCO

This reasoning highlights a distinctive phase and an important aspect of the general policy-making process, what I call the impasse. Policy advocates often identify, or look for, an impasse as a way to suggest their preferred breakthrough strategy. An impasse occurs when, despite discussion and debate, representatives of different interests in a debate have talked themselves into gridlock, their arguments and contending positions are blocking progress, and a third party is able to interpret this as a potentially fruitful opportunity for a new direction. To break through the gridlock requires some kind of circuit breaker, such as a radically new policy measure. In relation to bilingualism, the rhetorical roadblock is what unilinguists and bilingualists each allege are the adverse effects of the other party’s position, and because progress is not possible and the issue is becoming an “elemental conflict” or more dramatically “a running sore” (p. 311), a third party can represent the stalemate as dangerously corrosive for social cohesion. In the present case of Wildavsky (1992), and indeed the two decades preceding his article, in the same stable discursive format noted earlier for Australia, four serious social cohesion-eroding problems are identified. These are: student underperformance, student low self-esteem, political fragmentation, and minority oppression. Hence both unilingualism and bilingualism need to be cast aside and allow an innovation to break through.

4 Solving the Problem by Missing the Point The circuit-breaking innovation Wildavsky (1992) puts forward is the principle of a universal second language requirement. In an attempt to chart new discursive territory, he claims the requirement will have additional positive benefits for the nation by making “American education [. . .] tougher and more demanding” (p. 311). The requirement is therefore promoted as a solution to two language problems: “English-speaking students who need the advantage of a second language and Spanish-speaking students who need to view English as an equal opportunity language,” both of which would be “ameliorated in a way that adds to but does not subtract from the educational opportunities of other students” (p. 313). This formulation is a classic device of professional policy analysis that conducts risk analysis and cost-benefit calculations to avoid unintended negative consequences from a policy innovation. Wildavsky’s (1992) article shares the prevailing political concern that transitional bilingual education is a form of segregation, even if benevolent, and should be rejected in the overriding interests of the nation by promoting an “integrated society” (p. 314). Here we see a continuity with past discourse hostile to bilingualism by opponents of standard academic views of two-language proficiency. By the late 1970s, claims of minority

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

159

advocates and academics in favor of transitional or maintenance bilingual education for educationally disadvantaged ethnic minority children were increasingly repudiated by critics and gradually came to be incorporated into the bitter cultural literacy politics and culture wars of the 1980s in the United States. In this process two-language teaching came to be called “affirmative ethnicity” (Epstein, 1977). It is worth tracing briefly this trajectory of poisoned discourse as a premise to distinguishing between the policy meanings of terms and their specialist meanings. Critics of bilingual education were careful to distinguish foreign language education from immigrant or indigenous language maintenance. Foreign or prestige language teaching was conceptually separated and granted various kinds of positive recognition, and this was in turn conferred on students, so that the effort expended by learners of some languages was praised, bringing personal credit to them for skill accumulation. These languages were associated with the national interest and this too elicited praise to schools and students for engaging in nation-assisting activity. This separation clearly exposes both ethnic and racial prejudice, since ethnic minority children’s equally arduous linguistic trajectory to acquire English is denied similar praise for intellectual effort and citizenship virtue. This prejudice operates as a distinction between educationally admired and nation-supporting foreign language study (see Mun˜oz & Spada, Chapter 12, this volume, on foreign language provisions), on the one hand, and language maintenance, on the other. The ethnic and racial prejudice is generated through a social ranking order of bilingualisms, but also through a ranking of speaker populations (see Fuller, Chapter 6, this volume). This double standard is encountered in many contexts in the world. In policy terms, it marks support for the efforts of majority students as an acceptable educational investment, and the support for minority students as largesse, potentially divisive ethnicity promotion, and ultimately as threatening national fragmentation (see also Bigelow & Collins, Chapter 2, this volume, and Garcı´a & Tupas, Chapter 20, this volume, for related discussion). Returning to the Australian case, one characteristic of the same dynamics has been occasional wide-ranging alliances that have attempted and succeeded in fracturing this damaging dynamic. These alliances have formed across the four contending language policy interests, the four protagonist voices. They have permitted coordinated pressure on government and have led to the production of a comprehensive policy that embraces both minority population language needs and universal language study entitlements. These alliances have achieved considerable breakthroughs in language education but have succumbed to pressure, unraveling once these coalitions of interest were undermined. The political dynamic of unraveling coalitions is a subject for political anthropology, but relevant for bilingualism studies too, since the many

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

160

JOSEPH LO BIANCO

kinds of language sociologies in contemporary society require alliances across various interests to support and advance policy. Students of what is known as factionalism in political anthropology (e.g., Silverman & Salisbury, 1977; also Boissevain, 2013) have identified various types of erosion or breakup of previously organized structures or parties. Factionalism in policy struggles arises when alliances need to forge compromises. Since no one interest can totally prevail, once alliances are formed they become subject to inside and outside pressure, and they are often undermined when more powerful interests within the coalition perceive greater benefit in pursuing a separate agenda, or when administrators, funding authorities, or politicians interfere to promote comprador tendencies within the coalition, buying off one or more parties with separate rewards. This is possible because within language policy interests, as with any political constituency, the grouping of voices is heterogeneous and never completely internally uniform. Interests overlap and not all the components of each category align with the entire interest of that category. Often there are subgroups that are connected to other interests, either for short-term collaboration or in a longer-term way. Language professionals, for example, form a category that includes language teachers. But they teach many different languages – heritage, foreign, prestige languages. They teach these languages in different education sectors, in different kinds of programs, with different methodologies; and some identify with a native speech community, such as a native speaker teacher who might identify with the source community, compared to a nonnative speaker whose identification with the language might be limited to professional or scholarly interests. Academic linguists are similarly divided with potential factions related to research methodological preferences, publishing outlets, disciplinary allegiances, and so on. Translators and interpreters are also a highly differentiated subgrouping among language professionals, with differences arising from the specialist domain that they work with (health, legal, conference, etc.) or due to often personal and familial ethnic connections or more distant professional kinds of identification. When their work is with community groups strongly linked to ethnic and indigenous populations, interpreters and translators tend to operate with a community-based reasoning for the bilingualism they advocate. As we have seen with the Wildavsky example, policy is often premised on the underlying kinds of reasoning that diverse kinds of insider and outsider groups adhere to. The heterogeneous grouping of language professionals includes strands closer to professional/institutional settings and others located in community life. Within the language professional interest, a further dissolving tendency is whether individuals teach and research languages that are legitimized and promoted by trade, diplomacy, and security interests and favored by policy reports and government programs, or whether they teach in language maintenance programs of immigrant communities or language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

161

recovery activities of indigenous communities. In the Australian experience, this was a particularly sharp source of division, used by education administrators and government officials to fragment previously stable alliances that had favored inclusive language policy. The alliance had been led largely by ethnic minority political activism, which had been powerful enough in electoral terms to command political interest (Ozolins, 1993), but over time it came to be resented by some in government who wanted to direct language education toward commerce and trade-supporting activity rather than heritage language support. Just like Wildavsky’s (1992) universal second language requirement, proposals for universal second language teaching and learning have been made in many settings (see, e.g., Clyne, 2005). In the United Kingdom, these have often been termed an entitlement, meaning that all learners should be guaranteed a minimum amount of quality second language study. Language education provision of this kind is simply the norm in most non-English-speaking countries, but it is a regular point of debate in many English-dominant societies because of the failure of such societies to deliver what should be a fundamental expectation of all schooling everywhere: that all children experience serious, sustained, and supported language education that fosters the achievement of proficient bilingualism. By contrast, the bilingualism envisaged in schemes put forward in the United States or Australia as compromise positions to deflect attention away from lobbying for comprehensive and rights-based language policies tends to be a limited ambition bilingualism, essentially meaning the universal experience of language study. The Wildavsky requirement that promotes universal language study is valuable and offers some benefits, but it misses the point by missing the specificity of the educational needs of language-minoritized populations and fails to recognize the impact of the different language sociologies involved. No general universal entitlement to general language study comes close to responding to the specific sociolinguistic needs and language rights of minority populations, such as the threatened intergenerational vitality of their languages, their marginalized social status, and the frequently prejudicial attitudes of institutions to their social presence (Lo Bianco, 2014; see also Fuller, Chapter 6, this volume, for discussion of language ideologies that impinge on these issues). The year following Wildavsky’s (1992) proposal, the Journal of Policy Analysis and Management published a repudiating response from a language planning specialist, Susan Dicker. Dicker (1993) declared Wildavsky’s second language requirement for elementary public schools “an inadequate substitute for bilingual education” (p. 779), and, though lauding its general goal of second language study, pointed out that Wildavsky “demonstrates a significant lack of knowledge about the most effective methods of second language acquisition for both Englishspeaking and minority-language children” (p. 779).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

162

JOSEPH LO BIANCO

5 Knowledge(s) and Power . . . and Problematic Problems As I have been arguing, the dispute between minority population language needs and universal language study entitlements encapsulates some of the most important elements in public policy making. The central issue is often expressed in the binary formulation of knowledge versus power, essentially and crudely, that those who we might think of as administrators have power but little knowledge, while those who we might think of as practitioners have knowledge but little power. Several dynamics contained within this binary separation are now explored in this section and the next, in order to probe some of the ways that expertise (knowledge) is linked to policy (power), and specifically how different kinds of knowledge about bilingualism and the practices of language planning and policy making are connected. In exploring links between expertise and policy, knowledge and power, I draw from key poststructuralist and knowledgeutilization scholars in the wider field of public policy. Knowledge comes in different forms, such as practitioner reflections, professional experience, and research findings and wisdom, among others. Knowledge is organized, packaged, and presented in diverse ways, such as statistically presented data, persuasive arguments, personal narratives, or timely action-linked advice. These different manifestations of knowledge, with their different statuses, are enveloped in different discourses and framed around different constructions of what is to be the problem to tackle in policy, and the likely consequences and the priority or merits of doing so. In her work on how policy operates, Bacchi (e.g., 2009) focuses attention on how problems are represented in policy and she problematizes the assumption, increasingly popular today, that policy should or could be exclusively based on evidence. What is being critiqued is not evidence or supportive data as such, but the view that evidence can stand apart from the preferences, interpretations, and desires of citizens, at least in areas of policy related to human lives where subjectivity, interests, emotions, and communities are involved. Instead, by reviewing a large number of actual policy processes, Bacchi is able to show that policy formulation is a kind of meaning-making practice; in this Bacchi is extending a longer repudiation of policy as technique among public policy scholars, which is an important focus for this discussion (e.g., Majone, 1989; Weiss, 1983).

6 Persuasion We know, intuitively, in lived experience and through study, that the binary opposition that positions policy makers and administrators as unknowing powerholders and practitioners as knowledgeable impotents

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

163

is crude and inaccurate. Yet the question of power and knowledge has been the subject of philosophical reflection from Plato through Machiavelli to modern students of the political process. In a significant installment in turning public policy away from a highly technical and procedural approach, Majone (1989) showed that science and public policy are each oriented toward fundamentally different notions of argumentation. Academic research and public decisionmaking rely on different forms of reasoning. While science practices favor tests of falsifiability and legal settings look mostly for forms of robust evidence, in politics, the main rationality that underlies action is what is pragmatically achievable or which is most publicly accountable. Majone summarizes these three orientations to knowledge as evidence, argument, and persuasion. It is true to say that, even today, most approaches to training in policy analysis in mainstream institutions, and the practice of most policy work, rely on various technical procedures, such as options comparisons, cost-benefit analyses, etc. However, with increasing attention to the role of language (as rhetoric in the main), a shift away from exclusive focus on managerial operations toward an appreciation of the role of communication is occurring. Like Bacchi’s (2009) later focus on how problems are represented in policy, Majone’s approach “has less to do with formal techniques of problem solving than with the process of argument” (1989, p. 7). This shift within the policy sciences is an important dimension of how language advocates should think about impacting policy. Seeing policy making as essentially a kind of argument and narrative describes a large part of how policy making actually happens, and it has the additional benefit of validating citizenship participation. For language teachers promoting the value of bilingualism, this approach to policy also offers support for the role of practitioner knowledge as a valuable element of policy. In effect we can see language teaching as policymaking activity (Lo Bianco, 2010), through which teachers influence the linguistic repertoires and attitudes of learners. The language classroom is a “container” of language-shaping activity, whose verbal interactions are hierarchically organized so that teachers structure what it is possible for learners to say, to whom, and how. Teachers also influence, implicitly through behavior, and explicitly through verbal marking, attitudes that accompany the future language uses and choices of learners (see Simpson, Chapter 13, this volume, and Garcı´a & Tupas, Chapter 20, this volume). These ideas about policy as argument also apply more generally to the political jurisdiction in which formal policy making occurs, which can also be productively understood as an essentially verbal interaction. Here the knowledge–power dichotomy is clearly too simple. Based on a series of studies of actual policy-making processes, Weiss (1983) has shown how the participants in these verbal policy-making encounters, typically

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

164

JOSEPH LO BIANCO

administrators or politicians, community advocates, and academic researchers, all bring unique and therefore different stocks of information, ideology, and interests to their interactions. Each group has its own knowledge stocks and its own forms of power, as well as differential interests that align and misalign according to different issues and at different times. A professional’s picture of what bilingualism should actually be conceived to be, in policy-making settings, therefore involves active and informed participation in such verbal interaction, marshalling persuasive language directed at modifying the stocks of knowledge of the counterparts, aware of and shaped by knowledge of their interests. Some kinds of researcher knowledge and evidence can modify the presuppositions of the formally positioned policy makers; other kinds are unlikely to because of misalignment of interests, information, and ideologies.

7 Arguing for Bilingualism: Policy as Argument Wildavsky’s (1992) argument of bilingualism as an activity of minority grievance constitutes the central policy and political problem of bilingualism as a threat to civil harmony arising from the social status of minoritized languages (e.g., Spanish in the United States), a threat that needs to be deflected into and generalized as foreign language study in a process of nation-supporting skill accumulation. His aim was that education in languages should be seen to produce citizen subjects who are understandable according to current views of economic need, traced to human capital theorizations of entrepreneurial individuals, with attributes of selfimprovement, ready to enter competitive marketplaces in search of employment. Prior to his intervention, there had been two decades of construction of the question of how the linguistic socialization of immigrant children to US schooling was to occur. Within the celebrated culture wars of the 1980s and 1990s, this item of rather specialized education was taken up by the key exponent of cultural literacy discourse, E. D. Hirsch (1988). In dramatic prose, he argued that linguistic pluralism “enormously increases cultural fragmentation, civil antagonism, illiteracy, and economic-technological ineffectualness” (p. 91). How in-class pedagogical responses to out-of-class linguistic diversity is to be managed is here transformed from pedagogical or programmatic solutions to a series of society-threatening nominal categories. Similarly, Bernstein (1994) discusses a proposal for a Cherokee program in a US public school as an “act of rebellion against white, Anglo-cultural domination” with a “multicultural animus against European culture and its derivatives” (p. 245). Written after Wildavsky’s intervention, Bernstein’s argument shows the enduring power of policy voices.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

165

8 The Value (and Risks) of Becoming a (Public Policy) Problem In a significant article, Ruiz (2010) reflects on how his well-known typology of three implicit and recurring orientations to language in policy (as a resource, a problem, or a right), has fared in later bilingualism discussions. In his original work, Ruiz had advocated that language should be recognized as a resource, so that mother tongues other than English should be supported as an intellectual and cultural asset for learners. In his 2010 reflections, he shows how the language as a resource orientation proved a rather problematic construct, which was severely criticized for aligning with a neoconservative political agenda in which social goods are regarded as valuable only by economic or military utility. According to Petrovic (2005), language scholars or teachers should not support minority languages using a resource approach, which he sees as a strategic mistake. In his view, the language-as-resource argument was exploited during the 1990s by reactionary political interests “to preserve the inequitable” (p. 395). Petrovic warns that if researchers and teachers are too successful in convincing policy makers that minority languages are a resource, this will flow into policy as a language-as-tool simplification and displace community rights or intrinsic values for minority languages with narrow instrumentalism, because the language as resource orientation in policy “is, in large part, an economic defense of minority language maintenance” (p. 397; see Ricento, 2005, for a similar critique). Ruiz (2010) concedes there are risks, but he shows convincingly that language policy advocacy inevitably and always requires educators to engage with the categories, names, and understandings of nonspecialists, and he highlights several cases where language as a resource rationale achieved positive breakthroughs expanding language learning opportunities without sacrificing language rights for minority groups.

9 Toward a Better Model of Language Planning and Policy for Bilingualism For a brief period in its early existence during the 1970s, the field of language policy and planning believed that the then emerging domain would grow into a unitary academic discipline, with singular aims and purposes. It is clear today that no such aspiration is realistic. Language planning is engaged in by all groups in a society, and the methods of policy making in languages are used by interests with diverse and sometimes contradictory aims, whether they are consciously aware of doing so or not. The language learning contexts investigated in language planning and policy studies are highly diverse and intersect with bilingual interests in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

166

JOSEPH LO BIANCO

many ways. A useful research topic that should be developed in the future is to account for all the ways that teachers of languages, language researchers, and communities of speakers are positioned in policy debates. The teaching and learning of languages in formal education is only one of the contexts that language planners investigate, and teachers and researchers cannot assume any particular bilingual learning outcomes are intended in any setting targeted by policy makers. This is because language policy processes are available to those who aim to restrict and contain bilingualism, just as they are to those who work to advance pluralistic multilingual agendas. While education is a major focus of most language policy, not all language development or language problems are located in educational domains, so the special needs of hospitals and legal contexts, of citizenship determination tribunals, and a myriad of other administrative operations rely on language decisions or make use of language information. (For in-depth discussion of the legal domain, see Angermeyer, Chapter 7, this volume; see Simpson, Chapter 13, this volume, for discussion of language testing for citizenship and other benefits.) One of the key motivators for language planning is the social phenomenon of individual unlearning (Keijzer & de Bot, Chapter 14, this volume, and Ko¨pke, Chapter 18, this volume). This can be understood as collective unlearning, and then it is treated in language planning studies as intergenerational language loss, attrition, and language death. The aim is typically to reverse language loss (Fishman, 2001). Language planning processes, methods, and concepts are best seen as tools available for people in language-related disciplines (teaching and research for the most part) to engage with the wider sociolegal and political context of their teaching and learning, and to influence and shape it to promote inclusive and pluralistic acceptance of bilingualism, language rights, and opportunities. In this space they will interact with others with similar views, and many whose views and intentions are the opposite. Language planning processes are available to all of these interests and purposes. The forms that language policy may take will differ extensively according to the sociolegal environment of the container in which it occurs. A language law in one setting might be the equivalent of a set of public and inherited norms of practices in another. Spolsky (2012) describes the diversity of ways in which LP is possible as the “language policy of a speech community” and identifies three components of this: (i) practices, the habits and language usage patterns from among the languages or language varieties available in the linguistic repertoire of a community; (ii) values, the beliefs and ideologies which hold in a given environment about the linguistic repertoire; and (iii) management, the deliberate interventions made through formal policy, regulations, or laws in an individual setting. It varies considerably therefore how in different political jurisdictions, containers of government, but also in families, classrooms, and schools,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

167

people determine which languages and varieties individuals will be educated in and use within a given society or community. The actions and processes for making decisions in these different containers encompass formal, official governmental action, all the way through to informal norms, social pressure, encouragement, and teacher classroom modeling or management of the learning and attitudes of students. My view of language policy and planning is related to but varies from Spolsky’s (2012) model and is expressed as LP = A3 x P4 x G6. That is, LP is conducted under three kinds of Authority (containers), or rather three forms of power, with four modes of Participation, to achieve six kinds of language Goals. In observations of language policy, I have identified three broad power and authority sources (A3) for LP activity. These are, first, the source of sovereignty. This refers to the actions of political agents who work with their exclusive political authority in their setting, varying according to tradition and political regime, to direct state law toward some aspect of language. The second kind of authority in LP activity is jurisdiction. This refers to LP conducted through devolved authority, so that an education system or a municipal council has the authority to make language decisions within the constraints that the politically sovereign authority has allocated. Third is the realm of influence. Here political actors, who could be teachers of languages, or researchers, on their own as citizens or in alliances and coalitions, aim to influence policy. This is done in several ways. These individuals or groups can operate legitimately within the decisionmaking systems of the sovereign state as, say, citizens of the United States or Zambia. It is also available to these groups to act transgressively or subversively and to undermine legitimate political authority. In each of these ways, by being a sovereign, or by acting under delegation, or by exercising influence and pressure, LP can achieve the backing of executive authority. All involve the exercise of power, but influence is a means of changing the existing relations of power, through mobilizing knowledge, or force, in economic markets or in political and cultural domains of different kinds. My research has also found that people typically participate in these three kinds of authority in four characteristic ways of participation (P4). The first is to shape public texts, usually done by politicians and administrators in formal settings; second, LP can be shaped through public discourse, so that argument in public life, in the media, and in meetings of school boards can persuade authority holders to adopt a change or to resist another one; the third mode of participation involves the performance of people with language influence, especially teachers in the acts of teaching, but also cultural agents such as celebrities, whose models of language shape public usage through how their followers emulate their models; and finally, participation can happen through expert-led deliberation, in the rare cases where facilitation by researchers is allowed to shape and direct decision-making on the basis of research findings, but also through evaluations that impact practice.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

168

JOSEPH LO BIANCO

These kinds of authority and power, A3, and these kinds of participation, P4, are directed toward six characteristic language planning goals, G6, all six interrelated and connected, such that the names represent points along a continuum or a network. The most common in the literature on language planning is the status of languages, usually in law or constitutions, such as the postapartheid constitution of South Africa, which has possibly the world’s most extensive statements on language, or that of the Republic of Ireland or New Zealand, both explicitly including support for Irish and Ma¯ori, respectively, as official and national languages, despite being both dominated by English. Second is the corpus of languages, their body of words, spelling conventions, and so on. Often this is more technical work taken forward by linguists who assist an indigenous community to develop a writing system for a language so that it can be extended from home or village use to be included in schooling, or it can be the high-level pronouncements of official language academies, such as the French or Spanish language academies, making rulings on spelling norms or points of grammar, or which might seek to ban foreign words in print. Third is teaching and learning, and especially of the literate and advanced levels of language use, often delegated to education ministries who develop curricula, or taking the form of policy reports that aim to introduce new languages in schools or create schools to train interpreters for a newly emergent need. These three goals are the most common, but three others can be identified by studying cases of language planning and policy making. One is the usage of languages. Usage planning involves research to determine in which domains individual languages are encountered, permitted, and naturally present, and from which ones are they absent. On the basis of this documentation, usage language planning can take a series of measures to encourage speakers of a given language to extend the settings in which they use their language, for example, pushing it to be present in business or technology or higher education. The fifth goal of language policy is prestige, which involves many kinds of cultural activity aimed at raising feelings of regard, admiration, or esteem for a dialect or language, and often of contesting negative ways in which it is depicted. The last or sixth common goal, one not often recognized, is discourse, that is, how different activities of language planning aim to change existing attitudes toward a language, or to influence the views that people have about particular issues through changing how they use language about that topic. The most extreme case of discourse planning is the brainwashing that totalitarian states undertake, such as depicted in the 1949 novel Nineteen Eighty-Four by George Orwell, in which a political party is constantly brainwashing citizens. The aim of the Party is to gain universal love and loyalty to Big Brother; starting with the young, and working through spies, the Party manipulates the meanings of words and expressions to ensure that the thoughts of the citizens are controlled. Most discourse planning is not so extreme, but advertisers, political parties,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

169

and social reform movements, such as anti-racist or feminist discourse, operate to change past negative meanings by isolating prejudiced language and replacing it with more inclusive talk (see Fuller, Chapter 6, this volume, for studies of how ideologies shape and are shaped by discourses). The depictions I have used to account for the dynamic of policy and bilingualism in this chapter are only two, Australian and US, and uneven between these. Each regime of language (Kroskrity, 2000) is a particular container, a distillation of legal, political, and sociolinguistic features, making available different renderings of bilingualism and its possibilities in public policy. Even between these two examples, there are major differences. In every setting, there is a dominant language constellation (Aronin & Singleton, 2012), that is, a societal-level grouping of languages and their relative status, and therefore these are contingent on the particular histories that prevail there – histories of migration, settlement, indigeneity, colonialism, international relations, within the particular sociolegal envelope of the communication setting. In the United States, the dominant language constellation is not only a grouping of materially and socially organized languages and their communities, and the relationship among these, in legal and socioeconomic terms, but also the symbolic presence and material life of these languages. This US constellation is a radically different one from what prevails in Australia. For example, Spanish, and several kinds of Spanish, is a majorityminority presence in many but not all parts of the national territory of the United States whose names prevail in the toponyms of some territories of the country. Nowhere in Australia is there any such majority-minority presence and non-English place names are more haphazardly distributed. Indigenous language sociologies are radically different between the dominant language constellation found in the United States and the Australian counterpart, and the immigrant historical experience is only similar in a very broad way. The two countries have different proximal language experiences as well, due to their radically different international profiles: Australia has long been pursuing an energetic cross-political project of regionalism while the United States effectively exercises regional hegemony, with its proximal languages mostly being Spanish. Also deeply different are the national styles of policy making. What is shared is dominant English at a time of global English, and a common history of LP essentially enshrining English first or English only. How and if languages are taken up into public policy relates to the persuasive rhetorical capacity of professional language interests to fashion productive alliances and coalitions with others. These potential partners include other language interests, community groups, language professionals in noneducational areas, such as SLPs, translators, librarians, and so on, and strategic nonlanguage interests. Once alliances are formed, they must use the sociolegal instruments of each setting, the forms of political

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

170

JOSEPH LO BIANCO

power available and relevant in a particular dominant language constellation in a given society. The actual work typically follows the pattern of LP = A3 x P4 x G6, the forms of power and authority in each setting are the ones that must be targeted to make change, the kinds of participation are the ones that the institutions of that society allows, and the goals tend to respond to the specific language problems the alliance is seeking to address. The case to advance a more equitable bilingualism is almost universal. However, most societies privilege one or at most a few languages, and favor only a minority of prestige foreign languages in their education, and tend to provide minimal or no support in educational, medical, and legal settings for minority populations. Redressing inequities of these kinds requires concerted lobbying and persuasion for a rights-based language policy, defending a general case for universal language rights, while asserting the practical importance of language rights in the institutional power discourses relevant to a specific society (Skutnabb-Kangas & Phillipson, 2016). What is most crucial to understand is that the absence of formal policy on bilingualism does not mean the absence of language policy in a more general sense. The absence of policy when different languages are positioned unequally is in effect a longer-term, silenced, unplanned attrition of all languages other than the dominant one, or at least a permanent hierarchical inequality (cf. Fuller, Chapter 6, this volume). Whether this constitutes a problem or issue depends on the values and views of various players, positioned as they are close or far to the consequences of these societal linguistic regimes.

10 Conclusion Changing, influencing, and collectively shaping the direction of how the concept of bilingualism in a given society or institution is understood is a necessary role, both practical and intellectual, for language professionals, to counteract the historical legacies of official normalizing of monolingualism. Drawing from my decades-long work facilitating LP in diverse and conflict-ridden contexts, some points of advice could be the ones offered in what follows, formulated as questions for those who want to advocate for recognition of bilingualism and support for minoritylanguage communities, prior to organized participation in policychanging debates; the list is not exhaustive but rather an indication of key areas and points: 1. What would be the aim and purpose of policy action on behalf of bilingualism? 2. Is there a constituency for bilingualism? 3. Could one be generated? Who would comprise it? What would be its limits?

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

Language Planning and Policies for Bilingualism

171

4. Are existing constituencies inclusive of all language interests, including dialectbased, bilingual and bimodal, indigenous and immigrant, and international languages? 5. Do they aim to change language attitudes and information in general in society, and therefore encompass interests in the widest sense imaginable in the given context? 6. What internal groupings would comprise such a constituency? 7. Would it cohere? Would it aim to be temporary only and exist only for limited, strategic purposes? 8. What alliances could compensate for its absence, or strengthen its traction? 9. Are existing alliances/groupings and constituencies sufficient to impact policy in particular jurisdictions? Society wide? 10. What are the risks of constituting one? 11. What challenges of understanding, knowledge, and strategy face such alliances? 12. What role would teacher/researcher voices play? What role would professional voices demand for themselves, and concede to nonprofessionals? 13. Are they consonant with each other? 14. Are professional associations sufficient, loud, or heard in the containers in which they wish to increase or gain traction? 15. What are the stocks of interests, information, and ideology each party espouses or holds? 16. What alliances and networks are present/absent with policy makers? How can these be created or strengthened, and for what purposes? 17. Would the alliance understand and acknowledge the position/power and interests of its antagonists? Bilingualism cannot enter policy-making environments without its history of names, attitudes, ideologies, or experiences. Teachers, researchers, and other language professionals should not imagine that they occupy a neutral or even a privileged space from which to name and constitute bilingualism as an object of public policy attention. Instead, it is important for all interested in engaging in language policy and planning to study closely the ideologies, interests, and information stocks of all participants in policy debates. This information must be linked to how the institution or setting being targeted for policy change operates, since each container uses and understands language in particular ways. In schools or ministries of education, hospitals or courts of law, language and language issues are present and challenging but in radically different ways. To prepare for policy action relevant to these spaces, language advocates, whatever their professional background, often need to reshape the inherited understandings of language. Language professionals and scholars have an obligation to enter the space of public policy making due to their intimate acquaintance with bilingualism, the benefits of bilingualism for intellectual and cultural functioning, and the many inequalities in material, symbolic, and participation arrangements of the current communication order. The process

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

172

JOSEPH LO BIANCO

of public policy making is complex and relies on persuasive argument and mobilizing knowledge in political action. Language policy and planning is a body of concepts and methods for understanding how planned language change occurs and should be studied by language scholars and professionals and their associations to forge productive alliances to promote and defend language interests and equitable bilingualism as a human right.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:09:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.009

9 The Economics of Bilingualism François Grin

1 Introduction Economics has long been regarded as having very little to do with language, and the very mention of “economics of language” or “language economics” sometimes still elicits surprise. Yet the economics of language has progressively made its mark, and the economic perspective on language (whether language is approached in terms of “bilingualism,” “multilingualism,” or “plurilingualism” – more on these terms in what follows) has earned recognition as a legitimate component of any well-rounded approach to language in society. The progressive emergence of an economic perspective on bilingualism and related notions reflects concerns that are not unlike those of sociologists and political scientists. In essence, taking language into account can help to explain influences affecting some key dependent variables in these disciplines, such as “social groups” for sociologists, “power” for political scientists, and “value” for economists. At the same time, however, and particularly in the case of economics, the approach can be put to use in the examination of the reciprocal causation, that is, to investigate the effect on language variables of economic processes. Since its beginnings in the mid1960s, the economics of language has gone through a slow but steady growth, and it has recently reached important milestones with the publication of two edited volumes of considerable scope, namely Ginsburgh and Weber (2016b) and Gazzola and Wickstro¨m (2016). This chapter’s main focus is on how economic research can elucidate the complex dynamics of economic (dis)advantages of individual (and sometimes, by aggregation, collective) bi- or multilingualism. It does so by focusing on the question of value, which is the core issue of economics in general, and is therefore central to language economics as well. The notion of advantages and disadvantages points to differences in value. In language economics, a very normal way of talking about value is to use language-based earnings differentials, showing how language can

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

174

FRANÇOIS GRIN

be an explanatory factor in the determination of an economic variable that is central to people’s experience, namely, their labor income. The complex dynamics, then, operate mainly through the evaluation of certain characteristics of people on the labor market. This chapter emphasizes the advantages derived from language skills since for most people, the value of those skills (as distinct from socioeconomic traits that may be correlated with having those skills) is positive, or at least nonnegative. Before proceeding further, it is useful to warn against a frequent confusion between “language economics” and “language in the economy.” The mere fact that some pattern of language use may be observed in a context such as “work,” which is deemed by some to justify the adjective “economic,” does not automatically give rise to a question of language economics. Such a question only arises if economic variables or economic processes are actually involved in the phenomenon being examined. However, this is not necessarily the case in several contributions in the sociolinguistic literature whose title may include the word “economics,” but that contain no language economics, because neither economic variables nor economic processes are addressed in them. What they offer is, rather, an ethnography of language at work, which is a worthwhile endeavor in its own right, but does not address economic issues as understood by economists. Thus, an economic analysis of bilingualism is quite different from a socioeconomically oriented commentary on the conditions experienced, in certain settings, by certain groups of people who possess certain traits, some of which are linguistic ones. The fact that a process with linguistic features unfolds in a setting such as the labor market does not in itself make it an economic process. The overview of the economics of bilingualism proposed in the following pages is organized as follows. In Section 2, I clarify a few terminological points, circumscribing the sense in which bilingualism is used in this entry, and recalling the distinction between multilingualism and plurilingualism. Section 3 contains a brief overview of the historical development of the economics of bilingualism, showing that practically all economic work on language is about bi-, pluri-, or multilingualism. Section 4 goes to the heart of language economics with a discussion of the economic concept of value and its application to language. Section 5 turns to empirical results on a selection of major topics in language economics. Section 6 discusses a particularly exciting range of applications of economics to language, namely, those that pertain to the selection, design, and evaluation of language policies. Section 7 concludes.

2 Conceptual Distinctions and Terminology In economics, the difference between “bi-” and “multi-” lingualism is essentially a matter of degree. Far more decisive, because the corresponding

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

175

Table 9.1 Terminology of linguistic diversity adopted in this chapter

Individual level (person) Group level (e.g., country)

One Language

More Than One Language

Monolingual or unilingual

Plurilingual Multilingual

implications for economic analysis are substantial, is the distinction between the individual and the collective levels. To an economist, the implications of knowing one, two, three, or more languages as a person belongs to one class of questions; the implications for a society of using one, two, or three languages – and how and when they are used – are fundamentally different. This raises terminological issues that are not fully settled, but that must be addressed here for the sake of clarity in the subsequent discussion. For this purpose, it is useful to introduce the term “plurilingualism,” a calque from the French plurilinguisme, just as French, which normally only uses the word plurilinguisme, has imported “multilingualism” from English. The two terms now cover, for most authors, two distinct semantic fields. French plurilinguisme stands for the plurality embedded in a person’s linguistic repertoire, with two or more languages. Someone who speaks several languages, therefore, is a plurilingual, as opposed to a uni- or monolingual. French multilinguisme applies to larger units of observation, such as a city, region, or country, in which many languages are used. Switzerland, for instance, is a multilingual country in at least two distinct senses: first, because its constitution recognizes four national languages used as a first language (L1) by the members of its four historical language communities; second, because an expanding range of languages may be heard in the streets of Swiss cities, encompassing not just national languages, but immigrant languages as well. But all this does not mean that the country’s residents themselves necessarily are plurilingual. This multiversus pluri- distinction is beginning to turn up in the specialist literature published in English, and it is used in this chapter as well. Table 9.1 provides a bird’s-eye view of the terminology of linguistic diversity adopted here. The terms bi-, tri-, quadrilingual, etc. are still used in the literature with reference to individuals or groups who know or use two, three, or four languages. Very little of the literature in language economics concerns language in abstracto or in general, that is, independently of its embodiment in a specific language, although a few contributions (e.g., Marschak, 1965; Rubinstein, 2000) propose economically inspired theories of the evolution or structures of language. Putting it in Saussurean terms, language economics, in the main, is about langue in its various manifestations (that is, a code defined by a set of common rules shared by a community and acquired by actors of that community, where communities differ), rather

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

176

FRANÇOIS GRIN

than about langage as a capacity to communicate through language. In other words still, language economics is about languages that differ from each other. Hence, practically all language economics is about pluriand multilingualism, which both encompass bilingualism (used for individuals or groups). The rest of this chapter uses the hyperonym “language economics” to refer to the field of study as a whole. Individual bilingualism or plurilingualism can be the product of any type of personal biography: some people are familiar with two languages or more as a result of their own migration, or of migration by a previous generation, followed by more or less systematic transmission of a “heritage” language. Others have acquired their language skills through classical, formal language study, whether in traditional forms of school instruction or thanks to extended stays in a region or country where a language other than their L1 dominates. In all these cases, informal and non-formal modes of learning may have played an important role. It is important, however, to disentangle linguistic repertoires on the one hand from socioeconomic conditions on the other hand, and to account for them by using distinct variables, even if certain linguistic repertoires tend to be correlated with certain socioeconomic conditions. Of particular importance, in language economics, is the level of proficiency achieved in the various languages making up a person’s repertoire, as well as the actual languages in his or her repertoire, because different levels of proficiency typically have very different effects on economic variables such as earnings, with higher proficiency being typically associated with higher earnings, all other things being equal, and because a certain degree of proficiency in a given language is typically associated with higher earnings in the case of some languages but not in the case of others (see TreffersDaller, Chapter 15, this volume, for discussion of problems involved in assessing proficiency). This terminological discussion calls for a brief comment on some notions that have recently gained popularity in some quarters of applied linguistics. First, let us recall that practically no one in the social sciences and humanities would dispute the notion that languages are fluid and porous. Against that backdrop, however, economists have no problem with “named” languages like Swahili, Mongolian, or Spanish. Most economists working on language have never come across the term “(trans)languaging” (see, e.g., Garcı´a & Tupas, Chapter 20, this volume), but if they have, they view it, in line with Edwards’s (2012) critique of this notion from a sociolinguistic standpoint, with suspicion – particularly its variants in which the very relevance of “named” languages is questioned. The mere fact that social actors never ask to be trained in “(trans)languaging,” but in identified foreign languages, and are not interested in learning the largely imaginary construct called by some (e.g., Cogo & Jenkins, 2010) “English as a lingua franca,” but require instruction in a form of English aligned on

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

177

a native speaker norm (Mackenzie, 2014) echoes the fact that demand for language skills on the labor market is for skills in named languages. Unquestionably, named languages are a product of sociohistorical processes, but this duly acknowledged point makes no difference to the economic issues involved. Second, language economics, which is defined more formally in what follows, is about the interplay between economic and linguistic variables, as the latter are experienced in the real world of language users, and taking account of changing linguistic conditions in a globalizing world. Many of these changes hinge on the presence of diversity, a notion that needs to be properly understood. We can observe two apparently contradictory, but in fact simultaneous processes. The first is a decline in objective diversity, which is reflected, for example, in the demise of small languages: the total number of elements of diversity is tendentially decreasing (on the measurement of aggregate objective diversity, see Ginsburgh & Weber, 2016a). At the same time, we are witnessing an increase in subjective diversity, that is, the diversity with which a growing proportion of people are confronted in their everyday lives, particularly in those parts of the world most exposed to globalization and the associated migration flows (on the distinction and interplay between objective and subjective diversity, see Grin, 2017). This increasing subjective diversity can be described without any need to resort to the eminently fluctuating notion of “super-diversity,” as explained by Pavlenko (2018).

3 Definition and Emergence of Language Economics Early definitions of language economics tended to focus on “writings by economists on language questions” (in the original: “textes d’e´conomistes portant sur les questions linguistiques,” Vaillancourt, 1985, p. 13, my translation), but other definitions have been proposed since the 1990s in order to offer a more analytical perspective on the specialty, independently of the training or affiliation of the authors contributing to it. This class of definitions (e.g., Grin, 2003) has remained stable over the past two decades, and the following captures its essence: language economics is an area of research epistemologically anchored in mainstream economics, which uses the core concepts of the discipline, and which mainly focuses on three classes of questions: (1) how do linguistic variables influence economic ones? (2) how do, conversely, economic variables influence linguistic ones? and (3) how can choices regarding language issues be usefully informed by economic analysis, particularly in the selection and design of language policies? The development of language economics began in the 1960s and 1970s with the study of language-based earnings differentials, that is, differences in income from labor. Such differentials can be traced back to agents’ L1, to

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

178

FRANÇOIS GRIN

their skills in additional languages, or both. Since then, the specialty has branched out in many directions and covers a wide range of topics. The total number of scientific publications in the field of language economics may now be said to total around 500 (see Gazzola, Grin, & Wickstro¨m, 2016, in which these contributions are arranged by category depending on their subject matter). Language economics presents two noteworthy features. First, this specialty remains on the fringes of the discipline, which may be due to its necessarily interdisciplinary orientation. It is of course possible to formulate some propositions about language (for example, regarding the value of language skills on the labor market) without engaging in deep considerations about the nature of language. However, for some other topics in language economics, for example when language use is treated as a dependent variable influenced by economic processes, one needs a good grasp of what language is about. What makes this need even more evident is that these mutual influences are strongly enmeshed in social, political, and cultural realities. Second, language economics truly comes into its own thanks to its use in the selection, design, and evaluation of language policies. Ultimately, the chief usefulness of economics is in assessing the pros and cons of competing scenarios, including language policy ones (see further Lo Bianco, Chapter 8, this volume, on language policies and bilingualism). Language economics first emerged as a response to actual socioeconomic questions, particularly in Canada, where it was used to identify the source of earnings differentials between L1 speakers of English and L1 speakers of French in the province of Que´bec (where more than 80% of residents report French as their L1). This work showed that such differentials were related to individuals’ L1 and persisted even after standardizing for education, second language (L2) skills, and work experience. A person’s education, experience, and human capital are regarded as socially acceptable sources of differences in earnings, but, other things being equal, a person’s L1 is not, since differences in earnings that could only be traced back to people’s L1 would then reflect ethnolinguistic bias against those who have this particular L1. These early findings, therefore, established the political and ethical appropriateness of corrective public policies. Similar tools were then used to assess the role of language in explaining socioeconomic disadvantage among non-English-speaking immigrants to the United States, a country in which English is the dominant language (see, e.g., Chiswick & Miller, 2007). In that case, low levels of competence in English as an L2 were shown to be a prime cause of earnings differentials, thereby supporting the (unsurprising) notion that investment in learning English was crucial to immigrants’ socioeconomic advancement. The same approach has been used later, primarily in European contexts, to estimate the value of investing in foreign

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

179

language skills (that is, in languages other than the locally dominant or official language[s]); we return to these developments later. Studies of language-based earnings differentials all have one thing in common: their main concern is to explain an economic outcome (in this case: earnings), where language is merely one explanatory factor along with many other ones. Another line of research, which has gained momentum from the early 1990s, mainly in Europe, focuses on the reverse type of causation. It seeks to explain a linguistic outcome (such as patterns of language spread or decline) as a result of the complex interplay of various factors, including economic ones. Alternatively, it may use economics as a way of unraveling the workings of this complex interplay itself, without necessarily giving standard economic factors a particularly prominent role in the explanation. Such developments have had a strong influence on language policy analysis, because they are directly applicable to the evaluation of the effectiveness, cost-effectiveness, and fairness of competing language policy scenarios. Since the mid-2000s, language economics has been moving into a new phase marked by increasing and more structured interaction with other disciplines, along with the more explicit inclusion, in economic modeling or econometric work, of sociolinguistic information.

4 The Core Question: Language and Value “Value” is a central question in economics, and this also holds for language economics. Language economists are therefore often asked questions about value: “How much is a small (minority) language worth?” “How much does an economy gain by operating in one language or another?” “How much should we invest in the teaching and learning of foreign languages?” “How much does bilingualism cost?” etc. These questions are not always fully identified or clearly distinguished from one another by those who ask them, which is why a parsing exercise is useful in order to clarify what “value” means economically, and what exactly needs to be measured in order to assess value in various contexts. This parsing exercise is particularly helpful for questioning the myth of inevitable hardship for bilinguals. Bilingualism, indeed, is often seen as inextricably bound to lower socioeconomic status and economic disenfranchisement. Though widespread in some quarters, this perception is largely erroneous and due to a basic problem of confounds. Multivariate analysis shows that language does not statistically explain, let alone cause, disenfranchisement. Indisputably, some groups earn less than others on average, that is, they show what is known as gross earnings differentials with other groups. These groups may be characterized by a whole range of traits, including aspects of their linguistic repertoires. Differences in earnings

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

180

FRANÇOIS GRIN

obtained by simply computing and comparing average earnings by language groups do not mean that these groups earn less than others because of their linguistic characteristics. The statistically solid associations that withstand the test of multivariate analysis are not between “lower income” and “being bilingual,” but between “lower income” and other variables, which include, among others, lower education, absence from the networks that provide access to information about employment opportunities, or lack of connections that could increase the likelihood of being shortlisted for a job. Multivariate regression, generally using the standard technique of ordinary least squares (OLS), serves to disentangle the impact of various determinants of earnings, and with some exceptions (see later in this chapter), the fact of having certain language skills does not by itself entail lower earnings. People who belong to certain groups with a high prevalence of bilingualism do not earn less because they are bilinguals; they earn less for other reasons, such as those just listed. Even if those other reasons are correlated with language, making bilingual people monolingual would not per se improve their work prospects. In cases where bilingualism remains associated with lower labor income, other things being equal, it is because this alleged “bilingualism” is characterized by relatively low, or even insufficient skills in the dominant language. The difficulty, then, is not that the persons concerned are bilingual (or rather, described as such). In fact, the difficulty is that they are not bilingual enough. This situation is likely to occur among immigrants where inadequate command of the dominant language in the country they have moved to may indeed be associated with lower earnings, other things being equal (that is, after having controlled for as many determinants of earnings as data availability allows). Thus, in cases where language remains one of the statistically significant determinants of a lower socioeconomic status, it is lack of bilingualism, rather than bilingualism, that explains it. Conversely, learning the locally dominant language is consistently shown to have a strongly positive and significant effect on immigrants’ labor income (Chiswick & Miller, 2014; Esser, 2006). Learning the dominant language may not be a sufficient condition for socioeconomic success, but there is little doubt that it is a necessary condition. Summing up, a simple descriptive account and commentary on gross earnings differentials between people from different language groups (that is, simply computing and comparing average earnings by language groups) may lead to incorrect conclusions, hence the need to apply multivariate analysis in order to estimate language-based net earnings differentials. A few exceptions to this general result have been documented. The bestknown one has been mentioned in Section 3: L1 French speakers in the 1960s and 1970s in Que´bec earned less than their L1 English-speaking counterparts, even after taking into account the effect of education,

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

181

experience, and English language skills. Thus, it was the fact that their L1 was French that explained lower earnings. Similar findings are reported in a few other cases; see, e.g., Patrinos and Hurst (2007) on indigenous languages in Bolivia, where having one of these languages as an L1 is associated with lower labor income even with high levels of education and competence in Spanish. Another interesting, though symmetrical exception is that of “expats” in some countries, who can enjoy high earnings despite poor skills in the local language. Their situation generally reflects the presence, in the labor market, of a specific niche where the local language is not considered important by the employer. Our discussion so far has been referring to market value, of which earnings are a prime example, but ideally, the analysis should go beyond market value and also include nonmarket (sometimes also called “symbolic”) value. Lack of data usually prevents this, but it is nonetheless important to discuss these two types of value. Market values are those that can be read off data such as observed market prices: They capture the material or financial expression of the value of a good or service (from tomatoes to cars and, using somewhat more elaborate techniques than simple price observations, linguistic or computer skills, or environmental assets). However, nonmarket values are no less relevant, since people assign value to language for a variety of reasons, in particular as a repository of experience and as a stock of symbolic resources used for manifesting one’s individual identity and/or group affiliation. The complex interplay of processes resulting in such valuations can also be approached through the concept of language hierarchies (see, e.g., Fuller, Chapter 6, this volume). In economics, the identification of manifestations of nonmarket value often relies on parallels with the natural environment, because environmental economics is the specialty of economics where some of the most promising developments in the methodology of evaluation have developed. For example, an important component of nonmarket environmental value is the enjoyment one may experience from unspoiled landscapes and clean streams, and specific survey-based techniques have been developed to estimate them. In the realm of foreign language skills, nonmarket value may proceed from the enjoyment of direct access to foreign cultures and to the people with whom unmediated communication becomes possible. This first distinction (market versus nonmarket) corresponds to the two rows in Table 9.2. Table 9.2 Four types of value

Market Nonmarket

Private

Social

A B

C D

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

182

FRANÇOIS GRIN

At the same time, it is important to consider at what level, or from whose perspective of value, whether of the market or nonmarket kind, value is being examined. We may be concerned with value for the typical individual, in which case we would talk of private value, or with value for society as a whole, thus opening up the question of social value. This problem is conceptually and methodologically difficult, because owing to some crucial aspects of language, social value is not the mere sum of its private components. First, language is not just a tool for communication; it is also a carrier of culture, and it is bound up with politically and socially complex matters of individual and group identity. It follows that some aspects of language, including some that carry value, are only experienced collectively, and the corresponding estimation work must take these aspects into account. Second, language presents many of the characteristics of a network through which people are connected with one another. In this context, the value of one’s own language skills, to any given member of the network, can change (upward or downward) as a result of entry into and exit from the network by other people. For example, a Canadian with English or French as an L1 who learns Japanese can derive certain market and nonmarket benefits from it. If one more person learns Japanese, this will, for the first person, constitute competition and push the value of her or his Japanese language skills downward; at the same time, this means that there is one more person with whom these skills can be used, making them more relevant and thus pushing their value upward. There is no general rule to predict which of these two opposite effects on value will dominate. At this time, the problem of aggregating private values to estimate social value is, in the case of language, not adequately solved in scientific research. Nonetheless, the conceptual distinction between the two levels in Table 9.2 at which value can be assessed remains valid. Finally, little is known about the order of magnitude of nonmarket values. The estimation of market values, however, is quite straightforward if adequate representative data are available. Let us therefore turn to a survey of empirical findings, starting with figures on the rates of return to second/foreign language skills.

5 Selected Empirical Findings 5.1 Language and Earnings Different language skills are valued differently in different contexts. Obviously, Swedish is essential for most jobs in Stockholm, but of little use for most jobs in Shanghai. Thus, for people moving to a new linguistic environment, for example in the context of migration, knowledge of the local language is generally a key asset. As Isphording (2015) rightly observes, “this creates incentives to invest in the acquisition of languages” (p. 1).

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

183

However, surprisingly few countries have the data necessary for estimating the labor market value of language skills, and fewer still collect them on a regular basis. By way of illustration, let us consider the results of a survey in Switzerland (Grin, Sfreddo, & Vaillancourt, 2010) as regards the earnings differentials accruing to men living in the three main language regions there (French-, German-, and Italian-speaking). The original survey includes 2,400 respondents of both genders; it is representative in terms of age, gender, and language region. I focus on the results for the male subsample, since these results are typically more robust from a statistical standpoint. These men speak the locally dominant language (French, German, and Italian depending on their region of residence). Grin et al. (2010) report on their extra net earnings as a result of their competence in French, German, or English as a second or foreign language, controlling for education, experience, and experience squared (which are standard control variables in the estimation of earnings equations). Using multivariate analysis, the additional labor income in percentage terms accruing to people who have acquired an (approximately) B2 level in the foreign language concerned can be computed net of the effect of other standard determinants of labor income (a B2 level according to the Common European Framework of Reference is defined as upper intermediate; see further Mun˜oz & Spada, Chapter 12, this volume). The estimations reported here are all statistically significant at the 5% level. In Germanspeaking Switzerland, the average wage premium is 14.1% for knowledge of French and 18.1% for knowledge of English. Reciprocally, in Frenchspeaking Switzerland, the average wage premium is 13.8% for German and 10.2% for English. The most important point to make here is that second/foreign language skills are eminently profitable. For a given level of education and work experience not many investments result in wage premiums in the 10% to 18% range. Second, these rates of profitability vary not just according to the second/foreign language concerned, but also depending on the linguistic features of the labor market. Because of strong language territoriality in Switzerland, and also because the French- and the German-speaking language regions both include a wide panoply of economic activity in industry as well as in services, regional labor markets can be seen as primarily defined by the local majority language; we can then reason in terms of L1 versus L2/L3 combinations. In the two main language regions, the wage premiums are remarkably symmetrical for French and German, two of Switzerland’s official languages (see earlier in this chapter). The contrasting case of English, which is not an official language in Switzerland, is interesting, because it proves much more profitable in the German- than in the French-speaking region of the country, with premiums of 18% and 10%, respectively. That said, a 10% premium still amounts to a very attractive investment proposition. What is more, the rates of return for English are extremely robust in statistical terms, as

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

184

FRANÇOIS GRIN

shown by tests not reported here. The subsample collected in Italianspeaking Switzerland is too small to reveal statistically significant effects for the knowledge of English. However, the wage premium is 17.2% for residents of this part of the country who speak French, and 16.9% for those who speak German; these higher rates of return reflect the minority status of Italian nationally and the particular relevance, for people who live in Italian-speaking Switzerland, of expanding their linguistic repertoire. The survey findings from Switzerland illustrate private market value fitting in cell A of Table 9.2 in Section 4. But we can move from private to social value (cell C in Table 9.2) by combining the econometric results on wage premiums with data on spending on foreign language learning and teaching. Such data are difficult to come by, however, because education systems typically do not have numbers on spending by school subject, and estimates have to be derived from other school statistics. Data on private spending on foreign language learning, whether by individuals themselves or by their employers as part of continuing education, are no easier to come by, but if they are available or can be convincingly proxied by other information, they may be taken into account in order to fine-tune the estimations. In the case of Switzerland, the resulting “social rates of return” confirm that even when one factors in the amounts that residents spend on language learning and teaching, second or foreign language skills constitute a highly profitable investment proposition from the standpoint of society as whole. As reviewed in Gazzola et al. (2016), the relevant literature shows extensions to the foregoing results, as well as estimates of other aspects of economic value. For example, the macroeconomic counterpart of the share of Switzerland’s gross domestic product (GDP) that can be traced back to second/foreign language skills is in the region of 9% to 10%; this high rate reflects the fact that in the Swiss economy, many languages are used, both for international trade and for domestic trade across internal language boundaries. Similar calculations for Que´bec, with a main focus on competences in English and French, yield estimates of the contributions of second/foreign language skills to provincial GDP in the 3% to 4% range. For a near-exhaustive review of existing research up until around 2015 using various types of data from different countries and containing results on a wide range of other questions related to language economics, readers are referred to Gazzola et al. (2016).

5.2 Language and Access to Employment Generally, language skills have a positive impact, not just on wage rates, but also on labor market access. This section hence discusses the links between languages and employability. However, these links are not easy to discern, because hard evidence on access to jobs as opposed to earnings (once a person has a job) is scant and heterogeneous. Beyond (mostly) qualitative case studies relying on a clutch of interviews, there is often

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

185

little choice but to rely on indirect or circumstantial evidence, and there are only few studies working with samples offering a size and representativeness sufficient for reliable inference. Local-area surveys of employers and/or employees in multilingual regions or countries, for example in the Brussels area (Mettewie & van Mensel, 2006), indicate that skills in the local languages are much in demand, suggesting that the absence of such skills is a barrier to job access. It is also plausible that, in contexts where a given L1 is stigmatized (see Section 4), skills in the “wrong” language may have a negative impact on job access too, and several theoretical explanations can be drawn upon to account for this effect. Some rely on economic models of ethnic discrimination. Others are derived from models of wage rate determination that go beyond the classic explanation based on human capital theory, which is centered on the impact of skills on productivity. Among alternative theories of the labor market, signaling theory is likely to be particularly relevant to the understanding of languagerelated earnings effects. Skills in some particular languages may not be directly necessary for a particular job, but they signal, to the employer, an employee’s capacity to acquire skills in general, and will therefore be sought by the employer. This translates into higher wages for people who possess such language skills, even if these are objectively not very useful. This effect is more likely to occur for skills in prestigious languages or varieties. Alternatively, the mere presence of skills in another language may constitute cases of negative signaling and be negatively interpreted by the employer, possibly because of ethnic or racial prejudice. The question of languages and employability is related to another one, namely, that of the relevance of language skills, possibly skills in specific languages, in particular industries or for specific activities. The picture is a highly diverse one, with some sectors and jobs requiring high-level foreign language skills, often in English, and others where other foreign languages matter more, or others still where foreign language skills are not particularly relevant. Research on the matter is dispersed in a collection of case studies, often qualitative (see the annotated bibliography of language economics by Gazzola et al., 2016, for specific suggestions); recent work by Grenier and Nadeau (2016) investigates the language skills needed by industry and type of occupation. Clearly, L1-like (“native”-like) knowledge in a given language X (which does not necessarily mean having language X as an L1) may perfectly legitimately be requested for certain jobs. This is the case, for example, for work such as translation into language X, commercial or legal writing in language X, copy editing for an X-medium newspaper, or X language teaching beyond a certain level. Problems arise, however, if an applicant with fluent X language skills is excluded from a recruitment process for a job requiring such skills on the grounds that his or her faultless fluency in the language is the result of having learned language X as a second or third language, but not as

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

186

FRANÇOIS GRIN

a chronologically first (or “native”) language. This problem is akin to the one mentioned in Section 4 on the issue of language and value concerning L1-based linguistic discrimination.

5.3

Language and the Likelihood of Keeping One’s Job When Wages Go Up When wages go up, firms have an incentive to lay off staff in order to reduce costs. Do language skills make a difference in this respect? Estimating such possible effects requires a technically elaborate approach based on the reconstruction of three standard functions in applied microeconomics, namely the production, cost, and profit functions. Data from the manufacturing sector in the Swiss economy (the only one, to my knowledge, for which such calculations have been made) show that firms tend to retain bilinguals more than monolinguals, even if the former are more expensive. More precisely, when wages go up, firms will tend to let some of their staff go, but they retain bilinguals more than monolinguals, even if the latter are “cheaper,” because, all other things being equal, they tend to earn lower wages, as shown in Section 5.1). For example, an across-the-board 2% increase in the cost of labor leads to a 3.5% decline in the volume of the monolingual workforce, as opposed to a 1.5% decline in the volume of the bilingual workforce. All other things being equal, therefore, monolinguals are more than twice as likely to be laid off (Grin et al., 2010). Putting it differently, language skills can be seen as a form of insurance against being fired.

5.4 Language and International Trade Flows It is commonly assumed that language orients trade, in the sense that having a common language is expected to boost bi- or multilateral trade between the countries sharing it. More generally, the role of language as a driver of international trade has long intrigued economists (Egger & Lassman, 2012; Egger & Toubal, 2016; Me´litz, 2016; Me´litz & Toubal, 2014). The issue is whether having a common language induces trade flows over and above those that would have arisen anyway. This means that the role of other variables must be controlled for in order to tease out the specific impact of a common language. There again, a number of studies tend to confirm the presence of such an effect. However, the mere share of foreign trade taking place with countries with which there is a language in common could be a misleading indicator, because “linguaspheres” sharing a language greatly differ in size: in a two-country linguasphere (e.g., “Lusophonia,” encompassing Portugal and Brazil), bilateral trade is significantly boosted by the fact of having Portuguese in common, but this still represents a modest share of these countries’ total foreign trade. This latter share is tendentially bigger in larger linguaspheres. For this reason, an alternative

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

187

indicator is useful for making comparisons. One such indicator is the degree of openness to (international trade), defined for each country as the ratio featuring (i) in the numerator, the sum of imports and exports divided by two; (ii) in the denominator, GDP. Carre`re and Masood (2016, p. 87) report that linguistic commonality increases the degree of openness to trade of the countries concerned, on average, by 4.5% for French, 5.2% for Spanish, 6.1% for Portuguese, 6.5% for English, and 3.4% for Arabic. A striking result, however, is that having a first or second language in common (as distinct from having second language skills) proves particularly important in situations of economic crisis: bilateral trade between countries sharing a common language proves more resilient in periods of economic downturn, even after controlling for a wide range of other explanatory variables.

5.5 Language and GDP Growth in Developing Countries Oft-quoted results in the specialist literature in development economics insist that “linguistic fragmentation” (the negative pendant, as it were, of the more positive notion of linguistic diversity) has a negative impact on GDP per capita in developing economies (Easterly & Levine, 1997). This result has received such publicity that it has become quasi-axiomatic in much of development economics. However, it is based on what is, upon closer examination, an insufficiently prudent application of econometrics. It ignores the fact that the presence or absence of skills in a dominant language (usually the official language of a former colonizer) and GDP per capita may both be influenced by another factor. One example is the stability of political institutions in a particular developing country, the role of which is duly taken into account in other facets of the study of the determinants of GDP growth, but usually not in investigations of the role of language. When the possibility of such a link is taken into account using the relevant econometric procedures (Arcand & Grin, 2013), “linguistic fragmentation” turns out, in most cases, not to have any statistically significant effect whatsoever. In those cases where it has one, the effect is actually positive. Data show that this result, which applies to English, applies to other former colonial languages as well, like French and Spanish. The general lesson is that there are no economic reasons for advocating linguistic uniformity over linguistic diversity.

6 Applications to Language Policy This chapter’s emphasis has been placed on mainstream economic aspects, showing how linguistic variables may influence economic ones. However, mention must be made of other and no less interesting areas of application that focus on explaining linguistic variables, either as the result of economic processes, or as influenced by particular economic

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

188

FRANÇOIS GRIN

variables. This line of research can be split up in two orientations. The first contains work that looks at such economic effects in general, independently of any explicit, targeted intervention by the state. A typical example is the study of how minority language use may respond to changes in the relative prices of some goods and services. For example, a decline in the relative cost of access to minority-language-medium entertainment tends to result in increased consumption. The second contains work offering an economic analysis of language policy. This area is particularly exciting, because it opens up onto a vast range of politically and socially crucial issues, as well as throwing up stimulating methodological challenges. The economics of language policy is a thriving area of research (Gazzola & Wickstro¨m, 2016; Grin & Vaillancourt, 2015), and it addresses issues such as: 1. the effectiveness, cost-effectiveness, and fairness in minority language protection and promotion, and how these variables can be identified, operationalized, and measured; 2. the relative effectiveness of different channels of second language acquisition; 3. the relative costs and benefits of alternative language regimes for international organizations (which may designate as official a larger or smaller set of languages) or institutions such as universities (which make very different choices regarding the language[s] of instruction); 4. the study of patterns of language use, for example on the labor market or in other activities related to the production, consumption, and exchange of various goods and services in response to language policy decisions in those areas; 5. the investigation of language dynamics (why and how do some languages expand while others are in retreat, and what determines the rise of the next waves of dominant languages), and the extent to which these patterns actually respond to policy intervention. Limitations of space prevent discussion of all these avenues, but this non-exhaustive list provides a natural link to the discussion of an important conceptual contribution of economics to the understanding of language policies. Language policy, just like any other public policy, can be evaluated in terms of two main criteria, namely, efficiency and fairness (Gazzola, 2014; Gazzola & Grin, 2017). Most of this chapter has addressed the former, allocative criterion. We have seen that individual plurilingualism is associated with various measures of economic prosperity; moreover, the more macro-oriented analyses presented here suggest that this is also true of the resulting societal multilingualism. Furthermore, there are several reasons for thinking that these language skills contribute to prosperity, that is, there is a causal connection flowing from “language skills” to “positive economic outcomes.” However, this should not lead us to forget another side of language economics and in particular of the

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

The Economics of Bilingualism

189

economics of language policy about which little has been said so far, namely, the problem of fairness and the distributive implications associated with language. Of course, it is well known in the language disciplines that language is often associated with power, and critical sociolinguistics devotes considerable attention to revealing the ways in which patterns of language use as well as language policies themselves are enmeshed with the exercise and reproduction of power and influence (see Lo Bianco, Chapter 8, this volume). A similar concern exists in language economics, albeit in a somewhat different perspective that focuses on the identification and measurement of the ways in which language policies redistribute material and symbolic resources between groups. In the assessment of competing policy scenarios, therefore, what matters is not only how much aggregate prosperity they create (or, if misguided, how much prosperity they destroy). What also matters is how this prosperity is distributed among individuals and groups, where groups need not be defined solely in classical terms of income or socioeconomic status, but may also be defined according to other criteria such as age, gender, or first or main language. The latter criterion is of particular importance here and forms the conceptual bedrock of analyses of linguistic justice (Ricento, Peled, & Ives, 2014). Let us simply point out, by way of returning to the topic of bilingualism, that specialist research in language policy analysis mostly agrees that linguistic hegemony is not a desirable state of affairs. A policy of linguistic hegemony would imply favoring one particular language, and it makes no difference whether this language be English, French, Efik or Putonghua. It would abet processes of linguistic uniformization in various domains. Such processes are demonstrably socioeconomically regressive because dominant language skills tend to be more widespread among the more favored groups in society, whether in terms of education, age, or gender (Gazzola, 2016). Conversely, the marginalization of other languages sharpens inequality. These regressive effects appear not only between groups defined in terms of social class or socioeconomic status, but also between groups defined by their linguistic attributes, that is, their L1, as well as the L2, L3, etc. that they speak.

7 Conclusion Language economics, together with its application to language policies, makes essential contributions to the understanding of issues that have major social and political importance and have direct bearing on bilingualism as a social issue. Overall, the findings in various areas of economic research reviewed here tell a fairly consistent story, namely, that skills in “big” languages, through one or another channel, contribute to the creation of value and thus to prosperity. This applies to many “big” languages

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

190

FRANÇOIS GRIN

(not just to one dominant language), and sometimes to linguistic diversity per se, which depends on the continued vitality not just of big languages, but of small languages as well. Thus, from an allocative standpoint, converging evidence suggests that linguistic diversity is economically more efficient than linguistic uniformity. Symmetrically, research into the distributive issues involved suggest that linguistic diversity, rather than linguistic uniformity, guarantees fairness. This suggests that public policy at the local, national, and inter- and supranational levels should not only encourage individuals to develop a good portfolio of languages skills, but also ensure that the world of business, science, and culture remains highly linguistically diverse.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:11:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.010

10 The Nature of Exposure and Input in Early Bilingualism Sharon Armon-Lotem and Natalia Meir

1 Introduction The linguistic development of young bilingual children, whether simultaneous or sequential, and regardless of whether a single modality or two modalities are involved, is affected by both child-internal (e.g., level of maturation, working memory, chronological age) and child-external factors. The latter include age of onset of exposure to more than one language (whether in another modality or not) as well as the quality and quantity of the language children are exposed to and other aspects of children’s linguistic environments, such as whether parents or caregivers provide native or nonnative input and high or low levels of code-mixed input. Our aim is twofold: to contribute toward a better understanding of the complex factors that moderate the impact of exposure and input on bilingual children’s linguistic development, and to help improve measurement practices in this research domain. Rather than offer a historical review starting from the first studies published since Pearson, Ferna´ndez, Lewedeg, and Oller’s (1997) seminal study, we concentrate on the present state of understanding about the topic, highlighting methodological advances and key findings. We pay close attention to early sequential bilingualism and review many studies where exposure or input are examined. For a review of studies examining the influence of input and exposure on mainly bilingual first language acquisition with simultaneous onset of exposure to both languages from birth, readers are referred to De Houwer (2018a). Measuring young bilingual children’s language exposure and input is not only of interest to better understand the bilingual development process, but also to be able to evaluate whether any finding of delayed language acquisition by a bilingual child is due to limited exposure or to

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

194

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

language impairment. However, we do not review here the research that attempts to distinguish delays in acquisition caused by language impairment versus low exposure (for more on this, see Tuller, 2015; Hammer & Edmonds, Chapter 19, this volume). We show that there are advantages and disadvantages in measuring exposure and input via relative exposure (including cumulative exposure), absolute exposure, and the content in the input of specific linguistic constructions. We argue that ultimately it is the measurement of the latter that will be needed in the future. This is because, as our review reveals, measures of exposure, be it relative or absolute, are good predictors of lexical and phonological knowledge, but are not as good for predicting the acquisition of morphosyntactic and syntactic structures. This pattern of findings is consistent with the prediction that morphosyntax holds a special status in language. The study of how exposure and input influence bilingual development has also theoretical importance, in that it helps advance knowledge about two competing explanations for language acquisition: a usage-based or constructivist explanation, which accords input and exposure a central role in learning under the hypothesis that language (including morphosyntax) develops piecemeal via general cognitive mechanisms of statistical learning, abstraction, and analogy (e.g., Gathercole, 2007; Paradis, Nicoladis, Crago, & Genesee, 2011); and a generative-nativist approach, which stipulates that some parts of grammar (and more specifically some aspects of morphosyntax) are genetically endowed via Universal Grammar and thus learned through domain-specific mechanisms that are less sensitive to frequency of exposure (e.g., Egger, Hulk, & Tsimpli, 2018).

2 Terminology Bilingual children are learning at least two languages: their heritage language (HL), also labeled home language, first language (L1), or minority language, and the societal language (SL), also labeled second language (L2) or majority language. For bimodal bilingual children we use the term HL for the sign language they are learning, regardless of whether they learn it within the family or through a special school for the Deaf, and SL refers to the spoken and/or written language they are learning. Age of onset (AoO) of exposure to the SL is the most basic measure of exposure in early child bilingualism. It signifies the end of a period of monolingual acquisition for non-simultaneous bilinguals. Simultaneous bilinguals do not experience a period of monolingual acquisition (Serratrice, Chapter 1, this volume) and for them, AoO is the same for each language (namely, birth). From the moment of SL onset, the actual amount of linguistic exposure in the HL may drop as a function of the increase in SL exposure, as the number of waking hours and thus opportunities for interaction and language exposure of a bilingual child are then

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

195

divided between the HL and the SL. The AoO of bilingual exposure is easy to determine for simultaneous bilinguals and for sequential bilinguals who migrated to a new country, but it is not an easy task to establish the beginning of dual exposure for children who are born to a home where the HL is used while the broader community uses the SL. Thus, it is often crucial to explore children’s daily routines from a very young age in order to know how much SL was present in their language learning environment through siblings, playgroups, or television programs before (pre)school entry. Length of exposure (LoE) to the SL measures the time since the onset of exposure to the SL. The AoO and LoE correlate, since LoE is calculated as children’s chronological age minus their AoO (e.g., Paradis, 2011). For simultaneous bilinguals, LoE to an SL is the same as chronological age. Bilingual exposure at home is clearly different from bilingual exposure to one language at home and another at school. Neither LoE nor AoO take into account any variation in the amount of exposure to each language over the years or months, or the distribution of exposure over interlocutors or settings. LoE can be seen as a very crude measure of relative exposure, in the sense that it gives a rough indication of when HL exposure started being less than 100%. Additional quantitative measures beyond LoE that can be used for both simultaneous and sequential child bilinguals are those measuring the relative amount of exposure in each language compared to the other, and the absolute amount of exposure in any one language or both. Relative exposure is often determined by questionnaires in which parents are asked to assess the relative amount of exposure to each language at home or to give details of daily routines at different child ages and the language(s) used for each activity. Such measures yield an approximation of the exposure to each language relative to the other and are usually expressed in percentages. Absolute exposure, by contrast, is assessed by measuring child-directed speech (CDS) in each language regardless of the other. Different children with balanced relative exposure (i.e., 50% exposure to the HL and 50% to the SL) might actually be exposed to different numbers of words per time unit in each language, as may different children with, say, 70% of exposure to the SL: one child might hear an average of 500 words per hour in the SL, while another child might hear as many as 2,500 words per hour in the SL. Language samples of natural interactions yield a more accurate picture of children’s present language exposure, or, if used longitudinally, past exposure as well. Measures of absolute exposure can capture the density of global exposure to a particular language. They can also capture more qualitative aspects of exposure, that is, they can capture the presence and the frequency of specific phonemic strings, lexical items, inflections, or syntactic

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

196

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

constructions. Within one language, it has been shown that the monolingual acquisition of a lexical item, abstract categories (e.g., word order sequences) and abstract cues (e.g., animacy) can be accounted for by the frequency of these items in CDS (e.g., Ambridge, Kidd, Rowland, & Theakston, 2015). For example, the words baby and ball are acquired faster than coffee and computer, as the frequency of the words baby and ball is (much) higher in CDS. We here use the term input to refer to the availability and frequency of specific linguistic target units in CDS or signing. In addition to measures quantifying exposure, scholars have turned to investigating other aspects of exposure such as language use in different settings (e.g., home versus school), family language policy, and the linguistic proficiency of children’s interlocutors. Such measures refer to quality of exposure. Quality of exposure in turn may impact the availability and frequency of particular morphemes or syntactic structure in the input, given that in many cases bilingual children receive input from bilingual and bidialectal speakers that may differ from monolingual norms.

3 Length of Exposure and Relative Exposure LoE and relative exposure have been extensively investigated for bilingual children acquiring two spoken languages, whereas they have been focused on less in studies of bimodal acquisition. Typically, scholars rely on parental questionnaires to explore LoE, relative exposure to each language, and other aspects of the linguistic backgrounds of language-minority children. Several questionnaires exist (here also meant to include interview protocols). Some of these have been developed for use in both research and clinical settings. An example is the Alberta Language Environment Questionnaire (ALEQ) developed in Anglophone Canada (Paradis, 2011). It elicits oral information from parents of bilingual children from varied HL backgrounds. The ALEQ focuses on children’s experiences with their SL, English. It asks when children were first exposed to the SL, thus yielding an age of acquisition (AoA) and LoE measure, and asks to what extent children are exposed to the SL through, e.g., media, organized activities, and playmates. The ALEQ also asks parents to self-rate their fluency in the SL. Finally, it also asks about the languages that family members speak to children. The Alberta Language and Development Questionnaire (ALDeQ) (Paradis, Emmerzael, & Sorenson Duncan, 2010) was developed to gain information on potential language learning problems in young bilinguals and focuses on children’s HL, regardless of which it is. Both the ALEQ and the ALDeQ were the inspiration for the COST Action IS0804 (www.bi-sli.org) Questionnaire for Parents of Bilingual Children (PaBiQ) (Tuller, 2015), which was developed for use in the multilingual European context. As we see in what follows, LoE hides a number of complex exposure and input realities that cannot be captured with even

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

197

the best questionnaires, and thus the findings about this variable must be interpreted with caution. Using the ALEQ, measures of child-internal factors and measures of children’s comprehension vocabulary size and performance on a morphosyntactic test, Paradis (2011) studied 169 sequential bilinguals between ages four and seven with 3–62 months of exposure to English as an SL who had heard an HL since birth (there was a wide variety of HLs, a point whose importance becomes clear later). The study only focused on the SL but used information on selected morphosyntactic characteristics of children’s HLs as well and was designed to evaluate the relative contribution of child-external and -internal variables on SL proficiency. The external variables included LoE and the proportion of SL use at home as a quantitative measure of relative exposure, mother’s self-rated SL proficiency, and richness of SL use in after school activities as qualitative measures. Most of these factors turned out to have low predictive value. Instead, internal variables (chronological age, phonological shortterm memory, nonverbal intelligence) strongly predicted both vocabulary size and SL morphology. Whether children’s HL marked verbs for tense and agreement or not also significantly predicted children’s SL morphological outcomes. And while LoE and English richness in after school activities also predicted children’s SL performance, these factors were less important than the child-internal factors. In fact, regardless of the LoE, older children were more developed in their SL proficiency. Furthermore, SL use at home had only limited impact on SL proficiency. Armon-Lotem, Joffe, Abutbul-Oz et al. (2014) suggested that LoE effects might be different for different bilingual groups living in the same country. They compared 49 sequential bilingual children between ages 4.5 and 6 living in different communities in Israel, where Hebrew is the SL. They found effects of LoE on general SL language proficiency scores for the Russian–Hebrew but not the English–Hebrew bilingual children. Over and above LoE, differences in SL proficiency across the Russian–Hebrew and English–Hebrew bilingual children were explained by differences among the two groups of children in their ethnolinguistic identities and their attitudes toward their HLs and the SL. Armon-Lotem et al. explained these differences in terms of differences across the communities where the children lived. Children who performed better in the SL lived in a community where there was a lot of mixed use of the HL and the SL and that supports integration into the host society. This was the Russian–Hebrew community. SL performance was lower for children living in the English–Hebrew community, where the HL was used a lot in the neighborhood and whose members generally have lower levels of motivation for integration into the host society. The authors further argued that the relative prestige of each of the HLs can mediate the effect of relative exposure. Both Russian and English are minority languages in Israel, but English has the prestige of a lingua franca. Thus, its prestige might

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

198

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

negatively impact the acquisition of the SL. This impact of English (an HL in Israel) is a mirror image of the more common situation where English is the prestigious SL. Yet this impact is sensitive to the language combination that children are acquiring. If Armon-Lotem et al. (2014) are right, Paradis (2011) may have not found much explanatory power in LoE in part at least because the sample included sequential bilinguals with many different HLs, possibly hiding the variable effects on exposure of a wide variety of ethnolinguistic identities and attitudes toward a particular HL and vis-a`-vis the prestige status of English both in Anglophone Canada and as a lingua franca.

4 Relative Exposure and Cumulative Length of Exposure As an alternative and perhaps an improvement over LoE, researchers have turned to relative exposure, which estimates the amount of exposure to each language compared to the other. Relative exposure is a strictly quantitative measure, focusing on the proportion of time that bilingual children hear language. It is intended to offer a more precise measure of exposure than just LoE because it quantifies the relative amount of exposure within a given period rather than just the global length of time. For example, there is a clear difference between whether children who have been exposed to a new language (the SL) for three years have been exposed to it for just 5% of the time or for as much as 70% of the time during this same period. The LoE does not capture this difference, while relative exposure does. Sometimes just global relative exposure is assessed, but often different measures of relative exposure (e.g., home versus school, child-directed speech versus exposure via TV) are compared for their relative contribution to different linguistic abilities of children in each of their languages. This has been done for both sequential and simultaneous bilinguals and for morphological, syntactic, and lexical knowledge. The hypothesis that has driven researchers’ interest in relative exposure is that there is a relation between the relative amount of exposure to a language and bilingual children’s linguistic performance. Often studies include a comparison of relative exposure and AoO. The findings thus far have supported the basic hypothesis, but with a number of interesting qualifications, to be reviewed next. Relative exposure effects have been found for the development of phonological skills, morphosyntax, and lexis. Using questionnaire data from parents and teachers for calculating the relative amount of exposure that 91 Spanish–English five-year-old bilingual children in Texas received and children’s overall proportion of use of each language (output), Ruiz-Felter, Cooperson, Bedore, and Pen˜a (2016) showed that current exposure and output in the HL (Spanish) predicted better knowledge of HL consonants and vowels. The same was true for the SL, where current exposure and

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

199

child use was more predictive of children’s SL phonological skills than AoA. Paradis et al. (2011) explored the relation between the amount of exposure to each language at home in 23 French and English four- to fiveyear-olds and the acquisition of French and English verbal morphology (13 of the 23 children had heard both languages from birth). Children with more English exposure did better on the English regular past tense than children with more French, but no difference was observed for irregular past verb forms (all children had low scores). Similarly, children who heard more French than English were significantly better than children with more English in the use of French past tenses. This was true for both regular and irregular forms. Like Paradis (2011), but now in a longitudinal design, Dijkstra, Kuiken, Jorna, and Klinkenberg (2016) looked at the influence of language exposure both in the home and outside the home, but only considered adult–child interaction. Like Ruiz-Felter et al. (2016) and Paradis et al. (2011), Dijkstra et al. (2016) also considered both of the languages that 91 nonimmigrant early sequential Frisian–Dutch preschool-age children were acquiring. The focus was on comprehension and production vocabulary. Children were living in the bilingual province of Friesland in the Netherlands, where Dutch is the SL but Frisian as HL is also present in the public domain, though Dutch is omnipresent and has higher status. Frisian and Dutch are fairly closely related languages. Children heard either mainly the HL or mainly the SL at home and were 2.5 years old at the first wave of data collection. They visited preschools where either mainly Dutch, mainly Frisian, or both languages were spoken. Children were at preschool one to three days a week. Children’s nonverbal intelligence influenced vocabulary scores. When these were controlled for, several differences among children emerged depending on whether they were hearing mainly the HL or the SL at home and depending on whether comprehension or production was involved. SL-speaking children who were learning Frisian outside the home through preschool hardly learned to speak it, whereas HL-speaking children for whom Dutch was an L2 did learn to speak it, thus showing that preschool-aged children are already sensitive to language status differences (cf. Armon-Lotem et al., 2014). For Dutch comprehension vocabulary there was no home language effect: children performed equally regardless of whether they had Dutch as HL or SL. Outside home exposure affected comprehension vocabulary in both languages, but there was no outside home exposure effect for production vocabulary. Basically, the SL-speaking children did not learn much of the HL, but the HL-speaking children made quite a bit of progress in the SL (see Serratrice, Chapter 1, this volume, for discussion of similar findings for simultaneous bilingual children). Recognizing that outside home exposure to the SL may be an important factor in bilingual children’s development, Gutie´rrez-Clellen and Kreiter (2003) decided to use both parents and teachers as agents of knowledge to

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

200

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

fill in questionnaires designed to explore the role of the amount of exposure at home, years of exposure to a language, amount of exposure at school, and amount of exposure to literacy-oriented activities (e.g., reading and writing). They used this questionnaire with parents and teachers of 57 eight-year-old children who were growing up with Spanish as HL and English as SL in California, where English is the SL. Measures of exposure were compared to children’s grammatical performance based on narratives in one or two languages. Gutie´rrez-Clellen and Kreiter (2003) found that of all measures of relative exposure, the relative amount of home HL exposure was the only predictor of children’s grammatical performance in the HL. No associations were detected, however, between the relative amount of exposure to the SL at home and children’s linguistic abilities in the SL. These findings are consistent with what Dijkstra et al. (2016) found for lexical acquisition in younger children. In an unusually large study, Bedore, Pen˜a, Griffin, and Hixon (2016) studied 568 six-year-old and 287 eight-year-old Spanish–English bilingual children, comparing AoO to the SL (English), relative current exposure and output (that is, amount of children’s use of the language) with measures of children’s morphosyntactic and semantic knowledge. Contrary to their study from the same context that focused on phonology (Ruiz-Felter et al., 2016), Bedore et al. (2016) found that AoO robustly predicted performance in the SL, with the effect decreasing with age. Interestingly, for the eightyear-olds, HL current exposure and output explained much of the variance in SL performance. For the HL, relative exposure accounted for more variance than AoO. Crucially, for the children in this study AoO and relative exposure were highly correlated with each other, making it difficult to achieve the goal of separating the effect of both measures. In a study conducted in the French-dominant bilingual city of Montreal in Canada, Thordardottir (2011) calculated relative exposure on the basis of 49 five-year-old French–English bilingual children’s interactions with another speaker, creating a score for estimated proportion of exposure to each language at home and at preschool for each year since birth (children included simultaneous and sequential bilinguals). Relative exposure to each language changed as children entered daycare or changed their preschool provider. Generating a single variable by summing up the yearly proportions yielded a single measure of the relative time spent in a French or English communicative environment. A strong relation was observed between the relative amount of exposure to a language and children’s vocabulary scores in that language. When children were matched on the amount of relative exposure, comparisons of children with early and late AoO showed no AoO effect. These studies share attempts to find reliable measures of relative exposure, with the understanding that AoO and LoE, even when combined with reports on estimated home or school language use, are not accurate enough. Unsworth (2013) argued that relative exposure should be

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

201

measured by computing what she has termed cumulative length of exposure (CLE). Into this computation Unsworth introduced the proportion of time spent with different interlocutors on different activities as well as the proportion of target language use within these different situations on both weekdays and weekends. The calculation takes into account children’s total waking time. This made it possible to calculate the current percentage of time per week that children were exposed to Dutch, the SL. To capture the variation in relative exposure that may occur over time, cumulative length of exposure was calculated, applying the same procedure for all previous years. Once the percentage of exposure to the SL per week was available for each year, the proportion of the year in which the child was exposed to the SL was calculated. Summing up these numbers yielded the cumulative years of exposure to Dutch, from which the years of exposure to the HL were derived as well. While this computation did not address the quality of the language, it refined the quantitative measures available, yielding measures for present and past exposure. In addition to introducing the new CLE measure, Unsworth’s contribution has also been to suggest that the effects of exposure differ across linguistic domains, a position that supports generativist predictions that some parts of the grammar, namely those belonging to Universal Grammar, are more impervious to input and exposure effects than others. Unsworth’s (2013) study was unique in testing a wider age range than any other study reported earlier in this chapter (the 136 simultaneously bilingual children studied ranged between 3 and 17 years of age). It explored children’s production of lexical and grammatical aspects of gender in their SL, Dutch, as well as their grammaticality judgments related to gender. Given that the children were simultaneous bilinguals (i.e., in addition to Dutch, children had been acquiring English from birth), chronological age, AoO, and LoE for both languages were identical. CLE and current relative exposure were the best predictors for children’s gender marking on determiners (which is partly lexically based) but not directly for gender marking on adjectives, which is much less lexically based. Thus, this study suggests that the relation between relative exposure and development may not be constant across all domains and features of language. In a subsequent study, Unsworth (2014) further pursued the possibility to tease apart the impact of relative exposure and the impact of chronological age using CLE, while also addressing the potential differential impact on lexical versus syntactic knowledge. Unsworth did this in reference to the acquisition of gender on Dutch definite articles and the acquisition of scrambling (a morphosyntactic phenomenon involving compositional semantics and syntax). Based on data from 109 children between ages 5 and 17 acquiring Dutch and English from birth in the Netherlands, Unsworth demonstrated that the effect of exposure differs across linguistic domains. Exposure effects were observed for the

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

202

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

acquisition of gender marking on definite articles, while no such association was observed for the acquisition of scrambling. This suggests that the latter does not depend on the relative amount of exposure but on the availability of the right kind of input. Similar findings showing a different impact of relative exposure (but now just of LoE) on lexical knowledge versus syntactic knowledge are presented by Armon-Lotem, Walters, and Gagarina (2011) in a study of 65 Russian–German and 78 Russian–Hebrew bilingual children between ages 3.9 and 7.2. Children had the ability to repeat complex syntactic structures in the SL at the monolingual norming level after two years of SL exposure, while their lexical knowledge in the SL still was below monolingual norms after four years of exposure (see Treffers-Daller, Chapter 15, this volume, for discussion on using monolingual norms with bilinguals).

5 Moving beyond Relative Exposure: Absolute Exposure The studies in Section 4 show that relative exposure has differential effects on vocabulary, phonology, and morphosyntax. Yet it must be remembered that relative exposure is based on parental reports of the proportion of exposure. This presents two weaknesses. One is that proportional use of one language relative to the other may not be a faithful reflection of the actual, absolute exposure a child gets in each language. The other is that relative exposure, and in fact any self-reported measure, does not include any information about the linguistic content of the exposure children receive. Just by the proportion of exposure and without a focus on some specific aspects of language it is not obvious that relevant grammatical structures are available to the child. In what follows, we turn to a consideration of each weakness. The first weakness is well illustrated in a reanalysis of third-party observational data: De Houwer (2014) showed that the proportion of utterances that Inuktitut–English bilingual children heard in each language did not align with the actual number of utterances children heard in each language. Likewise, Nakamura (2015) found that a mother of Thai descent in Japan reported using solely Japanese with her child. However, actual observations revealed that 13.5% of maternal utterances addressed to the child at age 1.2 were in Thai, decreasing to 4.6% by the time the child was 2.5 years old. Thus, relative exposure to a language reported via parental questionnaires cannot be automatically transferred into the absolute amount of exposure to particular linguistic input. To address this problem, it is necessary to evaluate the absolute exposure children get in each language. In order to obtain richer data to elucidate the ties observed between actual quantity of exposure and bilingual children’s performance in their two languages, De Houwer and Bornstein (2003) developed the Language

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

203

Input Diary (LID). The LID is meant to yield an online record of who children interact with and in what language (main activities may be recorded as well). The LID is meant to be filled in by whatever caregiver happens to be present at a given time. Information is collected for 30minute blocks a day, between 6 AM and 9 PM. Rather than being a one-time assessment, the LID is meant to be used regularly. Based on LIDs for several Dutch–French bilinguals covering every eighth day between the child ages of 5–20 months, De Houwer and Bornstein (2003) showed that the amount of time that children heard each of their two languages fluctuated considerably over time. As done in many previous studies under Erika Hoff’s direction carried out in Florida, United States, Place and Hoff (2016) used the LID developed by De Houwer and Bornstein (2003) to assess input quality on top of quantity. They measured the number of 30-minute blocks in which 90 Spanish–English bilingual toddlers heard just the HL (Spanish), just the SL (English), or a combination of both. This made it possible to get a more accurate measure of children’s daily linguistic exposure to each language (Place and Hoff derived relative exposure based on this absolute exposure measure). Place and Hoff were further able to find out how many speakers used each language with children, which parents self-reported being native speakers of the HL and/or the SL, and how much mixed input the child received from main caregivers (their Table 3 shows just how much variation there was among families on all these measures). These rich measures showed that 94% of the exposure to the HL was provided by native speakers, while for the SL this was true for only 40% of SL exposure (half the children had two parents who were native speakers of the HL). Thus, even in cases where relative amounts of exposure happen to be equal, the quality may not be. The explanatory value of Place and Hoff’s findings demonstrates the importance of more detailed measures of absolute exposure for younger bilinguals. While LIDs give very detailed information, they are harder to collect than relative exposure or CLE. LIDs are time-consuming to fill in and require a longitudinal commitment from parents. In an attempt to obtain some of the precise information captured by language diaries, Thordardottir, Rothenberg, Rivard, and Naves (2006) developed a detailed questionnaire to capture how many hours a week children were exposed to each language in the bilingual home and outside. Parents were asked to provide an estimation (in number of hours) per week for every person children were communicating with. This measure was added to the hours spent in daycare every year, as well as in other extracurricular language-oriented activities, taking into consideration the proportion of time each language was used for these activities. Ideally, absolute exposure can further be captured through direct observation of caregiver–child interactions. Quantitative measures of absolute exposure concentrate on the amount of child-directed speech (e.g.,

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

204

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

number of turns, utterances, word tokens, syllables), while more qualitative measures grasp the richness of exposure that is available to the child (e.g., mean length of utterance, number of word types). Quality of absolute exposure further pertains to the availability of specific morphemes or syntactic structures, which can serve as input for acquiring a particular kind of linguistic knowledge. In a rare study utilizing quantitative measures of absolute exposure in bilinguals, De Houwer (2014) compared Dutch maternal child-directed speech to monolingual and Dutch–French simultaneous bilingual toddlers. Data were collected longitudinally at the ages of 13 and 20 months in free play sessions (with identical toys and picture books) and during feeding time; 10–15 minutes from each session were transcribed to form the corpus. Bilingual children received as much exposure as monolinguals for all 13 measures of absolute exposure investigated (e.g., number of syllables, average utterance length). Furthermore, average hourly maternal speech rates were similar across groups: 2,715 words per hour for bilinguals and 2,673 for monolinguals, with extensive variation across groups (bilingual: 1,314–5,640; monolingual: 723–4,407). De Houwer (2014) did not, however, report on children’s language use. Marchman, Martı´nez, Hurtado, Gru¨ter, and Fernald (2017) carried out the first group study to compare relative and absolute frequency of exposure and their effect on bilingual children’s language use. Using interviews with caregivers, information on relative measures of exposure was obtained for 18 Spanish–English toddlers born and raised in California. Data on absolute exposure were obtained via recorded naturalistic interactions between children and caregivers for a total of at least eight hours per child. Most families reported extensive exposure to Spanish (80%), with much less exposure to English (20%). These relative measures of exposure masked substantial variability between families in the number of actual words children heard in each language. Thus, for example, of two children whose parents reported 80% exposure to Spanish, one child heard 190 Spanish words per hour, while the other heard as many as seven times more Spanish words per hour (cf. the observation in De Houwer, 2014). Marchman et al. (2017) found only moderate correlations between reported relative exposure to each language and children’s language abilities in each, while absolute measures of exposure (the number of words per hour heard by a child in each language) more consistently predicted children’s language abilities in each language, thus demonstrating the explanatory advantage of measuring absolute exposure. In a detailed case study, Nakamura and Quay (2012) considered measures of both absolute exposure and absolute input in evaluating the linguistic performance of a simultaneously bilingual English–Japanesespeaking child between ages 1.3 and 2.3 growing up in Japan who heard English only from his multilingual Malaysian mother (who self-reported as having near-native English) and Japanese from other interlocutors. For

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

205

absolute exposure they showed a correlation between the number of words and utterances heard in each language and the child’s vocabulary in that language. The child heard double the number of words in English, which was the dominant language of the child, than in Japanese. For absolute input they demonstrated how the abilities of the child in each language in terms of different types of words and tokens is related to the number of types and tokens in the input. Thus, for example, the child showed noun dominance in his lexicon only in English, in which a naming game was frequent in her input. So far we have seen that relative exposure, and in particular cumulative exposure, is found to be a better predictor of bilingual development than AoO or LoE. The effect of relative exposure seems to be strongly associated with the acquisition of the lexicon while it is less so with the acquisition of morphosyntax. Currently, there is ample research using relative exposure as a predictor of bilingual development, but studies of absolute exposure are still underrepresented. The few studies that have compared two measures of exposure (relative and absolute) show that relative exposure does not always grasp the actual amount of exposure children receive in their HL and SL. Thus, studies using different measures of exposure are bound to provide better understanding of bilingual development in both languages and the impact of exposure at different linguistic levels.

6 Input: Quantity and Quality Having available actual recordings of speech addressed to bilingual children makes it possible to address the relation between absolute exposure to specific linguistic constructions and bilingual development. We use the term (absolute) input to refer to the exposure to specific linguistic constructions. Paradis and Navarro (2003) probed the question whether certain grammatical characteristics of bilingual children’s language use might be attributed to the token frequency of specific linguistic constructions they hear. They looked at subject realization in Spanish (overt versus null subjects) in the spontaneous speech of a bilingual Spanish–English child and two monolingual Spanish-speaking children. Compared to the monolinguals, the authors found a higher percentage of overt subjects in the bilingual child’s speech. The reason might be found in crosslinguistic influence from English, which does not allow null subjects. However, in the Spanish input from especially the bilingual child’s mother, who had Spanish as an L2, the proportion of overt subjects was significantly higher compared to the parental input to the two monolingual children. This finding led Paradis and Navarro (2003) to call for future investigations of the actual input available to bilingual children exposed to an HL.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

206

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

Chan (2010) considered the role of absolute input for the acquisition of Cantonese double object constructions by bilingual Cantonese–English children compared to monolingual Cantonese-speaking children. As argued by Chan (2010) on the basis of comparisons of earlier published findings, bilingual Cantonese–English children take longer to acquire canonical word order in these constructions than monolingual peers. Chan’s research design inspected evidence for absolute exposure to specific constructions but was based on the assumption that balanced bilingual children get half of monolinguals’ exposure in each language, a relative frequency logic (but see De Houwer, 2014). Based on 15 hours of speech in Cantonese directed at bilingual children (18,185 utterances) and 30 hours of speech directed at monolingual children (32,467 utterances), Chan found different absolute input frequencies for double object constructions, with a total of 496 in the bilingual corpus versus 728 tokens in the monolingual corpus (not reported by Chan is that the proportions across the corpora are very similar, with 2.73% in speech to bilinguals and 2.24% in speech to monolinguals). More interestingly, the absolute frequency of double object constructions with canonical word order was just 36 in the bilingual corpus but 107 in the monolingual one (not reported by Chan is that the proportions here are different as well, with 7.3% of double object constructions having canonical word order in the bilingual corpus but 14.7% having so in the monolingual corpus). Thus, the bilingual children heard canonical word orders of Cantonese double object constructions less frequently than their monolingual peers, which may be why the former were found to take longer to acquire the structure. Further but indirect evidence of the role of the specific characteristics of the input comes from studies like that of Bialystok, Luk, Peets, and Yang (2010), who found that primary school-aged bilingual children with different HLs acquiring English as an SL through school in an Anglophone part of Canada performed comparably to monolingual SL-speaking peers on a test of English comprehension vocabulary as far as “school” vocabulary went. However, there was a gap between monolingual and bilinguals for “home” vocabulary. The asymmetry between school and home vocabulary reflects the type of input that monolingual and bilingual children are exposed to in the SL. While monolingual children have ample exposure to English at home and their input therefore provides labels for homerelated concepts, the input to bilingual children provides exposure to home-related concepts in their HL rather than the SL. Like the child studied by Paradis and Navarro (2003), many bilinguals in today’s mobile world may receive exposure to the SL from one or both nonnative-speaking parents. Other children also get exposure to the SL (or the HL) through nonnative-speaking caretakers such as nannies. Place and Hoff (2016) among others have considered the nonnative status of parents and caretakers as a variable that may characterize the quality of input, and have suggested that the nonnative quality of the exposure at home is less

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

207

supportive of SL development (see also Serratrice, Chapter 1, this volume). However, this and other studies have simply classified parents and other caretakers as either native or nonnative. We have argued that LoE and selfreports of relative exposure, or even CLE, can never be a good substitute for direct observation of input in terms of specific features of language. The same must be said of a binary categorization into “native” and “nonnative” input providers. Nativeness status can hide large differences in levels of proficiency in a language and great variation in the actual amounts and quality of child-directed speech. It cannot be assumed that a nonnative speaker’s input is “faulty” just because they are classified as nonnative. As a further complication, a particular language may be spoken to children by native and nonnative speakers alike (as was the case for the bilingual child studied by Paradis & Navarro, 2003). These are all additional reasons for studying the actual input directed at children, regardless of preconceived notions of “nativeness.” While Place and Hoff (2016) have suggested that the nonnative quality of English exposure at home is less supportive of English development in very young simultaneous bilinguals, Dulay, Tong, and McBride (2017) found that the use of English as spoken nonnatively to older sequential bilingual children by foreign domestic helpers in fact had a beneficial effect on children’s English language development. In a rare study of the actual input to a bilingual child involving nonnative input at home, Nakamura (2015) found that (infrequent) morphosyntactic errors in Japanese that a Thai mother made in interaction with her young son were not picked up by the child. The error analysis revealed that the errors in the mother’s and the child’s speech had different characteristics. Several other case studies indicate that even when they are exposed to nonnative parental input, children can achieve high levels of proficiency in that language. Yet, for deaf children, nonnative input may have a negative effect. Lu, Jones, and Morgan (2016) found that deaf children who received sign input from hearing parents who were beginning nonnative signers had less welldeveloped sign and phonological handshape repertoires than deaf children who had native sign input from deaf parents. These parents, who are committed to learning a sign language as a response to the unexpected deafness of a child in their family, had developed only beginning levels of proficiency. This is likely to be the situation in many hearing families engaged in learning a sign language, at least initially. In addition, if cochlear implantation is eventually chosen, some hearing parents may reduce their use with their child of the sign language they are learning. Thus, studies of parents who reach good proficiency in a sign language that they continue to use with their children beyond a few months or even a few years would be needed in order to better evaluate the impact of nonnative signing input on children’s development of a sign language (see Tsang & Sze, Chapter 24, this volume, on sign language and bilingualism).

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

208

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

In summary, there are few studies investigating absolute exposure on a child’s linguistic abilities in the HL and the SL. Even fewer studies have considered quantity and quality of input available to children in the HL and the SL and how input characteristics affect the development of specific linguistic phenomena. Child-directed speech to bilingual children might contain qualitative differences not observed in the input of monolingual children. Thus, bilingual children’s apparent divergences from baseline monolingual grammars might merely be a reflection of the available input. Yet some studies demonstrate that such divergences do not always occur. The available evidence shows, furthermore, that nonnative input need not jeopardize SL acquisition.

7 Code-Mixing in Child and Adult Speech Code-mixing is a unique feature of bilingual speakers’ speech. Studies of code-mixing in child-directed speech provide a test case for the relation between relative exposure, absolute exposure, input, and children’s performance, while taking into account the impact of parental language choices. Code-mixing as understood here encompasses the alternate use of two languages within a conversation. The Modeling Hypothesis (Comeau, Genesee, & Lapaquette, 2003) suggested that children learn to code-mix from their parents and that the frequency with which they codemix reflects the input they get. This has been explored for different spoken language pairs and for bimodal acquisition (Tang & Sze, Chapter 24, this volume). Yet the relation between the rate of code-mixing in parental input and the rate of code-mixing in children is not clear. Some studies show that children’s code-mixing rates directly mirror parental codemixing rates. Other studies find little or no support for this. Conversely, some studies link the frequency of code-mixing in children’s speech to their level of language proficiency rather than to the characteristics of parental input (McClure, 1977). The relevant studies, however, consist mostly of case studies or, as indicated by Byers-Heinlein (2013, p. 43), of relatively small samples in the case of group studies, or homogeneity in parental mixing. These limitations make it difficult to see larger patterns of influence. In a study investigating the influence of relative exposure to code-mixed input on child SL English vocabulary for 168 children between ages 1.5 and 2.3, Byers-Heinlein (2013) found that parental use of code-mixing had a detrimental effect on children’s vocabulary learning. A Language Mixing Scale probed parents’ mixing behavior in interactions with their children. For example, parents were offered statements such as “I often start a sentence in English and then switch to speaking another language” and then were asked to provide a score for that statement on a scale from 0 to 6 (with 0 indicating “Not at all true of me” and 6 indicating “Very true of

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

209

me”). Most parents reported that they code-mixed while interacting with their children. Users of just a single language were in the minority: only 14% of parents reported using a single language at least 90% of the time. In addition to the Language Mixing Scale, relative exposure to the SL was measured using parental reports. Parental code-mixing was a significant negative predictor of children’s SL comprehension vocabulary at age 1.5 and their SL production vocabulary at the age of 2 years even after controlling for relative exposure to the SL. Thus, exposure to code-mixing was associated with smaller English lexical repertoires (production at age 1.5 showed no relation with any measures of exposure). Altman, Burstein Feldman, Yitzhaki et al. (2014) relied on sociolinguistic interviews with both parents and children to study the relation between parental orientations toward the HL Russian and the SL Hebrew and their children’s self-reported code-switching practices and linguistic abilities. Children’s ages were between 4.6 and 6.9. Parents’ language orientations were classified as strict pro-HL (with parents, e.g., insisting that children speak just the HL at home), mild pro-HL (with parents, e.g., preferring the HL but allowing use of the SL at home), and pro-bilingual (with parents actively supporting children’s use of both the HL and the SL). Like in the Byers-Heinlein (2013) study, code-switching was measured using the Language Mixing Scale, but now for children (e.g., “Has it ever happened to you that you started speaking in one language and then switched to the other one?”). Children who gave a positive answer were asked to mark on a “ladder” reflecting a 0–10 scale of how often they switched from the HL to the SL and vice versa. Parental orientations were associated with children’s self-reported language proficiency: Children from strict and mildly pro-HL families reported Russian to be the language they spoke best. Interestingly, however, regardless of their parents’ language orientation, all children reported switching more from the HL to the SL than the other way round. Mishina-Mori (2011) looked at the influence of quantitative and qualitative aspects of both maternal and paternal input in relation to code-mixing by two (nonrelated) bilingual English–Japanese two-year-olds residing in the United States, where English is the SL. The data were based on recordings of mother–child and father–child naturalistic interaction for a period of approximately one year. Both sets of parents used hardly any codemixing but spoke only a single language to their child, thus following the one-person one-language (OPOL) strategy. Although one child reflected her parents’ low proportion of mixing, the other child used a lot of English (47% on average) in speaking to his Japanese-speaking mother. Thus, one has to look beyond parental language choice to explain children’s codemixing rates. Following Lanza (e.g., 1992), Mishina-Mori (2011) suggested that parental discourse strategies can also impact children’s language choice, that is, if children are not socialized into using only a particular language they may use the other one (see also De Houwer, Chapter 17, this volume).

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

210

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

While most studies look at the influence of relative exposure of parental input on children’s code-mixing, Comeau et al. (2003) assessed the link between input characteristics of interlocutors who are not members of the family on children’s use of code-mixing. They manipulated the rate of code-mixing in research assistants’ speech in interactions with Frenchand English-speaking two-year-olds. The bilingual children were sensitive to the rate of code-mixing in the interlocutors’ input and adjusted their own code-mixing accordingly, on a nearly turn-by-turn basis. Bimodal acquisition (Tang & Sze, Chapter 24, this volume) offers yet another angle for investigating the impact of code-mixing in child-directed speech on children’s code-mixing. Petitto, Katerelos, Levy et al. (2001) found that hearing bimodal children acquiring Langue des Signes Que´be´coise (LSQ) and French from birth simultaneously spoke and signed as a function of the rate with which their caregivers did so. For example, a child whose mother spoke to her only in French used very little modality mixing (also known as code-blending); by contrast, a child whose parents often mixed sign and spoken modalities also showed high modality mixing rates. In a similar vein, Baker and van den Bogaerde (2008) compared the absolute input of four deaf mothers to deaf and hearing children with children’s language choice. Children were acquiring Sign Language of the Netherlands (NGT) and spoken Dutch. Mothers used only Dutch, only NGT, or code-blends combining both speech and signing. Mothers used relatively more code-blended utterances in addressing their hearing than their deaf children. This found reflection in the fact that hearing children used more code-blended utterances than their deaf peers. In an opposite design in which each of eight hearing children acquiring Finnish Sign Language (FinSL) and spoken Finnish interacted with different interlocutors, Kanto, Laakso, and Huttunen (2015) investigated language choice and code-blending. Similarly to many bilingual children who acquire two spoken languages (e.g., Comeau et al., 2003), the bimodal children adapted their language choice and the rate of code-blending according to their different interlocutors’ input characteristics. Children did not code-blend when communicating with hearing, non-signing adults, but while communicating with deaf adults they used both spoken and signed modalities. Similar findings were also reported in Lillo-Martin, de Quadros, Chen Pichler, and Fieldsteel (2014). This section explored whether children’s code-mixing is linked to properties of child-directed speech. We reviewed studies reporting on different indices measuring density of code-mixing (oral or sign) in child-directed speech: relative exposure (by means of questionnaires) and absolute exposure (by measuring frequency of code-mixing though direct observations). The studies reviewed here demonstrate conflicting evidence. On the one hand, code-mixing in child output seems to be tied to parental code-mixing strategies; on the other hand, not all studies find support for this.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

The Nature of Exposure and Input in Early Bilingualism

211

8 Conclusion Language acquisition crucially depends on the quality and quantity of the language the child is exposed to. Effects of exposure on development have been found to be robust in bilingual acquisition for both the heritage and the societal language. Yet it is not trivial to tease apart the quantity and quality of exposure to each language and the input that young sequential bilingual children receive. Different measures are used to capture variation in quantity and quality of exposure and input (e.g., age of onset, length of exposure, percentage of exposure to the heritage and the societal language, or the number of words in each language produced by a caregiver). This chapter promoted a sharper distinction between relative exposure as measured by parental questionnaires and diaries, absolute exposure based on recordings of caregiver–child interactions, and input, which refers to the availability and frequency of specific linguistic information. The method by which exposure and input are measured matters a great deal, with some methods offering advantages over others. Relative exposure and in particular cumulative exposure are better predictors of bilingual development than age of onset or length of exposure. Both relative and cumulative exposure take into consideration the linguistic experience children have in each language relative to the other, and not just the number of years from the onset of exposure to a (new) language. What relative exposure cannot explain, however, is the absence of a relation between relative exposure to a societal language at home and its acquisition. Studies of the quality of absolute exposure and the linguistic input have shown that nonnative input does not endanger societal language acquisition and cannot explain this puzzle either. However, studies of absolute exposure have been able to explain it by showing that parents’ reported use of the societal language at home does not reflect the actual exposure at home in terms of the number of words and utterances available in each language. It is clear that each measure of exposure on its own cannot fully predict bilingual language acquisition. Furthermore, the studies presented in this chapter demonstrate that measures of exposure, be it relative or absolute, are good predictors of lexical and phonological knowledge, but are not as good for predicting the acquisition of morphosyntactic structures. By contrast, input characteristics that capture the presence of a specific grammatical structure in childdirected speech provide better evidence on how these structures are acquired. Crucially, the effects of exposure appear to be different for the societal language and the heritage language: Continued good heritage language development is more likely in homes that maintain high levels of its use, in contrast to homes that make relatively more frequent use of the societal language. Furthermore, the impact of exposure on children’s

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

212

S H A R O N A R M O N - LOT E M A N D N ATA L I A M E I R

language development can be mediated by children’s own motivation for acquiring the societal language and maintaining the heritage language. While the present chapter addressed the possible impact of exposure to code-mixing, a unique bilingual feature, other unique features of bilingual acquisition that impact the exposure and input that bilingual children receive are addressed in other chapters in this volume: parental discourse strategies might determine not only family language policy (Gonc¸alves, Chapter 3, this volume) and relative exposure to each language but also attitudes to the use of code-mixed and blended utterances (De Houwer, Chapter 17, this volume). There are still some open questions that require the attention of researchers, especially in terms of the qualitative characteristics of bilingual child-directed speech and how these properties are linked to children’s linguistic development in the heritage and the societal language. The age ranges that are looked at might impact the relation between exposure and input and children’s linguistic performance, as does the consideration of continuity versus disruptions in exposure, which was not addressed in this chapter, but requires further research. Moreover, to the best of our knowledge, studies have yet to investigate the properties of child-directed speech in bilingual contexts longitudinally. Little is known about how bilingual parental input changes over time based on children’s ages, their communicative demands, or length of exposure to the societal language. Researchers also need to draw their attention to the input available to bilingual children in the heritage language and how this input is different from or similar to the one available in the country of origin. Furthermore, for children’s acquisition of the societal language, the effects of nonnative and code-mixed input are not clear and need further exploration. Currently, there is ample research using relative exposure as a predictor of bilingual development, but studies of absolute exposure and input are still underrepresented. Studies using two or three different measures are bound to provide an even better understanding of bilingual development in both languages and at different linguistic levels.

Downloaded from https://www.cambridge.org/core. Durham University Library, on 13 Nov 2018 at 15:13:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.011

11 Becoming Bilingual through Additive Immersive Programs Maria Juan-Garau and Roy Lyster

1 Introduction Parents who want to provide their children with opportunities for becoming bilingual through school-based programs have an increasing number of options at their disposal. These options include additive programs designed to foster bilingual outcomes such as immersion education, content and language integrated learning (CLIL), and study abroad (SA) programs. These three options are the focus of this chapter. We place particular emphasis on school-based programs for immersion and CLIL, which aim for additive bilingualism by offering content-driven curricular instruction in at least two languages. Typically, one of the languages of instruction is the dominant language in the community and likely to be the home language of many students, while the other language of instruction is an additional language that students might be learning as a second, foreign, heritage, or indigenous language. In addition to immersion and CLIL programs as potentially effective means of becoming bilingual in classroom settings, SA programs are increasingly seen as a promising complement to in-school language learning. We therefore include in this chapter a secondary focus on SA as a programmatically designed opportunity for an immersive experience. Immersion and CLIL programs have proven to be a more effective way of developing bilingual proficiency than foreign language programs that focus on language primarily as an object of study (for a review of the latter, see Mun˜oz & Spada, Chapter 12, this volume). One of their most attractive features from a learning perspective is their increased exposure to and engagement with the target language (TL). In addition to more time on task, the effectiveness of these programs has been attributed to their being content-driven, which provides a motivational basis for purposeful

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

214

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

communication and a cognitive basis for language learning. That is, the depth of processing required for learners to engage with subject matter taught through the TL is thought to enhance TL development. For these programs to be effective, however, the conditions for their implementation need to be favorable, and these include opportunities for professional learning on the part of teachers and the availability of instructional resources. The chapter is organized as follows. We first define immersion, CLIL, and SA programs. We then summarize research results on the language learning outcomes and affective factors influencing TL use across these different contexts. We conclude with some discussion of challenges and suggestions for future directions.

2 Program Options 2.1 Immersion School-based, one-way immersion programs aim for additive bilingualism by offering students at the elementary level at least 50% of the curriculum through the medium of an additional language that they are learning as a second, foreign, heritage, or indigenous language (i.e., the immersion language). The follow-up at the secondary level to such immersion programs is usually called a continuation program and includes a minimum of two subject courses in the immersion language. Immersion programs have been adopted in some countries to promote the learning of a second coofficial language. Examples of these include French immersion in Canada, Swedish immersion in Finland, Catalan immersion in Spain, Basque immersion in Spain, Welsh immersion in Wales, and Irish immersion in Ireland. Use of the term “immersion” to refer to a given language or program, however, may not always be sufficiently inclusive of all learners in instructional settings where both first language (L1) and second language (L2) speakers of the TL coexist. Such programs may be more aptly referred to, for example, as Welsh-medium or Irish-medium programs. There is considerable variation in the proportion of the curriculum devoted to the immersion language in one-way immersion programs. Those following the Canadian model typically begin in kindergarten, which in the Canadian education system means at age five, with 100% of the curriculum taught in the immersion language (called “total immersion”) followed by a gradual decrease to 80% around age seven in grade two, when majority-language instruction is introduced, and then to 50% around age 10 in grade five, when the other half of the curriculum is delivered through the majority language. Still other programs (called “partial immersion” or bilingual programs) begin with a 50/50 distribution of the two languages from the beginning in kindergarten. In Irish-medium and Catalan-medium programs, however, the proportion of instruction in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

215

the immersion languages remains much higher, at about 85%, throughout the program. There is also variation in entry points, with early immersion beginning in kindergarten or grade one (i.e., age five or six), mid-immersion beginning around grade four (i.e., age nine), and late immersion beginning around grade seven (i.e., age 12). Mid-immersion and late-immersion programs thus include students who are already schooled in L1 literacy and have usually been exposed to some instruction in the L2 as a regular subject. Early total immersion involves the teaching of literacy skills first in the L2 followed by instruction in L1 literacy around age seven or grade two, whereas early partial immersion often introduces literacy instruction simultaneously in both languages. In two-way immersion programs in the United States, a similar number of children from two different L1 backgrounds are provided curricular instruction in both languages: in English and a minority language, which in most cases is Spanish but also includes Chinese, Vietnamese, Russian, Japanese, French, German, and Italian. Two-way immersion programs typically begin by allocating either equal time to both TLs (50/50) or 90% to the minority language (e.g., Spanish, Mandarin) and 10% to English. By at least grade five, when students are 10 years old, most two-way immersion programs are 50/50. Also found in the United States are oneway developmental bilingual programs serving language-minority students (e.g., Spanish speakers) or simultaneous bilinguals who learn the curriculum through the minority language and English. Because a defining characteristic of immersion is its aim for additive bilingualism (i.e., learning an additional language with no threat of losing the L1), programs that do not maintain parallel instruction in at least two languages (such as the so-called structured English immersion programs in the United States) are not considered real immersion programs. Classrooms in which minoritylanguage students find themselves without any L1 support and the majority language is the sole language of instruction may be more accurately referred to as mainstream or even “submersion” rather than immersion (see Garcı´a & Tupas, Chapter 20, this volume). In France, school-based immersion programs (known as “classes bilingues”) have been designed to deliver at least half the curriculum through the medium of regional languages such as Occitan, Basque, and Catalan in the south, Corsican on the island of Corsica, Breton in the northwest, and German in the eastern regions of Alsace and Moselle. Immersion programs have also been implemented to revitalize and maintain indigenous languages such as Maori in New Zealand and, in the United States, Hawaiian, Ojibwe, and Cherokee, drawing on a range of program models with varying degrees of instruction in the indigenous language. Immersion programs are also offered in a range of foreign languages (e.g., Mandarin, French, German) in countries such as the United States and Australia. English as a foreign or international language is the TL of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

216

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

a variety of content-based programs ranging from early immersion in Japan and Brazil to English as a medium of instruction (EMI) in Hong Kong secondary schools and Chinese–English bilingual programs in China.

2.2 Content and Language Integrated Learning (CLIL) In school-based CLIL programs, students often study one or two subjects in the TL, usually in tandem with a foreign language or language arts class. Many CLIL programs begin in secondary school and usually offer less than half the curriculum in the TL, but there is a great deal of programmatic variation. Dalton-Puffer (2017) aptly describes CLIL as “a multitude of forms and models, ranging from individual didactic units lasting a few lessons to several subjects of the regular school curriculum taught in the target language over several years” (p. 153). She refers to (a) content-driven forms of CLIL, which are similar to one-way immersion programs, and (b) language-driven forms of CLIL, which are more theme based and have TL development rather than content learning as their main objective. Given the wide diversity in characterizations of CLIL, many now use “CLIL” as a generic cover term for bilingual, content-based teaching, and it is thus considered synonymous with terms such as “bilingual education” or “content-based instruction.” However, the extent to which languagedriven, theme-based forms of CLIL aim for additive bilingualism is questionable in light of Banegas’s (2014) content analysis of CLIL-related sections of English as a foreign language textbooks marketed for use in secondary schools in Argentina. He found that the CLIL components were “superficial supplements rather than a meaningful attempt to promote weak forms of bilingual education” (p. 345). He concluded by urging publishers to link content to the L1 curriculum in cohesive ways that take topic complexity and cognitive development into account.

2.3 Study Abroad (SA) The SA context typically combines TL instruction in a classroom setting with many opportunities for contact with native speakers in the TL community. SA shares some features of immersion and CLIL, with its emphasis on TL exposure and interaction. In contrast to immersion and CLIL, however, most of the research on SA has been conducted at the college level, comparing the learning outcomes of students participating in SA programs with the outcomes of “at home” (AH) students, whose exposure to the TL is often limited to formal language classrooms with little or no extramural contact. The relative success of SA programs can be related to their variable characteristics. To begin with, stays can vary in duration, ranging from short term (e.g., a month) to long term (e.g., a year). Longer stays have

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

217

generally been found to be more beneficial, although even very short programs have also been seen to advance linguistic competence in some respects (Churchill & DuFon, 2006). SA living and organizational arrangements constitute another variable to take into account. In sheltered programs, participants usually travel to their destination with an accompanying tutor, tend to stay closer together, and engage in activities designed exclusively for them. In non-sheltered programs, however, students are more likely to be housed and educated alongside host families and host country or international students, rather than mingle with fellow nationals. Generally speaking, the latter option provides more opportunities for authentic interaction in the TL with the local community, which may lead not just to the development of communicative abilities but also to enhanced intercultural sensitivity and a more enriching personal experience overall. Pre-departure language level may be another program feature to consider, under the hypothesis that a functional level of proficiency is needed for substantial gains abroad to accrue.

3 Learning Outcomes 3.1 Immersion With respect to L1 development and academic achievement, the accumulated evidence supports two general conclusions for early immersion programs serving majority-language students who wish to learn an additional language: the academic achievement of immersion students in subjects they study through the L2 is equivalent to that of nonimmersion students studying the same subjects in their L1, and their L1 development ranges from equivalent to superior to that of non-immersion students. These positive results emerged in evaluation studies of early French immersion programs in Canada (Harley, Cummins, Swain, & Allen, 1990) and have since been substantiated in Canada and also in the United States, where one-way and two-way immersion students perform as well as or better than their non-immersion peers on measures of academic achievement administered in English (Genesee & Lindholm-Leary, 2013; Tedick & Wesely, 2015). Similar results have been found in other early immersion programs for majority-language students learning an additional language such as Swedish immersion in Finland and English immersion in Japan. With respect to L2 development, research has demonstrated, as would be expected given the greater amount of time in the curriculum devoted to the immersion language, that early immersion students develop much higher levels of proficiency in the TL than do non-immersion students studying the L2 as a regular subject (i.e., for one period per school day). This is also true of students with learner characteristics that are disadvantageous with respect to academic and linguistic abilities: Genesee and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

218

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

Lindholm-Leary (2013) summarize research showing (a) that majoritylanguage students in one-way immersion who are at risk for academic difficulty achieve higher levels of proficiency in the TL than nonimmersion students with similar disadvantages studying the TL as a regular subject and (b) that English language learners (ELLs) who are at risk for academic difficulty in two-way immersion can develop proficiency in two languages and achieve in content areas such as mathematics to the same level as similar students in English-only programs at least by the end of elementary school. They conclude, however, with a call for more research on the suitability of additive immersive programs for at-risk learners and, more specifically, on the types of instructional practices that would best suit their needs. The L2 development benefits have been investigated in detail. In comparison to non-immersion students, one-way early immersion students develop (a) strong comprehension skills as measured by tests of listening and reading comprehension; and (b) high levels of communicative ability, but with lower-than-expected production skills in terms of grammatical accuracy, lexical variety, and sociolinguistic appropriateness (Harley et al., 1990). More specifically, in comparison to native speakers of French, Harley et al. found that French immersion students performed similarly on measures of discourse competence, including story retells, argumentation, and persuasion tasks, but were clearly less proficient on most grammar variables, which included verb and preposition usage, and on all sociolinguistic measures, especially in their use of singular vous and the conditional to express politeness. Comparisons of French immersion programs with different entry points show that early-immersion students tend to develop higher levels of L2 proficiency in comparison to late-immersion students (Wesche, 1993). However, the differences are not as great as one might expect. Advantages have been found for early-immersion students on measures of listening ability and fluency in oral production. Students from lateimmersion programs may catch up with early-immersion students in writing tasks and other measures requiring knowledge of formal language features. Some studies have shown that differences between early- and late-immersion students disappear altogether at the university level, although these findings need to be interpreted with caution, because lateimmersion programs attract a self-selected, academically successful group that may easily catch up with early-immersion students during secondary school (Wesche, 1993). For this reason, early immersion has been considered a more accessible option for a wider range of students. In other contexts, students in immersion-like programs with late entry points beginning at the secondary level have fared less well. Many factors other than only entry point may contribute to these outcomes, not the least of which may be the institutional constraints at the secondary level that militate against teacher collaboration and integration across content

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

219

and language classes. Specifically, research on the EMI program beginning at the secondary level in Hong Kong has shown that EMI students attain higher levels of proficiency in English than their non-EMI counterparts, but lower levels of academic achievement in science, history, and geography (e.g., Lo & Lo, 2014). In Malaysia, the EMI program implemented in secondary schools to teach math and science did not lead to higher proficiency levels in English and was discontinued in favor of a return to Malaymedium instruction in math and science (Hashim, 2009). Research on two-way immersion programs in the United States has compared the language outcomes of students beginning in programs with a 90/10 distribution of the two languages with those of students beginning in programs with a 50/50 language distribution. Interestingly, by the upper elementary grades, the English development and academic achievement for students in both types of programs are the same, whereas students whose program is initially 90/10 achieve higher levels of proficiency in the minority language (Tedick & Wesely, 2015). English L1 students continue to perform better in English than in Spanish, while Spanish L1 students tend to develop more balanced oral and written proficiencies in both languages, although some Spanish L1 students report greater facility in English owing to its societal dominance as a majority language. Studies of Mandarin immersion programs in the United States have also yielded positive results, in spite of the greater distance between languages. In their small-scale study of a 90/10 two-way Mandarin immersion elementary school in California, Padilla, Fan, Xu, and Silva (2013) found that, at ages seven and eight in grades two and three, non-immersion students from the same school achieved higher scores than the immersion students on the English language arts and math test, but by ages nine and ten in grades four and five, the Mandarin immersion students scored higher than their non-immersion peers. With respect to proficiency in Mandarin, the researchers found that students demonstrate what they deemed to be high-level performance in oral/listening, reading, and writing. In a largerscale study of one-way Mandarin immersion programs in four different school districts in Minnesota, Fortune and Song (2016) were able to examine the outcomes of programs that begin with 100% Mandarin in kindergarten or grade one (at ages six or seven), introduce English in grades two or four (at ages seven or nine), and then move to a 50/50 distribution of Mandarin and English by grade five (at age 10). The results showed that the Mandarin immersion students in grades three and five achieved as well as and, at times, better than grade-level non-immersion peers in English reading and math. Regarding their Mandarin, most grade five students demonstrated proficiency at intermediate levels with listening and speaking skills outpacing character writing and reading. Researchers and teachers alike have noted that early-immersion students appear to reach a plateau in their TL development around the same time that (a) more L1 instruction is introduced into the curriculum

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

220

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

and (b) preadolescents and adolescents increasingly use the L1 to communicate among themselves in class. This plateau effect was confirmed by Fortune and Tedick’s (2015) study of one-way Spanish immersion programs in the United States, which showed that the oral proficiency in Spanish of students in grade eight (aged 13–14) was not significantly different from that of students in grade five (aged 10–11). This suggests that, at least in the North American context, slowed rates of TL development result from a decrease in both exposure to and use of the TL, specifically in programs where the TL is a minority language such as French or Spanish and the L1 is English. In contrast, when students’ L1 is a minority language such as Spanish, spending more time in English in school does not necessarily result in higher achievement or proficiency in English, as shown by comparing the outcomes of Spanish L1 children either in allEnglish programs or in two-way immersion and one-way developmental bilingual programs in the United States (Genesee & Lindholm-Leary, 2013). Spanish L1 children in Spanish–English bilingual programs acquire proficiency in English as fast as or faster than their counterparts in all-English programs, despite the fact that they have significantly less exposure to English, while also attaining higher levels of functional proficiency in Spanish. As summarized by Genesee and Lindholm-Leary (2013): “The results of extensive research on the English language development of ELLs [English language learners] in a variety of programs do not provide evidence that the reduced exposure to English that ELLs experience in DBE [developmental bilingual education] and TWI [two-way immersion] compromises their English language development” (p. 16). This suggests that, when considering the distribution of time devoted to each language in additive bilingual programs whose languages of instruction include a minority language in addition to English as the primary societal language, more time can be allocated to the minority language without detrimental effects on English language development. This applies regardless of whether English is the students’ L1 or L2. All over the world, the demographics of today’s immersion programs are more linguistically diverse than ever. Canada offers an excellent illustration of this point. Whereas the first immersion programs included students from primarily English-speaking homes, the demographics of today’s programs reflect the growth of the country’s immigrant population and its motivation to learn both official languages. These demographic changes have led to research that seeks to find out what outcomes can be expected for transnational and diverse student populations. For instance, Mady (2015) compared the French and English proficiency of three groups at the grade six level (11–12 years of age): Canadianborn English-speaking students, Canadian-born multilingual students who also spoke the home languages of their immigrant parents, and immigrant multilingual students in the process of adding English and French to their home language repertoire. Interestingly, she found that

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

221

the immigrant students outperformed both the Canadian-born Englishspeaking students and the Canadian-born multilingual students on French proficiency measures of reading, writing, and speaking (but not listening), whereas there were no significant differences among the groups regarding English proficiency. These preliminary findings make an important contribution to the discussion of the suitability of immersion for all learners, responding to the concern of some educators that immersion might be too demanding for recent immigrants whose home language is not one of the school languages of instruction.

3.2 Content and Language Integrated Learning (CLIL) Evaluations of CLIL have mostly focused on its language outcomes and, to a much lesser extent, on content learning. Empirical findings regarding the former, which have proliferated since the 2010s, indicate that partaking in a CLIL program – similar to results of immersion and other duallanguage programs – generally helps learners reach higher levels of L2 attainment than if they follow a foreign language program on its own. Not all aspects of linguistic competence seem to be equally affected by CLIL instruction, however, as also found in immersion and SA contexts. Ruiz de Zarobe (2015) remarked that reading, comprehension vocabulary, speaking, writing (fluency and complexity), and some morphological phenomena are favorably affected by CLIL. By contrast, other aspects appear to be unaffected or insufficiently surveyed. They include listening, syntax, production vocabulary, writing (accuracy and discourse skills), informal/nontechnical language, pronunciation, and pragmatics. A substantial line of CLIL research has presented comparative data based on language test scores from CLIL and non-CLIL control groups. A number of these studies have been gathered by Dalton-Puffer, Nikula, and Smit (2010) and Juan-Garau and Salazar-Noguera (2015). The results of these evaluations reveal an overall advantage of CLIL learners over their nonCLIL counterparts. The findings must be interpreted cautiously, however, in recognition of certain methodological shortcomings. For example, CLIL participants tend to be convenience samples. It is often impossible to include a cohort that is properly matched in CLIL/non-CLIL comparisons, and the hours of instruction received in each program have not always been considered in such comparisons. Nevertheless, some researchers have begun to address these limitations, for instance, through procedures such as comparing learners with different ages but similar amounts of instruction. Regarding receptive skills, Admiraal, Westhoff, and de Bot (2006), among several others, found positive results for reading comprehension in English L2 with a sample of Dutch L1 CLIL students. These findings, nevertheless, have not always been paralleled in relation to listening comprehension (e.g., see studies in Dalton-Puffer et al., 2010), an area

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

222

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

that deserves more attention. Concerning productive skills, writing exhibits quite positive findings particularly in relation to overall competence, fluency, and complexity (Dalton-Puffer et al., 2010), and also in accuracy (Juan-Garau & Salazar-Noguera, 2015). Problems remain in textual competence and academic language functions, probably as a result of the little presence of writing activities in content lessons. Moving on to speaking, findings are quite positive too, indicating higher oral abilities for CLIL learners (e.g., Admiraal et al., 2006). The impact of CLIL programs on L2 pronunciation, however, appears to be limited. In terms of the language system, lexical development reaps the greatest benefits, although gains are more apparent in the technical than in the general and informal language registers, and in learners’ comprehension rather than production skills. As regards morphosyntax, research has produced inconclusive findings. A CLIL advantage has been shown in the use of tense and agreement markers and irregular verbs, while no advantage has been found in other components including null subjects, suppletive forms, and negation (e.g., Villareal & Garcı´a Mayo, 2009). Concerning pragmatics, research has mostly centered on classroom discourse. In recent times, interlanguage pragmatics in CLIL and non-CLIL settings has begun to be assessed as well, although more investigation is needed in this respect as well as in other aspects of oral production. Another fruitful line of inquiry in CLIL research has been the study of the contextual conditions under which language is used in CLIL classrooms and the discourse patterns that emerge from them. The analysis of classroom discourse in additive immersive programs is central to a more complete understanding of the interplay between language and content knowledge development. Examples of this line of research include DaltonPuffer’s (2007) volume on discourse in CLIL learning environments, which provides an in-depth analysis of teacher and student classroom interaction in secondary schools in Austria, bringing to light how students use and learn the TL. The author provides empirical data on the use of questions, academic language functions, directives, and repairs, among other discourse moves. She concludes that more explicit attention to language learning is needed in CLIL environments, a finding paralleled in immersion settings, so as to help students to gradually appropriate the target foreign language as a medium through which learning occurs. Similarly, Llinares, Morton, and Whittaker (2012) present interactional data collected in CLIL classrooms across four countries in Europe. They analyze the genres and registers through which meanings are enacted in different academic subjects at the preschool, primary, and secondary school levels. The interest in CLIL classroom discourse has also supported the adoption of genre-based curricula in many programs. One of the concerns most frequently voiced in relation to CLIL is the possibility of weak content learning derived from learners’ limited competence in the language of instruction. This issue remains

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

223

understudied. The little research that exists has produced contradictory findings depending on the educational and geographical areas considered (Ruiz de Zarobe, 2015). While some scholars have found no differences in the quality of content knowledge of CLIL students with respect to their non-CLIL counterparts (e.g., Admiraal et al., 2006), other researchers have uncovered negative effects of CLIL instruction on content learning (e.g., Dallinger, Jonkmann, Hollm, & Fiege, 2016). Even though the research findings with regard to content learning are inconclusive, it appears that CLIL learners are often able to attain content knowledge levels comparable to those of students learning through their L1, a finding that reflects the results of the immersion programs presented in Section 2.1. When studies have found comparable quality of content knowledge for CLIL and non-CLIL students, the explanation may be, at least in part, that in those programs CLIL teachers were able to deploy effective pedagogical strategies to help learners with emerging language competencies learn content better. This is noteworthy considering the scarcity of meaningful materials available to implement CLIL, as the study by Banegas (2014) on the inclusion of CLIL components in textbooks for young learners available to teachers in Argentina reveals.

3.3 Study Abroad (SA) In the early days of SA research with college students, noteworthy multiple-skill studies were conducted with North American and European university students that reported increased performance after SA. As Churchill and DuFon (2006) point out, however, most SA research has focused on learner gains in specific skills and forms. Attention has concentrated on various dimensions of oral skills, particularly speaking proficiency, reflecting expectations for gains in this domain. Overall oral proficiency has been found to register considerable gains abroad (e.g., Segalowitz & Freed, 2004). Oral fluency too has been examined with respect to various temporal and hesitation phenomena, revealing that SA learners increase the length and rate of their fluent speech runs (e.g., VallsFerrer & Mora, 2014), while reducing their pauses and disfluencies (e.g., Segalowitz & Freed, 2004). Listening has received little attention, revealing nonetheless progress in the ability to comprehend TL discourse. In the area of phonological development, findings from studies comparing AH and SA learners indicate that neither L2 pronunciation nor L2 speech perception benefits substantially from an SA period. By comparison with oral skills, very few studies have focused on the acquisition of literacy skills. Research on reading proficiency has shown that SA can have a positive impact on the development of this skill. When looking at writing development, recent studies offer evidence that SA can enhance this skill even in the case of a short stay (e.g., Pe´rez-Vidal &

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

224

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

Barquin, 2014), although a clear advantage for the SA context over AH settings has not been established. As regards the formal aspects of language, most comparative studies of grammatical development have found AH groups to be equal or superior to SA groups in the ability to monitor and accurately use grammatical forms. Concerning lexical development, research results point to larger and more nativelike vocabularies for SA than for AH learning contexts. Briggs (2015), however, found no relation between informal out-of-class language contact and vocabulary gain. The researcher concluded that learners should get guidance to manage informal language contact in order to maximize linguistic gains. The greater access to TL contact in SA environments, paired with the possibility to use it to interact in authentic communicative contexts, has been reported to have a positive effect on learners’ sociolinguistic and socio-pragmatic competences. With regard to sociolinguistic competence, for example, a study by Marriot (1995) investigated the variable use of politeness forms by L2 learners of Japanese, who exhibited an increase in their sensitivity to variation as a result of their SA experience in Japan. Regarding socio-pragmatic competence, knowledge of the appropriate contextual use of TL resources for the realization of various speech acts such as making offers and requests or providing advice has been shown to improve after SA experiences (Regan, Howard, & Leme´e, 2009). Results in this area also point to a greater use of TL formulations and awareness of their contextual appropriateness, despite considerable distance between TL and L1 norms (Regan et al., 2009).

4 Affective Factors and Their Influence on L2 Use in Additive Immersive Programs L2 development and use are mediated by the complex interplay between preprogram proficiency levels, instructional setting, TL community characteristics, and affective factors including attitudes, motivation, anxiety, and willingness to communicate (WTC). Accordingly, learners vary widely in their ability to benefit from a given immersive learning environment.

4.1 Immersion Social-psychological studies comparing immersion and non-immersion students demonstrated that French immersion students perceive less social distance between themselves and French native speakers, and develop more positive attitudes toward the TL and its native speakers (Genesee, 1987). However, this trend is short lived and has been more consistently documented with younger than with older students and early in students’ participation in the program but diminishing with

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

225

each grade level. In a comparative study of English immersion and nonimmersion students conducted in Japan, Downes (2001) found that in addition to more flexible cross-cultural attitudes the immersion students displayed a stronger sense of Japanese cultural identity than the nonimmersion students, indicating that bilingual programs can serve to strengthen rather than weaken one’s L1 cultural identity. Although many French immersion students in the Canadian context remain geographically remote from the target community, this is not the case in Montreal and Ottawa, where studies compared immersion and nonimmersion students with respect to TL use outside the classroom. In comparison to non-immersion students, immersion students in Montreal reported that they were (a) more comfortable and confident when using French with native speakers, (b) more likely to respond in French when addressed in French, and (c) less likely to avoid situations in which French was spoken. However, immersion students were not more likely than non-immersion students to actively seek opportunities for TL exposure by watching television, listening to the radio, or reading books in French (Genesee, 1987). A similar type of reactive use of French was observed among immersion graduates in the Ottawa area (Wesche, 1993). Also in this context, graduates of immersion programs that included contact with native speakers of French tended in their young adult lives to use French on social occasions and with neighbors, and to attend plays performed in French. Graduates of immersion programs that included access to French through activities outside the classroom reported having more positive attitudes toward using French and reported using it for reading and at work. Other studies, however, have noted that some French immersion graduates and junior high school students feel dissatisfied with their proficiency in French and are hesitant to use it, not seeing themselves as legitimate speakers of the language. A legitimate speaker of the language, explains Roy (2010), “means to be accepted by others in the group when you speak their language” (p. 543), yet she observed “a lot of anxiety among students, who believe that they will never speak as well as native speakers” (p. 556). This may explain the significant drop in participation rates in Canadian French immersion programs at the secondary level. As previously mentioned (see Section 3.1), there is both anecdotal and empirical evidence that French immersion students use increasingly more English as they progress through the program. Similar patterns of English language use have been found in Spanish immersion programs in the United States and in Irish immersion programs in Ireland. Research in Spanish–English two-way immersion settings in the United States has consistently shown a strong preference for the use of English by students regardless of their language background, particularly in interactions with peers. For example, Ballinger and Lyster’s (2011) study of two-way immersion at three different grade levels revealed that only the Spanish L1

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

226

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

students used their L2 (i.e., English) with peers and that their opportunities for L1 enrichment were mainly limited to interactions with teachers. Interestingly, the results of their study suggested that teachers’ expectations played a pivotal role in determining their students’ language choices. The Spanish teachers in a two-way immersion program described by Herna´ndez (2015) expressed concern about students’ use of English, which occurred in spite of their expectations, feeling that it affected the academic development of Spanish as well as the sociocultural identity and self-esteem of the Spanish speakers. In contexts where English is the majority language and is one of the school’s two languages of instruction along with a minority language (e.g., French, Spanish, Mandarin), there have been calls for students to be able to draw more freely on their L1 to support content learning, L2 development, and bilingual interactions (see arguments in Garcı´a & Tupas, Chapter 20, this volume). Such calls need to be cautiously interpreted with more concern for the specific sociopolitical contexts in which the proposal may be best applied or not. Whereas the benefits of L1 use by minoritylanguage students have been well documented, majority-language students (whose L1 is English or another language that is dominant outside the immersive program experience) are unlikely to benefit as much from L1 use. Their L1 enjoys a high status in the dominant society, which tends to militate against the use of other languages, including the TL these students are trying to learn by electing to enroll in immersion, CLIL, or SA. Because majority-language students already receive ample support for L1 development, both academically and socially, calls for use of the L1 in these contexts need to be tempered in favor of TL use. Ballinger, Lyster, Sterzuk, and Genesee (2017) argued that the sociopolitical context of school settings needs to be taken into account in order to distinguish between (a) L1 use as a cognitive support for learning content through the L2 and (b) L1 use as a manifestation of an existing societal language imbalance that favors majority language use. Because majority-language students in French immersion and other similar programs that aim for additive bilingualism still need to make strong L1–L2 connections for the purpose of biliteracy development, Ballinger et al. (2017) propose various crosslinguistic awareness activities – such as cognate instruction or the study of word families and patterns in derivational morphology across languages – and language awareness instruction as a means to foster bidirectional transfer between languages while maintaining separate spaces for French and English.

4.2 Content and Language Integrated Learning (CLIL) The affective dimension in CLIL has begun to gain some attention. Research carried out in the Basque country by Lasagabaster and colleagues (e.g., Lasagabaster & Doiz, 2017) has reported higher benefits for

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

227

CLIL instruction over non-CLIL settings regarding learners’ motivation toward English and their overall language attitudes. However, their findings also suggested that motivation toward learning English does not have long-lasting effects for CLIL learners, but rather tends to decline over time, whereas no such decline in motivation was observed in relation to content learning. Amengual-Pizarro and Prieto-Arranz (2015) also found that motivation reaches higher levels among CLIL students than non-CLIL students but, overall, these results are tempered by Sylve´n and Thompson’s (2015) finding that CLIL students tend to be already more motivated at the onset of the program than their non-CLIL counterparts. With respect to anxiety, non-CLIL learners report higher degrees of anxiety than CLIL students when required to speak English (AmengualPizarro & Prieto-Arranz, 2015). Results of Menezes’s (2014) study of anxiety among CLIL students revealed that they feel more anxious in the CLIL learning context than in the English-language class, probably as a result of their finding the former context more demanding. In a similar vein, CLIL learners have been noticed to lack selfconfidence in TL use as they are often presented with language beyond their current abilities. Notwithstanding, Menezes and Juan-Garau (2015) found WTC clearly higher in CLIL streams than non-CLIL streams. More research is needed to delve further into the affective domain in CLIL contexts, using longitudinal evaluations that incorporate both quantitative and qualitative research instruments, since generalizations are still hard to come by. In terms of TL use in CLIL classrooms, the use of English L2 does not appear to be as difficult to maintain as it is for English L1 students to persevere in using a non-English language. English L2 learners are possibly more prone to use the TL on account of the social prestige attached to English, while English L1 students may find it easier to switch into their native language given its high status and omnipresence in their community as an international language. For example, CLIL students in vocational secondary education in the Netherlands were asked which teacher behavior was most useful to help them learn subjects through English L2 (Denman, Tanner, & de Graaff, 2013). The most important teacher behavior scoring far above all the others was the teacher speaking English all of the time, while the second most useful teaching strategy was the teacher encouraging students to use English themselves. Nonetheless, the use of the students’ L1 as a pedagogical strategy in CLIL settings is well documented. CLIL teachers at the secondary level in Austria were observed using German L1 as a means to help students understand academic content and to scaffold learning (Gierlinger, 2015). However, they did so partly because they were not sufficiently fluent in English and did not possess the pedagogical knowledge required to help students understand concepts through English.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

228

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

4.3 Study Abroad (SA) Research indicates that the SA context appears to accentuate individual differences. Learners’ integrative motivation and WTC abroad can spur them to partake in situations that require interaction, thus facilitating the acquisition process. It appears that motivation and WTC abroad might be influenced not so much by a learner’s actual proficiency level as by his or her perceived communicative competence. Trenchs-Parera and JuanGarau (2014) reported that a three-month stay heightened their participants’ motivation – paired with a reduction of anxiety – and rendered listening abilities more important in their view. Although some studies, such as the one just mentioned, report an increase in motivation and other affective factors even after a relatively short stay, this has not been a consistent finding in the literature. The role of learner strategies for social integration in the TL community has also deserved some consideration. In this sense, it has been contended that the learners’ ability to be open to the TL culture might help them fit in the host context more than other factors, such as their actual proficiency or their assumed language aptitude. In that respect, Churchill and DuFon (2006) emphasize the importance of building a social network, which they relate to learners’ openness as well as to their ability and persistence in being socially outgoing and paying attention to unmodified input.

5 Challenges and Future Directions 5.1 Optimal Integration of Content and Language Learning A common finding across well-implemented additive immersive education programs is that students fare as well as their peers in regular mainstream programs in terms of academic achievement and L1 attainment, while learning much more of the additional language than their peers studying it as a school subject for one period a day. Other findings with respect to TL attainment are more variable, however, owing arguably to differences related to a range of factors, including sociopolitical context, program design, quality of instruction, learner motivation, teacher preparation, and TL status in the community. Overall, the high levels of comprehension and communicative ability attained by students in additive programs such as immersion and CLIL can be attributed to greater exposure to the TL as well as to increased cognitive and affective engagement as they process subject matter taught through the TL. Much is beginning to be known about how greater exposure to the TL drives language learning over the full lifespan and across naturalistic and instructed contexts (see Armon-Lotem & Meir, Chapter 10, this volume; Mun˜oz & Spada, Chapter 12, this volume). Much less is understood about how exactly increased language learning is also driven by cognitive

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

229

and affective engagement while processing subject matter taught through the TL. Perhaps herein lies the greatest knowledge gap to tackle more fully in future research. Filling this gap would go a long way in helping answer the most fundamental question that persists at the level of best program designs and practices: how can the simultaneous goals for content and language learning be best balanced in immersive programs? That students do not attain even higher levels of TL proficiency, especially regarding production skills in grammatical accuracy and lexical variety, may come as little surprise given the emphasis in immersive programs on content learning and the assumption that language learning will occur if the context is sufficiently meaningful. Years ago, Swain (1988) noted that students can understand content without having to engage at the same time in syntactic processing of the language. She proposed that content teaching on its own was not necessarily good language teaching and needed to be complemented and manipulated in ways that enable students to notice form-meaning mappings in the TL. Yet early observations of immersion classrooms found that it was relatively rare for teachers (a) to refer during content-based lessons to what had been presented in a grammar lesson and (b) to set up content-based activities specifically to focus on form related to meaning. As Lightbown (2014) argued, separating content and language in this way “may deprive students of opportunities to focus on specific features of language at the very moment when their motivation to learn them may be at its highest” (p. 30). Even in the case of CLIL, it may be argued that the coalescence of content and language has not reached its full potential insofar as the connections between the content and language classes have not yet been fully exploited in ways that ensure that the language addressed in the English as a foreign language class is language that complements or supports the content focus. Undoubtedly, the integration of language and content continues to be a challenging task for educators to accomplish both in instructional practices and program design, hence the importance of teacher education and ongoing professional development opportunities. Indeed, a common thread running through research on additive immersive programs is the important role played by teacher education and professional development in their continued success. A pivotal question that remains open for further investigation in this respect is how teachers can most effectively engage in immersion and CLIL approaches in ways that scaffold content learning while ensuring continued development in the TL. One way for teachers to integrate language and content is by emphasizing the ways in which linguistic features of discipline-specific language construe particular kinds of meanings (Llinares et al., 2012). This involves making students explicitly aware of (a) the academic language functions they need to understand and communicate in specific academic

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

230

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

disciplines (e.g., describing, comparing, explaining, hypothesizing, predicting) and (b) the conventional text structures or genres that are characteristic of particular disciplines (e.g., science reports, historical accounts, math problems, essays). Another way to integrate language and content is for teachers to more intentionally draw students’ attention to language in discourse contexts related to subject matter or literacy themes in ways that support students’ use of the target features in interactive situations. This involves giving content and language objectives complementary status and shifting students’ attention between language and content through a counterbalanced approach (Lyster, 2007). Counterbalanced instruction entails reactive and proactive approaches to integrating language and content implemented in tandem. A reactive integration approach includes scaffolding techniques such as questions and feedback in response to students’ language production that serve to support student participation while ensuring that classroom interaction is a key source of learning. A proactive integration approach entails preplanned instruction that interweaves noticing and awareness activities with opportunities for both guided and autonomous practice. A set of quasi-experimental studies undertaken in French immersion classrooms in Canada (see Lyster, 2007, for a review) yielded overall positive effects for such language-focused instruction on a range of challenging target features in French. Students participating in formfocused tasks improved more in their French proficiency than students left to their own devices to “pick up” the target forms from the regular curriculum. However, this research, as well as other classroom-based studies investigating the effects of language-focused interventions, has tended to measure L2 development more than content knowledge, leaving open many questions about the feasibility and effectiveness of focusing on language during subject-matter instruction. Specifically, we still need to know whether content knowledge is enhanced or possibly compromised by a greater focus on language during content instruction.

5.2 Understanding the Benefits of SA With respect to SA experiences, they have the potential to provide access in even greater quantity and quality to language that is varied, meaningful, and authentic than is possible in immersion and CLIL programs. From a research perspective, one of the main challenges is to continue to assess the magnitude and nature of L2 gains abroad, which have proven strong in some specific language domains but not others, and to do so with more longitudinal research to assess the benefits more adequately. Such research should ideally incorporate a qualitative component as well and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

Becoming Bilingual through Additive Immersive Programs

231

find appropriate methods to better fathom learners’ contact with – and use of – the TL abroad. In this regard, Ranta and Meckelborg (2013) revealed a general tendency for learners to make receptive rather than interactive use of the TL in the SA context, pointing to the need to further investigate the factors that condition the amount and type of language exposure and use overseas. From an organizational perspective, a good plan – often put in place by universities and other educational institutions – is most useful for learners to prepare prior to departure, to adopt in-country strategies enabling successful adjustment to the SA context, and to continue the learning experience once back home.

6 Conclusion We wish to conclude by underscoring the many positive outcomes of immersive, content-driven programs that aim for additive bilingualism, such as immersion, CLIL, and SA, while calling for further research and development with pedagogical objectives aiming to continue to improve learner outcomes in these programs. Their strengths lie in the flexibility they show to meet the needs and wishes for additional language learning of local communities, with variations in grade level entry point, proportion of the TL in the curriculum, and academic subjects associated with the TL. Their success is contingent on support from all stakeholders and covaries with multiple factors such as overall quality of instruction and program design. To be effective, these programs need to be well implemented. In the context of immersion and CLIL, specifically, this entails acknowledgment of the pivotal role played by teachers and the concomitant need for professional development to support them in meeting some of the challenges of teaching content through a language that their students know only partially. Specifically, teachers need to be adept at scaffolding content learning while ensuring continued TL development, and stakeholders responsible for program implementation need to appreciate this dual role as the sine qua non of program effectiveness. As for professional development, it needs to be of high quality and requires institutional commitments to provide teachers with sufficient time and sustainable opportunities for peer coaching in the spirit of crossdisciplinary collaboration. Finally, we stress that improving any educational initiative requires concerted efforts. Schools can do only so much to promote additive bilingualism and then learners themselves need to pursue their bilingual development beyond the walls of the school. It is for this reason that we have included in this chapter a secondary focus on SA programs, which have the potential to complement school-based programs while developing greater learner autonomy.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

232

M A R I A J U A N - G A R A U A N D R O Y LY S T E R

Acknowledgments We gratefully acknowledge funding to the first author from the Spanish Ministry of Economy and Competitiveness (FFI2013-48640-C2-2-P, AEI/ FEDER, EU). We would also like to thank the volume editors, Annick De Houwer and Lourdes Ortega, for their detailed and insightful comments.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:15:43, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.012

12 Foreign Language Learning from Early Childhood to Young Adulthood Carmen Muñoz and Nina Spada 1 Introduction Today foreign language (FL) learning, or the learning of a language that is used neither in the home nor in the public life of the society where a person usually lives, is an enterprise that engages individuals in many countries in the world. This is true over the full lifespan: in the learning lives of children, who have the option to begin learning an FL in kindergarten or earlier, in the busy adult lives of those individuals who study FLs for professional reasons, or during later adulthood, when some people may find themselves with more leisure time and decide to learn an FL in order to travel and explore the world. Indeed there is a wide range of contexts, motivations, opportunities, and paths for lifelong FL learning. How satisfactory are the outcomes of foreign language learning in formal classrooms, and on what factors might they depend? In this chapter, we discuss research that has investigated these questions, focusing on the very early ages of preschool and primary school and the ages covering young adulthood. We survey key opportunities and challenges in optimizing the outcomes of foreign language education across a variety of child and young adult ages and contexts of formal FL instruction, drawing in large part from examples in Europe and North America. These represent the geographies where many of the key findings have been generated, often within our own research programs. Our focus is on classrooms and programs characterized by limited instructional hours and limited language access outside the classroom, since these are the conditions under which the majority of students learn an FL throughout the world (for learning and teaching of languages in immersive conditions, see Juan-Garau & Lyster, Chapter 11, this volume). After establishing the fundamental tenet that a

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

234

C A R M E N M U Ñ OZ A N D N I N A S PA DA

focus on meaning and form is necessary for successful learning in the FL classroom, we pay particular attention to efforts that seek to optimize FL learning. The strategies we review are: concentrating instructional time; enhancing input through comprehension-based learning; using technology and media, and incorporating out-of-school learning and study abroad; and extending FL instruction through an early start.

2 Understanding the Contexts for FL Learning FL learning outcomes are dependent on the policies that each government may espouse and the resources that education systems are willing to invest in order to implement them (see Lo Bianco, Chapter 8, this volume, on language policies; see also Garcı´a & Tupas, Chapter 20, this volume). A key policy consideration is the instructional time that is made available for FL instruction. Overall, the research suggests that FL contexts are constrained with respect to what students, whether children or adults, can achieve, given the severely limited instructional time that is typical of these contexts. Government policies leading to resources that are provided to develop FL programs in public school systems are particularly notable in most European Union countries (for overviews, see Extra & Yag˘mur, 2012). Over the past two decades, many European regions have lowered the starting age for FL instruction to five to seven years and offer two (or more) FLs in primary and/or secondary school. But even when students begin FL learning early, the limits of what can be achieved when little instructional time is provided even after several years are evident. This was observed in the European Survey of Language Competences (ESLC) (European Commission, 2012b) investigating the FL proficiency of 53,000 students in the last year of lower secondary education or the second year of upper secondary education (mostly students aged 14–15) across 16 different educational systems. Students in 6 of the 16 educational systems had started learning an FL in grade one at the age of six. Achievement in their first and second FL was measured using the Common European Framework of Reference (CEFR) for languages, which defines six levels of functional competence from A1 (the lowest level) to C2 (the highest). In this survey the CEFR focused on levels A1 to B2. It was reported that the CEFR levels achieved (for listening comprehension, reading, and writing) varied widely across educational systems, and that they were always higher for the first FL (most commonly English) than for the second FL (with fewer years of instruction). For example, the proportion of students who had reached the B2 (upper intermediate) level in listening comprehension in the first FL ranged from less than 10% to almost 80%. However, a sizable number of students still did not reach beyond an A1 level in the first FL for reading (46%), listening comprehension (39%), and writing (33%).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

235

While testing experts estimate that 90–100 hours are required to reach an A1 level, some of these students had taken three to eight times more hours extended over many years to reach that level. Among the several factors that may account for the slow rate of progress that characterizes school FL learning, one can point to limitations in quantity and quality of exposure (e.g., opportunities for authentic language use), both of which are discussed later in this chapter. Exposure limitations are characteristic of FL contexts, which is why we refer to them as limited-input instructional contexts. Of course, the availability of FL programs for primary school-aged children is by no means universal. In the United States, for example, a national K-12 FL survey of more than 5,000 elementary and secondary schools reports that the opportunity to learn an FL in primary and middle school has substantially decreased over the past 15 years (Pufahl & Rhodes, 2011). Indeed, the vast majority of children growing up in the United States do not have the opportunity to learn an FL until secondary school. This is in part a consequence of the special status of English as a global language: generally, in places where English is the default FL there is more public pressure for an early start of FL instruction than in places where English is the national language for all or the home language of many. Adult FL learning in the United States, on the other hand, is more widespread than in compulsory education and has yielded considerable research – particularly as regards the learning of modern languages in university contexts. This research also attests to the limitations of language learning that come with restricted amounts of instructional time. For example, based on accumulated institutional research evidence, the Foreign Service Institute (FSI) of the US Department of State estimates that for an English speaker to obtain an intermediate level of knowledge of an FL (roughly equivalent to B1), approximately 600 hours of class time are required for closely related languages (e.g., Dutch) and 2,200 hours for languages that are typologically different (e.g., Chinese) (information retrieved from http://aboutworldlanguages.com/language-difficulty, last accessed January 25, 2018). This instructional time includes exposure to all four skills. In addition to the classroom hours the FSI also specifies another 400 hours of directed self-study outside class to reach this level. Yet the amount of time available in typical ab initio FL university courses in the United States is 2.5 hours of instruction per week over a 15-week semester, which over four years of university study would amount to only about 450 hours. That is still well below the number of classroom hours that the FSI estimates is needed to reach an intermediate level of proficiency. Moreover, the FSI estimates are based on intensive instruction (i.e., 25 hours per week), in classes of no more than six students who have a high aptitude for FL study as well as knowledge of several other languages (for a discussion of language aptitude, see Biedron´ & Birdsong, Chapter 16, this volume). These are ideal conditions and not representative of typical

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

236

C A R M E N M U Ñ OZ A N D N I N A S PA DA

limited-input programs for diverse groups of children, adolescent, and adult learners of FLs. While there are always exceptions to the rule, the possibility of achieving advanced levels of proficiency is highly unlikely in college-level FL programs in the United States. This was the conclusion reached in a recent survey by Tschirner (2016) based on 3,000 participants studying seven languages at 21 universities and colleges. Tschirner confirmed the conclusion other FL researchers had come to before him, namely, that reaching the low advanced level (equivalent to B2) on the American Council on the Teaching of Foreign Languages (ACTFL) proficiency scale for oral proficiency is “an elusive endeavor” (p. 201) for college graduates majoring in FLs. Low to mid-intermediate levels (i.e., B1) are more realistic. This is despite the fact that the development of oral proficiency has been the focus of FL instruction for decades. The survey results also indicate that while more advanced levels of reading proficiency are attainable at graduation, levels of listening proficiency lag considerably behind. In spite of the large differences in terms of government support and resources across contexts, the cases just discussed about FL outcomes for 14–15-year-olds in Europe and for adults in the United States serve to illustrate the constraints of limited instructional time. In the remainder of the chapter we unpack some of the challenges that limited time ultimately creates for FL learning: limitations in input quantity and its distribution in small amounts over extended periods of time; limitations in input quality (e.g., teachers’ FL proficiency) and diversity (range of functions and interlocutors); and limitations in opportunities for authentic use of the target language (TL). First, however, we discuss a prerequisite that research has shown to be important for optimal learning in the FL classroom.

3 A Fundamental Tenet for Optimal FL Instruction: Focus on Meaning and Form Historically, the focus of FL instruction has been on grammar and accuracy – whether this instruction took place through explicit decontextualized metalinguistic instruction associated with the grammar translation approach or through memorization, repetition, and drills associated with the audiolingual method. With the introduction of communicative language teaching (CLT) in the 1980s, many FL programs shifted from an almost exclusive focus on grammar to an equally almost exclusive focus on language use and communication. One of the major themes within CLT has been whether grammatical competence can be obtained through exposure to primarily or exclusively meaningful input and communicative interaction. Considerable research has been done to investigate this question. Most of it has been carried out

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

237

with adult learners of English in university settings at beginner or intermediate levels of proficiency. Far less research has been done with FLs other than English, with learners at more advanced levels of proficiency, with older adults outside university contexts, or with adolescents and children learning FLs in limited-input instructional settings. Recognizing these limitations in terms of populations studied, the overall findings indicate that meaning-based instruction that includes attention to language is more effective than instruction focused exclusively on either meaning or form (Ellis, 2006; Lightbown & Spada, 2013). Evidence from meta-analytic research indicates that instruction is more effective when it is designed to recruit learners’ explicit attention to language than when it attempts to engage their implicit attention (Norris & Ortega, 2000). It is important to note, however, that the majority of instruments used in the studies meta-analyzed consisted of discretepoint grammar tests, which are known to be biased toward measuring learners’ explicit knowledge. Far fewer studies employed language tests that are more likely to tap into learners’ implicit knowledge (e.g., oral communication tasks). This points to a need for future studies to do so. Explicit instruction includes rule explanation and/or directs learners to focus on language; implicit instruction does not include any rule explanation or direction to attend to language. It is useful to think about implicit and explicit instruction as two ends of a continuum with pedagogical activities reflecting different degrees of implicit or explicit teaching. For example, an implicit approach to teaching a particular language form (e.g., the passive voice) would include giving learners a text to read that contains examples of the passive voice and asking them to read the text and respond to comprehension questions. However, if learners are asked to complete the same comprehension task and given a text that is (a) seeded with a large number of exemplars of the passive voice or (b) seeded with a large number of exemplars that are underlined and italicized, then the instructional task would be more explicit in (b) than (a). Similarly, any given pedagogical approach can be modified to become less or more explicit. For example, input-processing instruction (VanPatten, 2004) is a psycholinguistically motivated approach that relies on reading and listening activities designed to help learners make form–meaning connections while processing meaningful language, and it can be made more or less explicit depending on whether learners are given metalinguistic rules prior to engaging in the processing activities. A related area of research comparing implicit and explicit ways of drawing learners’ attention to form is error correction, usually known as corrective feedback (CF) in the specialized literature. The overall findings for university-level learners of foreign languages, the group for which the bulk of this research has been carried out, indicate that CF is effective for learning and that benefits come with different types of CF (Lyster, Saito, & Sato, 2013). For example, it has been found that in communicatively

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

238

C A R M E N M U Ñ OZ A N D N I N A S PA DA

oriented classrooms, explicit CF (e.g., of a metalinguistic nature) is more effective than implicit CF (e.g., in the form of recasts or more grammatical reformulations of an utterance immediately upon a learner’s rendition). However, the few experimental studies that have investigated the effects of different types of CF on L2 learning with younger learners have tended to be conducted in intensive and immersion instructional settings (see Juan-Garau & Lyster, Chapter 11, this volume) rather than in limited-input classrooms. The overall message that emerges from the research literature on FL learning is that attention to both form and meaning is essential. Questions about the best balance remain, but given the time restrictions within most FL instruction settings and the fact that learners in these contexts have limited or no access to the TL outside the classroom, a primary focus on meaning and communication is advised, with attention to form included within it. Thus, all best evidence models of FL instruction will include the following elements: (1) plentiful opportunities to comprehend the TL via listening and reading for meaning, complemented with rich opportunities to engage in language communication in interactive speaking and meaningful writing; (2) fluency development in order to support automaticity in language use; and (3) targeted language forms that are explicitly attended to via instruction and practice. A number of theoretically divergent positions within the instructed second language acquisition literature support the fundamental principle of a primary focus on meaning with attention to form embedded within it. This includes task-based language teaching (Long, 2015), dynamic usage-based instruction (Verspoor, 2017), and systemic theoretical instruction inspired by sociocultural theory (Lantolf & Poehner, 2014). Considerations about the need for attention to form and meaning in limited-input FL instructional settings and the methods that can be employed to accomplish such attention depend on the age and proficiency level of the learner. With younger learners more experiential-based learning that is compatible with their cognitive and social development is recommended. This includes activities such as interactive games, songs, reading aloud, and storytelling. While a primary focus on meaning-based instruction is strongly advised, this does not mean that language-focused activities should be avoided with younger learners. On the contrary, there are many age-appropriate ways to draw young learners’ attention to language that can be motivating and challenging. With older learners, on the other hand, a focus on communication and meaning-based instruction is also recommended. However, adolescent and adult learners can draw on their metalinguistic knowledge, advanced cognitive abilities, and literacy skills, and this needs to be taken into consideration when making decisions about pedagogy. For example, in FL courses in North American universities, concerns have been expressed that the notion of communicative competence has been reduced to a pedagogy based on spoken

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

239

conversational ability (Schultz, 2006). This is considered inadequate and inappropriate for students who are taking combined language and literature courses and need to develop a more sophisticated competence in the FL. To accomplish this, FL instruction needs to focus on critical thinking, intercultural awareness, and metacognition about language and culture (Kramsch, 2014).

4 Optimizing Input by Concentrating Instructional Time One obvious response to the constraints that come with limited-input FL instruction is to increase the amount of time for language learning. Examples of this include French immersion programs in Canada, content-based instruction in the United States, and content and language integrated learning (CLIL) programs in Europe (see Juan-Garau & Lyster, Chapter 11, this volume). However, in most educational contexts, it is not feasible or desirable to create more time for language learning. Other options are to concentrate the available instructional time over shorter periods. In this section, we review research that has investigated the relative effectiveness of different concentrations of instructional time, often framed as the effectiveness of massed versus distributed instruction. Much of the research into different concentrations of language instructional time can be found in Mun˜oz (2012). Research has investigated high versus regular concentrations of instructional time in Canadian intensive English and French language programs. These programs are in addition to regular English and French language as subject-matter instruction offered in Canadian schools as well as L2 subject-matter instruction (e.g., French immersion). The research in intensive English and French language programs is summarized by Lightbown and Spada (1994) and Netten and Germain (2004), respectively. One intensive English model provides learners with a five-month experience in either grade five or grade six (age 10–12 years), during which time learners are immersed in the English language for five hours a day, five days a week. This contrasts with learners in the regular English as a subject-matter program, who receive only 20 minutes of English input a day throughout the school year. In an early comparison of the language development of learners in the intensive and regular programs, advantages were found on listening and reading comprehension and oral production for the intensive learners. Follow-up studies to investigate how the intensive learners compared with a group of learners who had accumulated the same amount of instruction but spread over a longer period of five years also indicated superior performance for the intensive learners on all language measures (Lightbown & Spada, 1994). Similar findings have been reported for intensive French-language programs, where it was observed that students in grades five and six could orally communicate as well as grade nine

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

240

C A R M E N M U Ñ OZ A N D N I N A S PA DA

and 10 students (ages 14–16) in regular French as subject-matter classes; moreover, the written production of the students in the intensive program was comparable to that of native French-speaking students in grades three to four (ages 8–10). Long-term studies indicate that students in intensive French programs tended to continue in French to the end of secondary school in greater numbers than students who did not have the intensive experience (Netten & Germain, 2004). One of the questions that has emerged from comparisons of massed versus distributed instruction is whether different concentrations of time may be more effective than others. This has been investigated with FL learners of varying ages across school and university settings in Canada and Spain. For example, research in core (i.e., regular, non-intensive) French programs in Canada has compared the progress of students in drip-feed programs, where they receive 40-minute lessons daily for the entire school year from grade four to grade eight (ages 9–13), with students who receive different degrees of concentrated instruction, for example, 80 minutes per day over five months, versus 150 minutes per day over 10 weeks (Lapkin, Hart, & Harley, 1998). Results have been mixed, showing advantages on some language measures for students in the more concentrated classes in some but not all studies. The different findings may be related to the fact that the degree or intensity of the concentration of time varied between studies. A consistent finding from this research, however, is that learners in the more concentrated classes express more positive attitudes about their instruction. Research with primary school learners in Canadian intensive English programs has also investigated whether different concentrations of time make a difference (e.g., Collins & White, 2011). The results indicate that the most concentrated groups make the most progress. However, the benefits are modest and, thus, the advantages for different ways of concentrating the instructional time remain uncertain. Research with university-level learners of English as a foreign language (EFL) in Spain compared intensive and semi-intensive English programs at home with study abroad experiences (Serrano, Llanes, & Tragant, 2011). These researchers report that students who received five hours of instruction five days a week over 4.5 weeks (i.e., intensive) outperformed learners who received the same total amount of instruction distributed over 2.5 sessions a week over three months (i.e., semi-intensive). It was also observed that students in the intensive program performed as well as students who had participated in a 15-day study abroad experience. Other advantages have been attributed to programs in which efforts are made to intensify the instructional time. For example, the concentration of time allows for more communicative activities, more project-based learning, learner-centered classes, literacy practice, and individualized instruction. It has also been observed that intensifying instruction leads to greater group cohesion and is thought to contribute to higher levels of learner motivation. The additional benefits suggest that it is not just

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

241

quantity and distribution of time that contribute positively to language learning but also the quality and range of pedagogical options that come with more time. In the next section, we discuss innovative pedagogical strategies that can help enhance the quality of input in FL contexts.

5 Innovative Strategies for Enhancing Input in FL Learning Because in FL instruction the input is limited to the classroom setting, the language to which learners are exposed is restricted to what is provided in the teaching materials, textbooks, and teacher talk. This is particularly the case in classrooms that reflect more traditional structure-based approaches to FL teaching. Even in highly communicative classrooms the linguistic input does not represent the full range of language features and functions that occur in natural settings. It is also difficult (if not impossible) in input-limited classrooms to provide adequate coverage of all the skill areas (i.e., listening, speaking, reading, writing) and to focus on the development of both accuracy and fluency in the limited time available. The quality of input available to students also varies from teacher to teacher. For example, in some FL situations it is difficult to find teachers with adequate levels of oral proficiency in the TL who can offer learners the rich and extensive input needed to trigger language development. Innovative approaches to providing learners with richer and more varied sources of input include comprehension-based instruction. Also important is the use of technology, digital and audiovisual media and out-of-school learning experiences, and study abroad. We review each area in turn.

5.1 Comprehension-Based Instruction The approach to language teaching known as comprehension-based instruction (CBI) emphasizes learners’ processing and understanding of language through reading and listening rather than through speaking and writing. CBI is based on the assumption that languages are learned by understanding meaningful language input and the content it helps convey. One of the first CBI programs was investigated by Lightbown, Halter, White, and Horst (2002). It was an experimental program developed for primary school learners of English in a French-speaking community in Canada where learners had virtually no exposure to English or opportunities to interact in English outside the classroom setting. Students in grade five (10–11 years old) did not receive any direct instruction in their daily 30-minute class periods. Instead, they read and listened to a library of books and tapes made available to them in a language lab-like setting. There was no production practice (i.e., speaking or writing) or interaction in English either with a teacher or other students. Evaluations of the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

242

C A R M E N M U Ñ OZ A N D N I N A S PA DA

English language development of the CBI learners compared to that of students in regular audiolingual classes indicated that they learned as much English as students in the comparison classes. This was true not only with respect to comprehension but also for speaking – a surprising finding indeed. A follow-up program evaluation three years later when the students were in grade eight (13–14 years old) indicated that learners in the comprehension-based programs continued to perform as well as learners in the regular program on comprehension measures. However, there were indications that learners in the regular audiolingual classes made greater progress in L2 production abilities. In similar research conducted recently in Spain, Tragant, Mun˜oz, and Spada (2016) investigated the effects of CBI on the learning of English with primary school children (10–11 years old) as well. In the comprehensionbased classes, students participated in relatively independent reading and listening activities by reading books that had an audio component. In the comparison classes, students received traditional EFL instruction that was teacher led and topic based and included both receptive and productive practice. Instruction took place over one school year and learners were pre- and post-tested on a variety of language measures. Learners in the CBI classes did as well as the comparison group students on listening comprehension, written production, and sentence imitation tasks, despite the fact that they had received considerably less teacher-led instruction. Students in the CBI group also indicated more positive attitudes toward English language learning. The researchers interpreted the results as encouraging for language learning in contexts where the linguistic input provided by teachers may be limited. CBI may be particularly beneficial for low-level proficiency learners. In a study comparing the effects of comprehension- and production-based instruction (PBI) on incidental learning with six–eight-year-old learners of English in primary school in Japan, Shintani and Ellis (2010) found that CBI and PBI were equally effective and that the greatest linguistic gains were among learners with less prior knowledge of English. The researchers suggest that less proficient learners were better able to attend to the input because they were not required to produce utterances that may have overtaxed their processing capacity.

5.2 Technology, Media, and Out-of-School Experiences The use of technology and media in the FL classroom provides increased access to TL input that is relevant to the purposes of today’s learners. One technological innovation that incorporates authentic content available on the Internet into classroom lessons, including primary school classrooms, is the interactive white board. In their review of technology types and their effectiveness, Golonka, Bowles, Frank et al. (2014) noted that although no studies have reported learning outcomes,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

243

there is some evidence from qualitative studies analyzing self-report data of the positive impact of an interactive white board on the process of learning and on learner affect. The use of technology can also open classroom doors to authentic language environments needed for taskbased language teaching. For example, a study by Guo and Mo¨llering (2016) of beginning learners of Chinese in an Australian university shows that the implementation of communicative task-based activities in a web-conferencing environment elicited learners’ online interaction and negotiation of meaning. Technology also provides opportunities to incorporate out-of-school interests into classroom practices and to capitalize on those practices to promote new learning experiences for students. This is the case with the use of subtitled videos, which can provide students with authentic input in the FL classroom and prepare them for watching similar material at home. The theoretical basis for the beneficial effects of this pedagogic practice relies on exposure to multimodal input, because L2 learning is enhanced when visual images are combined with verbal information (Mayer, 2009). Although this is a relatively new area of research in the field of second language acquisition, studies report benefits in speech segmentation and, mostly, in vocabulary learning and listening comprehension (see a metaanalysis by Montero Pe´rez, Van Den Noortgate, & Desmet, 2013). The benefits of the use of subtitled audiovisual material have been observed in different multilingual contexts, as shown in a study by Nely and Ayonghe (2015) in Cameroon, in which participating students, who spoke at least one of the indigenous languages, one lingua franca, and French, showed a significant improvement in a large number of areas of the English language. Providing resources for students and supporting students’ own learning are essential characteristics of teachers in FL classes that encourage autonomous learning. In turn, a distinctive trait of successful and autonomous language learners is that they look for opportunities to learn beyond the classroom. Nowadays, finding such opportunities is easier than ever with the help of technology and the Internet, and emerging empirical evidence has indicated that such leisure access to the TL accrues tangible language learning benefits (Sylve´n & Sundqvist, 2017). It has become an ingredient of good teaching practices to promote students’ awareness of possibilities for authentic input outside the classroom (e.g., extensive television viewing suggested by Webb, 2015). Besides actively promoting language learning awareness and autonomy, teachers can also productively integrate these out-of-class practices in the classroom. An illustration of such integration can be seen in the initiatives reported by Mercado (2015) in Peru, where students of English are provided with the means through which they can develop the strategies needed for effectively engaging in autonomous learning pursuits that complement what they have acquired through formal instruction. The outcome of such assignments is presented

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

244

C A R M E N M U Ñ OZ A N D N I N A S PA DA

in class and exemplifies the language skills and knowledge that students are able to develop on their own.

5.3 Study Abroad An increasingly favored way in which students in input-limited classrooms may achieve advanced levels of proficiency, especially in the areas of fluency and oral expression, is through studying abroad. In fact, at the college level, a long tradition exists of study abroad programs. These were first common in North America and have also become common in Europe (especially under the Erasmus program). They are also used in some Asian countries. These programs are seen as a way of breaking through the ceiling of input-limited programs and allowing students to progress to advanced levels of proficiency. Arguably, study abroad experiences that are fully integrated in the students’ home school program are most efficient. Relevant research questions about the integration of study abroad programs in the classroom include what the optimal pre-departure proficiency level is and how to help learners in their reentry into the regular FL classroom (DeKeyser, 2007; see Juan-Garau & Lyster, Chapter 11, this volume, for more discussion of study abroad research).

6 Extending the FL Learning Experience through an Early Start Based on the belief that it is best to begin learning a new language as early as possible, an early start is considered by many to be an optimal way of improving FL learning processes and outcomes. Another alleged benefit of an early start is the extension of the FL learning period, even though an early start has not necessarily implied a significant increase in the number of instruction hours in all cases. Consequently, the trend to introduce FL teaching in primary school (and even earlier) has grown in recent decades, particularly with respect to the teaching of English. However, there is relatively little research focused on young learners, and especially at the earliest starting ages that we increasingly see in EFL classes around the world. In this section we first review emerging research evidence from studies of very young learners. Then we focus on important issues that merit attention in research programs to inform policy making and teaching.

6.1 Emerging Research Evidence about Very Young Learners Children in preschool are often referred to as ‘very young learners’ and distinguished from ‘young learners’ in primary school. Because the beginning of primary school may vary in different educational contexts

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

245

(e.g., from age four to age seven), researchers and educators use different age ranges to identify very young learners (e.g., ages three to six, two to seven). The nature and aims of very early language learning are diverse, ranging from language awareness-raising programs to programs for teaching a particular language, usually English. Thus, in the policy handbook about pre-primary school-level learning of FLs produced by the European Commission (2011), early language learning is considered an enriching experience that opens children’s minds to multilingualism and different cultures and that “can shape the way children develop their attitudes towards other languages and cultures by raising awareness of diversity and of cultural variety, hence fostering understanding and respect” (p. 7). In contrast, the globalization of English in the past decades has spurred early teaching and learning of English worldwide for more pragmatic reasons, and this is nowhere more evident than in the great variety of early EFL programs that have appeared in Asian countries. In these early initiation programs, children are taught the language for a few hours per week. Research that investigates these learning environments is emerging. Sun, Steinkrauss, Tendeiro, and de Bot (2016) present one of a series of studies following a group of very young Chinese EFL learners for 1.5 years. Sun et al. examined the comprehension vocabulary, production vocabulary, and grammatical development of 71 children whose onset age of learning English was between 2 and 5.5. They found that the total amount of school input and level of the English media environment at home were significant predictors in the development of all three outcomes investigated, and explained more variance than age of onset, short-term memory, and nonverbal intelligence. Also with very young learners, the FLiPP project in the Netherlands conducted by Unsworth, Persson, Prins, and de Bot (2015) serves to illustrate a different regional context for learning English. The researchers aimed at examining the effects of amount of classroom exposure and teachers’ language proficiency in state schools where early teaching of English is very much a bottom-up development with limited governmental regulation and support. Three groups of very young learners (168 in total) starting English lessons at age four were compared. They had different weekly amounts of exposure to English (60 minutes or less, between 60 and 120 minutes, and 120 minutes or more). Teachers also varied in their English language proficiency. After two years, children in the two groups with more than 60 minutes of weekly classroom exposure scored significantly higher on receptive grammar and vocabulary than children with 60 minutes or less. However, teachers’ language proficiency was the best predictor of children’s scores on both language measures. In sum, what the research evidence emerging from studies with very young learners indicates is that the quantity of input (e.g., exposure to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

246

C A R M E N M U Ñ OZ A N D N I N A S PA DA

media outside the classroom) and the quality of input (e.g., teachers’ language proficiency) is crucial at this very early age.

6.2 Key Challenges for FL Learning at Primary School Ages Research into FL learning by primary school-age learners is needed in order to inform policy making and teaching. We highlight four challenges in particular. The first challenge is the scarcity of research-based assessments that gauge language outcomes in young learners, which is related to the general difficulty of assessing development at young ages (Nikolov, 2016). The reasons are diverse, from absence of control from educational authorities because early learning programs were launched as the result of bottom-up pressure from parents (Nikolov & Curtain, 2000), to lack of motivation to investigate the effectiveness of those programs (or the value of an early start) because decisions are taken on educational (political) grounds. In their review, Harris and O’Duibhir (2011) confirm the lack of research on process–product relations in many studies on primary school FL learning. Most studies focus on describing learning and teaching processes (Cable, Driscoll, Mitchell et al., 2010; Edelenbos, Johnstone, & Kubanek, 2006), with the result that less attention has been paid to the effects of instructional practices on learning outcomes. Harris and O’Duibhir (2011) observed the lack of research syntheses of studies in primary schools, which may be the result of authors not distinguishing between primary and secondary levels, or simply of disregarding possible age effects. Moreover, language outcomes in the primary classroom are difficult to assess. Crucially, children’s cognitive and personality characteristics impose important limitations on appropriate instruments that can be used to elicit language at an early age. One notable difficulty is the fact that language development is slow in young learners in input-limited classrooms (Mun˜oz, 2006). Another difficulty arises from the need to assess oral skills, those more commonly targeted in the primary classroom, which requires one-on-one assessment and hence is time-consuming and expensive. Second, the lack of continuity of early FL programs has long been a challenge related to the more general problem of the curricular transition between primary school and secondary school. With more and more children in preschool programs, continuity may be threatened even earlier, as noted, for example, in the policy handbook on FL learning at preprimary school levels produced by the European Commission (2011, p. 12). The lack of an explicitly articulated sequence of instruction means that in many cases students who have studied a FL in primary school may be placed in secondary school together with students who have had no prior FL instruction. There is also a lack of continuity in teaching methods, with significant changes toward a stronger reliance

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

247

on literacy in instruction at the end of primary and in early secondary school. FL instruction may also fail to rely on what children already know, which may result in learners’ perception of lack of progress and low perceptions of self-efficacy. All of these shortcomings may seriously undermine learner motivation. Indeed, decreasing motivation toward FL learning is the third biggest challenge in FL instruction with young learners. The observed decline starts in the last years of primary school (e.g., Cable et al., 2010) and is concerning, given that surveys on pioneer primary school modern language programs have concluded that the main gains in starting an FL early in childhood lie in the development of positive attitudes and motivation (e.g., Edelenbos et al., 2006). The negative consequences of the lack of continuity of early FL programs just discussed almost certainly contribute to the drop-off in motivation. Other reasons that have been suggested are lack of alignment between what students see as the goals of language learning and what they experience in the classroom, and lack of contact with the TL in real terms (Graham, Courtney, Tonkyn, & Marinis, 2016), obstacles that are to a large extent linked to FL classroom input limitations.

6.3

Input Quality, Quantity, and Distribution in the Early Years: Is Early Always Best? When FL instruction is introduced in the school curriculum from the very early years, the constraints of what can be achieved in the face of limitedinput classes and instruction that is distributed over many years are typically ignored (see Sections 2 and 4). And yet ample research shows that “early is best” does not hold for children who are learning an FL in classrooms where time and input are limited. On the contrary, there is increasing evidence that in limited-input instructional environments older learners learn faster and more efficiently than younger learners (Mun˜oz, 2006; see also Singleton & Pfenninger, Chapter 4, this volume). There are several reasons for this. Older learners have a more fully developed first language, more advanced cognitive abilities, and an ability to approach learning more analytically and explicitly. By contrast, younger learners are less cognitively and linguistically developed and approach learning more intuitively and implicitly. One of the conditions for implicit learning mechanisms to become activated is massive exposure to language and opportunities to engage in extensive communicative interactions with speakers of the TL, in much the same way that children learn their first language. Unfortunately, small amounts of instruction (e.g., 45 minutes three times a week) spread over a school year do not provide sufficient conditions for implicit learning to take place. Therefore, unless it is possible to substantially increase the time available for FL instruction earlier in the primary school curriculum, a more promising alternative may be to concentrate or intensify instruction in later school years.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

248

C A R M E N M U Ñ OZ A N D N I N A S PA DA

These caveats about the early introduction of FL learning notwithstanding, it is important to point out that other benefits can come with introducing an FL early in primary school. One such benefit may be to prepare school learners for engaging in study abroad experiences in adolescence, a time at which they have gained much autonomy and efficiency. Other benefits include the development of more positive attitudes toward other languages and cultures (Mihaljevic´ Djigunovic´, 2011) and the positive contributions that FL learning can make to L1 development (Murphy, Macaro, Alba, & Cipolla, 2015). More research is needed to further examine these issues as well as to investigate the effects of different amounts and distributions of instructional time with learners in a greater variety of educational contexts, at different ages and levels of proficiency.

7 Conclusion In our review of issues pertaining to children’s and young adults’ learning of foreign languages in classroom settings, we have brought to the forefront that in FL instructional contexts, the quantity and quality of input and opportunities for language use tend to be limited, and that this presents a main challenge for what can be achieved in terms of learning outcomes. This premium on the language input is consistent with observations made in the study of bilingualism in the family, where the evidence that bilingual children are sensitive to the differential exposure to each language is robust, and where research has also begun to suggest that the quality of the input (i.e., variation in and richness of the linguistic environment children are exposed to) may be an important factor of linguistic attainment, in addition to quantity (Gru¨ter & Paradis, 2014; see ArmonLotem & Meir, Chapter 10, this volume, on the role of input in early child bilingualism). The review of research in this chapter leads to three final conclusions based on studies carried out principally with learners of English as an FL (children, adolescents, and young adults) in formal settings (schools and universities), and who are at beginning or intermediate levels of proficiency. First, attention to meaning and form is essential in FL learning, and, given the time and input restrictions of typical FL classrooms, learners will benefit most from a primary focus on meaning and communication with attention to form included within it. Second, input limitations in FL classrooms prevent learners from reaching advanced levels of proficiency and may also present obstacles for long-term personal engagement with FL learning. As a result, there is a need for approaches that help to counteract input limitations, such as the concentration of instructional time, the integration of authentic and contextualized input from outside the classroom by means of digital media, the incorporation of students’ out-of-school learning opportunities into the classroom, and the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

Foreign Language Learning from Early Childhood to Young Adulthood

249

availability of immersion opportunities through study abroad programs. Third, there is a dearth of research into FL learning at the two ends of the lifespan, particularly with very young learners and with older adult learners. We have not reviewed the latter age population, but readers are referred to Singleton and Pfenninger (Chapter 4, this volume) and Goral (Chapter 5, this volume). Ultimately, all learner populations across ages should be included in the quest to understand lifelong FL learning as a site where many people experience bilingualism.

Acknowledgment Thanks are due to grants FFI2016-80564-R and 2014SGR1089 (Carmen Mun˜oz), and to the Social Sciences and Humanities Research Council of Canada (Nina Spada).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:54:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.013

13 Supporting Bilingualism in Adult First-Generation Migrants James Simpson

1 Introduction A feature of 21st-century globalization is the mass movement of people from one country to another. Around 244 million people in the world are migrants, representing roughly 3.3% of the world’s population (United Nations, 2016), and motives for their migration are far from uniform. People move because of a shortage of labor in certain sectors, or to be with their families, or as refugees to escape war, civil unrest, poverty, or fear of persecution. Host countries now accommodate multilingual and multicultural populations from potentially anywhere. Supporting bilingualism and multilingualism for adult migrants is therefore a complex global undertaking, though one treated inconsistently and unevenly in different parts of the world. Bilingualism for new arrivals involves the learning of the dominant languages and varieties of the new home as well as the use of established first languages (L1s). This is in some respects a human rights issue. Article 2 of the Universal Declaration of Human Rights (United Nations, 1948) has language as one of its categories for equal rights; the issue of linguistic human rights is further advanced in internationally constituted documents such as the Universal Declaration of Linguistic Rights (Universal Declaration of Linguistic Rights Follow-up Committee, 1998). The two fundamental linguistic human rights that apply to adult migrants are that they should be allowed to maintain the languages they grew up speaking, even as they and their families settle in a new country, and that they ought to be entitled to learn to communicate in the main language of their new country (see also Lo Bianco, Chapter 8, this volume). Hence, supporting bilingualism in adult first-generation migrants entails both L1 maintenance and enabling the development of competencies in a different language, a language of which, in many cases, they have no

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

251

prior knowledge. Policy makers, language educators, and academics working in the area of adult migrant language education are typically concerned with the second of these: the right of newcomers to learn the new language (however much this right is presented as a duty by politicians, the media, and educators) and the provision of opportunities for them to do so. Addressing L1 maintenance and development in practice and theory is also important, however. L1 use (as part of a multilingual repertoire) remains fundamental in communication in the personal and social spheres of many adult first-generation migrants, in the multilingual environments of contemporary life. Moreover, enhancing L1 literacy can promote effective L2 literacy development. From a pedagogical perspective, language education practitioners can use an understanding of their students’ language backgrounds, including their earlier experience of schooled literacy, to inform their L2 learning experience. This chapter is about adult first-generation migrants, salient issues in their learning and life experiences, and how their learning is supported (or not), first in policy, and then in pedagogy. Adult first-generation migrants are defined for the purposes of this chapter as people beyond school age who move from one (nation) state to another with the intention of staying more or less permanently and building a life in the new country. The term “migrant” – to or from a country – is used throughout in preference to the term “immigrant,” to avoid the negative connotations taken on by the term “immigrant” in public and media discourse in recent years. In the chapter the term “L1” is used to indicate the language or, indeed, languages, that migrants use to communicate with familiar people such as relatives and friends. The term “L2” refers to the new language that migrants may learn after migration, although for many migrants this L2 may actually be a third or fourth (or more) language. In order to emphasize the possible plurality of languages migrants bring with them, this chapter prefers to refer to migrants and the contexts they find themselves in as multilingual rather than bilingual, although the terms are seen here as interchangeable. The profiles of adult migrants are hugely varied: one might consider the affluent retired British couple who move to Spain, as well as the poor South Asian living and working in Saudi Arabia and supporting a family back home. One might think of the educated Syrian whose refugee journey takes her and her children to Northern Europe, or the affluent expat from France working in international business in Singapore. Likewise, people at different life stages face specific challenges, to which they bring their own singular life histories. A 20-year-old experiences arrival in a new country very differently from a 40-year-old, not least in terms of flexibility in language learning. Younger arrivals might have more recent experience of being a student to inform their current learning (see Gonc¸alves, Chapter 3, this volume, for more on young adult migrants, or, as she calls them, global hybrids). On the other hand, older people might have

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

252

JAMES SIMPSON

developed a richer linguistic repertoire, given more extensive life experience. Some new arrivals might need to work straightaway, and thus will not have the time to take classes: this is possibly more of an issue for younger than older migrants, without recourse to savings or financial resources. Often, newly arrived migrants experience a difficult time in settlement, but not all do, and often difficulties are only temporarily present. This chapter discusses issues surrounding formal language education for migrants from poorer and possibly unstable regions of the world who have moved to the postindustrial nations of Europe and the Englishdominant West (the United States, Canada, Australia, and New Zealand, in addition to the United Kingdom and Ireland). This limitation is in part due to the paucity of studies of the language learning needs of migrants in other parts of the world, and on how these needs are being addressed. Following this introduction, Section 2 sketches out the broad contexts of life and language learning for adult first-generation migrants at a time when increasing numbers of people are on the move. I relate the circumstances and challenges of migrant language learners’ lives, as they become more multilingual, to the current (and contested) notion of superdiversity, and to intersectionality (to be explained later) as a framework for understanding complexities in the lives of adult migrant bilinguals. Section 3 describes policy support for adult bilingual development, typically understood as education in, and learning of, the dominant language of the new country, for purposes of integration. This section notes the close relation between language education and immigration policy, and the gatekeeping role that language testing for citizenship plays in many parts of the world. Section 4 considers pedagogy in adult migrant language education classrooms, with first a focus on interaction in the new environment generally and then on specific areas of language education that are also of particular relevance to migrants: language learning for (and in) employment, L2 literacy development, multilingual language pedagogy, and critical participatory approaches to adult migrant language education.

2 Life and Language Learning Contexts of Bilingual Migrants 2.1 Super-Diversity and Intersectionality The movement of large numbers of people from diverse backgrounds from all over the world creates spaces where languages and cultures come into contact in new ways. Indeed, the mass movement of people associated with globalization, coupled with the mobility of linguistic and semiotic messages in online communication, now indicate cultural and linguistic diversity of a type and scale not previously experienced, and renders the overall environment of adult migrant language learning

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

253

inherently unpredictable. An understanding of adult migrant language education may benefit from new sociological and sociolinguistic tools equipped to cater for this unpredictability. For example, the concept of super-diversity, first coined by Stephen Vertovec as a description of the “diversification of diversity” (2006, p. 3), aims to capture the sense of the mass, rapid, and unpredictable movement of people that characterizes the current age. Super-diversity as a sociolinguistic concept is not without its critics, not least for its Anglocentric worldview (Piller, 2015; note that Vertovec was initially referring to the context of the United Kingdom in recent decades), and for its status in terms of its “unexamined normative assumptions about language” (Flores & Lewis, 2016). Nonetheless the term is retained in this chapter because it enables us to consider super-diverse practices that we might otherwise not have attended to. It also enables us to reconsider established understandings of language use and meaning making, including those that occur in language learning contexts. As Blommaert (2015) explains, the super-diversity perspective “enables us not just to analyze the messy contemporary stuff, but also to reanalyze and re-interpret more conventional and older data; now questioning the fundamental assumptions (almost inevitably language-ideological in character) previously used in analysis” (p. 4). Moreover, the concept of super-diversity also affords us an acceptance of a new paradigm of uncertainty, of movement, and of mobility, characteristic of the lives of adult migrants and their everyday language use. Given conditions of super-diversity, groups of adult migrants learning the dominant language of their new country will themselves often be diverse. This diversity is notable not only in terms of language background and geographical origin, but also in terms of educational trajectory and schooled experience, command of literacy in an expert language, immigration status and reasons for migrating, age and gender, and employment, inter alia. Individuals who share a similar background differ as well of course, in terms of personality, a sense of agency, motivation and investment in learning, and aspirations for the future. This suggests the relevance of an intersectional approach to the study of adult migrant language education. Block and Corona (2016) discuss intersectionality in relation to the dilemma that those examining language and individual and collective identities are confronted with, asking: “how can scholars in applied linguistics take on so many factors at the same time?” (p. 507). They conclude that it is not possible to account for everything. Nonetheless, researchers need to show “sensitivity, awareness and, ultimately, attentiveness to the necessarily intersectional nature of identity” (p. 507). Following Block and Corona, language education for adult migrants therefore cannot be considered in isolation from the migrants’ ethnic and gendered positioning, their social status (often as poor and sometimes unwelcome migrants), the circumstances of their migration,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

254

JAMES SIMPSON

the conditions in their new home, or the social, cultural, and political contexts through which they make their trajectories.

2.2 Characteristics of Adult Migrants in L2 Classes The characteristics of adult migrants who are developing their linguistic repertoires to encompass the dominant languages of their new homes vary from country to country (depending to an extent on global migration patterns), city to city, neighborhood to neighborhood, and of course from individual to individual. In the remainder of this section, I sketch out intersecting features that impinge upon language education for adult bilinguals, and that relate to characteristics of migrants themselves: their language and educational background and pre-migration language capital, their political status, their age, and factors associated with gender, family status, and employment. One clear difference between learners lies in the languages they speak. A study of adult learners of English (Baynham, Roberts, Cooke et al., 2007) found that 500 students in London and the north of England reported speaking more than 50 languages among them. Later studies have noted a similar range. A “census” view of languages, however, does not encapsulate the full complex picture of language use among such students, nor that many of the learners are already multilingual and multiliterate when they arrive in a new home. Multilingualism, as well as multiliteracy (including literacy in more than one script), is taken for granted by most adult migrant language learners. They are often surrounded by many languages; they use several languages themselves in a multilingual repertoire; they move between them (translanguage) as a matter of course; and use a lingua franca (e.g., English or another global language) with other speakers from diverse backgrounds. Moreover, the modes and media of communication are likewise diverse. It is common for individuals’ contemporary communication patterns to move fluidly between face-to-face communication and online communication using mobile wireless technology. For adult migrant bilinguals, that online communication will often be transnational: people belong to globally spread networks of diasporic populations. Diversity extends beyond countries of origin and first languages claimed. Educational backgrounds and previous experiences of literacy are far from uniform among adult migrants. In some formal languagelearning contexts for migrants, it is not unusual to find in the same classroom people who have received a university education together with people with very little schooling and therefore with little literacy in their first language(s). As explained in Section 4, the teaching of literacy for new readers and writers is considered by many teachers the most challenging area of adult migrant language pedagogy. This is not surprising considering that such students are learning to read and write for the first time, as

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

255

adults, and this in a new language. In a review of published research Collier (1989) established that a major factor correlating positively with learners’ literacy in the L2 was whether they were literate in their L1. Learners who were not took seven to 10 years to learn age-appropriate L2 literacy-related, context-reduced, and cognitively demanding academic language skills. A lack of foundational literacy also impacts other aspects of L2 language processing. In their studies of L2 development processes in non-literate second language learners, Tarone, Bigelow, and Hansen (2009) note that access (or not) to L1 literacy affects short-term memory, order of L2 acquisition, and grammatical form used in L2 oral narratives. Pettitt and Tarone (2015) have corroborated some of these findings in a case study of one multilingual adult English learner’s alphabetic print literacy development. They found that some syntactic elements of their participant’s oral production became more complex with increasing alphabetic literacy, though the development of alphabetic literacy did not appear to relate to oral fluency, lexis, or pragmatics. They conclude that limited formal school-based literacy instruction “is not necessarily a barrier to agency, to effective oral communication, nor to achieving lexical complexity comparable to that of L2 users at higher levels of education” (p. 36). The reasons for students not acquiring literacy when they were children vary. There are political, social, economic, and cultural barriers to schooling. The upheaval caused by military conflict and war is a reason why some children do not attend school, even in societies where the literacy rate was previously relatively high. Others may come from societies that do not have a strong literate tradition, or from a tradition that does not prioritize the education of girls. Others still may have been deprived of an education because of poverty. Lack of access to literacy has implications in the literacy-saturated world of the adult migrant in the postindustrial West, if not everywhere; for example, even the most unskilled manual work in Northern Europe now requires an ability to read and write. Students and potential students in language classes for migrants might be refugees (including those seeking political asylum), people from settled communities who may have been in the new country for many years, husbands or wives on spousal visas, so-called economic migrants, people who are joining family members, and people with work permits. In other words, adult migrant language learners represent a wide spectrum of people. The degree of welcome and the concomitant sense of belonging engendered upon settlement in a new country may depend on migrants’ political status. For example, refugees seeking asylum are increasingly unwelcome in certain parts of the world, and might have an uncertain future in a hoped-for new home. As discussed in Section 3, a further factor in the link between language and immigration policy is that many newcomers have to study a language specifically to fulfill language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

256

JAMES SIMPSON

requirements for gaining naturalization or permanent residence in the new country. The length of time someone might have spent in a new country before gaining access to formal education is an important factor in learning. Baynham, Roberts et al. (2007) found that relative newcomers (those who had lived in the United Kingdom for five years or fewer) made more rapid progress than long-term residents. Long-term residents often had had little chance to learn the dominant language when they first arrived. One of the classes in the Baynham, Roberts et al. study was for Hong Kong Chinese women who had been living in the United Kingdom for up to 30 years, but who had only recently started learning English formally (despite wanting to for a long time), because work and family commitments and constraints had prevented them from gaining earlier access to classes. A lack of childcare is a particularly acute problem for migrant women wishing to raise young children and attend regular language classes in the new country (Macdonald, 2013). Consequently migrant women’s learning may happen in a piecemeal way, over a longer period of time. Indeed, migration and asylum affect women in different ways from men, and this extends to their experience of language education, typically a gendered field. Migrant women’s chances of having received any formal education are generally lower than those of men. Additionally, people trafficking as part of forced prostitution affects women and girls almost exclusively, and the trauma associated with it will usually not be shared by men. A less obvious but important issue is the change in family patterns associated with movement across borders. Traditional family patterns can go through many shifts during and after migration. These are sometimes to the benefit of women but sometimes not; for example, many women migrants are single mothers who have been widowed due to war and conflict in their home countries and are therefore living in situations at odds with their traditional norms. It is clear, then, that L2 learning by migrants depends on a multitude of factors related to their specific circumstances and characteristics. One factor, age at the time of arrival in the host country, also distinguishes migrants from each other, but at this point it is not clear how age on arrival alone can explain any differences among adult migrants’ L2 learning (see Singleton & Pfenninger, Chapter 4, this volume, and Biedron´ & Birdsong, Chapter 16, this volume, for extensive discussion).

2.3 Employment A pressing reason for learners to engage with learning the new language is employment (see also Grin, Chapter 9, this volume). There is no doubt that a country benefits from a multilingual workforce, and one with competencies in the new language. As explained further in Section 4, employment is a key plank in national immigration policies relating to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

257

integration. Adult migrants bring with them a wide array of qualities and attributes that would normally give them status in society, including previous education, language and literacy, and a range of qualifications, skills, knowledge, and prior experience. But the linguistic capital that multilingual adult learners bring with them to the new context relates in complex ways to their integration and ultimately their sense of belonging. Some migrants who find work in their new country might be employed below their professional level and may remain in this position for years. Bourdieu’s (1991) work on the forms of social capital and its extension to language affords a link between the learning and the use of a new language and issues of power that are fundamental to the difficulties faced by bilingual language learners attempting to gain a foothold on employment ladders of their new country. Bourdieu’s notion of social capital as an index of relative social power suggests that the same forms and amounts of capital may result in different positioning vis-a`-vis different fields, so the forms of capital that are valued in one place (the home country, for example) may not be so in the new home. Interestingly, whether or not an individual newcomer has a partner from the new country also makes a difference to their economic as well as social position. Meng and Meurs (2009) studied the role of intermarriage in the process of what they term migrant economic assimilation in France. They examined the extent to which migrants who have intermarried (i.e., married someone born and brought up in France) have successfully joined the labor market, as measured by earnings, compared to their nonintermarried counterparts. Meng and Meurs found that people who had intermarried earned around 25% more than those who had not. Moreover, the “intermarriage premium” appears to be higher for individuals who already have a strong grasp of the French language before migration than for those who do not.

3 Policy Support for Adult Bilingual Development From the perspective of the nation-state, migration typically outpaces the development of policies and infrastructure that address the presence of new migrants and the linguistic diversity that their arrival entails. That said, national governments generally accept that new arrivals should use the dominant language(s) and language varieties of their new country. In language policy, understanding and using the dominant language of the new country is not only a proxy for national unity, but is often seen as a sine qua non of integration (the term “linguistic integration” is often invoked) and social cohesion. This understanding is shared by policy makers, language educators, and migrants themselves: acquiring good communicative abilities in the standard variety of a language is felt to equip newcomers with the means for navigating a fresh social context

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

258

JAMES SIMPSON

(see Treffers-Daller, Chapter 15, this volume, on bilingual abilities). This extends to competencies in reading and writing: an assumption is easily made that literacy in the standard variety is essential for getting by and is the route to a successful future. It is perhaps not surprising, then, that language education for migrants at the scale of national policy rarely embraces bilingualism or mutilingualism, that is, the development of competencies in the dominant language as part of a multilingual repertoire. On the other hand, there is some supranational policy interest in multilingual education, and in language education that recognizes languages other than the new language. For example, UNESCO (2003) stresses the importance of L1 instruction, and encourages UN member states to view such instruction as a strategy for promoting quality in education. The Council of Europe’s Linguistic Integration of Adult Migrants (LIAM) project is perhaps the most comprehensive supranational policy initiative concerning adult bilingual language support (see Beacco, Little, & Hedges, 2014). On the project website, the guiding principles are set out as follows: Languages are an essential instrument for building intercultural understanding and social cohesion. The language or languages of the host society into which migrants are seeking to integrate [emphasis in original], and the languages which are already part of their individual linguistic repertoire [emphasis in original], shape their identities as active, democratic citizens. A plurilingual and intercultural approach to the teaching of the language of the host society ensures that languages become instruments of inclusion that unite rather than segregate people. (Council of Europe, n.d.)

In their introduction to the LIAM guide to language policy development and implementation endorsed by the Council of Europe, Beacco et al. (2014, p. 12) propose that language programs designed to support linguistic integration should take into account the following: (1) the languages that adult migrants already know: programs should acknowledge these to help migrants learn the new language; programs should encourage migrants to value their L1(s), because this may help their self-esteem; and programs should encourage migrants to speak their L1(s) within the family, the reason being that these L1(s) will enrich the host societies; (2) the language needs of adult migrants: these should be identified but also discussed with the migrants themselves; and (3) the diversity of migrant populations: language education programs should adjust their approach to the particular situations of individual migrants (see discussion in Section 2). The LIAM project does not propose that there be L1 instruction, while other projects have. However, L1 instruction programs are instituted for

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

259

linguistic integration, chiefly as a pathway to the dominant language, and to literacy in that language. As Rivera and Huerta-Macı´as (2008) explain, L1 literacy programs are typically devised as a stepping stone to the target language, on the understanding that the development of L1 literacy will equip students with the skills and abilities to transfer to L2 literacy acquisition. However, L1 education is generally seen as too expensive and impractical to attract central government funding. Indeed, generally, policy discourses about migrant integration stress that it is the societally dominant (i.e., the new) language in which competencies should be developed. This is seen as crucial for employment (see also Lo Bianco, Chapter 8, this volume). Policy arguments relating linguistic capital to migrant integration suggest that being able to communicate in the host country language is one of the main drivers of successful economic and social integration of migrants (e.g., Isphording, 2015). Immigration policies as they relate to who may or may not enter a country are tightly intertwined with labor market mechanisms and language requirements. Highly skilled migrants with demonstrated competencies in the L2 tend to be welcomed, while those without accredited skills or certified L2 language capability tend not to be (see Treffers-Daller, Chapter 15, this volume, for discussion of issues related to measuring bilingual abilities). In some countries, including Australia, Canada, and the United Kingdom, a points-based system is in place for the granting of visas. For example, hopeful entrants to the United Kingdom are awarded points for qualifications, expected earnings, available financial assets for maintenance, and English language skills. Demonstration of proficiency in the L2 on a test and/or in an interview is also used as a key gatekeeper for the attainment of citizenship, naturalization, or permanent residence as part of the process of settlement in the new country once people are there. In this sense, language is a prerequisite for integration, rather than (as most language educators would see it) an outcome. Policies vary, though there has been a rising trend since the beginning of the 21st century for continued residence in a new country to be dependent upon reaching a certain level of language proficiency. For instance, by 2016, 28 (78%) of the 36 Council of Europe member countries had some kind of language requirement for migration purposes, up from 58% in 2007 (ALTE, 2016, p. 9). Proficiency is typically measured by a standardized language exam or a de facto language and literacy assessment in the shape of a citizenship test. The danger of language testing for citizenship and naturalization is that rather than fostering a sense of integration and inclusive citizenship, such testing regimes promote a feeling of exclusion and a message that some migrants belong more than others. To address this, and to ensure that testing does not impinge upon the civil and human rights of the test taker, the purpose of language testing for citizenship should be clear, fair, and commonly understood.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

260

JAMES SIMPSON

Once adult migrants have arrived in their new country and are settling in, education – including language education – is considered in social policy to be an effective tool to assure their better integration into their host countries. In their collection of studies of policy and practice in adult migrant language education in eight countries, Simpson and Whiteside (2015) noted, however, that “national policies concerning language education for new arrivals in most states [. . .] are inconsistent, contentious and contradictory, responding in uneven ways to the dynamic diversity associated with migration” (p. 1). How, then, can adult bi- and multilingualism be supported in pedagogy, given the multifaceted concerns of migrants’ lives and the complexities and inconsistencies of policy frameworks? This question is addressed in the next section.

4 Support in Pedagogy The advantages of having access to the dominant language(s) and the privileged varieties of the new home, and of developing a measure of competence in these, are more than apparent to the majority of migrants. Many (though not all) are highly motivated to learn. Key to successful learning is interaction: from a cognitive perspective, interaction is crucial in providing input, opportunities for negotiation of meaning, and the requirement to produce language output. Socially oriented perspectives on language learning view it as occurring through interaction as social participation, seen as central in the successful development of new identities and a sense of belonging. Success in this sense can be identified in many different ways. For example, in her analysis of interviews with 76 UK-based adult migrants about their language learning needs, Cooke (2006) notes that a feeling commonly reported by beginner migrant language learners is discomfort at their dependence on interpreters, friends, or even their own children to help with bureaucratic and medical encounters. Many of Cooke’s participants talk of their language learning achievements in terms of breaking this dependency. Drawing on her research with adult migrants in Canada, Norton (2006) proposes the construct of investment as appropriate to describe migrants’ language learning, to complement more established understandings of language learner motivation. Investment signals the relationship of learners to the target language and their desire to learn and practice it. For adult migrants, investment in language learning can be tinged with ambivalence, relating to the way they settle into life in a new country. As Norton writes, while adult learners (of English, in her case) “may strive to make a productive contribution to their new societies, unless the host community is receptive to their arrival, they will struggle to fulfil their potential” (p. 96). This view is echoed in Yates’s (2011) study of interaction and social inclusion for 152 new arrivals to Australia in the early months of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

261

settlement. Yates found that newly arrived migrants have very limited interactions in English outside classrooms in either social situations or in the workplace, and argues that: where social connections are not made through English, immigrants can lack a sense of affiliation and remain isolated and insulated in their “ethnic bubbles.” While such bubbles may support a sense of belonging to their ethnic community, they do not facilitate either the development of proficiency in English or access to broader social networks that will provide much needed connections to the local and global Englishspeaking communities. (Yates, 2011, p. 469)

Yates’s conclusions point to the importance of formal language lessons to equip new migrants with the language and cultural skills to participate in dominant language interactions. At the same time, Yates notes that interaction, like integration, is a two-way street, maintaining that it is also crucial to “equip speakers of the dominant language with the awareness, attitudes and skills that will help them to engage and communicate more successfully with new arrivals” (p. 469). How best, then, to address the language and interactional concerns of a diverse population of language learners in pedagogy? In the remainder of this section I focus on four disparate areas of concern where adult migrant language education can make a difference: (1) language learning for employment, (2) literacy development, (3) multilingual pedagogy, and (4) critical and participatory approaches in adult migrant language education. Each section includes illustrative examples of resources or programs that might be employed as models of practice.

4.1 Language Learning for Employment Migrants in language classes who are already workers need a complex set of competencies, including the specific institutional and occupational discourses of their jobs. In addition, accessing knowledge about – and negotiating – rights requires interactional competence, as does forming relationships with coworkers (Gumperz, Jupp, & Roberts, 1979). Language for work courses do not necessarily provide such richness and breadth, however. Sandwall (2010) describes a work placement scheme for adult migrants in Gothenburg, Sweden, the intention of which was to enable new arrivals to develop their competencies in Swedish as part of the basic Swedish language program for adult immigrants, Svenska for invandrare (Swedish for Immigrants). The student in Sandwall’s case study maintained very firmly that she learned more at school, highlighting “the need to discuss assumptions about language learning at work placements in relation to the student’s trajectory and the workplace on offer” (p. 542). A more critical concern is that practical work placements such as the one described in Sandwall’s study can be used as a mechanism for social

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

262

JAMES SIMPSON

exclusion, introducing migrants to low-grade work, whatever their educational background or work experience prior to migration. Companies with employees who are migrants might nonetheless enroll them in specially designed language learning courses. There are many examples of employment-oriented language programs worldwide that have been evaluated as successful. For example, partnerships between employers and professional educators are crucial for the success of the Deutsch am Arbeitsplatz (German at Work) initiative in Germany, whereby trained teachers work with employers and labor unions to develop appropriate workplace language-instruction programs (McHugh & Challinor, 2011). McHugh and Challinor (2011) recommend expanding language instruction to be contextualized for workplace use, to combine language pedagogy with work skills training, to encourage partnerships, and to work with employers and unions.

4.2 Literacy Learning for Adult Migrants As mentioned in the foregoing discussion of learners’ life and learning contexts, L2 literacy acquisition is a special concern for many adult migrant bilinguals. Meeting the needs of adult migrant language learners without well-developed access to literacy creates several challenges for teachers and organizations providing instruction. In their practical guide to teaching basic literacy to adult migrants in the United Kingdom, Spiegel and Sunderland (2006) point to a number of factors that complicate matters for teachers of basic English for Speakers of Other Languages (ESOL) literacy to bilingual students. Some students come to language classes with an ability to read and write another language that uses the same script. Others might be familiar with an ideographic writing system, a syllabary, or a non-Roman alphabet. Others still may have little or no knowledge of any writing system at all. Thus, students of basic literacy arrive in their classes with different starting points, and classifying students according to their literacy needs becomes problematic for teachers. One distinction that teachers find helpful is between those students with some foundational literacy in the L1 and those with none. Because of the circumstances that drove them to relocate in the first place, some migrants may have missed out on formal education as children, and consequently did not learn to read and write well. Those with some L1 literacy can be viewed as having skills to transfer onto literacy in their new language (see Section 3). In migrant language classrooms, teachers appreciate that progress is slower among those with no literacy skills to transfer (Bell, 1995). Teachers also recognize that people are able to transfer fundamental knowledge that they have about literacy, regardless of script. For example, people may realize that writing may depend on specific sound-symbol links (Spiegel & Sunderland, 2006). Building on this position, Vinogradov (2009) suggests a range of activities for beginner

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

263

adult L2 literacy learners, based on their own learner-generated texts, and with a focus on their bottom-up reading skills. These include sequencing, word recognition, phonemic awareness, and phonics tasks.

4.3 Multilingual Pedagogy Some researchers in bilingualism and biliteracy maintain that adults acquiring literacy for the first time will learn more effectively if literacy is taught in their L1: the stepping stone described earlier. This belief is based largely on a body of research carried out on children in the early grades of school. Furthermore, researchers taking a critical stance toward L2 literacy learning maintain that teaching students literacy in the L2 rather than the L1 is actually unlikely to be effective. For example, Auerbach (1993) suggests, writing about the US context: The result of monolingual ESL instruction for students with minimal L1 literacy and schooling is often that, whether or not they drop out, they suffer severe consequences in terms of self-esteem; their sense of powerlessness is reinforced either because they are de facto excluded from the classroom or because their life experiences and language resources are excluded. (p. 18)

In most places, L1 or bilingual literacy education for adults is controversial and is hardly ever used. In the United States, for example, the English Only movement fiercely lobbies against L1 literacy education. Teachers, however, are often aware of the massive task facing students with a low level of oral proficiency in the L2 who are attempting to learn literacy at the same time, and in places where there are large numbers of people from the same linguistic background, it would seem sensible to at least consider bilingual instruction as an option. In principle, there would also seem to be no necessary contradiction between supporting the maintenance and development of migrants’ L1s on the one hand and helping them to acquire the dominant languages and varieties of the new country on the other. In a publication supported by the Council for Europe’s LIAM project (see Section 3), Beacco, Krumm, and Little (2017) argue that both the L1 and the L2 may support each other through teaching activities “that give legitimacy to migrants’ linguistic repertoires” (p. 2) and that rely on students’ languages. Such a position creates a space for language pedagogy based on current sociolinguistic understandings of contemporary language use that are commensurate with life in super-diverse environments. A traditional view of bilingualism rests on the idea of two languages with two separate linguistic systems (an L1 and an L2). Sociolinguistically informed theories of translanguaging, however, take a different starting point, viz., they take a “speaker’s view” whereby mental grammar has developed in social interaction with others. Such a translanguaging

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

264

JAMES SIMPSON

perspective assumes that from a user perspective, there is just one linguistic system with features of two or more societally defined languages that are integrated throughout (Garcı´a & Li Wei, 2014). When people translanguage, they sometimes use these features – which are simply their own – in ways that align with constructions of “a language.” Often, though, they use them differently, for example to produce new practices in ways that emphasize the artificiality of boundaries between languages. This is most evident when languages and cultures come into contact, as in many if not most migration contexts. If multilingualism is seen as a resource, the inclusion of languages other than the dominant one in education can be viewed as productive (see Garcı´a & Tupas, Chapter 20, this volume). There has as yet been little research on translanguaging in adult bilingual learning contexts. An example from higher education involving multilingual academic sojourners in France, however, suggests the potentially broad applicability of a translanguaging approach. Mathis (2015) describes a literacy project in a French university that focuses on the expression of students’ plurilingual identities. Students from Lebanon, Morocco, Canada, and France were asked to carry out reflective writing activities where they concentrate on their own personal experiences with languages and migration journeys. They mostly wrote in French but also used other languages. In thus using their plurilingual abilities, “social actors take up, in their literacy practices, the positioning of learners to those of experts, and of being monolingual to plurilingual, while expressing tensions and creating new ways of conveying who they are in the world” (Mathis, 2015, p. 147). In a very different adult migrant language-learning context, Garrido and Oliva (2015) describe a multilingual workshop approach to teaching Catalan to migrants without official status, who are not entitled to state support, with a focus on translanguaging and intercultural debate. They explain how translanguaging as a pedagogical approach is appropriate in Barcelona, especially with learners who have learned languages in multilingual contexts and without formal instruction. In Catalonia, according to these authors, migrants have to learn to use both Catalan and Castilian Spanish if they are to fully participate. They will also be using their L1s and other lingua franca languages (e.g., English) as a matter of course. The translanguaging enabled in the workshops – that is, the fluid movement across the learners’ repertoires – both maximizes comprehension for a heterogeneous group and allows connections to be made between the classroom and life beyond its walls. Understandings of notions such as multicultural education, culturally responsive teaching, and culturally relevant pedagogy have much in common: they all refer to the use – and recognition of the value – of more than a single language in teaching and learning, to an awareness of different lived experiences and cultural worldviews of students, and to the importance of drawing upon prior knowledge of students with various linguistic, cultural, and ethnic backgrounds. There appears to be a critical flavor to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

Supporting Bilingualism in Adult First-Generation Migrants

265

the nascent research and pedagogic activity around translanguaging in adult migrant language-learning contexts, and it is to critical and participatory pedagogy that we finally turn.

4.4 Critical and Participatory Pedagogy If a multilingual turn has yet to reach mainstream adult migrant language education, language pedagogy for adult migrants still requires innovative responses to linguistic and cultural diversity and to the new mobilities of the 21st century. Critically oriented teachers recognize that many migrants are not only concerned with a wish to access the new language to enable them to operate effectively in daily life but are also engaged in a struggle for recognition and equality. Inspired by the writings of Brazilian Marxist educator Paulo Freire in books such as Pedagogy of the Oppressed (1970), participatory pedagogy has been practiced by some educators in Europe and North America since the 1970s, particularly in the teaching of adult literacy. Participatory pedagogy advocates that participants set their own agenda, devise their own learning materials, take action on the issues they identify as important, and evaluate their progress and the effectiveness of their programs as they go. The syllabus, therefore, is not brought along by the teacher but rather emerges from class to class, driven not by an external curriculum defined a priori but by the students themselves. An example of a participatory teaching initiative is the Whose Integration? project (Cooke, Winstanley, & Bryers, 2015), whereby teacher-researchers explored critical participatory ESOL pedagogy with their adult migrant students in London, England. The aim was to relate language and literacy learning to the critical concerns of students’ lives, on the students’ own terms. This can equip students with critical skills that can be transferred beyond the classroom to effect social action. Drawing on cognitively oriented understandings of interaction for language development, the project recognized first that adult migrants’ progress in speaking required the production of turns of talk that were longer and more sophisticated than are typical in many ESOL classrooms. At the same time, the content of classroom discussions needed to suit students’ out-of-class needs and interests on a personal, social, and political level. Whose Integration? addresses a contemporary concern – integration into a new society. This is an issue that centrally relates to adult migrant language learners, but is one upon which they are rarely consulted. In the project, the students’ discussions were so intense that they carried them on outside class with friends and at home with their families.

5 Conclusion This chapter began by noting that supporting bilingualism in adult firstgeneration migrants entails two things: the development of competencies

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

266

JAMES SIMPSON

in the dominant language of the new country, and the maintenance of, and possibly literacy development in, the L1(s), both as part of a multilingual repertoire. The diversity inherent in the adult migrant student body was then sketched out, noting that many adult migrant language learners are developing their bilingualism in conditions of super-diversity, and bringing in the notion of intersectionality as an appropriate empirical approach to the study of migrant language education. A number of characteristics of adult migrant bilinguals were covered, across a range of dimensions, stressing how each of these in its different ways has implications for their language development. Two areas of support were discussed. First, policy support for adult migrant language learners, where the suggestion was that there are gaps and contradictions in policy worldwide. Second, the chapter discussed how adult migrant bilingualism can be supported in practice, by focusing on language learning for employment, literacy learning, the use of multilingual pedagogies, and the use of critical participatory pedagogy. In sum, supporting bilingualism in adult migrants is a demanding area for policy makers and language education practitioners alike. This is unsurprising, as in addition to the sheer diversity of the field, this is the area of language education most closely tied with global shifts that lead to mass migration, and the difficult supranational, national, and local policy decisions that this entails. For many practitioners, the relation between language education and migration policy is uncomfortably close, and the inconsistencies and changes in direction of policy responses are unsettling. Nonetheless, this remains a rewarding if sometimes taxing area of pedagogy, one that typically inspires tremendous commitment and engagement on the part of both teachers and their students.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:27:41, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.014

14 Unlearning and Relearning of Languages from Childhood to Later Adulthood Merel Keijzer and Kees de Bot

1 Introduction Language development encompasses both language learning and unlearning. Throughout their lifespan, individuals experience periods of spurts and arrests in their language use: their language develops in patterns that are not linear and far from always predictable if one just considers the input. For years, cognitive scientists have studied how the human brain deals with incoming information: how we learn, and how and where this information is retained, including what we know about words, grammatical rules, and so on. This is reflected in the substantial work that has been done in second language acquisition (SLA) and bilingualism research regarding how languages are (best) learned. Also featuring prominently in cognitive science is the question of how memories are retrieved, and what happens in the case of retrieval failure (e.g., Anderson & Spellman, 1995). In contrast, notably less attention has been paid to the other side of the coin: how languages are unlearned. Most work in this domain has been done within the field of language attrition (see review by Ko¨pke, Chapter 18, this volume). One of the emerging consensus points among attrition researchers is that individual differences very much make up the attrition outcome. In other words, it is impossible to fully predict the conditions under which language attrition takes place. Individual differences notwithstanding, however, there are contexts in which bilinguals or multilinguals will almost certainly “unlearn” one or some of their languages, be it the language(s) they learned from early childhood onward, or additional languages learned as second or foreign languages at some later point over the lifespan. The present chapter explores language unlearning by mainly (young) children and older adults, who represent the two far ends of the lifespan

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

268

M E R E L K E I J Z E R A N D K E E S D E B OT

and concomitantly very different contexts of life. Most of the relevant research has studied these populations. More specifically, our review explores the nature of language unlearning and discusses empirical evidence suggesting that memories of any language once learned (whether the first or any subsequent one) always leave traces that can be recovered if they are probed with appropriate research methodologies. For instance, in our own work, we have underscored the importance of testing the extent to which a language may be completely unlearned and lost to bilinguals (or not): in testing we included a language retraining phase that probes the allegedly forgotten linguistic memories of the attrited language. We also identify gaps in our knowledge regarding the unlearning of languages pertaining to the middle of the lifespan. Pointing at theoretical and methodological causes of this knowledge gap can itself shed light on the nature of language development, including unlearning. It can also inform future investigations of language unlearning across the lifespan. The chapter is organized as follows. First, we explore what it means to unlearn a language. This is followed by an overview of the main findings that have shaped our understanding of language unlearning. Both unlearning and relearning are discussed as central facets of language development. We first focus on childhood and older adulthood. We then discuss the little that is known at present about language unlearning in the middle of the lifespan. While we are fully aware that researchers may differentiate between bilingualism and multilingualism, we use the term bilingualism to refer to the learning and unlearning of one of two or more languages within a single speaker. The ultimate goal of this chapter is to answer the question of whether it is actually viable to speak of language unlearning.

2 What Does It Mean to Unlearn a Language? The term language unlearning has been used in different contexts and with different meanings. As a term, it presupposes an almost active process of undoing earlier learning. As such, it has most often been used in reference to beliefs about language learning and teaching held by both language learners and teachers and what can be brought into the sphere of consciousness. Unlearning refers to changing habits, and to becoming aware of teaching and learning strategies that are in need of change (McLaughlin, 1992). Within the Universal Grammar (UG) tradition, the term unlearning has mostly been used to denote the unlearned language that children possess from birth, under the assumption that they come endowed with whatever abstract knowledge they need in order to learn natural language. This unlearned language is referred to as the “language of thought”: innate, inner, and private, it is an internal code that presents a blueprint or set of principles needed for setting parameters upon being exposed to natural

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

269

language (Malcolm, 1995, p. 73). Fodor (1975), building on the foundations of UG, explicates this by saying “one cannot learn a language unless one has a language” (pp. 63–64). In subsequent UG-oriented work on SLA, unlearning has also been used to refer to the effects of negative evidence in changing a previously erroneously used rule in the L2 learner’s mind. Thus, UG-based work (in both L1 and L2 learning) essentially builds on the existence of an innate representational system, and on computations that can modify this representational system, to learn or unlearn previously established rules (Malcolm, 1995). The innate representational system and computations “are not consciously accessible” (Fodor, 1975, p. 49). Largely due to the contribution that UG has made to cognitive science, the latter field has been permeated with the belief that the human brain processes information much like a computer. Let us give an example from the lexical domain: under the cognitive science premise of our brain as an information processor, newly learned words as well as the rules that tell us how to manipulate them are stored in the brain (see Epstein, 2016, for a more elaborate outline of this example). Once a new word is encountered, it enters a temporary buffer before being transferred to long-term storage, from which it can then be retrieved upon demand. In recent years, this idea has been fundamentally challenged. As Epstein (2016) notes, computers really do have physical memories that they store and retrieve and can replace upon demand. Organisms (humans), however, do not. Humans do not have a store of words or grammatical rules, not even memories. As such, linguistic memories cannot be consciously unlearned either. Indeed, in the radical embodied cognition framework (Chemero, 2013), the idea that the human brain mirrors a computer in storing and retrieving language in whatever form is completely rejected. Instead, such intelligent behavior as changes in language skills is best explained by the direct interaction between organisms and the world they live in. In other words, language learning, language unlearning, or even language relearning occur as a direct result of a speaker’s interaction with the environmental input. By extension, the same experiences shape organisms in different ways, making the results unpredictable; interactions with the environment affect us differently because they build on unique neural structures already in place that themselves were formed over a lifetime of unique experiences. This also means that the precise patterns of language unlearning are hard to predict. Radical embodied cognitive science integrates notions from Dynamic Systems Theory, the phenomenological tradition in psychology and ecological psychology. As Chemero (2013) mentions, “[radical embodied cognitive science] is skeptical of the explanatory usefulness of mental representations” (p. 145). Leaving out the notion of representations is indeed a radical change, since most current psycholinguistic theories lean heavily on this notion. While it has proven difficult to shake off the information-processing metaphor (Epstein, 2016), there is a general consensus that language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

270

M E R E L K E I J Z E R A N D K E E S D E B OT

unlearning results from interactions with the environment. Nevertheless, different terms have been used for the unlearned language that forms the end result. Past work has interchangeably used the terms language loss, language attrition, language unlearning, and language forgetting, contributing to the current situation where it is unclear what language unlearning exactly entails. In this chapter, we distinguish between those terms at the outset. Specifically, we view language attrition and forgetting as unconscious phenomena taking place within an individual. By contrast, we propose that language unlearning be used to denote a conscious process of changing habits or automatized skills. Language loss, then, is the overarching term to describe both processes. At the same time, we challenge the use of these very terms, and revisit terminology in our final discussion.

3 Can People Forget a Language? Language attrition is the loss of language skills in individuals. But can people forget a language? This is an interesting question to ask, because favoring the term “forgetting” over “loss” implies that language skills are not lost forever. At the same time, the question is imprecise: what is meant by forgetting (Schmid, 2011b)? The latter question bears relevance to the time frame in which attrition takes place. In order to allow a large enough window for forgetting to take place, earlier attrition studies typically set the threshold for participant recruitment at a minimum of 10 years’ immersion in the new language environment. However, this traditional approach to language and migration presupposes that a relatively homogeneous (language) group with a stable and fixed L1 moves to a different but equally stable L2 environment (as evidenced by the group comparisons that characterize the field of attrition). Recently, these very foundations have been challenged by the contemporary social sciences paradigm of mobilities. People, their attributes, and their ideas are never static, given, or fixed, but more fluid to begin with. Imposing a minimal time frame of 10 years following the move of attriters bypasses this fact in a top-down manner (see Brenner, 2004, for more discussion). Perhaps partly inspired by the mobilities paradigm, more recent investigations have shown that substantial – and often most – language forgetting takes place in the initial years following a change in the ecology of language exposure. An exemplary study is that of Chang (2012), who looked at L1 American English speakers becoming immersed in an intensive Korean as a foreign language course. The results showed that the participants underwent phonetic attrition in their L1 as little as six weeks into the experience. This was particularly noticeable in English stop consonants and vowels. Speakers restructured their English vowel space to assimilate to the Korean vowel space. Such short-term attrition is often explained by the need to avoid

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

271

interference from the L1 in the initial period of coming to terms with a new language, a need that is helped by a strong suppression of the L1. In his Activation Threshold Hypothesis, Paradis (1993) stipulates that L1 inhibition raises the activation threshold of L1 items, making it hard to access the L1 upon demand. Taking this a step further, subsequent studies have challenged the premise that attrition needs to affect the language system globally. Scholars have also pointed out local attrition effects in speakers who, like the participants in Chang (2012), just started learning a second or foreign language and who are not immersed in an L2 environment but instead continue to live and move in an L1 environment (e.g., Bice & Kroll, 2015). Such contributions place attrition within a broader framework of activation and inhibition of bilingual language use. Following such a perspective, Anderson, Bjorn, and Bjorn’s (1994) theory of Retrieval Induced Forgetting (RIF) is relevant, although it was developed in cognitive psychology without specific reference to language. Under this theory, it is assumed that when one retrieves a piece of information, related knowledge is inhibited in order to avoid interference. A classic RIF paradigm consisting of a familiarization, practice, and test phase can demonstrate this. Participants are first presented with a number of category–exemplar pairs (e.g., fruits - apple, fruits - kiwi, drinks - wine). A practice phase then follows in which subjects practice with half the exemplars of half the categories, being prompted by means of stimuli such as in fruits - a. In the final test phase, all exemplars of all categories have to be recalled. Perhaps unsurprisingly, the practiced items are most easily recalled thanks to a clear facilitation effect, but – more interestingly – exemplars from unpracticed (baseline) categories are more easily recalled than unpracticed exemplars from practiced categories (so in this case drinks - wine is easier to recall than fruits - kiwi). This is explained through inhibition mechanisms that are needed to keep exemplars from the same category from impinging on the practice item. Bringing all of this back to L1 attrition means that when a word for a certain concept is retrieved in the L2, subsequent retrieval of the corresponding (translation) L1 item should be harder to do. Levy, McVeigh, Marful, and Anderson (2007) looked at attrition data invoking RIF. Their goal was to test the theory’s prediction with a sample of L1 American English-speaking students taking a semester of Spanish in college. They found that these students became progressively slower in naming an L1 English item after the same item had been named 10 times in Spanish. Interestingly, when the stimulus word had a semantic link with the target word (e.g., being asked to name the word snake after seeing venom – s . . . . .), only a facilitation effect occurred (i.e., subjects named the target word faster). For the phonological prompt, on the other hand, retrieval-induced forgetting emerged. That is, retrieval (word naming) after seeing a written stimulus word that rhymed with the target word was markedly slower (e.g., the same word snake but this time seen after the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

272

M E R E L K E I J Z E R A N D K E E S D E B OT

phonologically related stimulus break – s . . . . .). Although the numerically small effect of this study has been critiqued, this research has offered an important step in understanding the nature of forgetting linguistic information. Indeed, with the introduction of RIF as a framework to study attrition, the threshold of 10 years needed for attrition effects to become noticeable has been challenged. This also means that the irreversible nature of forgetting can be examined more closely. The RIF framework allows for the position that the L1 will be reactivated upon more L1 input. However, this may not always be the case: Linck, Kroll, and Sunderman (2009) showed that RIF effects continued to exist in a group of L2 Spanish learners from an L1 American English background even up to six months following their arrival back in the United States after they had spent an exchange semester in Spain. Extending the insights from RIF and other general theories about forgetting further to different stages of the lifespan, one of the most robust predictors of L1 attrition for immigrants so far has been age at emigration. (Young) children invariably show more L1 attrition and do so earlier than adults undergoing an emigration experience. At the other end of the lifespan, in advanced age, there are also a number of phenomena that are invariably associated with language forgetting. Some of these changes are due to normal aging (see Goral, Chapter 5, this volume) as well as to phenomena like aphasia (see Hernandez, Chapter 23, this volume). A more detailed review of both these lifespan contexts can lead to a better understanding of what language forgetting is and does. Given the insights brought to this field through retrieval-induced forgetting approaches, however, it is crucial to not only examine language decline that results from inhibition, but also to relate such decline to how that language developed in the first place. In addition, one should investigate whether the language in question can be relearned and the circumstances that facilitate such relearning or reactivation.

4 Forgetting and Relearning of a Home Language in Children: International Adoptees and Heritage Speakers Language attrition studies invariably report the lowest levels of individual variation in young children. Moreover, age at emigration or, more accurately, age at onset of bilingualism is one of the most robust predictors of whether attrition will occur, with the general rule of thumb being that the younger (monolingual) children are when they moved and came into contact with a new language, the more L1 attrition they will show (Montrul, 2008; Schmitt, 2010). Theoretically, this has been explained as a reverse critical period: maturational constraints are at work so that a

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

273

gradual decline can be seen in the susceptibility to attrition as the age of onset of bilingualism increases (Bylund, 2009). While this seems to be a robust finding, the fact that maturational constraints may be at work insofar that as attriters get older, they are likely to show less attrition does not, of course, imply that language forgetting is inevitable in child migrants. When child migrants are tested in their L1 several years after they moved from their L1 environments, and if there have been either sudden or gradual decreases in exposure to the L1, a methodological conundrum is the question whether what is tapped is a manifestation of L1 attrition or of incomplete L1 acquisition. Indeed, especially at a preliterate age, the only substantial remaining source of L1 input a hitherto monolingual child receives after moving to a new L2 environment is generally from the parents and at home. Within a more nuanced view of language development that does not strictly adhere to the acquisition– attrition dichotomy, it becomes less interesting to talk about incomplete acquisition vs. loss. Moreover, the role of literacy in particular is more substantial than has previously been assumed. Ko¨pke (2007, see also Chapter 18, this volume) already noted that literacy – through the multimodal coding it induces – can attenuate attrition. Referring to so-called literacy anchoring, Schmitt (2010) sees literacy as an opportunity for both a constant as well as varied source of input. For the younger end of the lifetime spectrum, literacy pertains to how something is learned, or rather consolidated, and how this in turn impacts what is retained or forgotten. With these general L1 attrition or maintenance principles in mind, two main contexts are invariably associated with L1 attrition at young ages: the language loss found in international adoptees and in heritage language speakers, alternatively referred to as speakers of ethnic minority languages or community languages. In the most dramatic of these two contexts, international adoptees stop using their first language immediately following their adoption and subsequently show a very steep attrition curve of this L1. The literature often uses the term “birth language” for the language that international adoptees heard prior to adoption (Genesee & Delcenserie, 2016). The international adoptee literature mostly focuses on children’s learning of their new L2 or adopted language and distinguishes between very young children and children older than six. Although children make good progress in the adopted language, there is great inter-individual variation depending on factors such as children’s ages and their living circumstances prior to adoption (for comprehensive discussion of findings regarding adopted children’s new “adopted” language learning, see Genesee & Delcenserie, 2016). Very intriguing here is the question if traces of the allegedly forgotten language can be found years after adoption. One of the first and most oft-cited studies to have looked into this question is a study by Pallier, Dehaene, Poline et al. (2003) involving eight participants between 20 and 32 years of age who had been adopted

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

274

M E R E L K E I J Z E R A N D K E E S D E B OT

from Korea into francophone families living in France when they were between three and eight years old. The participants reported no recollection of the Korean language. This was corroborated by behavioral tests. More importantly, event-related functional magnetic resonance imaging evidence revealed no specific cortical activation in these subjects when they were listening to Korean stimuli (see Ko¨pke, Chapter 18, this volume, for explanation of some neurophysiological methods that are commonly used in psycholinguistic research). The adoptees were, in fact, no different from a group of native monolingual French-speaking controls. When tested using French stimuli, the adoptees again showed similar activated brain areas as the French controls, although the activation site for the French natives was more extensive. Pallier et al. (2003) used this finding as converging evidence against the existence of a critical period, as learning a language early in life should leave long-lasting traces in the neural circuit. However, in their study, the L2 French appears to have overridden the L1 Korean in the adoptees’ brains instead. Using more fine-grained follow-up tests of, among others, a recognition number series, Pallier and collaborators were able to differentiate between phonological and semantic memory, but only minimally so (Ventureyra, Pallier, & Yoo, 2004). The authors interpreted this as implying that the adoptees did possess some intuitions about the sound patterns of Korean, even in the absence of explicit knowledge of Korean words. Anecdotal reports of retraining in this subject pool, where previous experience with Korean should facilitate learning the language, did not yield any significant findings either: many of the adoptees had visited Korea (with visits ranging from a few days to a few months) and/or had taken Korean language courses. None showed any significant recovery of Korean in comparison with the French controls, who had never been exposed to Korean. In short, despite substantial early life exposure to Korean, virtually no traces of the language were found in this group of international adoptees. Especially the last facet, the reactivation-retraining component, was taken up later by two independent psycholinguistic investigations, but with different results. Choi (2014), also focusing on Korean as a birth language but in the context of international adoptees residing in Dutchspeaking families in the Netherlands, looked at how the birth language facilitated retraining of both phonological perception and production of Korean sounds. The adoptees were between the ages of 23 and 41 (mean age: 32) at the time of testing and had been adopted between the ages of 3 and 70 months, with a mean of 21 months. In an initial perception paradigm, the Korean adoptees and a group of Dutch controls were trained in identifying distinctive Korean alveolar aspirated (lenis and fortis) sounds. At pretest, the two groups did not significantly differ from each other, but midway through the training, the Korean adoptees showed superior performance both on the trained sounds and also on similar but untrained sounds. Although the Dutch controls did catch up, making the two groups

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

275

indistinguishable again at posttest, Choi (2014) interpreted this result as evidence that phonetic characteristics of a childhood language are saved as a linguistic memory. This was underscored by a follow-up experiment, where both groups were taught to produce the same Korean sounds that had been used as stimuli in the perception task. Their efforts were assessed by native Korean speakers. As a result of the training, both the adoptees and the Dutch controls markedly improved, as apparent from the native Korean speaker assessments, but only the Korean adoptees’ realizations were most often identified as target-like and were generally assessed more positively. This savings finding (see later in this chapter) corroborates earlier findings by, for example, Singh, Liederman, Mierzejewski, and Barnes (2011). Zhou (2015) carried out a similar investigation, but now with 4–11-yearold children adopted into the Netherlands from China. At adoption they were between 9 and 68 months of age. Prior to adoption, the children had lived in either a Mandarin- or a Cantonese-speaking region. Apart from occasional visits to China for some of the children, at the time of testing the adoptees had been cut off from exposure to their respective birth languages for an average of five years. Here too, following a perceptual training phase, the Chinese-origin adoptees were better able to recognize phonological contrasts in their respective birth languages (Mandarin or Cantonese) compared to age-matched Dutch controls. This study again provides evidence for first language traces, even when previously children had been only briefly exposed to their birth language (see Armon-Lotem & Meir, Chapter 10, this volume, for the role of exposure in young bilingual children) and with that contrasts with Ventureyra et al.’s (2004) findings for adults. The extreme cases of language loss and relearning seen in international adoptees can also be mirrored in a less dramatic setting, namely that of heritage language (HL) speakers. Following this volume’s editors, we define an HL as a language that someone learns as a child in the family and within the additional context of a community of speakers, often in competition with the dominant language of the majority society. Language loss in HL speakers is due to environmental influences typical in bilingual families more generally (see Armon-Lotem & Meir, Chapter 10, this volume; see also Gonc¸alves, Chapter 3, this volume). These can be directly linked to a geographical move, such as is the case with international migration. Often, though, they seem to organically arise out of the very bilingual situation that bilingual children find themselves in, where children’s exposure to their two or more languages is almost certain to be uneven and potentially subject to discontinuities due to the language of schooling, temporary travel, differential parenting roles (main language of communication used by which parent), and so on (see De Houwer, 2009). In studies of adult heritage language learners who have been exposed to that HL in early childhood but for whom exposure has ceased or greatly

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

276

M E R E L K E I J Z E R A N D K E E S D E B OT

diminished, it has repeatedly been found that they can attain higher levels of HL proficiency upon being re-exposed to the HL than speakers without previous HL experience (e.g., Au, Knightly, Jun, & Oh, 2002). Heritage language learning is interesting in this context, as it has often been classified as involving high degrees of attrition or incomplete acquisition. Rothman (e.g., Kupisch & Rothman, 2016) has long challenged interpretations of attrition or incompleteness, pointing out that the baselines to which heritage speakers are typically compared are obtained from literate and highly educated monolingual speakers of standard varieties from the countries of origin. Kupisch and Rothman (2016) point out that heritage grammars may be different from these monolingual grammars simply because the parents may be speaking to their children in a contact variety that has changed with regard to the varieties in the countries of origin, because the children may not be schooled in the minority language and therefore may not get sufficient exposure to certain aspects of grammar that are typical of written and formal language, or a combination of both. In other words, the main source of input in HL learners is not always a standard variety. This is reflected in HL speakers’ output (see also ArmonLotem & Meir, Chapter 10, this volume). While it is certainly possible for the main source of HL input to deviate from the input HL speakers would have received in an environment where their HL is the majority language, we question the usefulness of thinking in terms of incomplete acquisition or attrition. These distinctions do not do sufficient justice to the fluid nature of language development. It is the unique interplay between the environment and the bilingual individual and how this is reflected in the individual’s language output that is the really interesting issue. Indeed, in an early seminal paper, Gonzo and Salterelli (1983) already suggested that in many HL users there typically is a cascade effect: the generation who migrated undergoes attrition; this leads to them using a new variant that serves as input to the children, but that at the same time is also only partially transmitted to the next generation. Thus, any structural changes may be a combined effect of intragenerational attrition and intergenerational incomplete transmission. In many cases, the resulting linguistic reality may be more accurately described as a contact variety, worthy of investigation in its own right (see Aalberse & Muysken, Chapter 26, this volume, for more in-depth discussion). The studies reviewed here show the importance of combining the two crucial factors of input and age in predicting childhood language loss. At the same time, there are also less dramatic bilingual bases and contexts in childhood that are associated with attrition beyond those of international adoption and children of immigrants learning an HL. Children may unlearn their L1 as a consequence of new family arrangements like divorce or intermarriage, or because of travel. There is also a good portion of frombirth bilingual children growing up in bilingual families (bilingual first

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

277

language acquirers) who end up losing their speaking ability in one of their home languages without any apparent reason like travel or migration and in the absence of any abrupt change or major turning point in the lifespan (see discussion in De Houwer, 2009). Instead, language loss in such cases seems to evolve out of language use patterns and environmental influences within bilingual families. Indeed, while the terms L1 and L2 attrition, and also L2 acquisition, may be used liberally by attrition and second language acquisition researchers alike, for many bilingual families, these distinctions will not be so clear-cut. While important to point this out, the caveat does suggest that organic self-organizational mechanisms within bilingual families are highly unpredictable in terms of the language attrition they render (however, more and more research is being carried out that helps to predict in which circumstances attrition in children growing up in bilingual families is likely to occur). There is abundant evidence for unlearning in bilingual contexts in (early) childhood. What appears to be most pertinent for this end of the lifespan is how language unlearning is associated with type and contexts of learning (e.g., literacy anchoring, or the quality and amount of input in children’s two or more languages for heritage learners).

5 Forgetting and Relearning of Languages across the Adult Lifespan The other end of the lifespan, older adulthood, is another context in which language changes, including loss of language skills, are quite prevalent. There are overall intriguing mirror symmetries between childhood and older adulthood. These have been captured in Jakobson’s (1941) regression hypothesis. Also sometimes labeled “last in first out,” this hypothesis stipulates that those linguistic features that are acquired late in children are also the first to go in language-forgetting settings. Abundant evidence for regression has been found in the pathological domain, where similarities in agrammatical, that is, telegraphic, structures produced by both children and aphasics have been noted (Caramazza & Zurif, 1978; Kolk, 2001). For non-pathological language attrition in most notably migrant populations, evidence for regression has been found in the domains of morphology and syntax (de Bot & Weltens, 1991; Keijzer, 2007). Yet the occurrence of mirror symmetries between child language and either pathological or non-pathological language loss in older adults by itself does not explain why such convergences occur at both ends of the lifespan. Keijzer (2010) has argued that the regression hypothesis can best be theoretically approached from the vantage point of cognitive competition; she has proposed that competition can explain the similarities of surface forms or of errors between children and older individuals showing either pathological or non-pathological language loss, but that these stem from

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

278

M E R E L K E I J Z E R A N D K E E S D E B OT

different sources. Children’s cognitive systems are not yet fully mature, leading them to use simple language structures. In aphasia, likewise, cognitive resources are affected and this is the primary cause of changes in the linguistic repertoire of aphasic patients. In non-pathological attrition situations, on the other hand, competition mostly stems from two languages that compete for limited cognitive resources. In other words, the cognitive system is now fully mature and not damaged but is taxed due to demands in juggling two language systems rather than one. A confounding factor in language attrition research is, however, that many participants in the relevant studies are older adults. As mentioned earlier, most L1 attrition studies involving emigrated adults adopt a recruiting strategy of a minimum threshold of 10 years of L2 immersion, combined with a minimum age of at least 16 years at the time of emigration. A very pertinent issue then becomes how forgetting effects can be separated from normal aging effects (Goral, Libben, Obler et al., 2008), especially in the lexical domain, as this is most vulnerable to both attrition and aging effects (see overview by Higby, Lerman, Korytkowska et al., 2019; see also Goral, Chapter 5, this volume). Indeed, cognitive resources tend to decline in normal aging, as manifested in working memory and processing speed reductions as well as executive control compromises (see Higby et al., 2019). The context that then yields the most predictable language loss is ageassociated pathology. Degenerative diseases such as Alzheimer’s are prevalent in advanced age and are known to impact language use. Most notably, the semantic and pragmatic systems of Alzheimer’s patients are impacted, which frequently leads to communication problems (Ferris & Farlow, 2013). In recent years, a substantial amount of research attention has been given to the potential benefits associated with bilingualism in attenuating such degenerative diseases (compare, e.g., Mukadam, Sommerlad, & Livingston, 2017, and Grundy & Anderson, 2017). Substantial work has also been done on how aphasia differentially affects different language systems in bilingual patients (Paradis, 2001). Very interesting is the line of work done in relearning language once loss has taken place in bilingual aphasia. An important question is whether both of a bilingual individual’s languages recover in parallel or whether differential or selective recovery of one over the other language occurs (see also Hernandez, Chapter 23, this volume). Parallel recovery appears to be most common (Paradis, 2001), but Fabbro (2001) points out that recovery patterns in bilingual aphasia cannot be reliably predicted because of the intricate interplay between an individual’s lesion type, lesion site, the type of aphasia that results from it, and the contexts in and extent to which the languages had been used prior to the onset of aphasia. Interestingly, crosslanguage treatment has shown that rehabilitation of one language can positively affect the recovery of others in bilingual aphasics (Goral, Levy, & Kastl, 2010). Very much in line with recent discussions about a possible

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

279

bilingual advantage in attenuating aging effects, Alladi, Bak, Mekala et al. (2016) looked at the general cognitive recovery of stroke patients. They found that even though the age at which they were affected by a stroke did not differ for bilinguals and monolinguals, bilinguals recovered faster on the general cognitive level, irrespective of language recovery patterns. Compared to pathological language loss in degenerative diseases or following a stroke, non-pathological L1 attrition in older adults is far less certain to lead to language loss (see also Ko¨pke, Chapter 18, this volume). It does, however, see a similar trajectory as its pathological counterpart with regard to which aspects of language are affected, and what the recovery patterns are. In one of the few longitudinal studies on L1 attrition, de Bot and Clyne (1994) looked at changes in spontaneous speech in Dutch and German immigrants in Australia, who had Dutch and German as L1, respectively. These bilinguals had been tested by Michael Clyne in the early 1970s and were retested in 1987. The evidence suggested very little change over time. In particular, declarative knowledge (about, e.g., the gender of nouns) did not change over the 16 years of the study. De Bot and Clyne (1994) did witness some small but significant declines in lexical richness, however (operationalized as type token ratios). In 2005, some of the subjects who had been last tested in 1987 were visited again. By that time, as is evident from unpublished data by this chapter’s second author, many had moved into a Dutch-medium home for the elderly and now showed signs of reversion, in that their L1 Dutch seemed to have improved while their L2 English had deteriorated. This corroborates de Bot and Clyne’s (1989) twin hypotheses of L1 reversion and L2 attrition: as immigrants grow older, they tend to use the L1 more than they did in middle age; at the same time, they tend to forget L2 vocabulary and lose grammatical rules that they previously used in middle age. Although anecdotal evidence coming from elderly migrants themselves, their children, or caregivers abounds, very few empirical investigations have taken up these hypotheses. Clyne himself (2011) speaks of a reversion myth that has informed migrant aged care to great degrees; so far, however, without large-scale and robust empirical foundation. From the inhibition-activation perspective proposed by RIF and Paradis’s (1993) Activation Threshold Hypothesis, the reversion that is anecdotally reported can perhaps better be explained on the basis of the cognitive decline that characterizes normal aging and that makes it harder to suppress a language in bilingual contexts with advanced age. L1 interference in the L2 should then be reported as much as L2 interference in the L1. There is tentative evidence suggesting that this is indeed the case (Keijzer, 2011). Very much in line with what has been found for childhood unlearning, only the most extreme kinds of contexts make language unlearning predictable. For adults, those contexts are neurodegenerative diseases or aphasia brought about by a stroke. These often involve pathological language loss. In less dramatic bilingual contexts, not involving such

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

280

M E R E L K E I J Z E R A N D K E E S D E B OT

pathologies patterns of language unlearning in advanced age are much less easy to define and predict (see also Goral, Chapter 5, this volume). At this end of the lifespan, much can be gained from studies that actively attempt to disentangle effects due to normal aging versus language loss, and effects due to pathological versus non-pathological language loss.

6 The Great Unknown: Language Unlearning and Relearning in (Middle) Adulthood and the Savings Account Whereas in (early) childhood and older adulthood there are at least some contexts where language unlearning will occur almost without fail, the long period of time between these ends of the lifespan has offered less convincing or less consistent evidence of language unlearning. Instead, individual differences very much characterize attrition findings here (for an in-depth overview, see Ko¨pke, Chapter 18, this volume). In general, language unlearning comes about through an intricate interplay between speaker-internal factors and external environmental influences. It is especially the external factors, though, that appear to affect language unlearning and attrition in the years between childhood and older adulthood. Perhaps as a consequence, many bilingual investigations do not focus on the extended long period that is middle adulthood. In the years between childhood and old age, generally individuals’ cognitive systems are mature and well functioning. This offers the ideal framework in which to study language unlearning, because the cognitive confounds that exist for cognitively immature individuals (children) and cognitively declining individuals (in older adulthood) play much less of a role at these other stages of life. Thus it becomes much easier to focus specifically on language-related issues. The majority of what we know about language unlearning between childhood and older adulthood comes from L2 studies of middle-aged individuals looking at any residual knowledge of foreign and second languages learned later in life that had not been used for several years afterward. Despite self-reports by speakers that substantial L2 attrition had taken place, generally, studies in this area have found more retention than attrition of such skills (Hansen, 1999). Such findings have been taken as evidence for what has been labeled the savings/relearning paradigm. The savings or relearning paradigm was introduced by 19th-century German psychologist Hermann Ebbinghaus. In his 1885 book, U˝ber das Geda¨chtnis [On Memory], Ebbinghaus reports on experiments in which the only participant (viz., himself) memorized 169 lists of 13 nonsense syllables until he could reproduce them correctly. Next, after periods varying between 20 minutes and a month, he relearned a list he had learned

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

281

previously and additionally learned a similar, new list of nonsense syllables. His relearning reflected the advantage of relearning old words that he had in the meantime forgotten over new words. In Ebbinghaus’s experiment, relearning was operationalized as the difference between the number of trials needed to learn a previously learned list and a new one. In the beginning, forgetting took place rather quickly, but the rate of forgetting decreased as a function of time. This observation led to the assumption that once it has been transferred to long-term memory, information is never completely lost. Not until the 1970s did a number of experimental psychologists give new impetus to research on the relearning paradigm. Nelson (1978) investigated the properties of the (linguistic) information stored in memory traces and found that part of it is acoustic in nature. Nelson also showed that relearning serves a trace strengthening function: relearning trials resulted in an increased overall amount of memory strength that was sufficient for successful retrieval (Nelson, 1978). This is called the savings effect. MacLeod investigated the savings/relearning effect for translation equivalents (MacLeod, 1976) but also for pictures and words (MacLeod, 1988). Surprisingly, his experiments showed that savings was only detected in recall but not in recognition tests. This made MacLeod conclude that “relearning facilitates the retrieval of information, rather than (or perhaps in addition to) increasing its trace strength” (1988, p. 209). De Bot and Stoessel (2000) were the first to apply the savings paradigm to language attrition research. They compared the learning scores on a Dutch language test by two German adults who had spoken Dutch as children during a four-year stay in the Netherlands with those of a group of German control subjects who had not been exposed to Dutch before. On average, the experimental participants scored better on the (re)learning task than the control group. De Bot and Stoessel (2000) also administered a withinsubjects test consisting of previously acquired words taken from the original Dutch list of high-frequency words and a number of Dutch lowfrequency words. Both subjects appeared to have a significant relearning advantage for old over new items, confirming that residual lexical knowledge was still present in memory. De Bot and Stoessel’s (2000) study was the first of a series of studies of savings in vocabulary relearning of languages earlier acquired under natural exposure (see a review by Hansen, Umeda, & McKinney, 2002). Hansen et al. (2002) also conducted their own study on language attrition and savings of Japanese and Korean as L2s. Investigating a large sample of 304 native speakers of English after their return to North America from a stay in Japan or Korea lasting between 18 and 36 months, Hansen et al. (2002) found a considerable savings effect. They suggested that subsequent research into relearning should address the question of which variables affect the size of savings, with good candidate variables being, for instance, the original proficiency in the attrited language and age. A study focusing

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

282

M E R E L K E I J Z E R A N D K E E S D E B OT

on the role of one such potential variable is the one by Schneider, Healy, and Bourne (2002). In two experiments, Schneider et al. gave groups of non-French-speaking college students with English as a native language three trials of training on French–English vocabulary pairs and subsequently tested them on these translation pairs. In a second session one week later, participants were retested and then retrained on the same pairs, in both translation directions. The researchers found that the size of savings during relearning increased when during the first session students had been trained on the more difficult items, that is, on the items needing more repetitions in the learning phase. Research into savings was extended to the relearning of vocabulary of languages acquired under classroom conditions. De Bot, Martens, and Stoessel (2004) report on a series of experiments in which they compared the learning scores of newly acquired L2 words versus scores for “old” words students of an L2 are likely to have acquired in the past. In all three tests, the number of old words remembered was significantly larger than the number of new words in the list. Thus there appeared to be a significant savings effect for the old over the new words. The relearning paradigm has also been used with reference to phonetics. Bowers, Mattys, and Gage (2009) studied how a group of seven native English speakers tried to relearn the phonology of a language (Hindi or Zulu) they had learned in childhood and from which they had afterward been completely separated (see also our earlier discussion on internationally adopted children). At first there was no evidence that the participants had retained any knowledge of their childhood language, but after some practice participants under 40 years of age appeared to have regained sensitivity to phoneme contrasts in the language concerned. However, in participants over 40 years of age, there were no signs of relearning. Likewise, young control participants who had not been exposed to Hindi or Zulu before showed no learning. The authors concluded that even when adults have not retained any explicit memory of their childhood language, they may still have preserved traces of implicit knowledge of it. This finding is reminiscent of the international adoptees study results discussed earlier.

7 Avenues for Future Investigations of Language Unlearning Unlearning in the years between childhood and older adulthood is in need of further investigation. Beyond this, there are more learning, unlearning, and relearning arenas throughout the lifespan that deserve future research attention. Although in this chapter we have reviewed strictly bilingual contexts, many situations of both learning and unlearning involve more than two languages. Although there is work on multilingual

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

283

development in children as well as on change patterns that characterize multilingual aging, combining these lines of work could lead to valuable insights into language learning and unlearning as well as into the constraints under which they operate. In that same vein, it is imperative to understand more about language learning and unlearning in bidialectals (see Chevrot & Ghimenton, Chapter 25, this volume). Speakers who speak both a standard language variety and a dialect are often overlooked in bilingualism studies. Given the suggestion that a bidialectal background will facilitate the learning of additional languages (Han, 2015), this is unfortunate. Knowing how language learning and unlearning take shape in two language varieties that are as maximally close as possible on a typological level can greatly inform our understanding of the nature of language unlearning. Second, although literacy anchoring has been singled out as a safeguard against attrition effects because of the multimodal coding it entails (Schmitt, 2010), the loss of gestures or indeed attrition effects in bimodal language users is largely uncharted territory (but see a number of papers in Quinto-Pozos, 2007). Furthermore, gestures could play an important role in memory anchoring (Gullberg, de Bot, & Volterra, 2008). The study of gestures accompanying spoken speech is technically complicated since the coding of gestures in three dimensions (or four, if velocity of movement is taken into account) is particularly complex, as is the alignment with the spoken speech signal. In addition, it is hard to show the role of crosslinguistic influence in gestures and, as such, to assess the forgetting or unlearning of gestures. Constructing a workable paradigm here can be very important to understand the nature of language forgetting. More generally, there is much to be won if future investigations were to reflect more deeply on what it actually means to unlearn a language, while taking on board all that we have learned from language attrition studies in the meantime. An important question to pursue, also in view of what has been discussed in this chapter, is whether language unlearning can ever be conscious. We have argued that active unlearning only pertains to the undoing of earlier learned habits, not to knowledge itself. In fact, attrition is not conscious, although speakers may become consciously aware of their own language loss. Nevertheless, some studies suggest that even language forgetting can be conscious to some extent. Freud (1899) signaled that some memories are too painful to remember and are consequently suppressed to a level beyond conscious awareness. This behavior is mainly viewed as a self-defense mechanism. Freud’s ideas have been repeatedly contested and the brain mechanisms underlying motivated forgetting remain largely unknown, but there are examples where suppression does seem to be at work. For example, Schmid (2002) looked at the language attrition of L1 German in three groups of German Jews who emigrated to the United States from Germany before and around the time of World War II: one wave arrived in the United States before the Nazi regime

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

284

M E R E L K E I J Z E R A N D K E E S D E B OT

came into power, one during its reign, and the final group was unable to get away from Germany until after World War II. There was an incremental pattern of attrition, with those individuals who were most traumatized (i.e., those who had experienced most of the war) showing most attrition. Another example is a case study by Footnick (2007), who reports on an individual who had undergone an involuntary conflictual experience with his L1 (Mina, an indigenous language of Togo) during childhood. The use of that language was forbidden in school and at home. He apparently showed an almost complete loss of Mina: “The conflict may have resulted in the language becoming inaccessible [. . .] The loss of a hidden language is a rapid process in which the entire language, except a few words, disappears for a short time, due to psychological events involving conflict concerning the use of the language” (p. 171). Interestingly, through a regression-based hypnosis study, the suppressed language was brought back, and under hypnosis the person was able to produce it again, at the childhood proficiency level he had attained before exposure to the language stopped. This shows evidence of relearning. As a consequence of the hypnosis sessions, the language also started to come back outside of hypnosis. The example from Footnick (2007) directly speaks to the savings account. The relatively small investment and outcomes of relearning are a direct attestation of such residual knowledge (de Bot et al., 2004). This links back up to the relearning studies done in adoptees. It would also imply that language unlearning in essence is a fictional concept, a position underscored by the various studies reviewed in this chapter. Moreover, unlearning cannot be studied in isolation. Instead, language learning, unlearning, and relearning are part of the broader construct of bilingual language activation and suppression that organically follow from changes in a speaker’s input. This position fits in well with a usage-based perspective on language development that views language acquisition and language attrition or forgetting as two sides of the same coin that alternatively characterize a speaker’s language spurts and arrests throughout his or her lifetime.

8 Conclusion This chapter has explored contexts associated with the unlearning of language across the lifespan. Although there most certainly are contexts in which unlearning takes place, most notably in early childhood and older adulthood, this very fact does not answer the question of what it means to unlearn a language. This is underscored by the fact that most theories of forgetting that have been discussed in this chapter (i.e., retrieval-induced forgetting, regression, the activation threshold hypothesis) merely detail the observable details of language forgetting, but not their underlying bases. Based on the savings account, but also on what past

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Unlearning and Relearning of Languages from Childhood to Later Adulthood

285

studies have shown in terms of reactivation and relearning in even the most extreme cases of unlearning a language, such as by international adoptees or in pathological cases such as aphasia, it appears that language unlearning in and by itself does not occur. Speculating as to the reason for this, we can posit that language far exceeds the status of a skill. Contrary to the name that Lambert and Freed (1982) gave to the first conference on the topic of attrition, The Loss of Language Skills, language may well be markedly distinct from other skills in that it permeates all domains of life and it is always present in the environment, in whatever form. Its enormous complexity is likely to lead to a level of anchoring that subsequently becomes very difficult to lose. Echoing the claims made from a radical embodied cognition framework (Chemero, 2013), the complexity of the language system or systems for bilinguals cannot be captured in a simple shelf system where linguistic memories are stored and from which they can be retrieved. Instead, a complex network is formed, building on and relying on environmental input, in nonlinear and never fully predictable ways. The task of research into multilingual learning, unlearning, and relearning is to illuminate this complexity.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:29:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.015

15 The Measurement of Bilingual Abilities Jeanine Treffers-Daller

1 Introduction Scholars use the term “bilingual” to refer to individuals who learned their languages under very different circumstances and who may differ widely from each other in what they can do or usually do with these languages in everyday life. Scholars do not necessarily define bilingualism based on how well a person speaks a language. This can be rather confusing to outside observers who are faced with the complexity of the bilingual experience and may ask themselves: “Do I have to be able to fully understand two languages and speak them fluently to count as a bilingual?” or “Am I a bilingual if I can have an informal conversation in two languages?” Answering these questions is not easy, as researchers define bilingualism in a variety of ways. The contributions to this Handbook show that the concept and definition of bilingualism go well beyond the type of language abilities individuals have. Yet the issue of the measurement of bilingual abilities is of central importance to researchers in the field. The range of approaches to understanding and measuring bilingual abilities, however, makes it difficult to summarize the available literature and compare apparently contradictory results with each other. A key problem in tackling the issue of comparability among studies is that researchers use different terms for key concepts and that concepts used across the field only partly overlap (see also Section 2). Grosjean (1998, p. 48) points out that contradictory findings are often due to the lack of clarity of the concepts under investigation, including the concept of the bilingual individual itself, and to how concepts are operationalized in a particular study. Fortunately, the situation has somewhat improved over the years. This can be seen, for instance, in Hulstijn’s (2012) analysis of 140 articles published in the journal Bilingualism: Language and Cognition, one of the leading journals in the field, in the period from its first publication in 1998 up to 2011. Hulstijn found that just 19% of the papers published in the first half of this period

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

290

J EANI N E TR EF F ER S-DALLER

used an objective test to measure informants’ language abilities, while this percentage had increased to 54% of the papers in the second half of the period investigated. This means that during the period investigated in the study, a large number of papers still did not contain detailed information about participants’ language abilities. Many researchers now in principle subscribe to a holistic view of bilingualism (Grosjean, 1985), according to which all the languages spoken by bilinguals need to be included in analyses and not just bilinguals’ second (L2) or weaker language(s). This view is also shared by organizations of practitioners such as the Royal College of Speech and Language Therapists in the United Kingdom (2006). A more negative, fractional view of bilinguals is, however, still prevalent in much of the academic literature. In many studies comparisons are made between so-called native speakers (assumed to be monolingual) and bilinguals, which often result in a negative view of the abilities of the latter by comparison with the former. Differences between bilinguals and monolinguals are then labeled “the bilingual deficit,” and bilinguals who have a stronger and a weaker language are labeled “unbalanced bilinguals” or even “non-proficient bilinguals.” This deficit view is unfortunately still prevalent in both the field of second language acquisition (SLA) and that of bilingualism, in spite of Skutnabb-Kangas’s (1981, p. 194) warning many years ago that simple measurements on which monolinguals and bilinguals are compared do more harm than good. Considering the output of L2 learners as incomplete approximations of a purported complete monolingual native speaker target has aptly been labeled the target-deviation perspective by several scholars in SLA (see Ortega, Chapter 21, this volume). As it is high time to abandon this perspective, Ortega (2013) calls for a bi/multilingual turn in second language acquisition research in which the focus is not on what L2 learners cannot achieve but on what makes learning a language later in life unique. However, as Pearson (2010) notes, such a bilingual turn still needs to be accomplished even in the field of bilingualism itself. Comparing L2 learners with presumably monolingual native speakers is also problematic because such native speakers’ abilities differ widely from each other. Alderson (1980) not only found that what he termed native speakers did not always restore grammatical gaps in a cloze test but also that the differences between native and what he called nonnative (L2) speakers were very small, and that some nonnative speakers outperformed the native ones. These findings led Alderson to conclude that “native speaker proficiency, even on lower-order tasks, varies” (p. 74), and he considered the use of native speakers as criteria-setting for nonnative speakers on tests “misguided” (p. 75). In a similar vein, Hamilton, Lopes, McNamara, and Sheridan (1993) reported on the performance of presumably monolingual native speakers on a reading and writing test that is widely used with adults seeking to work or study in an English-speaking

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

291

country (the International English Language Testing System, IELTS; see www.ielts.org/). Hamilton et al. found this performance “far from homogeneous” (p. 348), and related to educational level and work experience: native speakers with higher levels of education or jobs that required them to engage in extensive high-level reading obtained significantly higher scores than respondents with lower educational levels (see, e.g., Street & Da˛browska, 2010, for similar findings). Such findings and his own study comparing large groups of native and nonnative speakers of Dutch led Andringa (2014) to posit that choosing a sample of native speakers that is representative of the full range of educational and socioeconomic and occupational backgrounds in any society is crucially important if one wants to compare native speakers against nonnative speakers. In many studies, native speaker samples have unfortunately been biased in favor of just highly educated speakers, often consisting of college students who are arguably from relatively affluent or middle-class families, thus adding an important variable other than just native speaker status. Comparisons between so-called native and nonnative speakers are much more valid if educational levels and socioeconomic and occupational backgrounds are controlled for. Hulstijn (2015) points out that very little research has been done into the abilities shared by (monolingual) native speakers. In his model of language proficiency in native and nonnative speakers, Hulstijn proposes that native speakers share the most frequent words and grammatical rules that are being used in oral/aural modes (basic language cognition [BLC]). However, they do not necessarily share any of the infrequent lexis and uncommon grammatical rules that are used in written (and occasionally in some oral/aural) modes (higher language cognition [HLC]). Therefore there is great variability in native speakers’ knowledge, which makes clear that using monolingual “norms” as a yardstick against which the performance of second language learners is measured is highly problematic. Finally, as Rothman and Treffers-Daller (2014) have argued, it is in fact not appropriate to reserve the label “native speaker” for monolinguals only. Given that a native language is “one that is acquired from naturalistic exposure, in early childhood and in an authentic social context/speech community” (Rothman & Treffers-Daller, 2014, p. 95), bilinguals who have acquired two or more languages in early childhood have more than one native language. Avoiding the deficit view inherent in much of the work in the field is extremely important. Therefore the key aim of this chapter is to identify what makes individuals with bilingual abilities unique speaker-hearers in their own right and how these abilities can be measured. In Section 2, I first explore the construct of bilingual abilities, how it relates to some other associated constructs, and how it can be modeled in different theoretical frameworks. Identifying the uniqueness of the bilingual experience will necessarily entail referring to code-switching and translanguaging and to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

292

J EANI N E TR EF F ER S-DALLER

the mutual influences between the languages spoken by bilinguals, which have been described with a range of different terms, including crosslinguistic influence and transfer (the older negative term interference has fallen into disuse). In Section 3, I move on to the operationalization and measurement of bilingual abilities in relation to how appropriate they are for different types of bilinguals. Attention is paid to the measurement of bilingualism across the lifespan, and wherever appropriate assessments for children are discussed as well as assessments for adults. The final section focuses on directions for future research in the field.

2 Bilingual Abilities 2.1 Defining the Construct Before we can discuss the different ways in which bilingual abilities can be operationalized or measured, it is important to clarify what the construct means and how it relates to other constructs. The construct of bilingual (language) abilities is an expansion of the notion of language ability, which was developed in the field of second language testing by Bachman and Palmer (2010). Bachman and Palmer (2010) see the construct as covering a wide range of abilities, including linguistic and sociolinguistic abilities and strategic knowledge. The term bilingual abilities comprises all these different components, but the choice of a plural form is more appropriate for bilinguals, as it reflects the fact that bilinguals possess these abilities in more than one language. The construct of bilingual abilities is not limited to the domain of syntax. This is signaled by the terminological choice of “abilities” over “competence,” as the latter term is strongly associated with the generative framework and is in that context often interpreted to be the ideal speaker-hearer’s underlying rule-based knowledge of her/his language (Chomsky, 1965). In the current chapter, I conceive of bilingual abilities as encompassing a wide range of uses of language and many different dimensions, including both traditional notions of proficiency that are narrow as well as sociolinguistic, pragmatic, or strategic competencies. This understanding contrasts with other existing delimitations of L2 ability in the field. On the one hand, it is broader than the construct of language proficiency proposed by Hulstijn (2015), who mainly refers to lexical, grammatical, phonotactic, and prosodic elements, but does not include sociolinguistic, pragmatic, or strategic competencies. On the other hand, it is not as comprehensive as the widely used construct of multi-competence, which originally referred to the knowledge of more than one language in the same mind (Cook, 1991), but was later expanded to cover not just language knowledge but also wider cognitive processes and has recently been redefined as “the overall system of a mind or a community that uses more than one language” (Cook, 2016, p. 3). In discussing bilingual abilities in what follows, I identify different

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

293

components, dimensions, domains, or skills within this construct from a range of different perspectives. It is not possible to review all of these here, and the points made are necessarily selective. Importantly, the construct of bilingual abilities also covers the ability to process two languages. However, modeling bilingual processing abilities is a complex task because, as Larsen-Freeman, Schmid, and Lowie (2011) point out, models of non-static scenarios, which include bilingualism, need to account for systems that are under development as is the case, for example, for children who are learning two languages from birth or in adult L2 learners. Another key issue for models of lexical processing is how to account for the ways in which bilinguals map meaning onto form (see Paap, Chapter 22, this volume, for a review of bilingual processing models). Particularly challenging for models of bilingual language processing is the case of bimodal bilingualism as found among bilinguals who use a spoken and a signed language. Contrary to unimodal bilinguals, bimodal bilinguals can produce simultaneous output in two languages through oral and visual channels (see Tang & Sze, Chapter 24, this volume, for extensive discussion). This simultaneous production through oral and visual channels presents challenges for models of language production designed to explain the production of single language elements at the time. In addition, while both bimodal and unimodal bilinguals can switch between languages, bimodal bilinguals’ production of code-blends (simultaneously produced words and signs) reveals that coactivation and inhibition processes in bimodal bilinguals are different from those found in unimodal bilinguals (Emmorey, Giezen, & Gollan, 2016). A few words also need to be said about comprehension versus production abilities, and the traditional division into four skills: speaking, listening, reading, and writing. For a given language, there can be vast differences between a bilingual’s level of comprehension and production. As discussed in De Houwer (2007), bilingual children can develop oral production skills in the languages they are exposed to. However, whether or not they actually do so depends to a large extent on the input patterns in the families in which they grow up. In her analysis of a dataset of 1,356 bilingual families in Flanders in which parents spoke both a minority language and the majority language, Dutch, at home, De Houwer (2007) showed that success rates in the intergenerational transmission of the ability to speak the minority language differed according to the distribution of the languages in the parent pair, and ranged from 34% to 93%. The most successful transmission occurred in families where one parent spoke the minority language and the other parent spoke the minority as well as the majority language. While bilingual families are understandably keen to have their children develop production skills in all languages spoken at home, receptive multilingualism (Bahtina & ten Thije, 2013) may be an asset both for individuals and society at large, because

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

294

J EANI N E TR EF F ER S-DALLER

bilinguals who have comprehension skills across different languages can communicate in settings in which they would otherwise face insurmountable barriers. Although many researchers have focused on the development of oral/ aural skills, the number of studies focusing on bilingual children’s reading skills is steadily increasing. These studies show that some of the skills that underpin the ability to read, such as phonological awareness (e.g., being aware of the sounds the words hit and bit have in common) and morphological awareness (e.g., being aware of the meaning of individual morphemes such as –er in worker) are transferable between languages. This is particularly the case if the language in which children have learned to read first has a transparent orthography, that is, is characterized by a regular or close to one-to-one correspondence between graphemes and phonemes (Kirby, Deacon, Bowers et al., 2012). The fact that certain skills transfer between languages can also be explained in the framework of Cummins’ (1979b, p. 222) interdependence hypothesis, which states that “the development of competence in a second language (L2) is partially a function of the type of competence already developed in L1 at the time when intensive exposure to L2 begins.” In their study involving 1,062 Spanish L1 children, Winsler, Kim, and Richard (2014) have offered very strong evidence for the validity of the interdependence hypothesis: children who had stronger skills in Spanish at age four made faster progress in learning English a year later. Cummins (1979b) furthermore proposed a fundamental distinction between basic interpersonal communicative skills (BICS) and cognitive academic language proficiency (CALP), where the former refers to everyday conversational language, and the latter to language used in educational contexts. This dichotomy bears some similarity to Hulstijn’s (2015) distinction between BLC and HLC (see the Introduction), although Cummins’s distinction was not proposed as a theory of language but mainly intended to offer guidance for policy and practice. Cummins’s distinction between BICS and CALP has been very influential in educational circles, in particular in those with a particular interest in the academic development of minority children, although it has also met with criticisms and modifications along the way (see Cummins, 2016). Going well beyond Cummins’s notion of interdependence, the relation between a bilingual’s languages has been studied in great detail.

2.2 Dominance and Balance Problematized Many researchers assume that bilinguals often have a stronger and a weaker language, and that it is rare for bilinguals to be balanced (SilvaCorvala´n & Treffers-Daller, 2016), that is, to have equal abilities in their two languages. In fact, Romaine (1989) is particularly strongly opposed to the idea of balance, which to her is largely an artefact of a deficit view of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

295

bilinguals, and does not do justice to the specific abilities of bilinguals. Scholars’ general preference for “balance” in bilinguals and treating it as a kind of ideal possibly also results from a general bias in favor of symmetry across a wide range of fields, including architecture, biology, and music (Treffers-Daller, 2016). However, as both Birdsong (2016) and TreffersDaller (2016) have pointed out, perfect balance in bilinguals is probably as rare as ambidexterity, a term used to describe the skills of a person who can carry out a variety of tasks with equal skill with either hand. Most individuals are predominantly right-handed or predominantly left-handed, and no one would qualify right-handed individuals as “unbalanced” just because they can carry out some tasks better with their right hand. For speakers of three or more languages, it is even less likely that the distribution of labor between the languages or the levels of ability achieved in each are completely balanced (Curdt-Christiansen, 2016). De Houwer and Bornstein (2016) have furthermore pointed out that bilinguals can indeed perhaps have similar levels of, for instance, word knowledge in each language, but that these levels are not necessarily high. In other words, “balance” is not necessarily a positive thing. Bilinguals who have a stronger and a weaker language are often labeled “unbalanced,” with all the negative connotations a term with un- implies. However, as many scholars have noted, different abilities in each language are the normal result of the fact that bilinguals learn and use their languages for different purposes in everyday life, for example, one language at home and one at work or school. Different abilities in each of their languages are simply linked to the specific purposes for which bilinguals need each of them. Different learning and usage profiles for each language make it extremely unlikely that bilinguals will develop exactly the same skills in each language. In fact, Grosjean (2016) estimates that bilinguals often have translation equivalents for only 30–37% of their words in the other language, which clearly shows that the communicative needs and abilities in one language are not necessarily mirrored by those in another language. Grosjean (2016) notes there is a need to dig much deeper into the different activities and functions for which a bilingual’s languages are being used as the skills and knowledge that a bilingual develops in each language are strongly domain specific.

2.3 Uniquely Bilingual Abilities The task of fully defining bilingual abilities calls for a consideration of abilities that are uniquely available to bilinguals. The uniqueness of the bilingual experience is perhaps most visible in the ability of bilinguals to seamlessly switch from one language to another, which is generally called code-switching, code-mixing, or, within some perspectives, translanguaging (for further discussion, see De Houwer, Chapter 17, this volume, and Garcı´a & Tupas, Chapter 20, this volume). Poplack (1980) was the first to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

296

J EANI N E TR EF F ER S-DALLER

show that code-switching is not a sign of lack of mastery of languages but is, in fact, highly skilled behavior, and that the most intimate forms of code-switching are found among bilinguals with high levels of ability in both languages. A detailed model of code-switching is proposed in Muysken (2013), who distinguishes four main types of code-switching and shows how the occurrence of these different types depends on (a) typological factors, (b) processing constraints linked to the bilingual abilities of the speakers, and (c) societal factors, such as the duration of language contact and power relations within society (see further Aalberse & Muysken, Chapter 26, this volume). Psycholinguistic approaches to code-switching that focus on switching from one language to another that is externally induced in a laboratory setting (Gullberg, Indefrey, & Muysken, 2009) are becoming more prominent. Such studies can provide more insights into the simultaneous processing of two languages in bilinguals. Work in this field shows in particular that bilinguals can never completely switch off any of their languages, and that words from different languages are probably not stored separately. When speakers need to retrieve a particular word from one of their languages, they need to search both lexicons, a phenomenon often referred to as nonselective access (see further Paap, Chapter 22, this volume). Spontaneous code-switching is different from crosslinguistic influence (CLI), which does not generally involve the importation of lexical material (words) from one language into another, but only a sound, a meaning or a grammatical pattern (Treffers-Daller, 2009). Examples of CLI can be found in the devoicing of final voiced obstruants such as [g] in English words such as bag, which are often pronounced [bæk] instead of [bæg] by L1 German users of English as an L2, or in the choice of the preposition on in on school by L1 Dutch users of English as an L2, where at school would be the conventional choice. The forms used by L2 users of English here are likely the result of CLI from German and Dutch, respectively. The existence of CLI across all levels of linguistic analysis (phonetics, phonology, morphology, lexicon, syntax, pragmatics, etc.) also means that bilinguals can never be two monolinguals in one person, and monolingual-likeness in either of the two languages is generally an impossibility (see Biedron´ and Birdsong, Chapter 16, this volume). CLI may be reinforced if speakers with the same L1 hear such L2influenced pronunciations frequently as members of the same bilingual community of, for example, Dutch–English or German–English bilinguals in the United Kingdom. The longer the duration of language contact between two communities, the more such language contact phenomena can become integrated into the language varieties spoken by individuals growing up in bilingual communities (in both directions), probably as a result of the constant coactivation of both languages in the minds of the speakers (these are studied in the field of contact-induced language change; see Aalberse & Muysken, Chapter 26, this volume). For researchers

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

297

in the field who take the notion of multi-competence (Cook, 2016) seriously, it is important to be aware of the long-term effects of language contact on the varieties spoken by bilinguals. This is crucial because the language varieties spoken and heard by bilinguals are often rather different from the varieties that are spoken outside the bilingual communities under study. This has important implications for qualifying the input to, for example, heritage speakers who learned their L1 (the heritage language) in childhood, but became more proficient or dominant in another language later in childhood (Benmamoun, Montrul, & Polinsky, 2013; see also Armon-Lotem & Meir, Chapter 10, this volume). Speakers of Turkish in Germany, for example, grow up with a variety of Turkish that is considerably different from that spoken in Turkey, and thus develop abilities in Turkish that are different too. Evaluations of speakers’ bilingual abilities must take into account the fact that bilinguals may not have had the chance to learn monolingual varieties of a particular language.

3 Measuring Language Abilities in Two or More Languages 3.1

Measuring a Multifaceted Phenomenon and Dealing with Variability As the construct of bilingual abilities is a very wide-ranging one, no single measure can assess a bilingual’s abilities in any comprehensive manner. Most researchers nowadays prefer to choose a particular domain, skill, or component that they consider important, and measure informants’ abilities in that area, in the knowledge that the measure represents only a fraction of the multifaceted nature of any bilingual individual’s language abilities. Measuring a wide range of abilities is generally not an option. This is why in some cases, the measurements used to assess specific phenomena (e.g., scores on vocabulary tests) are interpreted as a proxy for the construct of bilingual abilities itself. Especially vocabulary tests, and in particular tests that measure bilinguals’ vocabulary size in their L2, are frequently used as a proxy for ability in the L2. However, Kremmel and Schmitt (2016) point out that researchers who use such tests usually do not offer reflections on what exactly the tests measure. This means that the rationale for the choice of this particular component as a proxy for general language abilities is rather weak. Researchers agree that bilinguals’ abilities differ widely from each other. Those who learn two languages from birth are different from those who learn an L2 in childhood (De Houwer, 2013), and these are again different from those who learn an L2 in adulthood or those who are in the process of losing one of their languages (compare, for instance, Singleton & Pfenninger, Chapter 4, this volume, and Keijzer & de Bot, Chapter 14, this volume). As is apparent from many of this Handbook’s contributions,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

298

J EANI N E TR EF F ER S-DALLER

the variability in language abilities even within a single type of bilingual can be very large. The variability also increases with the number of languages that people know: each of their languages can start at different points in time and the scenarios in which individuals become multilingual may be very different from each other (see Quay & Montanari, Chapter 27, this volume). In order to gain an understanding of the variability in bilingual abilities, researchers often use other- and self-reports. It is common for researchers measuring children’s bilingual abilities during the first years of life to use parent report forms, where parents are asked to check lists that provide estimates of the words their children understand or produce (to be discussed later in this chapter). Researchers working with adults too have used reports, but in this case in the form of self-assessments, that is, they ask individuals to indicate their own abilities in some or all of the four skills, that is, gauging their own understanding, speaking, reading, and writing, often using a simple rating scale. Interestingly, however, researchers hardly ever ask bilinguals to assess their own proficiency in both languages in terms of more detailed components of language use (vocabulary, grammar, phonology, pragmatics, etc.). As Hulstijn (2012) points out, assessing one’s own vocabulary, phonological, or grammatical knowledge is quite difficult for non-experts, and therefore not really an option for lay self-assessments. Scholars who use self-assessments often correlate these with actual measures of performance such as a test of speaking, reading, or writing. In their meta-analysis of correlations between self-assessments and performance outcomes across a range of domains (e.g., “language competence,” sports, or academic ability), Zell and Krizan (2014) show that such correlations tend to be of moderate strength only. They also claim, however, that for language competence correlations between subjective perceptions and actual performance can be higher than in most other domains they investigated. Their claim is based on a high correlation coefficient of r = 0.63 found in the only study of language self-assessment they were able to include, namely an earlier meta-analysis of 60 studies carried out by Ross (1998), who looked at self-assessments and objective measurements of L2 abilities across the four traditional language skills. Zell and Krizan (2014) attribute the relatively high correlation for language competence to the existence of “ubiquitous” feedback and to language skills being “objectively defined” (p. 117). While it may be true that L2 learners get regular feedback on their performance if they are in a classroom setting, this is not normally the case for bilinguals who have learned their languages in uninstructed settings. Also, the separability of the four skills has been heavily criticized (Bachman & Palmer, 2010): writers need to read over what they have written, and speakers need to listen to their interlocutors. Bachman and Palmer (2010) instead suggest that more fine-grained approaches that specify what they call Target

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

299

Language Use Domains (such as making a compliment or a complaint) offer potentially more fruitful avenues for the (self-)assessment of learners’ skills. As bilingualism is such a multifaceted phenomenon, the key question is how we can measure bilingual abilities in such a way that the measurements are appropriate for the target groups and do not reinforce the deficit view. A large variety of language instruments and tests have been developed for monolingual children that are increasingly being used with bilingual children. The norms associated with these tests are generally appropriate for only monolinguals and not for bilinguals (Gathercole, Thomas, Roberts et al., 2013). Defining norms for bilingual children is particularly complex because of the great variability among them (Cruz-Ferreira, 2010). As bilingual children can be exposed to their two languages in very different proportions, and may have started from birth or later in childhood, Gathercole et al. (2013) propose that levels of exposure need to be taken into account in the measurement of bilingual abilities (see the contributions by Armon-Lotem & Meir, Chapter 10, this volume, and Hammer & Edmonds, Chapter 19, this volume, for further discussion). Considerable progress has been made in the development of new measurements focused on young bilinguals’ abilities. Some of these are summarized in what follows. Vocabulary has been a central area in the measurement of bilingual abilities, particularly in children (see also Serratrice, Chapter 1, this volume). This may relate to the fact that in the first years of life the comprehension and production of words vocabulary are the first clearly language-related skills to emerge, and that only later grammatical development comes in. In the first school years vocabulary becomes central to the acquisition of literacy and it is central to overall academic development. As discussed in the next section, measuring vocabulary well (across the lifespan) has its own methodological challenges.

3.2 Measuring Vocabulary An important new way to measure bilingual children’s vocabulary size across two languages was developed by Pearson, Ferna´ndez, and Oller (1993). In order to measure vocabulary knowledge in Spanish and English in children who were growing up with two languages from the beginning, Pearson et al. used the American English and the Spanish versions of the MacArthur-Bates Communicative Development Inventory (CDI), a standardized parent report form through which (monolingual) children’s comprehension and production vocabularies are noted (Fenson, Dale, Reznick et al., 1993). Pearson et al. (1993) introduced the notion of Total Conceptual Vocabulary (TCV), which refers to the counting of all concepts a child understands, regardless of whether s/he knows the concept in either one or both of her/his languages. More precisely, TCV

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

300

J EANI N E TR EF F ER S-DALLER

refers to the total number of lexicalized concepts (in fact, form–meaning pairings) known by bilingual children. In monolinguals, the TCV score is assumed to equal total word knowledge, called the Total Vocabulary Score. In bilinguals, TCV is calculated by counting cross-language synonyms that are both known by a child only once, and by accepting the child’s knowledge of the word in just a single language as knowledge of the relevant concept. Thus, supposing a child understood the two words Spanish perro as well as its English translation dog, TCV would be 1; it would remain 1 if the child only knew Spanish perro or only English dog. Thus, the TCV differs from a simple total score of all words known summed across both languages (the Total Vocabulary Score) as in that calculation perro and dog would each be counted. Pearson et al. (1993) compared the bilingual children’s Total Vocabulary and TCV scores to those of a demographically matched group of mostly English-speaking monolingual children with reference to (English) monolingual-based percentile norms, regardless of whether individual bilingual children knew more words in one language than the other. No bilingual–monolingual differences were found. Note, however, that at the time of the Pearson et al. (1993) study, monolingual percentile norms existed only for the American English CDI, but Pearson et al. took these percentile norms as a proxy for Spanish as well. As more recent work has shown, CDI-based monolingual percentile norms may in fact differ across languages. In addition, in order to compare children’s abilities in a single language only (for which Total Vocabulary and TCV scores are irrelevant), Pearson et al. determined which language bilingual children knew more words in (as they say, were more dominant in). They then compared only the bilingual children’s dominant language to the monolinguals’ only language, again with reference to monolingual-based percentile norms. The bilingual children’s production lexicon sizes in their dominant language were not significantly different from those of the mostly English-speaking monolinguals. Only for the Total Vocabulary Scores and the TCV score comparisons, then, were Spanish- and Englishdominant children combined and compared as a group to monolinguals. The methodology followed in this study reveals that a deficit approach to the measurement of bilingual abilities can be avoided by taking into account within-group differences among bilingual children. Different results emerge from a meta-analysis of studies involving 1,738 children (996 bilinguals and 772 monolinguals) between the ages of 3 and 10 by Bialystok, Luk, Peets, and Yang (2010). These researchers showed that bilingual children have smaller comprehension vocabularies in English as measured with the Peabody Picture Vocabulary Test (PPVT) (Dunn & Dunn, 2007) than their monolingual counterparts across the age range investigated. The authors found a nine-point difference between the results of the monolinguals and the bilinguals in their sample. Unfortunately, however, the children were not divided into dominance groups as in Pearson et al.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

301

(1993). The large discrepancies between the two groups seem to indicate that English was actually the weaker language of the two for many children, and probably a language they had started learning later in life (De Houwer, Bornstein, & Putnick, 2014). It is understandable that precise measurements of language dominance are difficult to make for a sample that includes such a wide range of languages other than English. However, the children or the parents could have been asked what the child’s stronger or weaker language(s) were. Separate analyses could then have been made of the English vocabulary of English-dominant children and children for whom the other language(s) was/were dominant. It is likely that at least for a subgroup of children in the sample, the English-dominant ones, knowledge of English vocabulary was similar to that of monolingual children. Interestingly, for words related to the school domain, no significant differences were found between bilinguals and monolinguals, which shows that looking at within-group differences is likely to provide a more positive, differentiated picture of the abilities of bilinguals. De Houwer et al. (2014) took up the challenge to control for the many sociolinguistic variables that affect vocabulary development in their study of the comprehension and production vocabularies of 31 Dutch–French bilingual children and 30 monolingual Dutch-speaking children, aged 13 and 20 months, using Dutch and French versions of the CDI. The authors ensured the two groups were matched in age, birth status, birth order, family socioeconomic status (SES), and gender. In addition, the bilingual children were exposed to each language for the same length of time as the monolingual children in the sample because they had grown up with both languages from birth. Thus, the authors avoided potential confounds in the study that could be the result of variables known to affect vocabulary development. After completing detailed analyses of monolinguals and bilinguals at the two different ages, De Houwer et al. (2014, p. 1209) conclude that there is “no evidence of consistent differences between young bilinguals’ and monolinguals’ vocabulary sizes” and recommend that researchers look at causes other than bilingualism to explain variability in lexical knowledge within and between groups. A key issue that makes it difficult to assess both languages of a bilingual is the fact that there is a lack of standardized assessments for a wide range of languages spoken by bilinguals. It is not possible to simply translate an English test or instrument into another language as the vocabulary items and the grammatical structures that need to be tested do not have a one-toone correspondence to those in another language. Also, intercultural differences make it likely that children or adults use very different vocabulary items in different languages. In fact, Gathercole (2013) suggests that translating a test from one language into another would be similar to giving a piano player a musical test created for the drums. While creating tests that are completely equivalent in two languages will remain an elusive goal, comparable assessment instruments are

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

302

J EANI N E TR EF F ER S-DALLER

urgently needed for researchers and practitioners alike. This is why researchers have adapted existing measures to other languages. The CDI, for example, now exists for close to 100 language varieties, including several sign languages, and new adaptations are being constantly developed (see https://mb-cdi.stanford.edu/adaptations.html). The PPVT has been adapted for a number of other languages too, including British English, French, German, Spanish, and Welsh, although concerns about possible cultural biases in its use are reported for some populations (Haitana, Pitama, & Rucklidge, 2010). An innovative way of measuring vocabulary knowledge among children across different languages can be found in Haman, Łuniewska, and Pomiechowska (2015), who started the development of Crosslinguistic Lexical Tasks (CLTs) for five-year-old children as part of the BilingualismSpecific Language Impairment (BiSLI) Cost Action IS0804 (www.cost.eu /COST_Actions/isch/IS0804). Because bilingual children as well as children with specific language impairment (SLI) sometimes present with smaller vocabularies in a particular language compared to monolingual typically developing children, it is important for researchers and practitioners to have tools to be able to measure vocabulary in preschool children, so that those with delays in vocabulary development and those with SLI can be distinguished from each other (see also Hammer & Edmonds, Chapter 19, this volume). The novelty of the Haman et al. (2015) approach is that they did not develop CLTs by adapting an existing English test, as is being done for the CDI and the PPVT, but that they initiated a new crosslinguistic assessment tool. They created a database of more than 1,000 pictures of actions and objects that groups of adult speakers of 34 different languages were asked to name. Participants were also asked to evaluate whether these could be named easily and unequivocally. Subsequently, stringent criteria were developed for the selection of 30 nouns and 30 verbs associated with the pictures to ensure that the words that were selected for the CLTs in different languages were comparable with respect to, for example, the complexity of the candidate words. A cultural bias was avoided because the tests were developed in parallel for different languages, and no single language was taken as the model for the CLTs. Variants of pictures (e.g., women with or without a head scarf) were created for different languages to ensure that picture styles were appropriate for the culture of the language under investigation. While the aim of the CLTs was to differentiate between typically developing and language-impaired monolingual and bilingual children, in future they can probably also be used as a baseline measure of bilingual abilities and/or language dominance in bilingual children, precisely because the measurements will be comparable across languages. The availability of reaction times for this task makes it an attractive option for researchers interested in lexical processing too. The vocabulary measurements considered here so far mainly tap into bilinguals’ knowledge of single words, which are often tested out of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

303

context. Such a focus on single words is typical of the research carried out with young bilingual children. Of course, vocabulary knowledge encompasses much more than knowledge about the link between a form and a meaning of an isolated word, as it primarily involves knowledge about the use of words in context (Nation, 2001). The roles of context, polysemy, and phraseology in shaping vocabulary knowledge increase and become particularly central to the measurement of the adult lexicon. Valid assessments of adults’ bilingual lexical abilities should therefore include measurements of bilinguals’ knowledge of the ways in which words are used. This also involves knowledge of different linguistic registers, as in, for example, knowing what the difference is between the use of help and aid. Such nuances are more difficult to measure than comprehension or production knowledge about the link between the form and meaning of isolated words. Much of the research conducted in corpus linguistics and SLA in recent years has focused on the patterns in which words are used in first language corpora (e.g., Hoey, 2005) and on the acquisition of such patterns and fixed expressions, often called formulaic sequences, by second language learners (e.g., Martinez & Schmitt, 2012). While tests of such fixed expressions have been developed for English and for some other languages, we are not yet in a position to assess bilingual abilities in the use of formulaic sequences. As is well known, there is no one-to-one translation of formulaic sequences in two languages, so that an expression such as pay attention translates as faire attention “make attention” in French, and comparing apples and oranges becomes A¨pfel und Birnen vergleichen “comparing apples and pears” in German. These examples serve to underscore the fact that measurements of bilingual abilities that are just based on comparisons of the knowledge of single words do not do justice to the highly complex abilities that are involved in using words in context in general and in fixed expressions in particular. Analyses of formulaic language could be made on the basis of transcripts of oral language, particularly if speech samples are elicited with a standardized tool, as free speech would be too variable to make analyses of formulaic sequences possible.

3.3 Select Approaches to Measuring Syntactic Abilities This brings us to the measurement of syntactic abilities in two languages. Researchers in bilingualism and second language acquisition use many grammar tests, some standardized and some homemade. Space allows only a brief discussion of some selected approaches here. Measuring grammar across a bilingual’s languages is a complicated undertaking because of the large typological differences between languages. Mean Length of Utterance (MLU) is a widely used measure of language development in young children. Some researchers (e.g., Schmeißer, Hager, Arnaus Gil et al., 2016) opt for comparisons of MLU in two languages, and compute language dominance indices on the basis of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

304

J EANI N E TR EF F ER S-DALLER

MLU differentials. However, many authors have pointed out that a key issue here is that for some languages such as Chinese, MLU is best computed in words, but for others, such as English, it is best computed in morphemes. Typological differences between languages make it therefore very difficult to compare MLUs across languages. Other measures of complexity are the percentage of multi-morphemic utterances or the upper bound (that is, the number of morphemes or words in a child’s longest utterance), but, again, these are difficult to compare across languages. Moreover, these measures do not really tap into the complexities of grammar as such and MLUs are not a valid indicator of syntactic ability beyond a tally of about four. Interesting approaches to the measurement of syntactic abilities in bilingual children have been developed by Chiat, Armon-Lotem, Marinis et al. (2013), who use sentence repetition tasks to measure sequential bilingual children’s ability to produce specific sentence structures such as passive constructions or relative clauses in their L2. This initiative was further developed by Marinis and Armon-Lotem (2015), who show that sentence repetition tasks can be used to diagnose language delay or impairment in bilingual children. Thus, if children have difficulties in both languages, this would be an indication of language delay or impairment, while for typically developing children, uneven development across languages is entirely normal and to be expected, that is, children may be more advanced in one language than the other, but with one language generally showing no signs of any kind of delayed development (De Houwer, 2009; see also Hammer & Edmonds, Chapter 19, this volume). Again as part of the Bilingualism-SLI COST Action mentioned in Section 3.2, Chiat et al. (2013) developed sentence repetition tasks for a wide range of languages, which makes these potentially interesting for future measurements of bilingual abilities and language dominance. For adults too, elicited imitation tasks have been shown to be useful as a shortcut to measure L2 proficiency for research purposes, for example, in order to distinguish foreign and heritage language speakers at two proficiency levels of Mandarin (Wu & Ortega, 2013) or foreign language speakers of L2 Spanish at three different levels of proficiency and experience with the language (Bowden, 2016). Currently, grammatical tests or instruments focusing on all the languages a bilingual knows are still lacking. The focus is usually just on a single language. This necessarily underestimates bilinguals’ grammatical abilities.

4 Conclusion In this chapter we have seen, first of all, that the construct of bilingual abilities is a broad, multidimensional one that cannot be adequately

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

The Measurement of Bilingual Abilities

305

captured by assessing one skill or knowledge domain only. While many researchers choose to measure bilinguals’ comprehension or production vocabulary knowledge in both languages as a proxy for the assessment of overall bilingual abilities during the early years of bilingual development, the choice of such a measure is not always well motivated. If we take the multidimensionality of bilingualism seriously, and we acknowledge the importance of context, polysemy, and registers most particularly present in adult language usage, a far wider range of skills and knowledge that bilinguals possess should be taken into account in the assessment of individuals’ bilingual abilities, such as their sociolinguistic and pragmatic knowledge. It has hardly been investigated, for example, to what extent child or adult bilinguals develop distinct politeness routines in two languages, or are able to use or understand standard varieties as well as regional varieties that are found in different parts of the countries (but see Chevrot & Ghimenton, Chapter 25, this volume). The difficulty is, of course, that measuring such knowledge is much more complicated than measuring knowledge of the form and meaning of isolated forms. It is difficult to specify which components of the construct are most relevant for the measurement of bilingual abilities as all depends on the research questions being asked. For some projects, lexical knowledge is most important, but for others, grammatical, phonological, or sociolinguistic knowledge might be more relevant. Few researchers offer principled reasons for their choice of a particular measurement. The emphasis on vocabulary in much child bilingualism research may be understandable, as measures of vocabulary have often been shown to be successful at discriminating between groups. In addition, by controlling for vocabulary, disadvantages that bilinguals may experience in lexical retrieval by comparison with monolinguals can disappear (Bialystok, Craik, & Luk, 2008). Nevertheless, in most cases, the choice of the PPVT as a proxy measure of language ability seems to have been made mainly for reasons of convenience (Section 3.2). Some researchers include a range of measures of bilingual abilities in the same study. Those who opt for this solution often find that bilingual abilities are task-dependent and that performance differs on each measure (Bahrick, Hall, Goggin et al., 1994; Kupisch & van de Weijer, 2016). This complicates the task of choosing which measure is to be used as a proxy for overall bilingual abilities and highlights once more the arbitrariness of decisions taken in some studies where the PPVT or MLU are adopted as measures of overall bilingual abilities. A more principled view of the aspects of the construct of bilingual abilities that should take priority particularly when investigating the abilities of late-timed sequential adults can be found in Hulstijn (2015). He suggests that the measurement of bilingual abilities in such individuals should cover BLC (see Section 1) in two languages only, as this is the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

306

J EANI N E TR EF F ER S-DALLER

dimension that is shared by all, irrespective of bilinguals’ levels of education, while language users differ in their mastery of HLC. As BLC is exclusively oral, it means that measures of bilingual abilities should also necessarily be oral rather than written. While detailed operationalizations of BLC and HLC are still under development, it is intuitively appealing that priority is given to oral/aural skills over written ones, as many bilinguals are literate in only one of their languages, and some do not possess any literacy skills (see Simpson, Chapter 13, this volume). An interesting question is also whether the distinction between BLC and HLC can be mapped onto Cummins’s (1979b) lower and higher thresholds of language competence. Knowledge of BLC in two languages could, for example, correspond to Cummins’s lower threshold and knowledge of (parts of) HLC to the higher threshold. Future research into these issues will need to show if operationalizing bilingual abilities in this way can shed new light on the impact of bilingualism on nonlinguistic cognition. An issue that will be important for the future is the measurement of domain-specific language use. While recently new questionnaires have been developed that aim to help researchers obtain further detail about bilinguals’ use of languages (Gertken, Amengual, & Birdsong, 2014; see also Armon-Lotem & Meir, Chapter 10, this volume), Grosjean (2016) argues that much more detailed information is needed about the different domains in which bilinguals use their languages. In order to make this possible, Grosjean (2016) proposes the Complementarity Index, which is an important new tool that can help researchers quantify the distribution of labor between languages across different domains of use. Although researchers interested in the concept of bilingual abilities or in the measurement of this construct have not often looked at perceptions of bilingual abilities by listeners, a focus on such perceptions could be highly interesting as it could provide further information about the dimensions of bilingual abilities that matter to listeners (whether expert or nonexpert) and on the relative importance of different features of a bilinguals’ speech that play a role in how that person is perceived. There are many avenues to be explored, and it is hoped that some of the challenges mentioned in this chapter will be taken up by researchers interested in further improving the measurement of bilingual abilities.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:32:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.016

16 Highly Proficient and Gifted Bilinguals Adriana Biedron´ńand David Birdsong

1 Introduction In every domain of human performance are exceptional individuals with remarkable talents. In the domain of language, we recognize Emily Dickinson, Gustave Flaubert, and Jorge Luis Borges as extraordinary writers, and find James Woods and Lisa Kudrow to be highly articulate celebrities. We tend to think of such individuals as having a masterful command of their native tongue. However, as it turns out, some of the most notable speakers and writers are not native users of the language of their fame. For example, actress Charlize Theron, singer-songwriter John Kay, and novelist Vladimir Nabokov gained their renown through the English language but were not brought up speaking English. Still other public figures command several languages to high levels of proficiency, among them actors Dolph Lundgren (Swedish, English, German) and Sandra Oh (English, Korean, French), and novelist and essayist Elif Safak (Turkish, English, French, Dutch). For most people, however, attaining fluent command of a foreign language is a long and effortful process. That is why even a cursory look at studies of individuals attaining very high levels of proficiency in one or more foreign languages inevitably leads to a question: are they all exceptionally gifted? That is, do they have a higher than normal foreign language (FL) aptitude that enables them to acquire languages faster and to higher levels than most of us? This chapter examines the relationship between linguistic giftedness and bilingualism. In particular, we focus on studies of gifted FL learners who achieved very high levels of proficiency. To start with, we define basic terms such as FL aptitude, ability, giftedness, and talent. Then we discuss the critical period hypothesis (CPH) as it applies to high achievement in second language (L2) acquisition. Next, we analyze the dynamic relationship between bilingualism and FL aptitude. This is followed by an examination of different populations of gifted foreign language learners who have learned several languages, including polyglots and savants,

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

308

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

along with their cognitive, affective, and neurological characteristics. Finally, our focus shifts to linguistically gifted children. Reflections on future research and concluding remarks close the chapter.

2 Terminology Ability and aptitude are terms close in meaning. Ability is defined as actual potential, that is, what a person is able to do, whereas aptitude refers to a cognitive ability that is predictive of future learning success. Both can be estimated on the basis of standardized test outcomes, such as the Modern Language Aptitude Test (Carroll & Sapon, 1959/2002). Giftedness is understood as untrained, outstanding innate ability, while talent, which is the most subjective of all the concepts, refers to a superior mastery of an innate ability (Gagne´, 2004). Gagne´ proposes the threshold of superior performance that would distinguish 1 in 10 individuals in the general population (i.e., scores at the 90th percentile) for mildly gifted or talented but 1 individual in 10,000 for exceptionally gifted or talented. Criteria for the identification of giftedness vary across age groups: adults and children must not be classified by the same standards. According to Housand (2009), gifted adults are recognized on the basis of their remarkable achievements in a domain of activity, which in turn is often reflective of high cognitive abilities. Their talents may be developed through educational opportunities. In addition, gifted individuals tend to share certain personality traits such as resilience and perseverance, a strong will and determination, a strong need to excel, motivation, ability to take risks, creativity, imagination, and innovativeness. Whereas some cases we discuss in this chapter refer to individuals who have learned just two languages, many others involve learners of several languages. In keeping with the terminology in the present Handbook, we use the term bilingual(ism) as the default term that includes any person who has learned more than a single language or language variety. We use the term multilingual(ism) whenever we specifically discuss individuals who have learned several foreign or second languages (see further discussion in Quay & Montanari, Chapter 27, this volume).

3 Aptitude for an L2 in FL Contexts Researchers in the field of second language acquisition (SLA) conceptualize high achievement from several viewpoints such as: nativelike pronunciation (Birdsong, 2007), near-native ability (Doughty, Campbell, Mislevy et al., 2010), giftedness (Biedron´, 2012), accomplished multilingualism (Biedron´ & Szczepaniak, 2012), high near-level language proficiency (Linck, Hughes, Campbell et al., 2013), talent (Novoa, Fein, & Obler, 1988;

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

309

Schneiderman & Desmarais, 1988), polyglotism (Hyltenstam, 2016), hyperpolyglotism (Erard, 2012), and savantism (Smith, Tsimpli, Morgan, & Woll, 2011). Some of the constructs would seem to imply high levels of FL aptitude (giftedness, talent, near-native ability, savantism), while some do not (nativelike pronunciation, accomplished multilingualism). However, all of them are associated with the potential for achieving a nearnative level of competence in one or more languages beyond the first language(s) learned during childhood. FL aptitude has been conceptualized as a combination of cognitive and perceptual abilities (i.e., aptitudes) (Skehan, 1998), whereas linguistic giftedness or talent refers to an extremely high level of aptitude (Biedron´, 2012; Biedron´ & Pawlak, 2016; Hyltenstam, 2016; Skehan, 1998). The concept of FL aptitude has evolved since its development by its main proponent, John Carroll (Carroll & Sapon, 1959/2002), thanks to major advances in the fields of SLA, cognitive psychology, genetics, and neurolinguistics. Today, the most influential theories of FL aptitude are those constructed by Skehan (the Processing Stage Model; 1998) and Robinson (the Aptitude Complex Model; 2002). Although different, both models involve multiple cognitive abilities and both are firmly situated in the research on cognitive factors and theories of language acquisition and processing. In the FL context, researchers also investigate explicit vs. implicit FL aptitudes that are relevant in different contexts of learning (Granena, 2013), macro vs. micro approaches in FL aptitude research (Skehan, 2016), domain generality vs. specificity of FL aptitude (Skehan, 2016), working memory (WM) as indicative of FL aptitude (Wen, 2016), and – most relevant for this chapter – aptitude subcomponents that predict attainment of very high levels of proficiency in FL learning, as measured, for instance, by the Hi-Lab, a test developed by Doughty et al. (2010) and Linck et al. (2013).

4 Critical Periods Consideration of limits of attainment in a foreign language is frequently linked to constraints imposed by critical periods (CP) (Long, 2013). Studies on ultimate attainment (i.e., the level of attainment reached by any given learner at the end of the learning process, however close to or far from nativelikeness that level may be) have provided evidence for a negative correlation between the age of onset of acquisition and ultimate L2 proficiency (Abrahamsson & Hyltenstam, 2009; DeKeyser, 2000). While age effects are generally acknowledged, the appropriate explanation for such effects has proven highly controversial and is far from resolved. Discussion of the critical period hypothesis for L2 acquisition (CPH/L2A) often revolves around some researchers’ claim that starting to learn an FL after the closure of a putative CP results in nonnative attainment. Proponents of

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

310

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

the CPH/L2A have advanced the position that in order to falsify the CPH/ L2A one would need to identify L2 learners who, under microscopic scrutiny, are indistinguishable from a monolingual native speaker in every measurable respect of language representation, production, and processing (see Treffers-Daller, Chapter 15, this volume, for discussion of variability among monolingual native speakers). Among some researchers the phenomenon of adult nativelike or nearnativelike attainment of certain features of the L2 is attributed to exceptional FL aptitude (Abrahamsson & Hyltenstam, 2009; DeKeyser, 2000; Novoa et al., 1988). For example, in DeKeyser’s (2000) study of 57 Hungarian immigrants to the United States, scores on a Hungarianlanguage version of the Modern Language Aptitude Test (Carroll & Sapon, 1959/2002) were compared with scores on an English grammaticality judgment (GJ) test. Younger-arriving participants, regardless of aptitude, scored at native or near-native levels on the GJ test. Among late arrivals, only those with above-average aptitude scored as near-natives. From the perspective of the CPH/L2A position, the only exceptions to the critical period would be those late learners with high aptitude. However, Birdsong’s (2013) reanalysis of DeKeyser’s data reveals that years of schooling significantly correlated with GJ scores among late learners, among early arrivals, among high-aptitude participants, and among low-toaverage aptitude participants. Years of schooling also predicted GJ scores over all participants. In sum, for the DeKeyser (2000) data, it was education, not aptitude, that was the more robust predictor variable of GJ scores. Thus, a focus on the narrow question of aptitude among late learners can distort the bigger picture of predictive factors in L2 proficiency, in this case, education. Finally, consider the view that nativelike attainment is requisite evidence to falsify the CPH/L2A. This position is undermined by a fundamental observation about the nature of bilingualism. Among individuals who are routine users of multiple languages, these languages influence one another in a bidirectional fashion. That is, in bilingualism, the L2 affects the first language (L1) just as inevitably as the L1 affects the L2 (see Ortega, Chapter 21, this volume). These reciprocal influences are attested across all levels of language knowledge – morphosyntax, lexis, pragmatics, pronunciation, etc. – as well as in real-time processing, where, for example, the lexicons of all of the various languages are potentially activated simultaneously in listening and speaking (see Paap, Chapter 22, this volume). Moreover, reciprocal L1–L2 influences may be found among individuals who are simultaneous (i.e., from birth) bilinguals, not just among later learners. What this means is that, quite simply, bilinguals are not to be seen as the sum of two monolinguals (see also Treffers-Daller, Chapter 15, this volume). In bilingualism, scrutinized across-the-board monolingual-likeness in either of the two or more languages is a near impossibility for reasons independent of age of acquisition, and thus the

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

311

criterion of nativelikeness to falsify or reject the CPH/L2A is untenable. From the perspective of this chapter, the special significance of the impossibility of multiple monolingual-likenesses resides in the fact that, no matter how gifted a given multilingual is, s/he cannot suppress the other language(s) in any absolute sense.

5 FL Aptitude and Multilingualism The current section turns to multilinguals specifically. Research on the relation between FL aptitude and multilingualism is scarce and in most cases based on indirect evidence. The discussion revolves around three questions: (1) Does language learning become easier and more successful with language learning experience, that is, after one first FL has been learned? (2) Are multilingual people privileged, in the sense of being more gifted language learners than most, and, if so, do the number of languages and the attained proficiency level differentiate the level of aptitude? (3) Does repeated experience with language learning contribute to an increase in FL aptitude? The first question was answered positively by Cenoz (2003), who reported that out of 18 studies she examined 13 yielded an advantage for third language (L3) learning over L2 learning. Schepens, van der Slik, and van Hout (2016) bolstered this conclusion with empirical evidence of an unprecedented scale, drawn from a nationally representative sample of 39,300 L3 Dutch learners in the Netherlands. The authors found higher levels of attainment for the multilinguals in the sample, for whom Dutch was an L3 (or L4, etc.) rather than an L2. Cenoz (2003) also suggested that literacy in two languages facilitates the acquisition of an L3, and Schepens et al. (2016) found that the benefits of multilingual learning are greater for smaller linguistic distances between the L3 and the L1 and/or best-learned L2 (see also Quay & Montanari, Chapter 27, this volume). In order to approach the problems associated with the other questions raised earlier we first have to decide whether FL aptitude is really a dynamic or rather a more static phenomenon. In the contemporary literature FL aptitude is often re-theorized as a dynamic phenomenon, that is, subject to evolutionary development in interaction with the environment (Sternberg, 2002). A number of studies indicate that aptitude can be increased as a result of training (Grigorenko, Sternberg, & Ehrman, 2000; Sternberg, 2002) or language learning experience (Thompson, 2013). These findings are not without counterevidence, however, as seen, for example, in Rogers, Meara, Aspinall et al. (2016), who reported no differences in FL aptitude for their 229 participants when they were compared by monolingual, bilingual, or multilingual background. On the other hand, with a new sample of 240 participants Rogers, Meara, BarnettLegh et al. (2017) found that on subtests measuring abilities related to

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

312

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

vocabulary learning and explicit grammar inferencing, participants with L2 formal classroom experience were superior to monolingual participants and no different from bilingual participants who had learned two languages prior to age five. Thus the authors conclude that at least the more explicit abilities implied in the construct of FL aptitude may be indeed trainable and susceptible to experience. According to Sternberg (2002), the learning of successive languages draws on linguistic meta-knowledge that is transferred from one language to another. This hypothesis accords well with Erard’s (2012) phenomenological report on very advanced multilinguals who know a large number of languages (hyperpolyglots), and who report a gradually increasing ease in learning of subsequent new languages, especially when they are typologically similar. Thompson’s (2013) study seems to confirm Sternberg’s dynamic view of FL aptitude. Using the Cognitive Ability for Novelty in Acquisition of Language (Foreign) aptitude test (Grigorenko et al., 2000), Thompson tested the correlation between language learning experience and the aptitude test results. As the correlation was positive, she suggested that FL aptitude is a dynamic factor. In line with this argumentation, the development of human ability is a constructive rather than a predetermined process. Human beings make sense of the world and develop in interaction with the environment; hence, the mechanisms of cognitive development are highly flexible. The concept of FL aptitude as a fixed trait is questionable because WM, a component of language aptitude, is believed to be modifiable as a result of experience and instruction (Gathercole & Alloway, 2008). Further, the instruments used to measure FL aptitude and language awareness appear to engage the same constructs to some extent, leading to uncertainty about the separateness of FL aptitude and language learning experience and/or awareness. The most prominent characteristic of talented individuals is their outstanding memory (Biedron´, 2012; Erard, 2012; Hyltenstam, 2016; Novoa et al., 1988; Schneiderman & Desmarais, 1988). Savants like Christopher Taylor (Smith et al., 2011) or Daniel Tammet (Treffert, 2011) have selective memory, which means that they can be excellent at retrieving linguistic items or numbers, even after a time interval, but are unable to remember nonverbal data such as faces (see Section 6.3 for more on savants). As Erard (2012) suggests, it is possible that hyperpolyglots have especially capacious WM. Recently, the concept of FL aptitude has been reconceptualized by adding WM to the array of cognitive abilities that may be decisive in achievement in L2 acquisition and processing (e.g., Wen, 2016). WM is a system containing a central executive and three buffer systems, that is, phonological, visuospatial, and episodic memory, which are all individually differentiated (Baddeley, 2015). Two of the systems, the phonological loop and the central executive, are of special importance in learning an FL (Wen, 2016). The phonological loop plays a crucial role in the development

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

313

of vocabulary, formulaic sequences, and grammar. The central executive directs attentional processes that create conditions for goal-directed behavior and seems to be significant in selective and resource-demanding language processes and the real-time performance areas of L2 comprehension, L2 interaction, and L2 speech production. Evidence has accumulated that deficits in both subsystems negatively affect language learning and that these deficits can be minimized as a result of training (Gathercole & Alloway, 2008). As many of the described cases of talented and multilingual individuals (Biedron´ & Szczepaniak, 2012; Erard, 2012; Hyltenstam, 2016) reveal superb memory abilities, it seems that this is their common feature. However, whether WM should be considered part of language aptitude and whether it is a prominent characteristic of talented individuals is still a matter of debate. Moreover, the chicken-egg problem of whether multilingual experience enhances WM capacity, or whether it is WM capacity that augments FL success, is unresolved. Linck, Osthus, Koeth, and Bunting (2014) note that it remains to be determined whether the relation between WM and ultimate attainment is shaped differently for simultaneous child bilinguals who by adulthood have achieved equally high levels of proficiency in both languages than for other populations. If this turns out to be the case, such a relation would be consistent with evidence for enhanced attentional control in executive function that according to some is associated with bilingualism over the lifespan (e.g., Bialystok & Craik, 2010; see Paap, Chapter 22, this volume, for more on this topic).

6 Talent and Language Learning An impressive number of studies in the literature describe exceptionally talented FL learners. Usually these are case studies, frequently anecdotal and not scrutinized scientifically, of multilingual individuals who are believed to speak six, eight, and even more than 50 foreign languages. Often there is little evidence for these numbers and the information is based on self- or other people’s reports. Because the phenomenon of nativelike or near-nativelike accomplishment in an FL appears to be quite rare, individuals who reach this landmark in several languages have evoked great interest. Along with numerous anecdotal cases presented in the literature (Erard, 2012; Hyltenstam, 2016) and on the Internet, there are a few scientific studies of high achievers in multiple languages. These can be divided into three groups: gifted multilinguals (Biedron´, 2012; Novoa et al., 1988; Schneiderman & Desmarais, 1988), polyglots and hyperpolyglots (see Erard, 2012, and Hyltenstam, 2016, for detailed reviews), and savants (Smith et al., 2011; Treffert, 2011). In this section we review what is known about each.

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

314

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

6.1 Gifted Multilinguals In early studies, linguistic talent was defined as an exceptional ability allowing achievement of L2 nativelike competence after puberty (e.g., Novoa et al., 1988; Schneiderman & Desmarais, 1988). The case studies described in the literature (see Biedron´ & Pawlak, 2016, for a review) ascribe exceptional language learning abilities to greater neurocognitive flexibility and a more bilateral language processing. One of the most detailed examinations of a talented learner was conducted in Novoa et al.’s (1988) study of CJ, the pseudonym of a 29-year-old male who knew German, French, Latin, Moroccan Arabic, Spanish, and Italian. Like other gifted individuals he revealed superior verbal memory and an enormous capacity for learning new vocabulary; however, his intelligence was not impressive. Biedron´ (2012) compared a group of 44 gifted multilingual foreign language learners to a group of 46 non-gifted learners on the following factors: various dimensions of FL aptitude, intelligence, WM, personality, motivation, learning styles and strategies, and biographical data. The age of the participants was between 20 and 35 years. They were all philology students and teachers recruited from universities in Poland. The number of languages they were learning varied from 1 to 11 (4.5 average) and included European and non-European languages. The level of proficiency for the gifted learners in at least one FL was advanced. Fourteen participants were highly advanced in one FL, 19 in two FLs, 8 in three, 2 in four, and 1 in five FLs. If they spoke more than two FLs, their level of proficiency in the additional languages ranged from basic to communicative. Both quantitative and qualitative data revealed that these gifted FL learners excelled in all cognitive ability tests, including FL aptitudes, intelligence, and memory. General intelligence as measured on the Wechsler Intelligence Scales was at a mean of 125 (which falls roughly at the 94th percentile of IQ scores in the general population), with the mean score on the verbal scale (130) being higher than the means on the nonverbal and memory scales (118 and 128, respectively). Also, the participants’ WM was well above average. The gifted L2 learners emerged as highly motivated, open to experience, hardworking, persistent, self-efficacious, creative, selfaware, and autonomous FL learners. The self-reported early ages of starting to speak and read in their L1 indicate high verbal ability from an early age, on the one hand, and environmental chances and choices, on the other.

6.2 Polyglots and Hyperpolyglots In what respects do polyglots differ from multilinguals? Overall, polyglots are assumed to be much more proficient across many more languages in comparison to multilinguals. Ultimately, however, in defining polyglots and hyperpolyglots the literature is neither consistent as regards the minimum number of languages nor precise with respect to observed or

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

315

assumed proficiency levels, nor indeed with respect to what is meant by proficiency. Accordingly, instances of alleged polyglots should be viewed with due caution, especially those based on anecdotal evidence. By Hudson’s (2012) definition, polyglots are highly proficient in at least six languages. Hyltenstam (2016) adopts this definition but limits it to adult learners by stipulating that polyglots are people who added at least six languages to their repertoire after puberty and attained a high level of proficiency in all of them. Erard (2012), based on his survey, introduced the notion of language super-learners, referred to as hyperpolyglots, that is, people who can speak or use in reading or writing at least 11 languages, thus well exceeding the classic minimum number of six proposed by Hudson and accepted by Hyltenstam. If 11 is the minimum, what is the upper boundary of languages a hyperpolyglot can learn? Johan Vandewalle – the winner of the Polyglot of Flanders Prize 1987 – was verified as being competent in 22 languages. There are also cases of polyglots from the remote past and historical figures such as Cardinal Joseph Mezzofanti, who, according to different sources, is said to have known around 60 to 70 languages, or Renaissance philosopher Pico della Mirandola, with 22 languages (Erard, 2012). Hyltenstam (2016) also lists such anecdotal cases as Dr. Harold Williams (58 languages) and Alexander Schwartz, who translated 31 languages into English. Not all of these languages can be expected to be known at equally high levels, however. In Erard’s (2012) estimation, the number of hyperpolyglots’ fluent and functional languages ranges from five to nine. Moreover, as Erard correctly observes, “[e]ach hyperpolyglot has a variety of uses for his or her languages” (p. 216), and seldom strives to achieve nativelike proficiency in all. Instead, hyperpolyglots strive for oral fluency in some of their languages, read or translate for pleasure in some others, or sometimes they just familiarize themselves only with chosen phrases or particular linguistic areas. Some languages are still active, others vanish or weaken as time passes, and some can be reactivated (see Keijzer & de Bot, Chapter 14, this volume, on the dynamics of language unlearning and relearning across the lifespan). Consequently, Erard proposed the term contextual and personally relevant level of proficiency to refer to their linguistic targets. This means that with hyperpolyglots one must be willing to relax the stringent proficiency requirements for polyglots implied in Hudson’s (2012) or Hyltenstam’s (2016) definitions. Erard (2012) and Hyltenstam (2016) discuss many cases of polyglots, and Hyltenstam provides an interesting list of 96 anecdotal cases of polyglots (see his Appendix 1). Among the cases of such linguistically talented individuals are famous public figures, such as Pope John Paul II (11 languages) and Pope Benedict XVI (10), whose ability to communicate in a number of languages is well recorded albeit never formally tested. There are many writers, and indeed Hyltenstam’s (2016) list contains 15 cases of polyglots who are fiction writers. Examples include Irish novelist

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

316

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

James Joyce, the author of Ulysses (14 languages); J. R. R. Tolkien, the author of Lord of the Rings (30); and Anthony Burgess, the author of A Clockwork Orange (12). Joyce was not only a writer, but also studied languages and was a language teacher for some time. Tolkien was also a poet, a philologist, and a university teacher. Burgess studied English philology and worked as a teacher. All of them were creators of artificial languages. The PolishJewish author of the artificial language Esperanto, Ludwik Łazarz Zamenhof, was fluent in 13 languages. Among other cases of polyglots are a number of professional linguists such as Rasmus Christian Rask (fluent in 25 languages and able to read in 35), Andre´ Martinet (12 languages), Donald Kenrick (communicative in 30 languages and able to translate from 60), Stephen Wurm (working knowledge of 50 languages), and Kenneth Hale, who studied indigenous languages of America and Australia (more than 50 of them). The notoriety of polyglots has also been exploited for commercial purposes. “How this perfectly normal guy learned 9 languages” is the headline of an article in +Babbel Magazine, an Internet portal for a commercial language instruction product. In the portal, the polyglot in question, Matthew Youlden, appears in a video monolog speaking fluent English, Irish, German, French, Hebrew, Spanish, Catalan, Portuguese, and Italian. A sociolinguist who studies minority languages, “Matthew has mastered a staggering number of languages by utilizing abilities that we all possess: persistence, enthusiasm, and openmindedness” (no page number, www.babbel.com/en/magazine/languagechamaleon-one, last accessed January 30, 2018). As our previous examples suggest, we see a tendency among polyglots to choose academic careers connected with languages; for example, they are professional translators, interpreters, language teachers, linguists, or language students (Abrahamsson & Hyltenstam, 2009; Biedron´, 2012; Hyltenstam, 2016). This suggests a link among bilingualism, linguistic giftedness, and professional career. Moreover, it also implies a high level of verbal intelligence, and good WM (Biedron´, 2012; Erard, 2012; Hyltenstam, 2016). Nevertheless, we can also find representatives of various other professions, like Finnish diplomat Pertti Laakso (14 languages), German diplomat Emil Krebs (60–65 languages), one gardener and one blacksmith, and the examples of higher-ranked clergy cited earlier. Surprisingly enough, on his list of 96 polyglots, Hyltenstam (2016) found only three women. The best-known woman polyglot is Kato´ Lomb (Erard, 2012; Hyltenstam, 2016), a Hungarian scientist, translator, and simultaneous interpreter, born in 1909 and deceased in 2003. She graduated in physics and chemistry and obtained a doctorate in the latter. Being an autodidact she attempted to learn about 70 languages, but was able to interpret in 9 or 10. A very enthusiastic language learner, Lomb did not believe in exceptional abilities but attributed her success to appropriate learning strategies.

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

317

The observation that females may be a minority among polyglots is fascinating, as compared to men women are believed to have either higher or equal verbal abilities, at least for some aspects of language acquisition and processing (see Wucherer & Reiterer, 2018, for a review). An interesting suggestion to explain the disproportion is offered by the Extreme Male Brain Theory of Autism proposed by Baron-Cohen (2002). The theory associates maleness and femaleness with empathizing vs. systemizing personality profiles and attempts to explain why the incidence of autism spectrum conditions is much greater in males than females. In the words of its originator: “The male brain is defined psychometrically as those individuals in whom systemising is significantly better than empathising, and the female brain is defined as the opposite cognitive profile” (p. 248). Baron-Cohen considers autism an extreme of the normal male profile. It may be that savants, who are usually individuals revealing autism spectrum characteristics, can possess extremely systemizing male brains. Indeed, the two savants described in the polyglot literature suffer from Classical Autism (Christopher Taylor) and Asperger’s Syndrome (Daniel Tammet), and both are extremely gifted multilinguals (see further discussion in what follows). Also, the reported lack of social skills in another polyglot, Emil Crebs, converges with a typical Asperger’s Syndrome symptom. There is also another plausible reason for the underrepresentation of women in the discussed population. All the polyglots described in the literature report that they devote at least a few hours a day to learning languages. It is likely that the disproportionate numbers of male and female polyglots can also be ascribed to social factors, such as time restrictions and household chores, which prevent many women even in Western societies from exercising any language learning passion to its full potential. Finally, what might explain the enormous talent for multilingualism that (hyper)polyglots seem to possess? A number of characteristics seem to matter. These are shared in the accounts by Hyltenstam (2016) and Erard (2012). Among cognitive abilities we can list extraordinary memory, especially for words, as well as efficient memory strategies. Moreover, polyglots seem to have high analytic abilities that enable them to discover regularities in grammar and pronunciation systems. Erard refers to them as linguistic multitaskers, as they can monitor their speech and switch between languages without interference, arguably thanks to superior executive functions. Additionally, their ability to imitate foreign sounds, indicating high phonetic aptitude, distinguishes them from many other FL learners. Needless to say, motivation is a vital factor for polyglots, which, along with intellectual potential, determines the extent of their achievements (Hyltenstam, 2016). Polyglots are very devoted to their passion and they are driven by what in gifted education is known as the rage to master (Housand, 2009), which is typical of many talented individuals. As autodidacts, they are masters of learning strategies and are enormously

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

318

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

inventive in discovering new techniques and devices to improve the effectiveness of learning. They also use a wide range of instructional materials, including those designed by themselves (see Biedron´ & Pawlak, 2016, for a review). Personality factors seemingly shared by (hyper)polyglots include openness to experience, good adaptation skills, curiosity, self-confidence, perseverance, and diligence.

6.3 Savants Linguistic savants are probably the most unusual group of talented language learners. According to Treffert (2009), “[s]avant syndrome is a rare, but extraordinary condition in which persons with serious mental disabilities, including autistic disorder, have some ‘island of genius,’ which stands in marked, incongruous contrast to overall handicap” (p. 1). One in 10 persons affected by autism reveals amazing abilities in calculations, music, art, or, more seldom, in languages. Not all savants have autistic conditions, but all of them, irrespective of their abilities, possess extraordinary memory (Treffert, 2009). Treffert explains savantism in terms of left hemisphere dysfunction and right hemisphere compensation, where motor and language functions are taken over by the right hemisphere because of deficits in the left hemisphere. Alternatively, right brain skills are not newly developed but represent latent and dormant skills that are released from what some call the tyranny of the left hemisphere. Moreover, there is a possibility that a savant gene is transmitted in some families. The two cases of linguistic savants described in the literature – Christopher Taylor by Smith et al. (2011) and Daniel Tammet by Treffert (2011) – provide evidence about people who possess extraordinary linguistic abilities accompanied by a variety of physical and mental impairments. Christopher, aged 55, is a multilingual savant who was able to communicate, write, read, and translate in more than 20 languages (Smith et al., 2011). He was diagnosed as having hydrocephalic brain damage and severe neurological impairment in his motor coordination. He was so severely mentally retarded that he was unable to look after himself. What is more, the profile of his cognitive abilities was very asymmetrical, with a significant dominance of verbal over nonverbal ability, with verbal ability falling within the normal range. His control over pragmatics and syntax was limited as compared to his lexical development. This limitation, coupled with the mental retardation, affected his language use. He translated languages word for word and was not able to evaluate the sense and cohesiveness of his text. Christopher possessed many autistic characteristics, which accords well with the Extreme Male Brain Theory of Autism (Baron-Cohen, 2002) that associates autism and male brain tendencies with high systemizing skills, narrow interests, and specific repetitive behaviors and habits.

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

319

Daniel Tammet is a linguistic savant who was 38 years old at the time of study and had been diagnosed with Asperger’s Syndrome (Treffert, 2011). He knew 11 languages, including French, German, Spanish, Lithuanian, Esperanto, Icelandic – which he was able to learn in a week – and Ma¨nti, a language he invented himself. Unlike Christopher, his intelligence level was normal. Tammet was diagnosed as having congenital child epilepsy. His outstanding abilities of memory, arithmetic computation, and languages appeared after a series of seizures in early childhood. As a result, he is considered an acquired savant, that is, one whose special abilities emerge as a result of an injury or disease. Tammet was also extraordinary due to his synesthetic ability, as he was able to see numbers and some words as colors. He was a high-functioning savant, who, like Christopher, possessed a superior memory capacity that was highly selective. Specifically, his phonological short-term memory measured by digit span was excellent; he was able to recall long strings of numbers or symbols and manipulated them with unbelievable speed (for example, he was able to multiply six-digit numbers within seconds). He was also a master of calendar calculations. On the other hand, his memory for faces was impaired. Birdsong (1996) examined the limitations of Christopher’s giftedness. First of all, Christopher’s command of multiple languages was far from perfect. For example, he tended to apply the rules of English syntax to other languages. In addition, his ability to interpret language in use – in his native English as well as in his other languages – was hamstrung by his autism-typical difficulties with pragmatics, nonliteral language, irony, humor, and rhetorical questions. Second, and more to the point of this chapter, Christopher’s linguistic virtuosity resided in his extraordinary capacity for memorizing vocabulary and surface morphology. In essence, he was an adept accumulator of facts that happen to be linguistic in nature. Arguably, his impressive but superficial knowledge of language reflected a highly developed intelligence in the domain of language, but not a faculty for language. What do the cases of savants tell us about aptitude for learning languages? From the preceding description it is clear that savant-like talent does not map directly onto the aptitudes recruited by highly proficient users of multiple languages (see also Erard, 2012). That is, neither the neuro-structural features nor the linguistic behaviors of savants are observed among the highly proficient FL learners that are the primary focus of this chapter. Moreover, qualitative and quantitative differences exist between the two groups in terms of experience and motivation. Thus, we are reminded that not all exceptional language learners are exceptional in the same ways. Further, in contrast to both polyglots and hyperpolyglots, the learning of multiple languages by savants appears to engage a unique drive to memorize and recall that is more like an obsession than a motivation. Pring (2007) suggests that obsession is what distinguishes savants from non-autistic

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

320

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

experts (which presumably would include polyglots and hyperpolyglots). Pring notes as well that savants appear to practice obsessively for the purpose of enjoyment rather than improvement of performance. Pring (2007) goes on to observe that learning among individuals with high ability is characterized by “goal setting, evaluation and feedback”; in contrast, “none of these factors [. . .] are evident in savant activity” (p. 502). Finally, as noted earlier, an additional dimension of qualitative difference between savants and others is their abnormal brain organization. In these various qualitative respects (and because, in the case of Christopher at least, his command of language features was superficial in important ways), there is no obvious place for savants along a hypothetical continuum of increasing aptitude for language learning. Rather, they belong in a category unto themselves.

7 Linguistic Giftedness of Children: The Case of Young Interpreters Generally, research into gifted education (e.g., Housand, 2009) reveals that gifted children learn more quickly, deeply, and extensively than their peers. They display high reasoning ability, creativity, an inveterate curiosity, a large vocabulary, and excellent memory, as well as the ability for sustained concentration. Owing to superior memory and concentration, gifted children are able to remember new concepts after a few repetitions. What is more, they demonstrate a preference for complex, abstract thought, often using higher-level thinking skills. They possess wide general knowledge and a vivid imagination. Gifted children can be physically and emotionally sensitive, perfectionist, and often tend to question authority. A well-known phenomenon that might involve linguistic giftedness is the case of young bilingual language brokers, who are selected from among their siblings to serve as informal interpreters for their families (see Bigelow & Collins, Chapter 2, this volume). Research into the linguistic abilities of these children is relatively scarce. A notable exception is a study by Valde´s (2002), who reports on young interpreters’ performance on a simulated interpretation task and concludes from this evidence that they are able to carry out the very complex task of interpreting under stressful conditions. According to Valde´s, young interpreters exhibit exceptional abilities recognized in prospective interpreters: online memory skills, analytical ability, speed of comprehension, and production, as well as stress tolerance. What is more, they possess strengths in areas connected with superior general intellectual ability, such as long-term memory, abstract word knowledge, and abstract reasoning. In a similar vein, Angelelli (2016) points out that because young interpreters’ interaction in adult contexts requires the use of advanced vocabulary and cognitive abilities, these early experiences of language brokering in bilingual

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

321

communities result in the development of higher cognitive functioning, an increase in their linguistic aptitudes, and the improvement of their interpersonal and social skills. It may be that the gift for languages these children have is enhanced due to their expertise in interpreting. However, given the lack of studies independently measuring language talent or FL aptitude in children who then become informal interpreters for their families, we are left with a similar chicken-egg problem as the one discussed in Section 5 for multilingual adults, only now for child bilinguals: are naturally superior linguistic abilities the reason that these children are selected as informal translators by their families? Or does language brokering experience at a young age enhance their linguistic abilities? In the absence of research specifically investigating children’s bilingual aptitude, answers must remain hypothetical.

8 Reflections on Future Research Researchers have investigated the etiology and sources of giftedness across many domains of human performance. As a noteworthy example, Vinkhuyzen, van der Sluis, Posthuma, and Boomsma (2009) conducted an analysis of giftedness of 1,685 Dutch twin pairs with the purpose of determining the relative contributions of nature and nurture to exceptional performance in the domains of music, arts, writing, language, mathematics, sports, memory, and general knowledge. On the basis of participants’ selfassessments of their talent in each domain, the researchers concluded that genetics contributed between 50% and 92% of the variance of performance across the domains in question, with the highest contributions of nature to music, mathematics, and sports, and the lowest to arts, memory, and language. However, the authors dismissed the responses for the language domain because of the fact that almost all people who attend high school in the Netherlands would respond similarly to the questions about the language domain, and would fall into the highest two levels of talent for language learning assessed in the study, viz., being able to speak and read one FL fluently (the next to highest level) and being able to speak and read three or more languages (the highest level). Based on these definitions, there are many gifted language learners in the twins survey, but statistically they are not exceptional within the Dutch population at large. This issue notwithstanding, the Vinkhuyzen et al. (2009) study triggers a set of fascinating questions that are worthy of future study. To what degree are various types of experiential factors and motivational factors associated with exceptional talent? How do these proportions vary across domains of human performance? At the same time, one would wish to know in what domains high levels of giftedness are observed most frequently, and where exceptional FL learning is situated along a continuum of domain-relevant outcomes. Further, it would be of special interest to conduct this type of

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

322

A D R I A N A B I E D R O N´Ń A N D D A V I D B I R D S O N G

research across different pairings of languages and cultures. With respect to the interplay of nature and nurture, for example, how does gifted Anglophones’ learning of Mandarin Chinese compare to their learning of, say, Moroccan Arabic? And to what extent does this comparison resemble Francophones’ learning of Mandarin Chinese vs. Moroccan Arabic? In the spirit of Vinkhuyzen et al., one could also consider correlations of aptitude in one domain with aptitude in a different domain. This approach would add to the literature that examines, for example, the relation of talents for music and math to talent for learning languages. Another study that prompts further reflection on aptitude and high proficiency is that of Winke (2013). In Winke’s model of FL aptitude, phonological WM, rote memory, grammatical sensitivity, and phonemic coding ability are four constituents of aptitude. Additionally, aptitude is hypothesized to be mediated by strategy use and motivation. Winke was interested in the relative contributions of these factors to the attainment of advanced proficiency in L2 Chinese by adult Anglophone learners who received 30 to 40 hours of instruction per week over 63 weeks at the Defense Language Institute in Monterey, California. For a variety of reasons, mostly relating to design and analytic features of the study, the factors that Winke (2013) examined did not explain much of the variance in scores on the Defense Language Proficiency Tests for listening, reading, and speaking in Chinese. These findings led Winke to speculate on why language learning aptitude is such a poor predictor of proficiency at advanced levels of L2 Chinese. She proposed that aptitude is a weak factor in later stages of L2 learning (cf. DeKeyser, 2000) – perhaps not because aptitude levels and types vary as learning progresses, but because of the way learners orient themselves to the learning experience (see Thompson, 2013, and earlier in this chapter). At higher proficiency levels, “[t]he learner’s actions in the social environment, the amount of time spent outside of class learning, and other personal reactions and choices of what to focus on ultimately affect learning” (Winke, 2013, p. 121). Winke’s observation is an invitation to future researchers to consider the connection of aptitude to high FL proficiency in terms of necessary and sufficient conditions. Although high aptitude may be necessary for high proficiency, by itself it is not sufficient; it must be accompanied by adaptive behaviors. In a related vein, recalling our earlier discussion of the scope of the education versus the aptitude effect in DeKeyser (2000), going forward it is important to keep in proper perspective the proportional roles of aptitude and other factors in the attainment of high L2 proficiency across all ages of learning.

9 Conclusion In this chapter we hope to have shown that “exceptional” in the context of bilingualism is a notion that must be understood insightfully. We believe

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

Highly Proficient and Gifted Bilinguals

323

that nativelikeness – in any of the languages of a bilingual – should not be framed strictly as an exception but, especially at granular levels of examination, as an implausibility, due to the reciprocal influences of the languages represented in the bilingual mind. Further, it is important to remember that attainment of unusually high levels of proficiency in more than one language, as with extraordinary performance in other domains, has both intrinsic and extrinsic components, and that education, practice, dedication, and purpose are essential to observed expertise for polyglots and hyperpolyglots as well as for run-of-the-mill bilinguals. Returning to the title of this chapter, we emphasize that gifted is not synonymous with proficient: to become highly proficient, the natural abilities of a gifted individual must be engaged with sound choices for training and experience, along with a strong and persistent will to achieve.

Downloaded from https://www.cambridge.org/core. Edinburgh College of Art, on 13 Nov 2018 at 15:34:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.017

17 Language Choice in Bilingual Interaction Annick De Houwer

1 Introduction Humans are a social species, and interaction is a fundamental form of social action. This social interaction is often realized through discourse, that is, through interlocutors using language in a given setting. I use the term interaction here to refer to social interaction through discourse, rather than through nonverbal interaction. Across ages and lifespans, interaction can include, but is not limited to, dyadic one-on-one conversation; small group interactions, for example, at family mealtimes or cocktail parties; classroom interactions; and more formal and ritualized interactions at, for instance, town hall meetings, conferences, and religious ceremonies. This chapter focuses on how bilinguals make use of their language repertoire in real-time interaction with others. In particular, I discuss the range of language choice options that bilinguals have when they interact with others. I show that these choices are fundamentally related to a person’s linguistic repertoire in comparison to that of potential interlocutors, to interpersonal and identity dynamics, and to the constraints of a given setting, with the attitudes, ideologies, and affective attachments they engender. I present examples from different stages of the lifespan, starting with one-year-olds, and discuss both private and more public contexts of language use. I use the term “language” as shorthand for “variety of language”; this includes (varieties of) sign languages. Throughout, I focus on real-time, face-to-face language use. Real-time bilingual interaction may, however, also take place through digital media in the written mode (e.g., in chatroom conversations or web forum discussions; for the latter, see, e.g., Androutsopoulos, 2007), but I leave such digital interactions aside. Likewise, there is a large literature on the cognitive processes engaged in switching between languages as measured in highly structured experimental trials (see Paap, Chapter 22, this volume).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

325

These studies usually do not involve social interaction, and are not reviewed here. To set the scene I present foundational theories and key concepts in Section 2. These include a characterization of what bilinguals’ linguistic repertoires and bilingual interactions consist of. In Section 3, I discuss the foundational role of language socialization and introduce the convergent choice principle for bilingual interaction. Contexts, attitudes, and ideologies play a large role in language choice as well, and these are discussed in Section 4. Section 5 then expands on the social meanings constructed through particular language choices. Section 6 explores the more psycholinguistic aspects of language choice. Before concluding, I discuss in more detail bilingual interactions in what are likely the most private and the most public of contexts for language use, that is, within the family in Section 7, and in the political arena in Section 8.

2 Foundational Theories and Key Concepts Theories of interaction apply to all humans, regardless of the number of languages they know. The central feature of social interaction through discourse is that the participants have an effect on each other (Schiffrin, 1994, p. 415). Following many (e.g., Gardner-Chloros, Sebba, & Moyer, 2007), I consider the utterance to be the basic structural unit for discourse. Social discourse interaction normally involves both real-time language comprehension and language production by at least two participants who take turns between attending to, processing, and trying to understand the other participant, on the one hand, and producing language themselves, on the other. Thus, comprehension and production are interwoven forms of joint action and action perception (Pickering & Garrod, 2013). Social interaction is essentially cooperative (even if people are arguing), and interlocutors accommodate each other as part of their desire to be understood and reduce psychological distance (e.g., Giles, 2016). In interactions people also worry about being perceived as polite and saving face (Brown & Levinson, 1987). I explore how these foundational principles of interaction apply to bilingual interactions. These I define as any discourse interaction involving at least one bilingual interlocutor. In turn, I define a bilingual interlocutor as a person who is in principle able to understand two (or more) language varieties at levels that are minimally appropriate for a given life stage, regardless of whether elements from multiple language varieties are actually used in a particular stretch of discourse. I do not discuss language interactions involving bilinguals who have suffered a stroke or are otherwise language impaired (for discussion of language choice in aphasics, see Hernandez, Chapter 23, this volume). For people to have become bilingual interlocutors, they must have learned to understand utterances belonging

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

326

ANNICK DE HOUWER

to each of at least two different language varieties, that is, to two distinct ways of speaking, signing, or a combination of these modalities (see Chevrot & Ghimenton, Chapter 25, this volume, on bidialectalism; see Tang & Sze, Chapter 24, this volume, on the use of sign languages by bilinguals). These different ways of speaking are not necessarily labeled (indeed, they often are not). By virtue of having a choice between or among languages, bilinguals can understand and (potentially) produce unilingual and mixed utterances (see later in this chapter). Bilingual interlocutor status can already be gained at one year of age or even before, namely when children are able to understand several single words and a few phrases in at least two varieties. It also obtains when adults are able to understand the kind of language use common in informal adult-to-adult conversations in at least two languages (even if they can only produce one language). Utterances with morphological material from just a single language are unilingual utterances (morphological material includes free morphemes). In addition to unilingual utterances in each of at least two languages, bilinguals may also have heard mixed utterances (De Houwer, 1983, 1990). These contain morphological material that linguists can describe as belonging to different languages. The combined use of morphological elements from two languages within a single utterance has also been referred to as mixing, code-mixing, code-switching, language alternation, or borrowing (Auer, 1998; Gardner-Chloros, 2009; MacSwan, 2016; Muysken, 2000; Myers-Scotton, 1993; Poplack, 1980). Often the distinction between these different terms is unclear. One reason is that these terms (except borrowing) are often also used to refer to units above that of the utterance. For instance, code-switching may be used to refer to utterances with morphological material from two languages (i.e., mixed utterances), but also to structurally different levels of conversation, that is, to the use of unilingual utterances from different languages within a conversational turn and/or between turns, and to the alternation between mixed and unilingual utterances in a larger stretch of discourse. As further explained by Aalberse and Muysken (Chapter 26, this volume), there has been considerable controversy about which terminology is best, and this controversy has involved opposing theoretical frameworks. For clarity’s sake, I prefer to use the descriptive terms unilingual and mixed utterances. Later analyses can explore the precise structure of mixed utterances, and any limitations and language-specific or universal processes involved that help explain their precise structure (e.g., Myers-Scotton, 1993). I do not pay attention to the structure of mixed utterances (but see briefly later in this chapter). Later analyses can also focus on the alternation among unilingual utterances in distinct varieties, or between unilingual and mixed utterances, in any stretch of discourse, and on how such alternations are structured and why they occur. This, in fact, is the focus of the present chapter: what kinds of utterances, unilingual in any relevant variety, or mixed, are used in bilingual interactions, and why?

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

327

It may not always be possible to clearly identify an utterance as unilingual or mixed. A categorization of free and bound morphemes in terms of a specific language variety will not always be easy. Also, linguists may make categorizations that laypeople would find puzzling. As I generally take an ethnographic, bottom-up approach to language use, I consider the people who are actually using particular varieties as the final arbiters of what variety specific word forms and phrases belong to (even if they have no name for the varieties they speak, people will be able to say that a particular word or phrase corresponds or does not correspond to how they normally speak). Laypeople (including bilinguals) often comment on the use of mixed utterances. This suggests that mixed utterances are sociolinguistically salient. Many language users hold quite negative attitudes toward the use of mixed utterances, but many others do not. In a large (n = 2,070) multinational online survey of mostly bilingual rather than monolingual respondents, Dewaele and Li Wei (2013) found that greater experience with living and working in linguistically diverse environments was associated with more positive attitudes toward the use of what was called “language switching” in the survey. Being a more emotionally stable person was also a factor associated with more positive attitudes. Respondents’ own levels of proficiency in their different languages, on the other hand, turned out to have little or no association with their attitudes toward language switching. In a bidialectal setting, De Schryver (2012) reports negative attitudes toward “tussentaal” – literally, inbetween-language – which refers to the use of utterances combining elements from standard Dutch and dialectal features in Flanders, Belgium (see Chevrot & Ghimenton, Chapter 25, this volume, on bidialectalism). The apparent sociolinguistic reality of mixed utterances warrants their treatment as a separate category of utterances, contrasted with unilingual utterances. In my personal experience, fluidly switching between unilingual utterances in different languages may also elicit comment from observers, but typically these people are not themselves bilingual. It is possible that the sociolinguistic salience of mixed utterances has influenced linguists’ long-standing theoretical interest in them (after all, linguists are language users too). There is a large literature on the structural aspects of mixed utterances (e.g., Gardner-Chloros, 2009). Mixing patterns can be quite simple, involving the insertion of a free morpheme (often a noun) from language X into an utterance that is otherwise in language Y. There are also highly complicated mixing patterns involving several combinations of free and bound morphemes from each of two languages (e.g., Muysken, 2000). There has been much controversy over how such mixing patterns can be theoretically explained (MacSwan, 2016). Most recently, interdisciplinary approaches have begun to appear that hold great promise for a better understanding. For example, some scholars have turned to studying large bilingual

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

328

ANNICK DE HOUWER

interaction-based speech corpora in order to better understand mixing patterns (Gries & Kootstra, 2017), mainly with the aim to explore the role of priming, that is, the influence of structures that recently occurred in a stretch of discourse. Other scholars are aiming to integrate psycholinguistic, structural, and sociolinguistic aspects (see the contributions in Stell & Yakpo, 2015). Such interdisciplinary efforts are likely to yield important new insights about the use and nature of mixed utterances in bilingual interaction. Bilingual interlocutors who are able to speak two languages on top of understanding them have the following options for every utterance they produce: forming (i) a unilingual utterance in language X, (ii) a unilingual utterance in language Y, or (iii) a mixed utterance combining morphological elements from languages X and Y. For each utterance, then, by necessity, a choice has to be made among these three possibilities. The term choice, however, is not meant to indicate a conscious decision (although it may be). Which language choice is actually made depends on many different factors. As I discuss in the next section, the role of language socialization here is fundamental.

3 Language Socialization Processes and the Convergent Choice Principle Individuals’ ability to successfully communicate through language in real time with others largely depends on their previous and ongoing engagement with others in face-to-face interaction. Through that social interaction, people socialize others into speaking or signing in a particular way. Such language socialization practices are usually thought of as involving a novice and an expert, a distinction that is socially and dynamically defined (Ochs & Schieffelin, 2011). Though they may not realize it, even linguistically proficient and mature community members (normally considered experts) never quite cease to be socialized into using specific linguistic forms rather than others. This occurs, for instance, when an interlocutor criticizes the choice of a particular word, asks for a clarification, or laughs in response to a particular turn of phrase. When such metalinguistic responses are generally lacking, the implicit message is that one’s linguistic choices are being accepted and, what is more, approved. Clear evidence that language socialization continues to play a role well beyond the novice speaker state consists of the pathologization of senior expert speakers whose language use may no longer live up to community standards of fluency or semantic accuracy (Goral, Chapter 5, this volume). Language socialization processes concern linguistic elements at all levels of structure and use. A well-known example is the socializing of young children into using culturally polite ways of requesting foods (child:

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

329

apple! father: say please! child: please apple! father: here you go [giving apple]). Here, language socialization involves pragmatics. A perhaps less common example occurs when an accomplished adult speaker of a language tells an incipient adult learner of that language during an informal conversation that they should use another form of a verb. Both these examples involve explicit language socialization. Language socializing processes generally tend to be more implicit and indirect, however (Ochs & Schieffelin, 2011), such as simply engaging children in conversation, not responding to a person’s request, or repeating a verb with the “correct” morphology after the interlocutor had used an “incorrect” form, as in implicit otherrepairs or so-called recasts. More explicit corrective feedback is also possible, especially in educational contexts. Whether implicit or more explicit, language socialization processes express specific sociocultural norms, that is, shared expectations, for linguistic behavior. These expectations also include norms for language choice, that is, norms for what kinds of varieties should be used in a particular context. For any longer stretch of discourse going beyond single utterances the choice is between (i) only unilingual utterances in language Y, (ii) only unilingual utterances in language X, (iii) only mixed utterances with morphemes from X and Y, or (iv) a combination of some or all of these. Norms for language choice include norms as to what kinds of mixed utterances, if any, are appropriate – the very simple insertional ones, or the more complex ones. As further explained in Section 6, on the processing level, language choice norms set up expectations for comprehension as well as for production, leading bilinguals to be in a specific language mode (e.g., Grosjean, 2001) and to have a specific language orientation (De Houwer, 2009). A very general norm for language choice relates to the fact that bilinguals choose to speak a language they think their interlocutor will understand. There is little point in talking to someone who cannot understand you (except perhaps in interactions with infants and in explicit language teaching settings). Even the youngest of children able to use words and phrases in two languages generally follow this comprehensibility norm (De Houwer, 2009; Marcon & Coon, 1983). Thus, the comprehension abilities of the persons involved in bilingual interaction are a major determinant of bilingual speakers’ language choice. Speaking a language one’s interlocutor will understand is part of accommodating communication behavior as stipulated in Communication Accommodation Theory (CAT) (e.g., Giles, 2016). A major insight of CAT is that speaking in the same way as one’s interlocutor reduces the perceived psychological distance between interlocutors. I call this the convergent choice principle (CCP) when it is applied to bilingual interaction. The principle helps explain why in the bulk of dyadic conversations interlocutors tend to speak the same way – either (i) both using just or mainly unilingual utterances in a single same language,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

330

ANNICK DE HOUWER

(ii) both using mainly mixed utterances combining the same two languages, (iii) both using mixed utterances and unilingual utterances in one language, or (iv) both alternating between mixed utterances, unilingual utterances in language X, and unilingual utterances in language Y. Patterns (ii), (iii), and (iv) are usually only possible when interlocutors are able to speak the same two languages. Conversations where bilingual interlocutors each consistently use different languages are the marked case (Grosjean, 2001), and, in CAT terminology, are examples of nonaccommodation (Gasiorek, 2016). The limits and possibilities of the range of language choice options are fundamentally related to a person’s linguistic repertoire in comparison to that of potential interlocutors. An Uzbek-speaking prisoner in Guantanamo who only spoke Uzbek and did not speak any English or Arabic remained silent because the other prisoners and prison staff did not understand Uzbek (Honigsberg, 2014). The prisoner in question did not understand English or Arabic, either, and so could not engage in verbal interaction at Guantanamo. However, even when the level at which interlocutors are able to speak a particular language is quite low, they can still accommodate others in their language choice, and thus adhere to the CCP. An example of this can be found in Rampton’s (e.g., 2014) rich ethnographic and micro-analytic study of bilingual interactions among adolescents from different ethnic backgrounds in the United Kingdom in the 1980s. For instance, a boy from a Pakistani immigrant family might use some utterances in Caribbean Creole, a language he does not otherwise speak, and a girl from an Anglo family might use mixed utterances with both English and Panjabi words, without knowing how to speak much more Panjabi. Rampton has called this language crossing. He emphasizes that this crossing takes place in the context of ordinary activity and that the participants are first and foremost intent on dealing with that everyday activity rather than on presenting themselves in terms of any particular ethnic or cultural identity (Rampton, 2014, p. 7; for corpus-based analyses of other examples of language crossing among adolescents, see Nortier & Svendsen, 2015). Another example where interlocutors who are not (yet) accomplished users of two languages show adherence to the CCP comes from the very youngest of bilinguals in the second and third years of life. As I showed for Dutch–English bilingual Kate, between the ages of two years and seven months and three years and four months, she more often than not replied in English when an adult who normally spoke Dutch with her happened to address her in English; when an adult who normally spoke English with her happened to address her in Dutch, she more often than not replied in Dutch (De Houwer, 1983). Children raised with two languages from the earliest days onward vary considerably from each other in the ages at which they first start to show evidence of the CCP, though. This is likely a partial result of the large variability among very young children in the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

331

ages at which they can produce words and phrases in two languages that are clearly identifiable as belonging to a specific language. Also, not all young bilingually raised children actually learn to speak two languages (De Houwer, 2009). When they do, the earliest age evidencing adherence to the CCP is 15 months (Sinka, 2000; see Serratrice, Chapter 1, this volume, for further discussion of language choice in very young bilinguals). Very young bilingual children’s ability to adhere to the CCP has been interpreted as a sign of early metalinguistic awareness (De Houwer, 2017a). When they are just a bit older, say four or five, children start to be able to reflect on their own language abilities. To avoid social embarrassment because of their perceived insufficient language proficiency in a particular language, some preschoolers may stop using that language (De Houwer, 2017b). Here we see the politeness principle at work, where individuals try to save their positive face. At the same time, if there is a willingness and motivation to communicate, the CCP again plays a more important role. This is exemplified by a trilingual six-year-old in my circle of acquaintances with varying abilities in Dutch, French, and English. Her self-identified preferred language was English, and, as she described to me, she spoke both English and French fluently, and understood Dutch but hardly spoke it. I used to address her in Dutch and she responded in English. Yet when she met up with a somewhat older unfamiliar girl who was identified as speaking Dutch, the six-year-old attempted to talk to her in Dutch.

4 Constraints on Language Choice: Contexts, Attitudes, and Ideologies As seen so far, both the comprehension and production language repertoires of the interlocutors in bilingual interactions are crucial in setting the stage for possible language choices. However, depending on the setting in which bilingual interactions take place, these choices are further constrained by several factors. Language choice norms may be dictated by local official rulings. For instance, for any medical insurance reimbursement of diagnostic language testing for developmental language disorders by speech and language professionals to be possible in Flanders, Belgium, it must take place in Dutch, regardless of whether client and clinician share another language. This rule makes an adequate diagnosis for children with potential language learning problems whose home language is not Dutch, or who hear both Dutch and another language at home, a near impossibility (De Houwer, 2018a; see also Hammer & Edmonds, Chapter 19, this volume). Many countries have legislation regulating language choice in the public domain (Angermeyer, Chapter 7, this volume). As with the clinical practice example, the actual language abilities of the people engaged in public domain communication will usually not be

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

332

ANNICK DE HOUWER

relevant. For instance, protocol normally dictates what language political leaders and important officials are to use in official statements in the public domain, thus precluding, for instance, Russian leader Vladimir Putin from speaking German at a press conference in Germany, even though he speaks fluent German (and even volunteered to translate a German guest’s German into Russian at a public event in Russia on April 7, 2016; see www.rt.com/news/338957-putin-translates-germanwimmer/; see further Section 8 on language choice in the political arena). Governmental institutions everywhere may impose their own rules concerning language choice. For instance, in a Russian prison, a Tajik inmate’s family members who were not able to speak Russian were not allowed to speak Tajik when they were visiting the inmate (Honigsberg, 2014). Official rulings and laws about language choice form part of language policies (Lo Bianco, Chapter 8, this volume) and relate to language attitudes and ideologies (Fuller, Chapter 6, this volume). Many schools have implicit or more explicit language policies that dictate what language(s) are allowed to be used in the classroom and elsewhere in the school. Teaching staff may actively try to enforce these policies, as in the example of a preschool teacher at a Dutch-speaking school in Flanders, who told me without any shame or guilt in 1995 that she had finally had to punish a four-year-old, who after being warned three times still was speaking French to a friend during the break. She locked the child up in a dark closet so he could reflect on his “sins” (primary school teachers in Flanders hold similarly negative views toward Turkish, see Ag˘irdag˘, Jordens, & Van Houtte, 2014). Such cruel actions (and worse) are likely still happening in many places in the world today. The ignoring of linguistically diverse students’ home languages or, indeed, punishing them for using these signals to children that their home languages are worth less than the school language(s) and may make children feel excluded (De Houwer, 2015b; see further Bigelow & Collins, Chapter 2, this volume; see also Garcı´a & Tupas, Chapter 20, this volume). Norms for language choice that are not dictated by top-down policies may likewise relate to attitudes toward specific languages and likely also to language ideologies. For instance, Belgian Dutch–French bilingual customers who were raised with Dutch might prefer to speak Dutch rather than French in service encounters in officially bilingual Brussels. When they do, however, they open themselves up to possibly hostile reactions, given that Brussels is a linguistic battleground where people have very different attitudes toward Dutch and French. Such hostile reactions are extreme examples of language socialization processes. For Dutch–French bilinguals with Dutch as a first language (L1), language choice in Brussels thus becomes a highly strategic, even politicized, affair. Furthermore, politeness considerations as modeled in Brown and Levinson’s (1987) politeness principle may interfere with preferred language choices, resulting in L1 Dutch customers’ surprisingly more frequent use of French than Dutch in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

333

service encounters in Brussels (Danblon, De Clerck, & van Noppen, 2004). Specifically, given that L1 French speakers in Brussels are less likely to have adequate levels of Dutch proficiency than the other way around, L1 Dutch speakers who accommodate French speakers in Brussels by speaking French to them can be seen as trying to save their French-speaking interlocutors’ positive face, at the expense of their own negative face. In doing so, they are styling themselves, that is, they are constructing a particular social identity for themselves (Preece, 2016), in this case, one of being easygoing, nice, and pleasant. (In fact, all strategic language use represents a type of styling.) What is more, customers are in a way concealing their identity as a speaker of Dutch, temporarily and strategically shifting to covert bilingualism (for an insightful discussion of such covert bilingualism, see Hult, 2014). In a language-political climate that is less conflictive, the dynamics of customer language choice may be different (Holmqvist, van Vaerenbergh, & Gro¨nroos, 2014). These convergent choices that are often seen in interactions between Dutch L1 customers in Brussels who also speak French contrast with what happens when the same Dutch L1 speakers interact with international visitors in officially Dutch-speaking Flanders. When non-Dutch-speaking visitors to Flanders attempt to speak Dutch, Dutch L1 speakers will often switch to English, French, or German; they may even try some Spanish or Italian, depending on how they assess the visitor’s origin. In doing so, they are not using the same language they were addressed in. This can be seen as non-accommodation from the visitors’ perspective. Some visitors experience this change to another language as a face-threatening act, that is, as an impolite rejection of their use of Dutch (see, e.g., www.flanderstoday.eu /living/no-textbook-needed, last accessed February 1, 2018). From the Dutch speakers’ perspective, though, they are being accommodating by using a language they feel both interlocutors will be able to use with ease and by sparing the visitors the difficult task of producing correct Dutch utterances (I am here expressing what can be assumed to be the underlying ideology, which I do not necessarily share, even though I am a Belgian L1 speaker of Dutch myself). It appears that Dutch speakers in Flanders are claiming their language just for themselves. The ambiguous dynamics of accommodation and non-accommodation as indexed in language choice in Flanders are not unlike those described for Japanese by Siegal (1996). Siegal reports on an interaction between a learner of Japanese, Mary, and her male Japanese professor at a university in Japan. Mary was in her mid 40s and the professor in his late 30s. Mary was a teacher of Japanese in New Zealand who was in Hiroshima on a fellowship to improve her Japanese proficiency, which on the basis of tests was judged to be at an advanced level. However, Mary found advancedlevel courses in Japan too difficult and took them at the intermediate level. During the Japanese interaction, she was consistently pragmatically inappropriate, but her professor did not correct her, leaving Mary unaware of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

334

ANNICK DE HOUWER

pragmatic infelicities, thus keeping her from the chance to improve. The same professor also persistently switched to English with other Western women in other data Siegal collected. Siegal attributed the professor’s interactional behaviors to face considerations, but also to widely held low expectations in Japan that foreigners can learn pragmatically appropriate Japanese. This became a self-fulfilling prophecy because, in addition to Japanese people not speaking Japanese to foreigners in naturalistic settings, at the time of Siegal’s writing, Japanese courses for foreigners offered few possibilities for learners to learn and practice the Japanese honorific system necessary for polite face-to-face interaction. Some foreigners may, however, speak excellent and pragmatically appropriate Japanese. Even in interactions with such foreigners, Japanese speakers in Japan switch to often “bad” English (Coulmas, 1987). Coulmas attributes this ubiquitous switching to four factors that converge with Siegal’s (1996): (i) Japanese people are historically not used to the idea that foreigners might in fact be able to learn Japanese, (ii) there is a low tolerance of foreign varieties of Japanese, (iii) Japanese speakers are “willing to spare foreigners the trouble of trying to make themselves understood in Japanese” (Coulmas, 1987, p. 104), and (iv) speaking English avoids getting into socially awkward situations because foreigners may speak Japanese in pragmatically inappropriate ways. All these points may be said to apply to Flanders as well. They certainly apply to the use of Gaelic in the Western Isles region of Scotland. McEwan-Fujita (2010) carried out an anthropological study with English L1 speakers who after settling in the Western Isles as adults had learned Gaelic up to various degrees of proficiency. In this region, Gaelic is still used by non-immigrated adults who have learned to speak it from early childhood onward and who also speak English, but Gaelic is undergoing strong language shift. McEwan-Fujita describes how, generally, nonimmigrated adults in the Western Isles did not speak Gaelic to people (including older adults) who did not learn Gaelic as children. This not only limited these adult learners’ opportunities for Gaelic language practice, but at the same time socialized learners into not using Gaelic in public spaces, thus further contributing to language shift. Also, like would-be users of Dutch in Flanders, the adult learners of Gaelic experienced the use of English to them by Gaelic non-immigrated adults as an impolite rejection of their own willingness and desire to speak Gaelic.

5 Language Choice and Social Meaning The contours of language choice with foreigners in Flanders, Japan, and Scotland may reflect protectionist in-group language ideologies (see Fuller, Chapter 6, this volume, on language ideologies). In any event, the examples in Section 4 show that language choice is never neutral. Language choice is always embedded in a particular context and creates a social

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

335

meaning within that context. For instance, Rampton (e.g., 2014) showed how adolescents fluidly negotiated their ethnic and peer group belonging through adopting a style of speaking in which they incorporated elements of the language use normally associated with other people (language crossing, see Section 3). Even much younger individuals already give social meaning to particular ways of speaking. For instance, 14-month-old infants, who understand language as used in very contextualized informal conversations but do not yet use much language themselves, already categorize people on the basis of whether these speak a language they have heard or not, and are more prone to imitate actions by someone who speaks a familiar language (Buttelmann, Zmyj, Daum, & Carpenter, 2013). This suggests that from very early on humans have a deeply seated desire to do things the way perceived group members do. In fact, this deep desire may lie at the basis of people’s overall tendency to verbally accommodate others. Social meaning relating to language choice is created against the quite universal backdrop of communication accommodation and politeness considerations that operate within a more locally defined sociopolitical and sociocultural environment in which official dictates (if any), attitudes, and ideologies will influence specific linguistic choices. So far not included in this complex equation is the emotional, relational component. Three-year-old Dutch–English bilingual Susan cried when her mother inadvertently addressed her in English, a language she had never spoken to Susan before (De Houwer, 2009). Over the years of their relationship, many married couples with at least one bilingual partner continued to use the language or languages they first happened to use with their partner (Piller, 2000). That was the case even if the partners’ language repertoires changed, and a previously monolingual partner became proficient in the bilingual partner’s first language. Many people interviewed for Piller’s study mentioned that changing language choice patterns would change the relationship. The language choice habits built up in a personal relationship, then, help lend meaning to that relationship. Pizer, Walters, and Meier (2013) report on the views of 13 adult hearing children of Deaf parents in relation to language choice. All the interviewees attributed strong social meanings to the extent to which they themselves, their siblings, and their parents used oral language (English) or sign language (American Sign Language), or a combination of both (for code-blending, see Tang & Sze, Chapter 24, this volume). For instance, many disapproved of hearing siblings who, according to their accounts, did not invest sufficient effort in using sign language with their signing, hard-of-hearing parents for communication to run smoothly and instead used only speech; the interviewees experienced this divergent language use as showing a lack of care. Indeed, in the family domain, the frequent or sole use of divergent language choices may create negative feelings. Parents of children in bilingual families often feel betrayed by their preschool-aged children if the latter do not speak the language parents speak to them

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

336

ANNICK DE HOUWER

(De Houwer, 2017b). Adolescents who speak English with their parents while their parents speak another language with them feel more emotionally distant from them than adolescents who converse with their parents in the same language (Tseng & Fuligni, 2000). In contrast, in the business or public domain, divergent discourse choices may be an asset and allow more easy communication between adult bilinguals who are not proficient speakers of two languages but who understand two languages well (e.g., ten Thije & Zeevaert, 2007). The force of habitual language choice can conflict with other norms for language choice. Many feel it is impolite for people in triadic or smallgroup conversational settings to speak a language that not everyone can speak or understand (cf. the general CAT norm discussed earlier). If participants in such interactions have already built up a habitual language choice relation, they will have to decide whether to stick to their own patterns of language choice or whether to adjust to a third party who cannot participate in those patterns. Social meaning will be attached to whatever option interactants select (cf., for example, Pizer et al., 2013). It is generally acceptable to adults who have developed habitual language choice patterns with each other to temporarily suspend them when others are present who do not share those patterns. An example, based on personal experience, is when family members who are used to just speaking an Antwerp dialect with each other speak in a more standard way at family parties that include family members who do not speak the dialect (the dialect speakers may still switch to the Antwerp dialect in personal asides). Continuing to speak the Antwerp dialect throughout would be seen as impolite, even though the other family members understand the dialect. There are many examples of bilingual schoolchildren who do not wish to speak their home language to their mothers when other schoolchildren might overhear them, but who switch to using the home language once they are out of earshot of others (De Houwer, 2017b). Such negative attitudes may be due to young bilingual children’s experiences at school with teaching staff’s language socialization practices signaling different norms for language choice than the ones they know from home (see Section 4). Schools can play a role in fostering and maintaining positive attitudes toward home languages by allowing and encouraging the use of any language at school (Garcı´a & Tupas, Chapter 20, this volume). Swanwick, Wright, and Salter’s (2016) study of language use by five deaf multilingual children in interaction with one of their parents at school shows the children choosing the language(s) they would use with their parent at home. This may be thanks to the fact that their deaf education school recognized and worked with the full communicative repertoires of all involved. A study by Byers-Heinlein, Behrend, Said et al. (2017) of fourto six-year-old bilingual and monolingual children’s social preferences for adult bilingual vs. monolingual speakers found that monolingual children

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

337

preferred monolinguals, whereas bilingual children showed no such preference. Such early social language-based preferences may explain why children who do not (yet) speak the school language or do not (yet) speak it well may become the victim of bullying by monolingual children (von Gru¨nigen et al., 2012). Bullying may also occur in cases where children newly arrived at school do not use the same regional accent as the other children (Elwert, 1959). Language choice and violations of norms for language choice, then, carry much symbolic, social meaning. These norms are not set in stone, but can be negotiated in interaction (Auer, 1998; Gafaranga, 2010). Through their language choice, interlocutors negotiate power relations and their social identity (e.g., Hua, 2008). This brings us back full circle to language socialization processes. All of us have been socialized through language from soon after we were born. Our very first interpersonal relationships were child–caregiver relationships. Through these child–caregiver relationships, we were socialized into using language in a particular way while at the same time we learned to speak or sign. Even though not all child–caregiver relationships are emotionally satisfying, we generally build up deep emotional connections with our first caregivers, and likely with their language(s), even though the latter can become conflicted as well (e.g., McEwan-Fujita, 2010; Tseng & Fuligni, 2010). At any point in life, we may find ourselves confronted with new socialization contexts through which we are expected to develop new ways of speaking (e.g., through being immersed in a new language at preschool, or through moving to another country). In McEwan-Fujita’s (2010) study, a few L1 speakers of Gaelic “took some pleasure in assisting adult Gaelic learners” (p. 57) and bolstered positive affect toward adult learners’ imperfect command of Gaelic, countering the much more frequent practice of other Gaelic L1 speakers who contributed to learners’ linguistic insecurity by refusing to speak Gaelic to them. These socialization-cum-learning contexts will have a different emotional charge than the ones we met up with in our earliest years (Harris, Gleason, & Ayc¸ic¸egi, 2006), giving rise to different emotional stances toward the languages we have learned. Thus individually experienced language socialization contexts contribute to the development of personally held attachment relations toward a particular way of speaking (or more than one). These affect-based attachment relations, combined with an increasing realization of sociocultural norms as individuals grow older, are likely a major foundation for the emergence of specific language attitudes. Hult (2014) discussed and analyzed several examples of his own language choice in Sweden at a time when he was doing fieldwork there as a visiting scholar from the United States. Hult is a highly proficient Swedish–English bilingual. Because of his language learning history, the kind of Swedish he speaks is a local variety that indexes a particular town in Sweden. However, up until the time Hult was a visiting scholar in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

338

ANNICK DE HOUWER

Sweden, he had mostly lived in the United States. This meant he was not familiar with certain mundane things such as bus routes, and he had met up with frustration, annoyance, or anger when he had shown his lack of competence in such matters while speaking his local kind of Swedish. His social gaffes did not match the Swedish local insider identity that others attributed to him. In order to save himself further embarrassment of this kind, Hult decided to speak English at the Swedish university library when he needed to find out how to obtain library materials. In using English, Hult was aiming to present himself as a foreigner, for whom it would be perfectly legitimate not to know how the library worked. The help desk person knew enough English to help him out, and there was no need for Hult to feel embarrassed. This example shows many different things. I chose it primarily because it shows that language choice creates social meaning not only as interpreted by others but also by oneself. Hult’s strategic use of English was only possible because of his awareness of the effects his language choice had both on other people and on himself. In this case, his choice of English despite being perfectly capable of using Swedish satisfied his own need not to feel ill at ease. Avoiding feeling ill at ease might have been an additional reason for the choice of French by L1 Dutch speakers in the service encounters in Brussels mentioned in Section 4. Regardless of the social meanings that language users construct for themselves or that others construct for them, in actual interaction, bilinguals must have the individual cognitive means to interpret unilingual and mixed utterances as well as changes in language choice; when they themselves produce language, they must be able to select the way of speaking they find appropriate in a particular setting, that is, one that takes into account the social norms for language choice discussed earlier. The following section explores these more psycholinguistic aspects of language choice.

6 The Psycholinguistics of Language Choice: Language Mode and Language Orientation At any point in face-to-face, real-time language comprehension, bilinguals could be hearing utterances in a language they do not understand, unilingual utterances in either or any of the languages they do understand, or mixed utterances. Depending on the specific communicative setting, bilinguals will have specific expectations about which kinds of utterances they will be hearing. They may be taken by surprise when addressed in a language they did not expect (e.g., Grosjean, 2001). They may not even understand a language they did not expect, even though they are very well capable of understanding it (De Houwer, 2009, pp. 117–118; see further later in this chapter). I have explained these phenomena through the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

339

concept of language orientation (De Houwer, 2009, pp. 116–119), which is a psychological state of attention toward a particular way of speaking in a particular sociolinguistic setting. Grosjean (e.g., 2001) explains a bilingual person’s readiness to perceive (and speak) a particular language or any of two or more languages on the deeper basis of neural activation, which he has called language mode. While there is much overlap between language mode and language orientation, in the sense that both involve the usually unconscious processing of language based on various sociopsychological and sociolinguistic factors, the notion of language orientation does not make any claims about neurological underpinnings. These are extremely difficult to study in realtime social interaction. Language orientation in comprehension can in principle be examined through analyses of clarification requests and hesitations (and other signs of not understanding) in real-time bilingual interaction. In combination with interview data on expected language choice from the same bilingual participants, such analyses can yield insights into the nature and role of language orientation for adequate comprehension in bilingual interaction. Both language orientation and language mode operate on a continuum from bilingual to monolingual. In a bilingual mode, bilinguals expect to communicate in two languages; in a monolingual mode, they expect to communicate in just one. As Grosjean (e.g., 2001) has repeatedly emphasized, researchers carrying out experiments with bilinguals should pay much more attention to respondents’ language mode. Ignoring the role of language mode in the experimental design may influence the way the results can be interpreted. This is already true for studies with young infants. For instance, if one is interested in studying whether monolingual infants are able to discern specific phones in the same way as infants raised with two languages from birth, care must be taken to test the latter twice: once in a completely monolingual setting where infants are greeted in testing language X and where everything else that they hear is just in language X, and once in another completely monolingual setting in language Y. If bilingual infants are tested in a setting where both their input languages are used, we must assume they are in a bilingual mode, which renders comparisons with monolinguals baseless. Language orientation and language mode also play a role in real-time speech production. In interactions with people they do not share all their languages with, bilinguals who can speak two or more languages will limit themselves to a single language, namely the overlapping one that their interlocutor understands (e.g., Baetens Beardsmore, 1982). This indicates a monolingual language orientation. Such a monolingual language orientation may be accompanied by the neural activation of just a single language, that is, a monolingual language mode, although according to Grosjean (e.g., 2001), the other language(s) a bilingual knows are never fully deactivated (for discussion of bilingual language processing, see Paap,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

340

ANNICK DE HOUWER

Chapter 22, this volume). A bilingual language orientation (and bilingual language mode) is more likely in bilingual interactions involving participants who are able to use the same set of languages. Given that here there are no a priori limitations on language understanding, these interactions are much more flexible in terms of what the language choices can be. Within the same stretch of discourse, changes from unilingual utterances in language X to unilingual utterances in language Y to mixed utterances and back can be made without any loss of comprehension, if speakers both operate in a bilingual language orientation. A monolingual language orientation for at least one of the interlocutors is still possible here, though. Take the example of two good friends, Tim and Tanya, who usually speak German together, but who are both quite proficient in Spanish as well, and who are both aware that the other friend speaks good Spanish. If, contrary to habit, Tim says something to Tanya in Spanish, she may at first not have understood, given that she is expecting him to speak German, and she may ask him to repeat himself, after which, with more focused attention, she will likely understand his Spanish (Tim is, however, more likely to switch back to German, following the habitual choice norm built up with Tanya). Generally following the convergent choice principle and language orientation are not necessarily congruent. For instance, as described in Section 3, although regardless of interlocutor Dutch–English bilingual Kate followed the CCP and more often than not responded in the language she was addressed in, she showed a difference in language orientation depending on her interlocutors’ linguistic skills and on setting. Specifically, Kate nearly always initiated conversations with her English-speaking mother in English (Kate’s mother understood some Dutch but spoke it very poorly; in fact, Kate chided her mother for making errors in Dutch). However, in initiating conversations with her Dutch-speaking but proficient bilingual father and babysitter Kate used mainly Dutch, some English, and an occasional mixed utterance (De Houwer, 1983). Kate’s near exclusive use of a single language with her mother shows a monolingual language orientation; her relatively more frequent use of two languages and mixed utterances with her father and babysitter shows a more bilingual language orientation. In De Houwer (1983), I explained these different ways of speaking in terms of language monitoring: Kate did not feel the need to monitor her language choice quite as much with her father and babysitter, who she knew to be proficient bilinguals, whereas her mother was not a proficient bilingual speaker. Kate was also able to change from one monolingual orientation to another: when she was four, she changed from speaking just English with me at her English-medium preschool to speaking just Dutch once we had left the school (De Houwer, 2009, p. 118). Indeed, regardless of who they are speaking to and regardless of the specific choices available in a particular interaction, bilinguals who speak two languages are able to easily and fluidly change from producing

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

341

utterances in one language to producing mixed utterances or unilingual utterances in another language (e.g., see many examples from Spanish– English-speaking children in McClure & Wentz, 1975, and see examples from five deaf multilingual children in Swanwick et al., 2016, who were able to switch seemingly effortlessly between sign language(s) and several spoken languages). Mixed utterances are usually also produced quite fluidly, without extra hesitations. This is seen as a hallmark of high bilingual proficiency (Poplack, 1980). There is a large experimental literature that seeks to explore the level of activation of bilinguals’ two languages in processing and (neuro)cognitive correlates of their ability to switch between languages. However, many of these use reading paradigms or picture-naming tasks and typically do not study language use in social interaction (see further Paap, Chapter 22, this volume). A notable exception here is Festman’s (2012) study, which in addition to several non-interactive tasks also included a long interview between Russian– German-speaking adult bilinguals and each of two unfamiliar interviewers that was designed to elicit both changes in language choice and the use of just a single language within five-minute chunks of time. The study also collected extensive language background and attitudinal information, including information about how participants generally felt about switching languages. Festman found inter-individual variation among participants in the extent to which they exhibited a more bilingual rather than a more monolingual orientation (my terminology) in the interviews. She explains these differences through conflict resolution and monitoring abilities that were found in a separate bilingual picture-naming task, and attributes participants’ sociolinguistically more appropriate monolingual orientation to higher levels of cognitive control. The study of self-repairs in natural conversation can lend insight into the planning processes underlying language choice during real-time production. Self-repairs in environments where speakers change the language of an utterance midway or between utterances are, however, quite rare. Often those self-repairs constitute the point at which a change in language choice is made (Hennecke, 2013). More detailed corpus studies of bilingual interaction are needed to further explore the use of self-repairs.

7 Bilingual Interaction in Young Bilingual Families Very young children growing up with two languages from birth in the home can be socialized into speaking both languages, or into speaking just one. Lanza (e.g., 1997) has uncovered some of the socialization practices with the very youngest of bilinguals that are relevant here. She proposed a continuum of parental discourse strategies ranging from monolingual to bilingual ones. Monolingual strategies encourage children to use just a single language within a given exchange. Examples include parents

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

342

ANNICK DE HOUWER

asking children (in language Y) to repeat what children just said in language X. Often children will interpret this as a request to change languages, and they will often repeat their original utterance in a translated version. Parents may also repeat children’s utterance by offering its translation. Often children will then repeat this translation. In using bilingual strategies, on the other hand, parents do not pay any specific attention to children’s language choice. For instance, they may just continue speaking language Y to children, even if children answer in language X. Another bilingual strategy occurs when parents switch to using utterances in language X after children used utterances in language X in response to parental utterances in language Y. Monolingual discourse strategies, if used in two languages, encourage children to speak two languages. Bilingual discourse strategies allow children to use any language, including just a single one. Together with other aspects of the input to very young bilingual children (see Armon-Lotem & Meir, Chapter 10, this volume), the use of monolingual vs. bilingual interactional strategies can likely help to explain why in many bilingual families (viz., in 25%) children speak only one instead of two languages (De Houwer, 2009). In such cases there is intergenerational transmission (in production) of just a single language rather than two. With continued exposure to two languages X and Y, children will, however, likely continue to develop comprehension in two languages. The start of (pre)school has often been identified as a turning point in home language choice patterns in bilingual families, with children who until then had spoken two (or more) languages at home suddenly limiting themselves to just the school language (De Houwer, 2009). Most likely, the language attitudes that children are confronted with at school (Section 4) play a role in this change in language choice. Many young bilingually raised children, however, never really start speaking two languages, but speak just a single one. The sole language that bilingually raised children speak is usually the language used in the broader community and at (pre) school (De Houwer, 2009). Minority-language families who live in school districts that offer choices in terms of which language(s) children are schooled in may have a better chance at intergenerational language transmission (Slavkov, 2017). Many bilingual parents switch to only addressing children in the language X that the latter happen to speak (De Houwer, 2017b), thereby effectively creating a monolingual environment in which children will no longer continue to develop bilingual skills in language Y as well. By switching to using language X instead of language Y to children speaking language X, parents are following the convergent choice principle. The power of this principle is such that it is self-evident for most bilinguals who are able to speak two languages: if they are addressed in one of the languages they speak, they automatically speak that language back. Parents may not even realize that they are doing this. They may also not

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

343

realize that through doing so, they are in the long term no longer offering children the opportunities for learning language Y. Thus there will be a lack of intergenerational language transmission within families. This greatly upsets many parents from different cultural and socioeconomic backgrounds all over the world, from Papua New Guinea (Kulick, 1992) to North America (Iqbal, 2005; Kennedy & Romo, 2013) to Europe (De Houwer, 2017b), and often leads to communication difficulties with family members outside the nuclear family. As Fishman (1991) famously claimed, the lack of intergenerational language transmission within families may have repercussions for the continued use of a language in the wider community, and may ultimately lead to language loss at the societal level. Kulick’s (1992) study in Gapun, a village in Papua New Guinea, has laid bare the importance of language choice in caregiver–child interactions in bilingual families, with parents lamenting the gradual loss of the local language, Taiap, in favor of the lingua franca, Tok Pisin. Parents did not understand why their children almost exclusively spoke Tok Pisin. The answer may lie in the fact that in addition to not speaking much to young children (see Armon-Lotem & Meir, Chapter 10, this volume, on the role of input frequency in early bilingualism), parents in Gapun tended to respond to their children in Tok Pisin, thus limiting their children’s opportunities to learn any more Taiap. Kulick explains how Gapun parents believed that their children could not be taught anything but had their own strong will, and that one could not do anything against that. As I proposed in De Houwer (1999), this position led parents in Gapun to what I have termed a no-impact belief, that is, a belief that they could not influence their children’s language development and use. Using monolingual interaction strategies, then, in which parents attempt to control their children’s language choice, and that will support early bilingual development if used in separate language contexts, was not an option in Gapun. As Pe´rez-Ba´ez (2013) shows in her longitudinal study of the gradual loss of Zapotec in Oaxaca, Mexico, parents’ weak-impact beliefs undermined their ability to increase the use of Zapotec in their families and to counter the ever-increasing influence of Spanish. Pe´rez-Ba´ez stresses the great importance of the field of family language policy for better understanding processes of language shift and language loss (Gonc¸alves, Chapter 3, this volume). Weak or nonexistent parental impact beliefs are not limited to non-industrialized contexts. As described in De Houwer (2009), even highly educated North American and European parents may hold such beliefs. Even if parents have a strong-impact belief regarding language, that is, they believe that their socialization actions can affect their children’s linguistic behaviors, they may not realize how important frequent language input is for children to be able to learn to understand and speak a language. Also, it requires much conscious effort for bilingual parents not to follow the convergent choice principle, an effort that many

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

344

ANNICK DE HOUWER

parents with young children and busy jobs and/or difficult life circumstances may simply be too overburdened to put into practice. Furthermore, as van Tuijl, Leseman, and Rispens (2001) have demonstrated, culturally defined childrearing practices may stand in the way of parents interacting with children in ways that are known to optimally support early language development. Finally, parents may live in settings that generally do not encourage the use of a particular language Y, and hence may feel it is too much of an uphill battle to try to do the emotionally difficult socialization work of trying to have children speak language Y (statements to this effect appeared in many of the European studies reviewed in De Houwer, 2017b). Just how difficult it can be to manage or control children’s language choice is shown in a study by Kheirkhah and Cekaite (2015) of a multilingual family in Sweden with two school-aged daughters. The family’s younger daughter, Mona, had been raised with Kurdish and Persian from birth, and had started to learn Swedish through preschool from 1.5 years of age onward and later through school. The parents were very keen to have their daughters speak all three languages, but especially Persian and Kurdish. Mona spoke mainly Persian with her mother, mainly Kurdish with her father, and Swedish with her older sister. However, by age seven, Mona was using many Swedish words in her Kurdish and Persian, or saying short Swedish utterances instead of fully Kurdish and Persian unilingual utterances. Her parents would insist (often playfully) that Mona translate the Swedish word or phrase she had used into Persian or Kurdish, depending on whom she was talking to. Also, the parents often said they did not understand Mona when she spoke Swedish to them. While such monolingual strategies often work well with much younger children (e.g., Juan-Garau & Pe´rez-Vidal, 2001; Kasuya, 1998), as a primary school-aged child Mona was quite reluctant to comply with her parents’ requests for translation, said she could not speak Persian or Kurdish well, or just left the interaction, leaving her parents stranded, as it were. The politeness principle likely played a role here: as children grow more and more into autonomous individuals, they do not necessarily wish to be reminded that they are considered imperfect users of a language. Parents asking their somewhat older children to translate something that they probably understood attacks children’s negative face. Finding the balance between adhering to politeness principles and parents’ desire for their maturing children to speak the heritage language is not easy.

8 Bilingual Interaction in the Political Arena Patterns of language choice may symbolize both power and solidarity. This is particularly clear in the political arena, where protocol often demands particular language choices in ritualized settings such as international press conferences, sessions of the United Nations, or official meetings at

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

345

the European Commission. These choices can be “pushed through” because of general acceptance of that protocol (or applicable legislation), and are recognized through the making available of interpreters. As already briefly illustrated in Section 4, in less ritualized or scripted settings, political figures and officials have more freedom in their language choice. For instance, former US Republican presidential candidate Jeb Bush was interviewed on the American Spanish-speaking television c h a n n e l Te l e m u n d o i n J u l y 2 0 1 5 ( w w w . y o u t u b e . c o m / w a t c h ? v=ns_dZuoQciM, last accessed February 18, 2018). Apart from the topics he was discussing, his language choice was notable: Jeb Bush spoke fluent Spanish in the interview. A subsequent article in the National Journal archived on the website of the magazine The Atlantic on August 3, 2015 (retrieved from www.theatlantic.com/politics/archive/2015/08/how-welldoes-jeb-bush-habla-espanol/432456/, last accessed January 27, 2018) makes the link between Bush’s language choice and gaining more Latino votes (the article referenced here used the term “Latino”; I am aware there are alternative terms). The article also points out how Bush did not use any “Mexican slang,” except for nin˜o popis, or “Cubanisms,” but “kept his Spanish mostly neutral” during the interview. The implication is that in speaking the way he did, Bush was trying to appeal to a nationwide audience of Latino voters. Starting from a theoretical position including reference to accommodation theory, Alamillo and Collingwood (2017) showed that Anglo-American political candidates like Jeb Bush who engaged in what they called social identity bridging through, among others, language choice, can indeed garner much higher shares of the Latino vote than Anglo candidates with similar policy stances who did not so engage. The symbolic value of politicians’ language choice in settings not controlled by protocol was also in focus in a July 2016 comment made on the German television channel ZDF by German-Turkish actor and comedian Fatih C¸evikkollu, who said that the Turkish-origin population in Germany felt like second-class citizens, and that to make that feeling less poignant, maybe then Chancellor Merkel could say something in Turkish perhaps – for instance, she could say Ihr seid willkommen [you’re welcome] in Turkish. C¸evikkollu’s choice of phrase was likely related to the newly coined word Willkommenskultur [welcoming culture] that infused much of Germany in the second half of 2015 when then Chancellor Merkel opened the borders to Syrian refugees. Of course, before then, Germany had already received many immigrants from many other countries, leading to new dynamics between linguistic practices, new political discourses, and new demographic constellations (Stevenson, 2015). The strategic use of the symbolic power of language choice was, arguably, a factor in the establishment of the Eurozone. France and Germany were the main players. In 1988, Frenchman Jacques Delors was made chairman of what later became known as the Delors Committee. Many committee members, including the head of the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

346

ANNICK DE HOUWER

German central bank (Deutsche Bundesbank), Karl Otto Po¨hl, were reluctant to accept Delors as chairman. Delors was well aware of this and wanted to do something about it (Mazzucelli, 1997, p. 52). As protocol demanded, and regardless of members’ proficiency in other languages, each member of the committee was speaking his country’s or region’s official language (there were only male members; cf. www.cvce.eu/content/publication/1999/10/8/2be3e1de-1e9b-47eab7c6-cbdc70dcd460/publishable_en.pdf, last accessed January 29, 2018). The languages that other members spoke were being simultaneously interpreted, so members were using headphones. This was quite cumbersome and created a social distance. The atmosphere at the beginning of negotiations was very tense. Delors then broke the tension by ripping off his headphones and suggesting everybody speak English (Mazzucelli, 1997, p. 52). This made a big difference, and especially the important player, Po¨hl, was mollified by the fact that a Frenchman would suggest speaking English at the negotiating table. The negotiations continued in an open and constructive spirit. While I have no information on the language proficiency of the committee members, we must assume that all members, who were bankers in senior management positions or high-ranking politicians, had sufficient English proficiency to take part in English negotiations, and that thus Delors’s initiative did not conflict with the general principles underlying language choice as explicable through Communication Accommodation Theory. Breaking protocol here had substantial symbolic power.

9 Conclusion A single chapter cannot do justice to the abundant and rich literature on the many different facets and complexities of bilinguals’ language choice in real-time interaction. My discussion has necessarily been selective. For instance, I have not been able to cover the growing literature on the way bilinguals use each of their languages differentially when it comes to expressing affect (for an overview, see Koven, 2017). Also, my discussion of bilinguals’ use of mixed utterances has been cursory. I did not discuss factors such as topic and setting that can help explain language choice among bilinguals who regularly use each of two languages with each other. Yet I hope this chapter has been able to explain the main features of bilingual social interaction as they pertain to the linguistic categorizations all bilinguals must make from the moment they are able to understand utterances in more than a single language variety, and the linguistic choices they must make from the moment they are able to produce utterances in more than a single language variety. As noted in Section 3, the range of language choice options that people have are fundamentally related to their linguistic repertoires in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

Language Choice in Bilingual Interaction

347

comparison to those of potential interlocutors. Furthermore, in settings where there are no official, top-down explicit norms for language choice that one must observe to avoid loss of public face, financial loss, punishment, arrest, or worse, language choice options are basically guided by the general principles for interpersonal communication through language as outlined in Communication Accommodation Theory (e.g., Giles, 2016) and the politeness principle (Brown & Levinson, 1987). This explains why bilinguals who can speak two languages X and Y will normally limit themselves to speaking either language X or language Y to monolinguals or other bilinguals who understand only language X or Y, respectively. This is true for children as well as adults. Bilinguals will have learned to act this way through being socialized into it. The tendency to linguistically accommodate and align one’s language use with that of others (and to lend social meaning to situations where this is not happening) is not only found in conversations at any specific time, but can also explain language choice patterns over a period of time: the history of past language choice patterns that language users have developed with particular people will continue to influence their later language choice patterns, thus following an interpersonal norm. Any deviations from previous language choice patterns will have social meaning. There appear to be no fundamental differences between age groups in their ability to use language choice in socially meaningful ways. Children are already able to position themselves socially through their language choices. As individuals gain more life experience and as the range of their linguistic practices grows wider, however, they may learn more different ways of speaking, which they then can incorporate in their language choice options. As Rampton (2015) has shown, stylistic choices are not unique to a particular age group: depending on who they interact with, not only adolescents but also middle-aged speakers may engage in language crossing, that is, in the adoption of language features that normally are used mainly by people with a different cultural-ethnic background. Such crossing and other patterns of language choice can be used in less or more conscious ways. Metalinguistic awareness may in fact play a large role in language choice, and appears to do so at all life stages. Avoiding social embarrassment because of perceived insufficient language proficiency in a particular language may keep people from using that language. Bilingual interaction, then, dynamically combines metalinguistic awareness, speakers’ language proficiency, and norms for utilizing this proficiency in real-time social interaction. Fundamental in shaping bilinguals’ ability to effectively interact in socially acceptable ways are their personal experiences with the specific language socialization contexts they have been engaged in. These personal experiences may lead to patterns of language choice that privilege one language variety and ultimately lead to family, community, and even societal language loss.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

348

ANNICK DE HOUWER

The longitudinal study of language socialization and bilingual interaction is key in our understanding of these processes.

Acknowledgments I thank Lourdes Ortega and Wolfgang Wo¨lck for insightful comments on earlier versions of this chapter. I am especially grateful to Lourdes Ortega for suggesting improved structuring of this chapter, specific sources, and for her keen editorial feedback.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 14 Nov 2018 at 02:56:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.018

18 First Language Attrition From Bilingual to Monolingual Proficiency? Barbara Köpke

1 Introduction When researchers started to investigate language attrition in the 1980s, the conception of what linguistic competence is and how it develops was very different from the one that prevails today. In second language acquisition (SLA) studies, the influence of the first language (L1) on the second language (L2) was of interest from the outset, and for a long time fully acquired L1 competence was assumed to remain stable, except in the event of a pathological deterioration caused by a stroke or degenerative disease. Sharwood Smith (e.g., 1983), one of the first scholars to acknowledge that the influence between L1 and L2 is reciprocal, preferred to talk about L1 influence rather than transfer, to avoid the latter’s behaviorist conception. It was he who developed the concept of crosslinguistic influence (CLI). But while this term is now widely used, accepting the idea that this influence is indeed crosslinguistic and not unidirectional has proven rather less straightforward. This is where research on language attrition comes in. Early studies, including a groundbreaking volume by Lambert and Freed (1982), were generally motivated by personal experience of L2 or foreign language attrition. The basic idea was that, just as people learn languages they are exposed to, they unlearn languages they are no longer exposed to (see Keijzer & de Bot, Chapter 14, this volume, on unlearning). This could be seen today as quite a modern view of things. As far as L1 attrition was concerned, however, most early publications actually dealt with traditional research questions concerning language shift and dialect death, and reflected sociolinguistic approaches to language contact (see Ko¨pke & Schmid, 2004, for more details). Research on L1 attrition at an individual level thus remained scarce, apart from Sharwood Smith (e.g., 1983), who was clearly a pioneer. Not until 1991, the year that saw the publication of another groundbreaking

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

350

BARBARA KÖPKE

volume by Seliger and Vago, were minds ready to address the specific issue of L1 attrition. While one of the earliest definitions of language attrition talks about “the loss of any language or any portion of a language by an individual or a speech community” (Freed, 1982, p. 1), more recent definitions use language loss as a generic term, and reserve the terms shift for intergenerational loss and attrition for intragenerational loss (de Bot, 2001). Accordingly, I here use the term shift with reference to bilingual communities, while attrition is a characteristic of individual speakers. Another aspect mentioned in many definitions of language attrition is its non-pathological nature, allowing us to distinguish language attrition in healthy bilinguals from acquired language disorders such as aphasia following stroke, head injury, or the onset of neurodegenerative disease. What seems to be common to all definitions of language attrition is (1) the idea of change manifesting itself as a decline in a speaker’s proficiency in the attriting language, and (2) the idea that the degree of decline in proficiency must be linked to nonuse or less frequent use of the language (see Ko¨pke, 2004, for an overview). It is, of course, rare for people who have been bilingual for a period of their life to become wholly monolingual. Even in the case of attrition of an L2 (see Keijzer & de Bot, Chapter 14, this volume) or another later learned language, complete monolingualism is the exception, and the experience of learning a new language, even at a very basic level, generally leaves its mark on the language system. In L1 attrition, monolingualism is even less likely. Something akin to a reversion to monolingualism may happen when, in early childhood, bilinguals find themselves in language contexts where one language is no longer present. Another possible case is when adoptees in international contexts quickly switch from L1 monolingualism to monolingualism in the language of their new family (see Keijzer & de Bot, Chapter 14, this volume). Very strong attrition has also been shown to occur in people who have experienced traumatic events (e.g., persecution) in one of their languages (Schmid, 2002). In most bilinguals or multilinguals, however, L1 attrition will, at most, lead to some kind of functional monolingualism, where speakers prefer to use their L2 because they no longer feel comfortable speaking their L1. Since the foundational publications by Lambert and Freed (1982) and Seliger and Vago (1991), research on language attrition has continued to make an important contribution to our understanding of the dynamics of language proficiency in bilingual or multilingual speakers (henceforth, following this Handbook’s perspective, I use the term bilingual to encompass multilingualism). The simultaneous and spectacular development of psycholinguistic studies of language processing in bilingual speakers has helped us gain a far more precise picture of what exactly language attrition is. Both the attrition and processing fields have clearly contributed to the conception of bilingualism as a dynamic and multidimensional concept

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

351

that is dependent on language use – a conception many researchers espouse today. In the present chapter, I expand on several of these contributions, focusing on L1 attrition in postpubertal migrants, i.e., in people who grew up monolingually and moved to an L2 environment after the age of around 12. First, I address some of the big-picture questions about the nature of language loss: why does attrition arise, what aspects of language are affected by attrition, and how does attrition take place? Next, I consider some of the methodological issues concerning the populations under investigation and the type of data considered. Finally, I discuss how these advances affect the definition of language attrition. All throughout this chapter, directions for future research are suggested.

2 What Questions Does Language Attrition Research Raise? The causes of language attrition have mostly been examined through three lenses. Many attrition researchers have interrogated the extralinguistic factors involved. The question of what is affected by attrition has been the emphasis of more narrowly focused linguistic investigations. Finally, the processes underlying attrition have been addressed in psycholinguistic research. I examine what is learned from each lens in turn.

2.1 Why Does Attrition Arise? One important question with respect to language attrition is why language proficiency changes. Researchers have probed the explanatory power of extralinguistic factors such as length of residence, frequency of L1 use, individual characteristics like educational levels, language attitudes, age, and more. Research on language attrition has long been based on the assumption that attrition results from lack of use, in conjunction with a long period since immigration (operationalized as length of residence, or LoR). This is why lack of use is often referred to as part of the definition of attrition itself (Ko¨pke, 2004). This is also why most studies have used an LoR of at least a decade, often longer, as an inclusion criterion, along with reduced L1 use. Rather surprisingly, however, empirical research has failed to demonstrate any direct relationship between the degree of attrition and either LoR or frequency of use. As I discuss in what follows, there are reasons to conclude that both LoR and frequency of use matter in attrition, just not necessarily in the ways initially assumed in the early research (see Armon-Lotem & Meir, Chapter 10, this volume, for discussion of the role of a related concept, length of exposure, in young bilinguals’ learning of a new language).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

352

BARBARA KÖPKE

The finding for LoR is particularly unexpected, as time is an important factor in frameworks such as the Activation Threshold Hypothesis (Paradis, 2007) or memory decay theories that have been put forward to explain how attrition arises (see Section 1.3 and Keijzer & de Bot, Chapter 14, this volume; on the role of familiarity with a language, a concept related to LoR, see Hernandez, Chapter 23, this volume, who, however, focuses mainly on aphasics). That said, findings do not necessarily indicate that time has no impact whatsoever. Rather, they show that there is no direct linear relationship between LoR and attrition. In other words, languages do not simply fade away if they are not used. Instead, time works in nonlinear ways, and there may be some periods in language development when changes are more likely to occur. The few longitudinal studies of L1 attrition that have been carried out in adults so far suggest that the first years of immersion are particularly vulnerable. This might also explain the results of a recent study by Kasparian, Vespignani, and Steinhauer (2017, see Section 2.2.4), which is one of the rare studies to have found an effect of LoR. These authors included participants with a mean LoR of 11 years, which is lower than in most L1 attrition studies. Clearly, more thorough investigations of timescales in attrition should be one of the priorities of research in the coming years. As far as frequency of L1 use is concerned, results have not been any more straightforward. Despite the development of tools to quantify frequency of use, many of the recent studies that have used these tools have found little or no impact of frequency of use (e.g., Cherciov, 2013; Keijzer, 2007). Ko¨pke’s (2004) comparisons of individual migrant profiles, on the other hand, suggested that L1 use in a migrant setting does not prevent attrition; far from it – in fact, high L1 use seemed associated with L1 attrition in that sample. The study by Kasparian et al. (2017) just mentioned yielded slightly different results for frequency of L1 use as well, reporting effects of amount of exposure on lexical-semantic processing in Italian–English bilinguals with an intermediate LoR. These were interpreted by the authors as reflecting the influence of experiential factors linked to type of L1 use. Thus, as with LoR, the mixed findings for frequency of L1 use are not to say that use does not matter. What the accumulated evidence suggests is that type of use may be much more important than mere frequency of use. A number of studies actually support this view. When Schmid (2007) compared uses of L1 in colloquial versus formal settings, she found that those migrants who used their L1 in colloquial settings with other bilingual migrants exhibited greater attrition. Meanwhile, based on their analysis of data from a case study, Mayr, Price, and Mennen (2012) suggested that convergence of the L1 system with an L2 may be induced by intensive code-switching. The importance of type of use suggests that attrition may be due in part to frequent coactivation and use of the languages in the same contexts, whereas language use contexts where each language remains completely

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

353

separate are less conducive to attrition. This account is also compatible with more recent findings by Schmid (2011a), who observed more attrition in migrants who used their L1 very rarely, but also in those who used it quite frequently. This observation can be interpreted as evidence for different kinds of attrition processes: in migrants with very little L1 use, the L1 becomes more and more inaccessible, whereas in migrants who frequently use the L1 with other migrants, frequent code-switching may lead to interference between the languages. Moreover, exposure to L1 input influenced by the L2 may favor the development of a contact variety. More recently, the interplay of LoR and frequency of use in the attrition process has again been questioned. It has been suggested that it is not so much the duration of a change in language use but its intensity that is decisive. Several studies suggest that periods of immersion in a different linguistic environment (even in the context of intensive language instruction) may be a factor leading to (sometimes temporary) restructuring of the L1 system at the level of phonetic categories (Chang, 2012) or syntax parsing (Dussias & Sagarra, 2007). Immersion contexts are characterized by intensive exposure to a new language with LoR being typically low or even inapplicable as in Chang’s study. Such studies demonstrate a strong influence of the immediate language use context. Similarly, in a recent single case study, Genevska-Hanke (2017) observed that the use of overt/ null subjects in the L1 of a Bulgarian–English bilingual migrant varied according to the country where the recording was done: Great Britain or Bulgaria. Further investigation of the immersion factor should clearly be a priority for attrition research. In sum, the investigation of extralinguistic factors like LoR and frequency of use indicates that attrition is not something that automatically arises when people spend long years in another country and only rarely use their L1. This realization has opened the way to a great deal of speculation about which individual characteristics may be conducive to attrition. Education level, attitudes toward the languages, general linguistic aptitude, and emotional factors, among others, have been investigated as independent variables in most of these studies. Two studies investigating attrition effects at the group as well as at the individual level (Cherciov, 2013; Opitz, 2013) contain in-depth discussion of the language proficiency of individuals performing outside the norm established with a control group, and report that only a small number of them actually exhibited perceptible L1 attrition. Results in these and other studies, however, are generally inconsistent with respect to most independent variables studied, and point to complex interactions of factors at the individual level. There is nevertheless one extralinguistic factor that is recognized as making a difference in all studies, namely age of onset (AoO) of bilingualism or attrition. There is broad and consistent evidence that children who are immersed in a new linguistic environment (i.e., prepubertal migrants) experience deep and persistent changes in the L1 system that may involve

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

354

BARBARA KÖPKE

complete forgetting of a language (see Keijzer & de Bot, Chapter 14, this volume, for discussion of attrition in heritage speakers and international adoptees). Nothing similar has ever been reported in the more profoundly entrenched L1 of people who migrated as adults (Schmid, 2014). This suggests that children and adults (a) follow very different timescales in the development of languages, and (b) have very different needs with respect to input and use. Keijzer and de Bot (Chapter 14, this volume) offer a review of the differences in attrition that result from age and comparisons of children versus older adults. However, even in prepubertal migrants, attrition varies according to the type of linguistic structures investigated, the topic to which I turn next.

2.2 Which Linguistic Domains Are Affected by Attrition? The issue that has probably motivated the largest number of studies, particularly from the linguistic side, is which aspects of language are affected by language attrition. First of all, researchers have asked whether L1 attrition involves a genuine loss of linguistic knowledge or whether it is a temporary problem affecting access to and processing of linguistic information. While Sharwood Smith (1983) proposed that attrition at the performance level might be a developmental stage, possibly leading to permanent changes in linguistic skills, research with adult migrants suggests that postpubertal migrants mainly suffer from changes in online L1 processing. In a recent study, Chamorro, Sorace, and Sturt (2015) compared Spanish monolinguals and Spanish–English attriters – some of whom were reexposed to their L1 before testing – in online and offline measures of antecedent preferences for subject pronominals. Results showed that attrition affects online sensitivity to interface structures, but this effect is dispelled by re-exposure (see Section 3.2.3. for more details on this study). These data were interpreted as evidence that no permanent changes take place in the knowledge representations of adult migrants, a view shared today by the majority of researchers studying L1 attrition in adults (see also the case study by Genevska-Hanke, 2017, described earlier). Furthermore, the use of the notions of competence and performance has changed substantially over the years, and more recent accounts tend to include them both, as linguistic representations are, in any case, only available through the study of processing. Another question that has been asked from the very beginning of L1 attrition research is whether some linguistic domains are more affected than others. Despite all the research on linguistic aspects of attrition, this question still has to find a satisfactory answer. Early research on attrition showed that lexical access difficulties are common in adult migrants, and attriters have also been shown to have a tendency to use a somewhat smaller range of vocabulary than monolingual control groups for the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

355

respective languages. These obvious difficulties at the lexical level are, however, perhaps only the most visible effects of attrition. While morphosyntactic errors in L1 production seem to be less frequent, many studies have found evidence of modifications in the processing of morphosyntactic information, as demonstrated, for example, through increased variation in the interpretation of binding properties in sentence comprehension shown by Gu¨rel and Yilmaz (2011). Recent studies in the phonetic domain have demonstrated assimilation patterns in bilingual migrants for the production of liquids (de Leeuw, Mennen, & Scobbie, 2013) and for voice onset time (Mayr, Price, & Mennen, 2012), sometimes emerging after only a few weeks of L2 learning (Chang, 2012). There is also a growing body of evidence in favor of restructuring in the semantic domain. This has been found for the conceptualization of time, events, motion, and emotion, to name but a few (for a recent review, see Schmid & Ko¨pke, 2017). Such observations have led to the claim that bilinguals tend to adjust their conceptual frames to accommodate all their languages as far as possible (Vanek & Hendriks, 2015; see Paap, Chapter 22, this volume, for further discussion of relations between bilingualism and conceptualization). However, in other linguistic domains, empirical research is still either absent (e.g., in the domain of pragmatics), or remains isolated, as in the study of gesture use, which has been shown by Brown and Gullberg (2008) to be sensitive to attrition effects (Keijzer & de Bot, Chapter 14, this volume, also briefly discuss the potential role of gestures in attrition). Finally, most studies have tried to find out whether some linguistic structures are more vulnerable to attrition than others. The selective nature of language attrition has never been called into question, and it has also been suggested that attrition takes place in sequences (Schmid, 2002, p. 6), similar to the developmental trajectories observed in L2 learners. The focus on differential attrition across linguistic structures is promising for the investigation of theories of linguistic structure. Many linguistic hypotheses, mostly within generative accounts, have been tested in the context of the question of differential attrition across linguistic structures. For example, researchers have investigated pronoun interpretation within the Principles and Parameters framework (Gu¨rel, 2004), or the interpretation of null subjects and use of pre- and postverbal subjects within a Minimalist approach (Tsimpli, 2007). Like in studies of SLA, a common starting point was the idea that areas where the two languages differ are particularly vulnerable to attrition. However, the findings of empirical studies suggest that attrition does not particularly affect strongly contrasting areas between languages. For instance, difficulty with gender marking and agreement are not reported in many of the studies involving L1 speakers of gender-marking languages who are immersed in an L2 English environment where they no longer use gender marking. By contrast, and resonating with similar findings for CLI on the L2 in SLA, research on L1 attrition (and other studies with bilingual participants) clearly shows that

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

356

BARBARA KÖPKE

similarities between linguistic systems facilitate interaction, and that language change preferentially occurs in areas where L1 and L2 show crucial crosslinguistic similarity – a conclusion compatible with different theoretical accounts (Schmid & Ko¨pke, 2017).

2.3 How Does Attrition Happen? The last big question refers to the nature of the psycholinguistic and neurological mechanisms underlying attrition. The mere assumption that languages attrite raises the question of the representation of language in memory. Strangely enough, however, few links have been drawn between studies of language attrition, mostly conducted by researchers in applied linguistics, and theories about forgetting provided by cognitive psychology. Studies investigating the forgetting of language skills have rarely gone beyond the learning of word lists. Instead, the neurological mechanisms underlying attrition have been theorized in frameworks based on the activation and inhibition processes intended to account for between-languages competition in bilingual language processing. Competition is, of course, also a major issue in attrition. Without delving further into the different theories of forgetting (interested readers may refer to Keijzer & de Bot, Chapter 14, this volume, and to Ecke, 2004; see also Hernandez, Chapter 23, this volume), one distinction arising from these theories has been repeatedly applied to language attrition and should be discussed here, namely the distinction between externally and internally induced language attrition. The former involves influence from the L2 (i.e., CLI), whereas the latter arises from internal language constraints (i.e., universal constraints such as markedness). The first authors to discuss external and internal processes in L1 attrition were Seliger and Vago (1991), who adopted a linguistic point of view. They considered that the main forces of externally induced attrition involve CLI such as rule generalization from one language to another, meaning extension or loan translation. Internally induced attrition, on the other hand, involves, for example, the replacement of irregular features by regular ones. Inspired by psychological theories of forgetting, several authors (e.g., Ammerlaan, 1996; Ecke, 2004) have taken up the idea of externally and internally induced attrition with a slightly different interpretation. Their definition of externally induced attrition is similar to that put forward by Seliger and Vago (1991) and based on interference theory, implying that interference is most likely from similar items of information or information systems: in the case of L1 attrition, such a system will be the L2. In this account, however, the main source of internally induced attrition is explained through memory decay, where owing to lack of rehearsal a memory trace is gradually extinguished. This is a less rule-governed process than that suggested by Seliger and Vago

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

357

(1991). What is common to both interpretations is that they stress the twofold origins of language attrition, that is, in the case of L1 attrition, lack of L1 use leads to modifications due to language and memory internal constraints, and interference arises from L2 structures and processing. What is specific to language attrition, and what makes it different from other kinds of forgetting, is the fact that it is not just that some kind of knowledge is no longer being used, but that it has been replaced by another symbolic system that fulfills the same functions in the bilingual’s life. This results in competition between the two language systems – a concept that was given prominence in early research on L1 attrition and that is now one of the pillars of research on bilingual processing. When in 1991 Seliger and Vago stated with respect to L1 attrition that “the languages spoken by the bilingual may be said, metaphorically, to coexist in a state of competition for a finite amount of memory and processing space in the mind of the speaker” (1991, p. 4), the idea that there is competition between a bilingual’s languages had only just emerged in Green’s (1986) model of control, activation, and resource. From then on, however, research on the neuropsycholinguistic aspects of L1 attrition and research on bilingualism were both based on theoretical accounts involving activation (e.g., the Activation Threshold Hypothesis, Paradis, 2007) and inhibition (e.g., the Inhibitory Control Model, Green, 1998; see Paap, Chapter 22, this volume, for extensive discussion of bilingual control models). The purpose of these models was to explain how the bilingual mind and brain manage competition between languages – an issue henceforth common to both attrition and bilingual processing research. With such a focus on competition between languages, two factors gained increasing prominence. First, this focus made clear that attrition is above all characterized by externally induced changes. This conception does not, however, exclude the idea that internal language and memory factors also play a role, as L2 influence may be particularly important in domains where internal language and memory constraints render the L1 particularly vulnerable to attrition. Second, based on the idea that competition from the dominant L2 is one of the main contributors to L1 attrition, frequency of L1 use (and even more importantly, quality of L1 and L2 use, as discussed earlier) should play a major role in determining attrition.

3 How Is Attrition Investigated? In this section, I discuss methodological issues in attrition research. An exhaustive consideration of these is beyond the scope of the present chapter, but some aspects are well worth considering, as language attrition research is particularly complex with respect to issues such as the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

358

BARBARA KÖPKE

comparison of presupposed attriters with other populations and the type of data to be collected.

3.1 Who to Compare? One of the major methodological issues in attrition research is the question of a baseline for comparison. In investigations of language attrition, the assumption is that attrition will be evidenced through a decline in performance between Time X (when the language was fully used and mastered) and Time X + 1. In most L1 attrition studies, X + 1 is situated several years after arriving in a different linguistic environment, and only migrants who have been living in an L2 environment for at least 7, 10, or even 15 years will be included (see also Section 2.1). Such timescales preclude any kind of longitudinal design in L1 attrition research, even though longitudinal research designs are frequently used in studies of L2 attrition. Hence, the most popular research design in L1 attrition studies is cross-sectional, with control groups that include L1 speakers living in an L1 environment. Differences in verbal behavior between monolingual control groups and bilingual migrants are interpreted as attrition effects. This means that any change arising in the L1 after immigration is attributed to attrition. However, the use of cross-sectional designs in L1 attrition research can be problematic for several reasons, even if we only consider postpubertal migrants (for a discussion of L1 attrition in children, see Keijzer & de Bot, Chapter 14, this volume). Comparisons between bilinguals and monolinguals were criticized early on by Grosjean (1989). Since then, research on bilingual processing has demonstrated interactions between different languages in bilinguals at different levels. It has shown that even the most proficient bilinguals behave differently from monolinguals of each language. Additionally, personal life history factors have to be taken into account in comparisons of migrants with people who have stayed in their home country. Then again, comparing potential attriters with other bilinguals who are assumed to be balanced or at least non-attrited is no mean feat either. This was the challenge Opitz (2013) took up when she compared German long-term migrants who had lived in Ireland for a mean of 19.5 years and rated their English proficiency as very good or excellent with German speakers of English who had spent less than one year in an English-speaking country and rated their English proficiency as at least good. Her results actually showed very few differences between the two groups, suggesting that it is not easy to tell the difference between potential attriters and other L2 users and bilinguals. One of the main ways in which these differences could be demonstrated might be to focus research on the most sensitive markers (see Section 3.2). Given the difficulty of establishing the ideal control group for research on L1 attrition, it has been suggested that it might be interesting to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

359

compare L1 attriters with any other bilingual or monolingual population, in order to examine parallels and divergences between different language contact or developmental situations (Schmid & Ko¨pke, 2017). This has been done, for example, by Keijzer (2007), who explicitly investigated Jakobson’s (1941) regression hypothesis through a comparison between late bilinguals with L1 Dutch and L2 English, and Dutch adolescents in the late stages of L1 acquisition. Overall, studies comparing different types of bilinguals (e.g., early and late) in attrition studies have only very recently developed in the field (e.g., Polinsky, 2016). The same is true for comparisons with data from patients with language disorders. Earlier volumes on language attrition include a number of studies of aphasia in bilinguals (e.g., Dressler, 1991). Since then, it has been stressed that attrition is of a non-pathological nature, stemming from natural processes of nonuse and CLI. But just as there may be parallels between attrition and language change or acquisition, comparing the dismantling of a linguistic system in healthy and non-healthy populations with acquired language disorders (e.g., aphasia or primary progressive aphasia) may yield further insights into the processes involved in language deterioration. First of all, even in aphasia, a shift to monolingualism is rare. In some cases, bilingual patients with aphasia may exhibit selective recovery of only one language, but this remains exceptional and most bilingual aphasic patients are affected in both of their languages (see Hernandez, Chapter 23, this volume, for further discussion of recovery in aphasic bilinguals). Second, both pathological language deterioration and language attrition in bilinguals are characterized by increased variability in performance. This variability is of specific interest in psycholinguistic research, as it points to the existence of vulnerable domains of the language system. The question of whether these are similar or different across attrition and acquired language disorders has not yet been investigated. Finally, both populations find themselves in a situation of communication handicap (Nespoulous & Virbel, 2007), with reduced proficiency in a language they used to be fluent in. Comparatively investigating the strategies developed by both monolingual and bilingual aphasics in order to compensate for the loss of L1 proficiency might therefore be another interesting perspective for attrition research.

3.2 What to Look At? As in other domains of psycholinguistic investigation, research methods used in language attrition research have progressed a great deal over the past three decades, moving from mainly descriptive studies to experimental designs with neurocognitive measures. As will become clear in this section, different methods show differential sensitivity to attrition effects and have allowed researchers to proceed to more fine-grained analyses in recent years.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

360

BARBARA KÖPKE

3.2.1 Errors Early studies of language attrition looked almost exclusively at error rates. These were mostly recorded in formal tasks involving language production, such as those eliciting the production of morphosyntactic structures, fill-in or cloze tests, and picture naming. Language comprehension was mainly assessed through metalinguistic tasks such as grammaticality judgments or truth value judgment tasks. In more recent studies, the focus has been on more naturalistic data gathered in either free speech or narratives. Schmid (2002), who analyzed extensive spontaneous speech data gathered in the context of historical interviews, established the total number of obligatory contexts in the sample for each feature being investigated (e.g., noun morphology), and then established the error rates for each feature. Striking in the results of all the studies based on error description is that error rates in production tasks are typically low. Even in linguistic domains that have been shown to be vulnerable to attrition, error rates rarely exceed 5% in postpubertal migrants. However, despite such good performance with respect to errors – which may indicate preserved awareness of what is grammatical in the L1 – participants still give the impression that they do not really feel comfortable in their L1. 3.2.2 Fluency In order to try to capture effects of attrition that do not necessarily lead to errors, more recent studies have used a variety of fluency measures. One of these is the verbal fluency task, where participants have to utter all the nouns they can find in a given category as quickly as possible (e.g., animals). Yag˘mur (1997) was one of the first to use this task, and found that Turkish migrants in Australia had clearly reduced lexical fluency on this task, compared with a control group in Turkey. Similar results were found by Cherciov (2013) with Romanian migrants in Canada, suggesting that the verbal fluency task is more sensitive in potential attriters than other language tasks. Other studies have focused on discourse fluency as evidenced by disfluency markers (empty and filled pauses, repetitions, retractions) produced in free speech, film narratives, or picture description tasks. All these studies but one (Opitz, 2013) systematically found evidence for reduced fluency in a variety of migrant groups (e.g., Bergmann, Sprenger, & Schmid, 2015; Cherciov, 2013), despite the fact that these populations did not necessarily differ from control groups on error analyses in spontaneous speech data and other more formal tasks such as C-tests. Once again, these studies demonstrate the sensitivity of fluency measures in the case of attrition. Such findings encourage the pursuit of research on both lexically and discourse-related fluency in attriting bilinguals. An additional means that has not yet been exploited could be to ask native L1 speakers to judge the fluency of migrants, as has been done in some studies investigating foreign accent in attrition (Hopp & Schmid, 2013).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

361

3.2.3 Processing Many recent studies have shown that language processing in bilinguals is different from language processing in monolinguals, even when bilinguals are tested in their dominant L1 (e.g., Friesen, Jared, & Haigh, 2014; Sanoudaki & Thierry, 2015). With respect to attrition, a number of authors have suggested that L1 attrition in postpubertal migrants is largely a matter of online processing (e.g., Sharwood Smith, 1983). Accordingly, it has been shown that the lexical difficulties long-term migrants encounter are usually due to temporary problems in lexical retrieval, as performance is much better on picture-matching than on picture-naming tasks (Ammerlaan, 1996). However, except for the studies on discourse fluency described earlier, there has been scant research on L1 processing in production tasks, and chronometric approaches involving reaction times (as in lexical decision tasks) are still strikingly absent. A notable exception is Hulsen (2000), who administered a timed picture-naming task to three generations of Dutch speakers in New Zealand. She found no differences between the naming latencies of first-generation immigrant speakers and the monolingual control group. This finding, however, may say as much about L1 attrition as about the limits of directly comparing bilinguals and monolinguals (Grosjean, 1989). Many of the studies of language processing in attriters feature interpretation tasks involving different grammatical features. Of particular interest are studies that have begun to examine this question with eye-tracking evidence. In a series of studies, Dussias and colleagues (see Kroll, Dussias, Bice, & Perroti, 2015, for a review) have extensively studied syntactic processing in proficient, balanced bilinguals. These studies not only document CLI in syntactic processing, which is generally thought to remain under the influence of L1 even in very proficient L2 speakers, but they also explore the contexts and conditions that favor the influence of L2 on L1 in such populations. Dussias and Sagarra (2007) showed that, compared with Spanish monolinguals or Spanish–English bilinguals with only limited exposure to L2, Spanish–English bilinguals with extensive exposure to their L2 modify their strategies with respect to relative clause attachment. However, as stressed by the authors themselves, these studies documented not so much attrition but CLI, as participants were balanced bilinguals who used both languages on a daily basis. In a study of antecedent preferences for pronominal subjects in Spanish–English bilinguals using eye-tracking, Chamorro, Sorace, and Sturt (2015) explicitly investigated whether L1 attrition involves online processing or representational deficits. Participants included two groups of 24 L1 Spanish speakers who had been living in the United Kingdom for a minimum of five years. One of these two groups had been exposed to only Spanish for at least one week prior to testing. A third group consisted of Spanish speakers with very little knowledge of English who

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

362

BARBARA KÖPKE

had only recently arrived in the United Kingdom (mean LoR eight weeks). The linguistic material was tested in a naturalness judgment task and an eye-tracking-while-reading experiment. While there were no group differences on the judgment task, non-exposed attriters showed a lack of online sensitivity for pronoun mismatches in the eye-tracking measures, which distinguished them from both the control group and the recently exposed migrant group. The authors concluded that attrition affects online sensitivity without causing permanent representational changes.

3.2.4 Brain Signals Research based on different brain imaging techniques has considerably contributed to our understanding of language acquisition and processing in bilinguals. Among the most important, event-related potentials (ERPs) extracted from neurophysiological data measured by electroencephalography (EEG) reflect the temporal course of language processing with much precision and have given important insights into the nature of the processes used (e.g., semantic vs. syntactic) in different bilingual populations. Functional magnetic resonance imaging (fMRI) measures a brain region’s level of local processing based on blood flow response, and allows the comparison of the brain regions activated for the processing of each language in different language tasks and different types of bilinguals. Additionally, MRI measures provide anatomical brain data about the structure of gray and white matter that have been shown to vary with increasing expertise in a task. Language attrition research based on neurophysiological and brain imaging data is just developing and remains rare. However, findings from recent ERP studies are promising, and suggest that such data may in some cases be more sensitive with respect to attrition effects than previously thought. The first study to take ERPs into account was conducted by Bergmann, Meulman, Stowe et al. (2015), who compared 26 L1 German speakers in North America and a control group of 27 German L1 speakers in Germany. The attriters were all postpubertal migrants (arrival in the United States or Canada at 21 to 39 years) with an LoR of 17 years on average. They showed significantly lower performance than controls on a C-test used as a general proficiency measure but not on a gender assignment task that was used as a baseline. Verb agreement and determiner–noun gender agreement were investigated in a behavioral task, and ERPs were recorded during stimulus presentation. Both groups differed only in the neurophysiological but not in the behavioral evidence. Specifically, in the verb agreement task, while controls seem to rely only on syntactic processing, attriters also showed evidence of semantic processing. This was interpreted by the authors as evidence of the influence of contact with English, since speakers of English tend to use both semantic and syntactic processing in similar tasks. However, with respect to gender agreement, L1 attriters showed fully

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

363

German-native-like ERP signatures, indicating the robustness of gender agreement processing even in long-time migrants. Verb agreement was also investigated by Kasparian et al. (2017) in 24 first-generation Italian migrants to Canada, who were all postpubertal bilinguals (mean age at arrival = 28 years; mean LoR = 11 years) and 30 controls in Italy. Participants performed a judgment task on sentences containing violations of verb–number agreement, and neurophysiological responses were recorded during sentence reading. Results showed differences between attriters and controls on both the behavioral task and on the ERPs, which the authors interpreted as reflecting L2 influence and shallower online repair processes in the attriters. In another study involving the same participants plus a group of L2 speakers of Italian with L1 English, Kasparian and Steinhauer (2016) tested lexico-semantic processing of what they call confusable words (e.g., cappello, hat vs. cappella, chapel) and showed that attriters and L2 speakers of Italian may differ on lexico-semantic processing. Interestingly, these differences were directly correlated with L1 proficiency. In both studies, the authors distinguished between participants who scored above the median (high proficiency) and those who scored below it (low proficiency) on a series of general behavioral measures in Italian (C-test, error detection, semantic fluency). The low-proficiency attriters and L2 learners differed from the highproficiency attriters and controls on the ERP components associated with semantic processing. Further ERP evidence was interpreted by the authors as increased conflict monitoring in attriters. Taken together, these results provide evidence for modifications with respect to L1 neurocognitive processing strategies in attriters that are related to subtle variations in L1 proficiency. These initial ERP results clearly demonstrate the usefulness of neurocognitive measures in research on L1 attrition. With respect to functional neuroimaging, fewer studies to date have collected fMRI data for L1 attrition, all in the context of international adoption. There is, however, no doubt today that fMRI could contribute to the understanding of language attrition. In their review, Rossi, Prystauka, and Diaz (2019) provide examples of a number of fMRI studies that have demonstrated changes in the functional networks dedicated to language processing due to the learning of another language and to bilingual processing. They also suggest that other measures such as structural MRI that can assess greyand white-matter density changes could also be used to investigate attrition, as has been done for L2 acquisition. The study of brain oscillatory signals (i.e., rhythmic or repetitive neural activity among firing neurons) collected with longitudinal or cross-sectional designs could also improve current understanding of the dynamics of linguistic changes and their interaction with domain-general cognitive processes such as memory or attention. This opens up a great many fascinating research avenues.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

364

BARBARA KÖPKE

Two conclusions can be drawn from this survey of approaches to attrition. First, attrition research should not be reduced to studies comparing migrants who are assumed to have attrited and monolinguals in the home country, as is generally the case. In order to gain the fullest possible picture, comparisons should involve all kinds of participants, including migrants with different LoRs, L2 learners, late bilinguals, early bilinguals, and monolinguals, as well as individuals with language disorders. Second, research should clearly not focus on errors alone. In 2004, discussing the error-based attrition literature available at that moment, Ko¨pke and Schmid stated that “we would thus strongly advocate that future studies on language attrition combine the analysis of ‘what went wrong’ with data on ‘what went right’” (p. 29). However, so far, not a single study has done this. A number of interesting studies have nevertheless focused on language processing and allowed for the combined analysis of the processing of what goes wrong and what goes right.

4 Conclusion The question is how we can define attrition today. More than 30 years of research on language attrition and the concurrent development of research on a wide range of aspects of bilingualism have considerably changed our understanding and view of language development in bilinguals. During this period, usage-based theories have gained in importance in SLA and have been instantiated in theoretical frameworks such as that of multi-competence (e.g., Cook & Li Wei, 2016), Complexity Theory (e.g., Larsen-Freeman, 2015), and Dynamic Systems Theory (e.g., de Bot, 2007). The development and spread of such theories has facilitated acceptance of the idea that usage may have consequences on well-entrenched knowledge such as that of the L1 in late bilinguals. Furthermore, results from research on bilingual processing have also shed new light on attrition processes. Studies of language processing in bilinguals have demonstrated that different languages interact at every level of processing. For example, the spread of lexical activation and lexical competition has been extensively documented, not only for isolated word processing but also for words in a sentence context providing strong cues to which language should be used (e.g., Friesen et al., 2014). Nonselective syntactic activation has been observed, especially for structures sharing the same word order (e.g., Sanoudaki & Thierry, 2015). Transfer and convergence have been documented in a variety of semantic domains, as has convergence for a variety of phonetic features (see Schmid & Ko¨pke, 2017, for a review). All these phenomena arise in different kinds of bilinguals, meaning that CLI is not limited to cases of reduced use of one of the languages, as in attrition, but also occurs in active bilinguals, L2 learners, and even beginners. Hence, it

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

First Language Attrition

365

is extremely difficult to conclude that differences observed between monolinguals and migrants are due to attrition, and the border between whatever must be considered normal interaction between languages in bilingual language processing and attrition is impossible to draw. The notion of language dominance is now well established in bilingualism research, although it is still not clear how to establish dominance in a bilingual person. Dominance is a relative measure of skills in one language compared with another, or of asymmetries in skills (see the volume edited by Silva-Corvala´n & Treffers-Daller, 2016, for in-depth discussion). While a number of authors have defined dominance as proficiency in one language compared with another, I would like to suggest a psycholinguistic definition of dominance, referring to the availability of each language for language processing. However, applying this definition of language dominance to bilingualism signals the beginning of the end for language attrition, for if the dominant language is the one that is most available for ongoing language processing, then language attrition can only be a matter of degree. Consequently, attrition can be said to occur when the availability of the nondominant language decreases so much that fluent language processing in that language is no longer possible. It is thus highly unlikely that we will ever be able to identify a cut-off point between a nondominant language and an attrited one, just as we are unable to determine where L2 acquisition ends and bilingualism begins. Perhaps the question of whether a language is just nondominant or clearly attriting is not the most important issue and we should just retain what we have learned from studies on attrition, namely that the L1 is not set in stone and will adapt to changing life situations, even in late bilinguals. Moreover, while attrition appears today, in the light of usage-based theories, to be the natural consequence of a decrease in the (quality of) use of a language, the question we should be asking is not so much “Why is there attrition?” as “Why is there not more systematic attrition in all types of bilinguals?”

Acknowledgment I would like to thank Lourdes Ortega and Annick De Houwer for their very insightful comments on a previous version of this chapter. All remaining errors are mine.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:46:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.019

19 Bilingualism in Clinical Linguistics Carol Scheffner Hammer and Lisa A. Edmonds

1 Introduction Speech-language pathologists (SLPs), also referred to as speech therapists, clinical linguists, or logopedists in different parts of the world, serve increasing numbers of bilingual children and adults in clinics, hospitals, and schools. This is undoubtedly a result of the fact that an estimated more than half of the world’s population speaks two or more languages and that countries across the globe are receiving unprecedented numbers of migrants. When working with bilingual clients across the lifespan, SLPs are faced with great challenges that are due to a number of factors including: (a) the inherent complexity within and across bilingual individuals with respect to language exposure, use, and proficiency; (b) limited research on the nature of language disorders in bilinguals; and (c) a scarcity of assessment and intervention materials and guidelines. This paucity of information can lead to misunderstandings of language disorders in bilingual populations and implementation of less than optimal assessment and intervention practices. In this chapter, we highlight findings on selected language disorders in bilingual children and adults, focusing on oral language use. The literature on children targets disorders in the first 10 years of life whereas the literature on disorders in adult bilinguals primarily focuses on disorders that are acquired in adulthood and not on disorders that were present earlier in life. This is reflected in our discussion. We also discuss key practices in assessment and intervention and conclude by emphasizing parallels that exist in dealing with language disorders in bilingual populations across the lifespan. For bilingual children, we define bilingualism broadly to include children who are in the process of learning two languages, often a minority and a majority language. This includes children who were regularly addressed in two languages from birth (i.e., simultaneous learners) and children who were regularly addressed in one language from birth and a second language at a later age in early childhood, at

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

370

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

age three or later (i.e., early sequential learners) (see Serratrice, Chapter 1, this volume, for further discussion of these different populations of child bilinguals). As is convention in the literature on children’s language disorders, we refer to children without any known impairments as typically developing (TD). For adults, we adopt the view that bilingualism over the lifespan is fluid and that, as we explain further later in the section discussing assessment, bilingualism (considering relative proficiency and usage across languages) lies along a continuum (e.g., Paradis, 1998) and cannot be defined in terms of binary or strict criteria.

2 Language Disorders in Bilingual Children Research on language disorders in bilingual children has focused on disorders that are specific to language such as specific language impairment, as well as language disorders that are secondary to cognitive and developmental disorders such as Down syndrome and autism spectrum disorder. We summarize key findings on these disorders. Note that we do not discuss phonological disorders in bilingual children. Readers are referred to McLeod and Goldstein (2012) for additional information on this topic.

2.1 Bilingual Children with Specific Language Impairment Specific language impairment (SLI) is estimated to affect approximately 7% of monolingual populations in the United States and is assumed to affect the same percentage in bilinguals (Kohnert, 2013). SLI is defined as an ongoing language learning deficit in the presence of cognitive and hearing abilities that are within normal limits and in the absence of significant deficits in other areas of development. Children with SLI are thought to have an underlying deficit in processing language input and have difficulty with both language comprehension and production. Because of this underlying impairment, SLI in bilingual children will be observed in both languages (Kohnert, 2013). Deficits in morphosyntactic skills are considered a hallmark of SLI. However, the type of grammatical errors produced by children with SLI is dependent on the characteristics of the specific language involved. For example, a language such as English has a sparse morphology. As a result, English-speaking children with SLI tend to have difficulties with verb morphology, likely because the few English inflectional verb morphemes are not highly salient. Relative to English, Spanish has a more rich morphology. Spanish-speaking children with SLI tend not to have deficits in verb morphology, likely because each verb is saliently marked for person, number, mood, and aspect. Instead, Spanish-speaking children with SLI have difficulties with articles and clitic pronouns, which are less salient (Restrepo & Gutie´rrez-Clellen, 2012).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

371

Research on SLI in bilingual children has focused on comparing the characteristics of bilinguals with SLI to those of their monolingual peers with SLI and to children with typical (non-disordered) language development. For further extended discussions of SLI in bilingual populations in the United States, Canada, and Europe, see Armon-Lotem, de Jong, and Meir, (2015), Bedore and Pen˜a (2008), Kay-Raining Bird, Genesee, and Verhoeven (2016), Paradis, Genesee, and Crago (2011), and Marinis, Armon-Lotem, and Pontikas (2017). Most studies have targeted the production of grammatical morphemes. In general, they indicate that simultaneous learners with SLI produce errors that are similar to those made by monolingual children with SLI who speak the children’s respective languages. Although fewer studies have been conducted on sequential learners, these studies demonstrate that sequential learners with SLI display similar grammatical errors and language skills as monolinguals with SLI once sequential learners have had sufficient exposure to their second language (Paradis et al., 2011). Thus, regardless of whether children are simultaneous or sequential learners or are monolingual, children with SLI have weaker grammatical abilities than monolingual or bilingual children with typical language development. Research has also focused on areas other than morphology, although few studies have been conducted in any given area. With regard to semantic development, a small body of research has shown that the semantic abilities of bilingual and monolingual children with SLI are similar. In general, children with SLI have limited depth in their semantic representations, and their semantic networks are poorly linked (Bedore & Pen˜a, 2008). In terms of narrative abilities, differing results have been found. A body of work by Gutie´rrez-Clellen (e.g., 2012) has shown that SLI in Spanish–English bilingual children results in difficulties with producing narratives, similar to SLI in monolinguals. SLI in both Spanish–English bilingual and monolingual children affects children’s ability to provide sufficient background information and impacts coherence or the expression of temporal, causal, referential, and spatial relations. However, work with Greek bilingual children with SLI (Tsimpli, Peristeri, & Andreou, 2016) and English–Hebrew children with SLI (Altman, Armon-Lotem, Fichman, & Walter, 2016) has shown that SLI does not affect narrative macrostructure, but does impact narrative microstructure (e.g., lexical diversity, length of utterances, and verbal productivity). In addition, studies have shown that SLI in bilingual and monolingual children results in poor performance on nonlinguistic cognitive processing tasks (Kohnert, 2013). However, SLI in bilingual and monolingual children does not appear to affect children’s pragmatic abilities nor does SLI in bilingual children appear to impact children’s pragmatic and grammatical skills in code-mixing (Gutie´rrez-Clellen, Simon-Cereijido, & Leone, 2009).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

372

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

Overall, the research demonstrates that SLI in bilingual and monolingual children affects children’s language abilities in similar ways, but depends on the characteristics of the particular languages spoken. Research has shown that bilingualism does not cause language impairment in children. Children with SLI would have SLI regardless of whether they are acquiring one or two languages (Kohnert, 2013). Additionally, SLI in both bilingual and monolingual children results in lower lexical/semantic, morphosyntactic, narrative, and language-processing abilities as compared to bilingual and monolingual children who are typically developing. The gap between the language skills of bilingual children with and without SLI is similar to the gap in skills between monolingual children with and without SLI (Kohnert, 2013; Paradis et al., 2011).

2.2 General Developmental Disorders in Bilingual Children The knowledge base on the language skills of bilingual children with general developmental disorders is extremely limited. This work has mostly focused on children with Down syndrome and autism spectrum disorder. Down syndrome (DS) is a genetic disorder that affects all areas of development. In general, children with DS develop language at a slower rate than children who are developing typically, that is, without any known impairments (TD). The sequence in which language development proceeds is highly similar in both populations. Children with DS may score lower than their nonverbal skills would suggest in some areas such as morphosyntax, but for language comprehension, their skills are generally equal to or comparable to their nonverbal cognitive skills. This makes comprehension a strength in comparison to the use of morphosyntax (Kay-Raining Bird, 2016). There is consensus that the language development of bilingual children with DS is highly similar to that of monolingual children with DS. Bilingualism does not appear to change the expression of the disorder. Additionally, the predictors of language disorders in children with DS are similar to those of monolinguals with DS. In monolinguals with DS, children’s mental age (MA) and chronological age (CA) positively predict language comprehension and language comprehension positively predicts language production. In bilingual children with DS, MA and CA predict children’s skills in their stronger language; MA and children’s skills in their stronger language are positively related with vocabulary, comprehension, and production syntax in children’s weaker language (KayRaining Bird, 2016; Kay-Raining Bird et al., 2016). Several case studies have shown that children with DS can become bilingual, and that bilingualism does not delay their language development more than would be expected if they were growing up monolingual. When matched on nonverbal cognition, bilingual and monolingual

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

373

children with DS performed similarly on English measures of vocabulary, and on the comprehension and production of morphosyntax (Kay-Raining Bird, 2016). However, like for TD bilingual children, the relative amount of language input affects the degree of bilingualism. Kay-Raining Bird (2016) estimated that children with DS who received less than 20% input in a language did not make progress in that language, whereas all children made progress in the language to which they were exposed at least 50% of the time. These estimates must be considered tentative in view of the complexities of establishing the minimal amount of language input for language development, thoroughly reviewed for TD bilingual children by Armon-Lotem and Meir (Chapter 10, this volume), who also evaluate the merits of exploring input quantity in proportional, absolute, and featurespecific terms. There are also children who are diagnosed with autism spectrum disorder (ASD). ASD is a pervasive developmental disorder that affects children’s social and communication abilities. Approximately 10–25% of English-speaking children with ASD do not acquire functional speech or language. The few studies that have compared the language skills of bilingual and monolingual children with ASD have found no differences in the two groups’ comprehension and production abilities (MarinovaTodd & Mirenda, 2016). For example, Hambly and Fombonne (2012) examined the language abilities of 75 children with ASD between the ages of 36 and 78 months. Thirty children were learning English, French, or Spanish only; 45 children were learning two or three languages simultaneously or sequentially. No differences in language comprehension or production were found; however, children whose mothers were nonnative speakers of the languages spoken to them had lower language scores than children whose mothers were native speakers (see Armon-Lotem & Meir, Chapter 10, this volume, for more discussion of the role of the quality of input addressed to children). Additionally, Peterson, Marinova-Todd, and Mirenda (2012) compared the comprehension and production vocabularies of Chinese–English bilingual children with ASD to those of Englishspeaking monolingual children with ASD. Bilingual children had higher total vocabularies than monolingual children, but had similar vocabulary sizes in English. Overall, researchers who study bilingual children with ASD have concluded that children with ASD are capable of learning two languages and that bilingualism does not impede their language development (Marinova-Todd & Mirenda, 2016). Taken together, the research reviewed on bilingual children with SLI, DS, and ASD demonstrates that they display language skills that are similar to the skills of their monolingual peers with their particular disorder: “Bilingualism does not appear to change the behavioral phenotype of a particular disorder” (Kay-Raining Bird, 2016, p. 52). Additionally, this research shows that children with impairments can and do learn two languages without detriment to their development. Removing one

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

374

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

language does not improve children’s language abilities in the other language nor does it cure the disorder. On the contrary, as is discussed in the intervention section, maintenance of both languages is critical for bilingual children.

3 Language Disorders in Bilingual Adults One of the most common acquired adult language disorders is aphasia, a language disorder predominantly caused by left-hemisphere stroke (though it can result from other causes, such as traumatic brain injury and brain tumors). Aphasia causes impairments in the production and comprehension of spoken and written language. These impairments can have a significant impact on individuals’ abilities to communicate basic needs, wants, or emotions and can compromise independence and overall life participation. Bilingual aphasia refers to the presence of aphasia in a person who was premorbidly (before the stroke or brain damage) bilingual (e.g., Paradis, 1998). In addition to potential impairment in language comprehension and production, bilinguals’ translation abilities can be affected. Also, the ability to control language switching can be impacted, leading to pathological code-mixing, a non-volitional mixing of languages that can compromise communication (Abutalebi, Miozzo, & Cappa, 2000), especially when the conversational partner does not share the languages spoken by the individual with aphasia (see also Hernandez, Chapter 23, this volume). While both languages are affected in bilingual aphasia, they may exhibit different patterns of impairment and recovery, which generally refer to both production and comprehension. Parallel recovery occurs when both languages are equally affected, with recovery of both languages proportional to premorbid relative proficiencies. Differential recovery occurs when the languages recover differentially from premorbid proficiency patterns – either one language is better than the other within the context of relatively equal premorbid abilities, or both language proficiencies are similar despite unequal premorbid proficiencies. Selective recovery refers to resumption of production in only one language, though comprehension may be retained in both. Successive recovery occurs when one language recovers after the other, though sometimes the recovery of the previously inaccessible language will correspond with proportional regression in the first recovered language (antagonistic recovery). If this pattern repeats over any period of time (e.g., days, weeks), it is considered antagonistic recovery. Blended recovery refers to unintended mixing of the two languages. It is important to note that these recovery patterns are not mutually exclusive (Paradis, 1998). Their actual incidence is unknown; however, it is estimated that parallel recovery is most common (Roberts, 2008). The clinical implication of parallel recovery is that, while aphasia may

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

375

exist in each language, the patient’s relative ability in each language is consistent with premorbid abilities. However, if some type of differential or antagonistic recovery occurs, relative language abilities may be inconsistent with the language environment, usage patterns, or needs. Despite potential differences in aphasia severity across languages, general language impairments as related to various types of aphasia are typically similar across languages. However, the nature of specific impairments may reflect language-specific characteristics (e.g., Gitterman, Goral, & Obler, 2012). Currently, impairment and recovery patterns cannot be predicted by any specific factor (e.g., lesion site, aphasia severity, language learning history, premorbid language use, or proficiency), though there is growing evidence that a breakdown of language control mechanisms may account for the seemingly diverse patterns of improvement (see further Hernandez, Chapter 23, this volume). Another type of language disorder in bilingual adults is the result of neurodegenerative disorders that progressively impact language and cognition. One of the most common neurodegenerative syndromes is primary progressive aphasia (PPA), which is characterized by a progressive language loss with relative preservation of cognitive and motor function within the first two years of symptoms (Gorno-Tempini, Hillis, Weintraub et al., 2011). At this time, there are few studies of PPA in bilinguals. Over and above the need to know more about the disorder to help patients, the study of PPA in bilinguals has implications for theories of bilingual language processing in terms of the relation between neural representation and processing of each language on the one hand, and the age and manner of language learning, attained proficiency, and usage of both languages at the time of disease onset on the other. The very limited research on bilingual PPA so far has focused primarily on case reports evaluating language changes relative to these variables (e.g., Zanini, Angeli, & Tavano, 2011) with preliminary findings suggesting that the manifestation of PPA in bilinguals appears to be consistent with findings in the monolingual literature. Larger, more methodologically rigorous studies are needed to formulate any generalizations, however. As with children, bilingualism itself does not change the basic manifestation of adult aphasia symptomatology. Of primary relevance is the interpretation of aphasia symptomology within the context of individuals’ premorbid language acquisition, proficiency, and usage history across languages. There is currently no definitive mechanism to account for the variability of impairment profiles in bilingual aphasia, though the study of neurological control of language has provided key insights (see Hernandez, Chapter 23, this volume). More research with consistent methodologies across large numbers of bilingual participants is needed to develop a comprehensive framework that can aid in the prediction of impairment and recovery patterns in bilingual aphasia and language changes in PPA, and thus inform assessment and treatment.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

376

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

4 Assessment of Language Disorders across the Lifespan A language assessment of bilingual individuals serves several purposes. For bilingual children, a primary purpose is to determine whether there is a language disorder in the absence of significant cognitive impairments, or to determine the extent of a language disorder when children have more pervasive developmental disorders such as DS or ASD. For adults, a key purpose is to discern the effects of aphasia in both languages within the context of premorbid proficiency usage and proficiency. Additional purposes for assessment for both children and adults include gaining an understanding of the contexts in which they communicate, informing the goals of intervention and treatment approaches used, and evaluating the effects of treatment over time. In the following sections, we discuss approaches to assessing bilingual children and adults. Before doing so, it is important to point out that SLPs who do not speak one of an individual’s primary languages should enlist the services of an appropriately trained translator or refer to another SLP who has the appropriate language skills. Discussing bilingual service delivery is beyond the scope of this chapter. We refer interested readers to MarinovaTodd, Colozzo, Mirenda et al. (2016) and to related guidelines like the ones developed by the American Speech-Language-Hearing Association (www .asha.org/practice/multicultural/) and by the Royal College of Speech and Language Therapists in the United Kingdom (www.rcslt.org/members/pub lications/publications2/linguistic_minorities) for more information.

4.1 Assessment of Language Disorders in Children Diagnosing a language disorder in children is complex. Because bilinguals vary greatly in the amount of exposure to and usage of their two languages, there is great intra- and inter-individual variation. Thus, it is difficult to find a so-called typical reference group to which the performance of a particular bilingual child can be compared (see also Treffers-Daller, Chapter 15, this volume, for discussion of assessing bilingual abilities in general). In addition, diagnosis of language impairment in children without a known developmental disorder is challenging, because some language characteristics of typically developing sequential bilingual children without language impairment resemble the characteristics of monolingual children with SLI (see Paradis, 2010, for a review). Thus, SLPs are charged with differentiating typical bilingual development from a language disorder. However, SLPs must keep in mind that children cannot have a disorder in just one of their languages. For a diagnosis of language impairment, children must exhibit difficulties in both languages. Throughout the assessment, best practice dictates that SLPs recognize that children are part of a family system, and that children and their families interact in various educational, work, and community settings

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

377

and are influenced by those settings and affected by their policies. Additionally, SLPs need to understand that children and families hold individual, family-based, and cultural beliefs that influence their behaviors and practices, including those related to language and communication (Hammer, 1998). To diagnose a language disorder, a comprehensive assessment is required that involves multiple sources of information. Key sources include semi-structured interviews, questionnaires, observations in authentic contexts, direct assessment of children’s skills in both languages through standardized tests, informal measures or tasks and language samples, and dynamic assessment. We discuss each briefly. Semi-structured interviews of children’s family members and the children themselves, in the case of older children, allow SLPs to gather essential information about children’s language abilities as well as the settings and contexts in which children communicate. Semi-structured interviews differ from traditional interviews in that more emphasis is placed on establishing relationships with families and/or children and on having conversations with those being interviewed, in order to arrive at an understanding of families’ views and perspectives (Westby, 1990). Before the interview, SLPs develop a set of questions to ask. During the interview, SLPs do not ask the questions in a preset order, but rather ask questions as the conversation unfolds. Through semi-structured interviews, information is gathered on (a) families’ views of language and communication as well as children’s language disorder; (b) children’s current language and communication abilities; (c) families’ language and communication goals; (d) contexts in which children communicate (e.g., home, school, community), topics of conversation, individuals with whom children interact, as well as the languages used; and (e) goals for assessment and treatment. The use of questionnaires is particularly helpful for gathering information about children’s experiences with their two languages. As the combined chapters in this Handbook show, these are not static and are influenced by myriad variables, including acquisition history, education, and cultural and linguistic interactions. Variations in language experiences result in differences in individuals’ language proficiencies as well as in differences among bilinguals (see particularly Armon-Lotem & Meir, Chapter 10, this volume). Questionnaires offer the opportunity to gather information about children’s language experiences. We discuss two that were developed for use with bilinguals living in North America. The Language Exposure Assessment Tool (LEAT) (DeAnda, Bosch, Poulin-Dubois et al., 2016) was designed for Spanish–English or French–English infants and toddlers. When completing the LEAT, parents list key interactants and estimate how much time children spend with those individuals, and how much time these individuals use the children’s two languages. The Bilingual Input Output Survey (BIOS) (Pen˜a, Gutie´rrez-Clellen, Iglesias et al., 2014)

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

378

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

elicits information about children’s history of language exposure and current usage of Spanish and English at home and at school. By asking about the languages used each hour of the day, the percentage of time children are exposed to Spanish and English can be determined. Whenever possible, it is highly recommended that SLPs conduct observations of children at home and at school. Such observations can provide valuable information about the language and communication abilities that children may not exhibit when interacting with SLPs during an assessment, which is an unfamiliar communicative context. Information from the observations can also be used to identify culturally specific practices and values of children and families. For example, many parents of Latino descent in North America place emphasis on teaching children to be respectful, well mannered, and obedient. Observations may inform SLPs whether this is a value a particular family holds and if it is, how the family instills this value in their children. SLPs need to gather this information to better understand how a child may interact with the SLP during intervention. For instance, children taught to be obedient to adults may be more reserved and less interactive than other children during an assessment, because of the respect of adults that children have acquired. Thus, without more background information, the SLP may think that the child has a problem with producing language. Additionally, observations can assist with the identification of key communicative partners, activities, and topics of conversation that can be incorporated into therapy sessions, classrooms, and/or home programs (Hammer, 1998; SLPs often collaborate with teachers and parents to promote children’s language skills). Naturally, direct assessment of bilingual children’s language and communication abilities is an essential source of information when diagnosing language disorders in children. As noted earlier, best practice dictates that bilingual children be assessed in both languages so a comprehensive picture of children’s language skills is obtained. However, SLPs face great challenges when attempting to do so, because few valid and reliable standardized tests are available for use with bilingual children. If a test is not standardized on bilingual children, the normative data cannot be used to diagnose a disorder. At times, tests have been translated into or adapted for other languages. Tests that are adapted rather than translated are much preferred for bilinguals, because they can take into consideration the unique characteristics of each language (e.g., lexical items, morphosyntax, etc.); however, many times these tests have not been standardized. In all cases, the cultural and linguistic appropriateness of test items should be taken into consideration and adapted or omitted as needed (for a discussion of assessing bilingual abilities in general, see Treffers-Daller, Chapter 15, this volume). The assessment of children should evaluate children’s phonological, semantic, morphosyntactic, narrative, and pragmatic skills in both

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

379

languages. In the United States, standardized tests normed on bilinguals are most commonly available for bilingual children who speak Spanish. These tests include the Preschool Language Scales-5 Spanish version (Zimmerman, Steiner, & Pond, 2012), the Clinical Evaluation of Language Fundamentals (CELF)-Preschool-2 Spanish (Wiig, Secord, & Semel, 2009), and the Clinical Evaluation of Language Fundamentals-4 Spanish (Wiig, Secord, & Semel, 2006). However, the English versions of these instruments are not standardized on bilingual children. The only assessment instrument normed in Spanish and English is the Bilingual EnglishSpanish Assessment (Pen˜a et al., 2014) that can be used with Latino children regardless of which language they know better. Also, the Receptive and Expressive One Word Picture Vocabulary Tests – 4 (Brownell, 2012) have Spanish-bilingual versions that can be used to obtain a Total Conceptual Vocabulary score based on children’s lexical knowledge across their two languages combined (for in-depth discussion of the notion of Total Conceptual Vocabulary, see Treffers-Daller, Chapter 15, this volume). Given the general lack of tests available for bilingual children in other languages, SLPs need to consider using assessment instruments developed for monolingual speakers of other languages. For example, the MacArthurBased Communicative Development Inventory (a parent-report measure of infants’ and toddlers’ vocabulary and early language abilities) is available in many languages (see also Treffers-Daller, Chapter 15, this volume). Information can be found through the Wordbank website (http://word bank.stanford.edu). Additionally, the Multilingual Children’s Speech website (www.csu.edu.au/research/multilingual-speech/speech-assessments) provides access to speech assessments in more than 40 languages (McLeod, 2014). Note that SLPs should not score these tests that were developed for monolinguals, but they can do an analysis of the items that are passed and failed. In addition to assessment instruments, conversational and narrative language samples elicited from children in both languages can yield highly valuable information about children’s production vocabulary and grammatical abilities (Patterson, 2016). When eliciting these samples, it is important to consider the effect of the context (e.g., home or school), interlocutors (e.g., an individual with a similar cultural and linguistic background as the child vs. an individual from the majority culture), and the materials used (e.g., culturally meaningful to the child or representative of the majority culture). If children’s cultural and linguistic backgrounds are not taken into account, it can be difficult to elicit samples that reflect children’s abilities in their two languages. Computer programs including the Systematic Analysis of Language Transcripts (SALT) (Miller & Iglesias, 2012) and programs available through the Child Language Data Exchange System (CHILDES, https:// childes.talkbank.org/) (MacWhinney, 2000) can be used to analyze language samples. Such analyses may include measures of lexical diversity

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

380

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

(i.e., the number of different words and/or type-token ratio), syntactic complexity (i.e., mean length of utterance; see Treffers-Daller, Chapter 15, this volume, on comparing this measure across languages), and macrostructure analyses (e.g., Heilman, Miller, & Nockerts, 2010), in the case of narrative samples. Also, analyses of children’s morphosyntactic errors yields highly valuable information, particularly when SLPs focus on areas where children with specific language impairment may have difficulty in their respective languages (Restrepo & Gutie´rrezClellen, 2012). SLPs may also want to consider using dynamic assessment to assist them in the diagnostic process, particularly when they are not completely certain that an individual child has a language disorder. The goal of a dynamic assessment is “to establish the amount of change that can be induced during interactions with the examiner during the assessment process” (Gutie´rrez-Clellen & Pen˜a, 2001, p. 212). Essentially, a dynamic assessment is a measure of modifiability. During a dynamic assessment, which occurs over several sessions, SLPs test children to establish baseline functioning in a specific area of language (e.g., a specific syntactic structure). They then “teach” that structure over several sessions and later retest children in that area. SLPs rate children’s learning on how modifiable they were, how much effort was required to teach the children, and how responsive the children were. Children who require much effort to learn the target and who show limited change after “teaching” them are likely to have language impairment whereas children who show significant changes and maintain those changes likely do not quality for services. It is thought that bilingual children who demonstrate meaningful change are exhibiting differences due to bilingual language development, whereas children who do not learn the target and require much effort to do so likely have an underlying language disorder (see Gutie´rrez-Clellen & Pen˜a, 2001, for additional information on dynamic assessment). Using information gathered through interviews, questionnaires, observations, direct assessment, and potentially dynamic assessment, SLPs triangulate the data to determine whether children have specific language impairment and/or determine the nature of language disorder in the case of children with developmental disabilities. Once a diagnosis has been made, SLPs use the information from the assessment to identify goals for the intervention, in conjunction with families, and determine the context(s) and activities in which the goals will be addressed.

4.2 Assessment of Language Disorders in Adults An assessment with individuals with bilingual aphasia, as with children, requires inquiry into and understanding of individuals’ language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

381

background, usage needs and patterns, communication partners, and contexts of usage. Additionally, it is important to inquire, rather than assume, how individuals with aphasia and their family and support network view the stroke, the aphasia, and rehabilitation as well as how the stroke and aphasia have affected individuals’ abilities to participate in desired and necessary activities. Other potential considerations include motor impairments, depression, or psychosocial challenges. With respect to more direct linguistic assessment, the primary task is not always to determine the presence of aphasia, but rather to assess the relative effects of aphasia in both languages within the context of variable premorbid language proficiencies and use patterns. In essence, the assessment should be conceived within the World Health Organization (WHO) International Classification of Functioning, Disability, and Hearing (ICF) model, which provides a holistic view of assessment and treatment that considers the effect that health condition, environmental, and personal factors have on individuals’ participation in life activities (see www.who.int/classifica tions/icf/en/). In addition, it is important for SLPs to broaden their idea of assessment beyond Western views of disability and rehabilitation, and anthropological perspectives can help (e.g., Goodwin, 2004; Penn, 2012). Following the WHO principles, we here discuss multiple sources of information that can facilitate a person-centered and comprehensive assessment. With respect to languages tested, it is ideal to test in both languages. However, in some cases individuals with aphasia and/or their families will request testing and treatment only in one language (often the culturally dominant language). In such cases, it is important to understand the motivations for this choice in order to provide information regarding patient options. The variability in proficiency in bilingual adults can be viewed, in part, within the context of a multidimensional continuum that includes comprehension and production of spoken and written language. This is relevant to bilingual assessment in that a client’s premorbid proficiency essentially serves as a baseline against which aphasia assessment results are compared (e.g., Gitterman et al., 2012). However, premorbid abilities are typically not empirically known, making it difficult to assess the effects of the aphasia across languages: are reduced language abilities due to aphasia or to premorbid proficiency levels? By contrast, premorbid proficiency is assumed in monolinguals (within the context of education, literacy, etc.) and can be compared to normative samples in standardized tests. In the absence of empirically known premorbid language abilities, a structured questionnaire can facilitate collection of information on language acquisition, history, usage, and proficiency. One useful questionnaire is the Language Experience and Proficiency Questionnaire (LEAP-Q) (Marian, Blumenfeld, & Kaushanskaya, 2007; it is available in multiple languages from www.bilingualism.northwestern.edu/leapq/). The LEAP-Q

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

382

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

has been shown to have good validity and reliability with questions broad enough to capture the variability that might be observed across bilinguals who vary in language backgrounds (e.g., simultaneous, sequential, later L2 learners), though the instrument was developed with participants who all had at least high school literacy in one language with reported functional communication abilities in both languages (Marian et al., 2007). The LEAPQ has not been evaluated with clinical populations, but its authors suggest it can be used with clinical populations to augment information from families, and its questions are similar to those used with people with bilingual aphasia on other questionnaires such as the Bilingual Aphasia Test (Paradis, 1987), which we discuss later in this chapter. It is critical to gather quantitative estimates of premorbid usage of each language in different contexts (e.g., work, home, social) from individuals with aphasia and a close bilingual communication partner (e.g., spouse, adult child), though if individuals with aphasia and/or family members are not comfortable with the detailed questioning that collection of such data requires, some estimates of relative proficiency will already be helpful. This information is valuable because research with bilingual populations has shown that participant reports of language proficiency and usage can be reliable predictors of language performance on a variety of constrained language tasks. Among others, these include verbal fluency (Mun˜oz & Marquardt, 2008), reading fluency (Marian et al., 2007, Study 2), reading and oral comprehension (Marian et al., 2007, Study 2), confrontation naming of nouns and verbs (Edmonds & Donovan, 2012), and the measurement of micro- and macro-level discourse variables (Kong, 2011). These can help to estimate premorbid abilities against which to compare language abilities evaluated during the aphasia assessment. After premorbid information is gathered, it is important to ask how proficiency and usage has changed post-aphasia. For example, if a person with aphasia rated herself with a “10” (perfect) in English and a “9” in Arabic before the stroke, but afterward rated herself as a “7” in English and a “3” in Arabic in spoken production, this would reflect a differential impairment. No valid and reliable standardized tests exist for bilinguals with aphasia. Some tests have been translated into or adapted for other languages. Tests that are adapted rather than translated are much preferred for bilinguals, since they take into consideration unique language characteristics (e.g., lexical items, morphosyntax, etc.); however, tests are often not standardized. If aphasia batteries or other tests developed for monolinguals are administered, the normative data cannot be validly used, even with proficient bilinguals, and the test results cannot be used to diagnose type or severity of aphasia (Roberts, 2008). In all cases, the cultural and linguistic appropriateness of the test items should be taken into consideration and adapted or omitted, as needed. An alternative to using tests for monolinguals is the Bilingual Aphasia Test (BAT) (Paradis, 1987). The BAT was originally designed to assess

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

383

each language of a bilingual in an equivalent manner while considering the cultural and linguistic uniqueness of individual languages. That is, the different language versions of the test are not translations, but adaptations (Paradis, 1987). The test is comprehensive, with materials that evaluate production, comprehension, repetition, translation, semantics, grammatical judgment, and reading and writing in more than 100 language combinations. It also contains a language use questionnaire. The BAT is not designed to diagnose aphasia type or severity, but it can provide a crosslinguistic comparison of language performance, though it is important to note that even monolinguals do not always perform at 100% on the test (Mun˜oz & Marquardt, 2008). Test materials and more information can be downloaded through www.mcgill.ca/linguistics/ research/bat (see also a collection of BAT studies in Miller Amberber & Cohen, 2012). In addition to constrained tasks as used in the bulk of test batteries, discourse should be evaluated, because connected speech better represents functional language use. In bilinguals, relying on discourse has also been shown to be less biased and more valid than relying on standardized testing alone (Gitterman et al., 2012). When using discourse in assessment, it is critical to consider the context in which the discourse is collected and to have a solid understanding of the linguistic, pragmatic, and cultural norms within a person’s community (see the contributions in Gitterman et al., 2012; Goodwin, 2004). Relying on discourse allows for examination of micro- (e.g., word retrieval, sentence production) and macro-level (e.g., topic maintenance, conveying main ideas, organization) aspects of communication, person’s overall ability to express their ideas, pragmatic aspects of communication, and whether there is pathological code-switching and, if so, how it affects communication (Mun˜oz, Marquardt, & Copeland, 1999). Discourse is also critical as an outcome measure (i.e., to evaluate before and after treatment to gauge improvement at the discourse level). The details of discourse assessment are complex and a discussion of these is beyond the scope of this chapter. However, it is useful to evaluate connected speech across different types of discourse (e.g., personal, narrative, procedural), tasks (e.g., conversation, picture description – sequential pictures are preferred over a single picture), and contexts (e.g., home, clinic), if possible. It may also be useful to involve multiple interlocutors (e.g., the SLP as well as a family member; see De Houwer, Chapter 17, this volume, on the importance of interlocutor roles in bilingual interaction). Collecting multiple language samples is important because cognitive, linguistic, and pragmatic demands differ across discourse type, task, context, and communication partner. This information will aid in understanding the person’s communication strengths and limitation in a more ecologically valid way than obtaining a discourse sample through just a single observation. If possible, discourse should be video or audio

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

384

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

recorded (with consent) for future analysis and/or to help compare before and after treatment. The BAT can help with obtaining and analyzing discourse samples. It contains a personal narrative prompt and a structured narrative with culturally appropriate picture sequences with basic guidelines for analysis. An additional resource is the Aphasia TalkBank (http://aphasia .talkbank.org), which was developed to allow SLPs and researchers access to videos and transcript samples of stretches of discourse from people with aphasia. The site makes available its testing protocols, stimuli, coding, and scoring guidelines, with some materials provided in languages other than English (viz., Japanese, Spanish, Chinese, and French). Computer programs such as the Systematic Analysis of Language Transcripts (SALT) (Miller & Iglesias, 2012) can be used to analyze transcribed language samples in English and Spanish. Similar to assessments in children, SLPs integrate their assessment findings from all sources and use their clinical judgment to make a diagnosis. After a diagnosis has been made, SLPs combine information from the assessment with input from the individual with aphasia and the family to develop goals and identify methods for the intervention as well as the context(s) and activities in which the goals will be addressed.

5 Intervention Practices across the Lifespan Research on intervention practices for bilingual children and adults with language disorders is extremely limited. However, best practice for both children and adults dictates that individuals (as age permits) and their families participate in the setting of goals as well as in the planning and implementation of therapy. Related to this, best practice requires that individuals’ and families’ cultural beliefs and practices be given key consideration during this process (Patterson, 2016). It is critical that SLPs understand and incorporate the individuals’ and families’ beliefs about language, communication, and language disorders; styles of interaction; daily activities and who participates in those activities; and goals for therapy and the extent to which they want to be involved and who should be involved (Hammer, 1998). Without integrating families’ beliefs and practices into the process, a mismatch can occur between SLPs, the clients, and their families, which may lead to a lack of follow-through on therapy recommendations and in some cases poor attendance to therapy. For example, SLPs in the United States regularly provide parents with home activities that involve play to reinforce the work that is done during therapy sessions. However, in many cultures, parents do not view themselves as peers who engage in play with their children. Play is for children and not adults. As a result, some parents may not perform the activities at home, because the recommendation does not fit their beliefs and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

385

practices. To achieve congruence, SLPs may instead train older siblings to conduct the play activity and/or may look for language-intense times during the daily routine when parents spend time interacting with their children and integrate the therapy targets into these naturally occurring activities. As recommended for the assessment process, best practice includes involvement of professional interpreters in the therapy sessions when SLPs do not speak an individual family’s primary language. This is because therapeutic interventions have specific terminology and concepts that may not be familiar to families in their less dominant language. Also, use of interpreters enables individuals and their families to fully participate in the intervention.

5.1 Intervention with Bilingual Children with Language Disorders Best practice dictates that intervention services for children should be provided in both languages, as opposed to only in the majority language. This is because children with language disorders have the capacity to learn two languages, their abilities from one language may transfer to the other, and, most importantly, children need their two languages in order to successfully learn and communicate at home and school as well as in the community (see Bigelow & Collins, Chapter 2, this volume; and Garcı´a & Tupas, Chapter 20, this volume). Without support of their families’ heritage language, children may lose their ability to take part in family conversations, which can affect their relationships with family members, and they may lose or reject their ethnic identity. Additionally, parents may not have the ability to communicate in the majority language and thus cannot provide quality language models in that language. As a result, if the minority language is excluded from the intervention, parents may communicate less with their children and may participate in fewer cultural and religious activities in their heritage language. Also, when children are not able to speak the heritage language, parents may feel a sense of sadness and loss, because language is a key component of one’s cultural identity (De Houwer, 2017b). In terms of effective interventions, relatively few intervention studies have been conducted to assist SLPs in providing evidence-based services for bilingual children. Studies have primarily compared the outcomes of bilingual and monolingual therapy (Kay-Raining Bird, 2016). These studies demonstrate that intervention in a particular language supports the development of that language and that bilingual interventions generally support both languages (Thordardottir, 2010). For example, Restrepo, Morgan, and Thompson (2013) investigated the gains made in young bilinguals’ English and Spanish language abilities when children received either a bilingual or a monolingual intervention. They found that children in the bilingual

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

386

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

and monolingual intervention groups made gains in English vocabulary and that children in the bilingual intervention group also made gains in Spanish vocabulary. These gains were greater than those of children in a mathematics intervention group and in a control group that did not receive intervention. Thordardottir, Cloutier, Me´nard et al. (2015) found similar results for bilingual children who spoke French and, a minority language. Gains in French vocabulary occurred when the intervention was provided in French and similarly, gains occurred in the minority language when therapy occurred in that language.

5.2

Language Intervention with Bilingual Adults with Language Disorders Language intervention with bilingual adults is similar to that with monolinguals insofar as clinical decision-making requires careful consideration of individuals’ cognitive-linguistic strengths and impairments within the context of their communication goals and needs. In addition, medical (e.g., pain, depression, paralysis, fatigue) and environmental factors (e.g., support system, transportation issues) must be considered. It is beyond the scope of this chapter to discuss different intervention approaches, though in general there are two main approaches: impairment based and socialfunctional. Impairment-based approaches address some underlying deficit, such as a difficulty in retrieving words that are sufficiently informative to convey an intended meaning. Social-functional approaches focus less on the impairment and more on participation. SLPs must consider the range of possible approaches and how they might be implemented to help individuals with aphasia toward their communication and participation goals. Since a fundamental goal of treatment is to maximize individuals’ level of life participation, a primary goal of language treatment with bilingual adults is the improved ability to improve communication in both languages. Thus, fundamental questions about language therapy in bilingual adults are whether treatment is effective in both languages and whether treatment in one language will generalize to the untrained language. Such generalization is termed crosslinguistic transfer (CLT). There are currently no clear treatment guidelines, as there are few methodologically sound bilingual aphasia treatment studies, with existing studies varying considerably with respect to participants, treatment approaches, outcome measures, and other methodological considerations. However, as outlined in what follows, some preliminary clinical implications have emerged from the limited literature. Aphasia treatment in monolinguals has been shown to be effective. Likewise, the bilingual literature shows that treatment provided in the first learned and/or premorbidly more proficient language (e.g., Ansaldo &

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

387

Ghazi-Saidi, 2014). Further, positive outcomes are also observed in the second learned or premorbidly less proficient language with treatment in that language (see Faroqi-Shah, Frymark, Mullen, & Wang, 2010, for a review). However, there is no clear effect of proficiency or age of onset of L2 learning on treatment outcomes in the later learned language (L2) in bilingual adults with moderate to high premorbid L2 proficiency (FaroqiShah et al., 2010). These findings suggest that treatment may be effective in either language regardless of age of L2 learning onset or premorbid proficiency, so SLPs can choose to treat in either language based on individuals’ communication abilities, needs, and/or goals. Some researchers have suggested that treating within bilingual environments draws on the preserved abilities of both languages and provides more communicative strategies to bilingual clients (e.g., Ansaldo, GhaziSaidi, & Ruiz, 2010). A potential contraindication for simultaneous treatment is the presence of pathological code-switching (see Section 4.2), for which unilingual treatment has been recommended to avoid crosslinguistic interference (e.g., Fabbro, 1999). However, treatment can potentially address non-volitional language mixing through theoretically driven assessment and intervention that does not require inhibition of one language. For example, Ansaldo et al. (2010) designed a treatment for a Spanish–English bilingual male patient with aphasia who used involuntary language mixing. The treatment, termed Switch Back through Translation, was designed to overcome involuntary switching through the use of translation. Semantic Feature Analysis (SFA), a treatment approach that promotes semantic activation of trained words or concepts through the generation of a concept’s features (e.g., a spoon is a utensil, is used for eating, is found in the kitchen), was provided with pictures and within discourse in Spanish, but the participant was encouraged to produce whichever language he chose. When involuntary language mixing occurred, the clinician cued the participant to use translation to get back into Spanish. Ansaldo et al.’s (2010) results showed improved naming and translation to Spanish without evidence of improved lexical retrieval in the non-treated language (i.e., there was no CLT). However, languageswitching intrusions were reduced in both languages. The findings of this study are encouraging and emphasize the importance of modeldriven treatment that considers the specific language profile of bilingual adults. A potentially efficient means of improving both languages is through promoting CLT. For example, Edmonds and Kiran (2006) attempted to promote CLT in three Spanish–English adults with chronic bilingual aphasia by administering SFA one language at a time. The basic hypothesis was as follows. SFA activates the semantic system, which is presumed to be shared across languages (for the types of concrete items that were trained), and that semantic activation, regardless of language, could potentially activate the phonological output lexicons of each language, which in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

388

CAROL SCHEFFNER HAMMER AND LISA A. EDMONDS

turn could promote lexical retrieval of the trained words in both the trained and untrained language as well as to semantically related words in both languages. That is, does training the word celery generalize to the untrained translation of that word (apio) as well as to semantically related words in both languages (cabbage/repollo)? The results showed a relation between premorbid proficiency and CLT such that training in the premorbidly more proficient language did not result in CLT, whereas training in the less proficient language resulted in improvement in that language as well as in the untrained language, showing evidence of CLT. While this study suggested a possible relation between premorbid proficiency and CLT, findings across the literature are mixed with respect to (premorbid and post-morbid) proficiency in the trained language and CLT (Ansaldo & Ghazi-Saidi, 2014). As with predicting recovery patterns, there are currently no known predictor variables for CLT, though evidence suggests that comprehension skills may promote greater CLT (Faroqi-Shah et al., 2010). Some researchers have suggested that the structural similarity of languages may promote more transfer than in those that are more structurally distinct (Ansaldo & Ghazi-Saidi, 2014), while others argue there is not enough data to substantiate this position (Faroqi-Shah et al., 2010). One potential manipulation that could be used to support CLT in lexical treatment is the use of cognates. Cognates are words from two languages that substantially overlap in form and meaning such as English lion/Spanish leo´n, and Afrikaans meisie/Dutch meisje (girl in English). Since approximately 25,000 English cognates are shared with languages such as Spanish, Portuguese, French, Romanian, Catalan, and more (see www.cognates .com), using at least some cognates in treatment may potentially facilitate CLT given the crosslinguistic engagement of phonology, orthography, and semantics. While the treatment literature is not clear on the use of cognates, SLPs can be cognizant of the use of cognates in treatment and potentially evaluate CLT of both cognates and non-cognates to gain insight into the potential effectiveness with individual patients (see Paap, Chapter 22, this volume, for issues related to the processing of cognates in non-impaired bilingual populations).

6 Conclusion As demonstrated throughout this chapter, there are numerous parallels between the state of knowledge about language disorders in bilingual children and adults as well as between the best practices for and challenges to serving bilingual children and adults with language disorders. In particular, there is a paucity of research on the nature of language disorders in bilinguals across the lifespan and on evidence-based practices for assessment and intervention. Yet there is consensus that language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

Bilingualism in Clinical Linguistics

389

disorders affect both languages in children and adults and that bilingualism in and of itself does not cause or contribute to core language impairment. However, the nature of the languages spoken by bilingual individuals may affect the manifestation of disorder characteristics or errors produced in each language. In general, the same sources of data are used to gather information about the language abilities of bilingual children and adults during the assessment process. However, language disorders are difficult to determine and characterize across the lifespan due to a number of factors that include differences in age of exposure, the amount and nature of exposure and usage of the two languages, and changes in exposure and usage over individuals’ lives. All these factors result in differences, and potentially fluidity, in language proficiencies within and between bilinguals. These differences among bilingual individuals make the diagnostic decisionmaking process highly challenging. Added to this challenge is the lack of valid and reliable standardized tests for use with bilinguals who speak specific pairs of languages. Yet these are needed in order to better serve the very diverse populations of bilingual children and adults in clinics, hospitals, and schools all around the world. While the treatment literature has some limitations, SLPs can find reports of case studies and theoretically driven treatment approaches that can inform the development of individualized yet evidence-based treatment materials and approaches. Such principled yet individualized assessment and treatment approaches are warranted in bilingual children and adults given the inherent variability of bilinguals’ communication backgrounds and needs in addition to the variability of bilingual language disorders themselves.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:48:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.020

20 Doing and Undoing Bilingualism in Education Ofelia García and Ruanni Tupas

1 Introduction Nation-states usually control education to ensure that only the society’s authorized values, histories, and ways of languaging (i.e., the social act of using language) are learned (or not). Thus, most schools throughout the world continue to build their programs and instruction on monolingual models. In the few cases when bilingualism in education is promoted by nation-states, the complex multilingual practices of youth are mostly ignored, and sometimes even stigmatized. This chapter explores the ways in which educators, policy makers, and researchers in different societal contexts understand bilingualism, and the ways in which bilingualism and multilingualism are done and undone, learned and unlearned in various educational settings. In keeping with the conventions adopted in this Handbook, when we speak about bilingualism and bilinguals from now on, we are also encompassing multilingualism and multilinguals. Throughout the chapter attention is paid to the ways in which the monolingual ethos continues to be upheld in schools, as well as the reasons for doing so. We point to the consequences that this monolingual ethos has had for bilingual learners, and we propose taking up a more dynamic understanding of bilingualism and adopting what we describe as translanguaging, or using language not as a system with socially and politically defined boundaries, but as a dynamic and fluid linguistic repertoire that draws from a unitary system instead of the named languages of nation-states. We propose that translanguaging might be a way of working against the ways in which the complex linguistic practices of bilingual learners have been ignored in schools. We first describe the competing understandings of the bilingualism of learners in schools, as we also look at the difference between schools’ expected performances of bilingualism and those in society and bilingual communities. We then turn to the ways in which bilingualism is done and undone, learned and unlearned in schools. Throughout the chapter we

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

391

point to the role that education scholars and researchers have had in contributing to the construction of our present understandings of bilingualism in education and the reasons for that construction.

2 Competing Understandings of the Bilingualism of Learners in Schools What is bilingualism and who are bilingual learners? Depending on the sociopolitical educational context and the power of the language and learners, these definitions could be totally different. In this section we look at how bilingualism in education is differently understood depending on the social characteristics of the learners. We also describe how performances of bilingualism in school differ from those of bilingual communities and individuals.

2.1 Types of Learners and Bilingualism The ways in which the bilingualism of learners is understood most often has to do with the social standing of the learners and the dominance, and thus market value, of the languages that make up their bilingualism (see Grin, Chapter 9, this volume). Throughout the world, the bilingualism of elite speakers (what Fishman, 1977, called elite bilingualism) is accepted and praised in schools as an accomplishment. This elite bilingualism is understood as being composed of two different languages, at least one of which is a dominant European language, today most likely English. Schools not only celebrate this type of bilingualism but also promote it. Thus, language education programs are developed to ensure that this type of bilingualism flourishes. Bilingualism for white learners of the powerful class is considered a resource (see Lo Bianco, Chapter 8, this volume, for further discussion). In contrast, the bilingualism of indigenous, conquered, colonized, refugee, or immigrant youth in schools (what has been called by Fishman, 1977, folk bilingualism) is often perceived as a problem. To eradicate the bilingualism of these poor students of color, schools most often impose monolingual education. Sometimes schools enact more benevolent forms of extermination – transitional bilingual programs, as well as some forms of mother tongue-based multilingual education programs – where the students’ home languages are used only until they learn the dominant language. Likewise, deaf youth bilingualism in schools is considered a problem, but this time it is seen as a medical condition. The solutions for eradication are not educational, but technological ones, most recently in the form of cochlear implants. Deaf bilingualism is recognized as being bimodal, meaning that deaf youth can use a sign language and can write an oral

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

392

OF ELIA GARCÍA AN D R UAN N I TU PAS

language (Swanwick, Wright, & Salter, 2016; for further discussion, see Tang & Sze, Chapter 24, this volume). It is elite bilingualism that is most valued in schools and promoted. Bilingualism considered folk and bimodal is not only stigmatized in schools, but one of the functions of schools is to eradicate it. The next section focuses on the two tools that schools use to privilege elite bilingualism over the bilingualism of the many, viz., language standardization and a monoglossic ideology.

2.2 Bilingualism of Learners in School One of the most important functions of schools has always been to teach a standardized form of language, especially written language, which is considered academic. Thus, it is not surprising that bilingualism in schools is accepted only as the use of two or more standardized dominant oral languages. The many sign languages, and the languages of minoritized groups that have not been standardized and are not written, as well as the complex language practices of bilingual speakers, are excluded from the authorized definition of bilingualism in education. Sociolinguistic and psycholinguistic research has shown that bilinguals do not use their different language practices for the same purposes and are not in any way linguistically balanced. Yet schools expect students to have “equal” proficiency in two or more languages, and to use them separately, as autonomous wholes. Bilinguals in schools are expected to be two monolinguals in one (Grosjean, 1982). Moreover, monolingual assessments perpetuate this view of bilingualism (see Treffers-Daller, Chapter 15, this volume, for related discussion). Garcı´a (2009a) has referred to this ideology about bilingualism as monoglossic (see also Fuller, Chapter 6, this volume). An exception to this monoglossic understanding of bilingualism in education is the Council of Europe’s (2001) efforts to promote plurilingualism, defined as the ability to “which all knowledge and experience of language contributes and in which languages interrelate and interact” (p. 4). But although this flexible plurilingualism is promoted for university students in Europe, bilingual learners in primary and secondary schools all over the world, including Europe, are most often subjected to educational programs with monolingual monoglossic ideologies. For example, many plurilingual refugee youth in European schools first have to demonstrate “full” proficiency in the dominant language in order to have access to universities. This is the case too among the youth in minoritized language communities in Asia, whose plurilingual practices, which are themselves products of multilingual forms of education, are devalued; full proficiency in the dominant national language is expected from them. For example, Mandarin is a compulsory examination subject for admission into higher education in China (Adamson & Feng, 2009).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

393

Bilingualism in schools continues to be viewed as additive (Lambert, 1974). The idea is that schools add a second or third language to students’ first language. Scholars working with this additive bilingualism research perspective have made important contributions. For example, they have shown that the use of the learner’s first language in instruction is important to develop a second language. And this research has conclusively shown that a bilingual additive education is superior to a monolingual subtractive education (see, among others, Baker, 2011; Collier & Thomas, 2017; Cummins, 1979a; Lindholm-Leary, 2014). And yet this monoglossic view of bilingualism as being simply additive ignores the complex fluid language practices of bilingual people throughout history (Canagarajah & Liyanage, 2012) and throughout the world. Learners’ bilingualism is not simply additive, but dynamic (Garcı´a, 2009a) and language practices are heteroglossic (Bakhtin, 1981), containing within them not one or more unitary languages, but features that “interanimate each other” (p. 47). Eastern scholars like Khubchandani (1997) have long held this view, but it is only recently that Western scholars of bilingualism have acknowledged this. Garcı´a (2009a) has referred to two types of bilingualism where fluid linguistic practices are evident today: recursive bilingualism and dynamic bilingualism. Minoritized groups who through processes of colonization and conquest have experienced a great deal of loss of their linguistic practices of origin exhibit a bilingualism that we can call dynamic recursive. This has to do with the ways in which speakers bring in bits and pieces of their language practices of origin into their present language practices, as they recover them. But all bilingual speakers throughout history, and especially in Africa, Asia, and Latin America, experience what has been called dynamic bilingualism, a way of languaging that responds to speakers’ own agentive way of using their language repertoire, instead of using what society calls one language separately from what is named as another one. Western scholars have recently begun to pay more attention to dynamic bilingualism and heteroglossic language practices, as their own language practices have been affected by globalization and a neoliberal economy that supports the free flow of capitalism in ways that benefit transnational corporations and economic elites, and that results in the commodification of bilingualism (Flores, 2013). Bilingual speakers are now said to have mobile resources (Blommaert, 2010) or flexible bilingual practices (Blackledge & Creese, 2010). This type of dynamic bilingualism has been called by many names with slightly different meanings: polylanguaging/polylingual languaging (Jørgensen, 2008), metrolingualism (Otsuji & Pennycook, 2010), translingual practices (Canagarajah, 2013), and translanguaging (Blackledge & Creese, 2010; Garcı´a, 2009a; Garcı´a & Li Wei, 2014; Otheguy, Garcı´a, & Reid, 2015). Although sociolinguists increasingly acknowledge the fluid

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

394

OF ELIA GARCÍA AN D R UAN N I TU PAS

dynamic practices of bilinguals, these are mostly ignored and stigmatized in schools (Pulinx, van Avermaet, & Ag˘irdag˘, 2017). The monoglossic ideology of bilingualism in education as only consisting of standardized languages performed equally well for all purposes leaves out most learners, but most especially those who have been minoritized. Learners who have been conquered or colonized, or who have had to migrate to a different country, or who are seen as clinically deaf, most often have not been given the opportunity to learn what are considered their home languages in schools, with the exception of those fortunate enough to be enrolled in bilingual/multilingual programs or so-called heritage language classes. And yet, in many of these programs, only the standardized version of the home language is legitimized, often leaving these bilingual learners feeling that their home language practices are also deficient. Their home languages are also blamed for interference with the dominant language. Bilingual learner’s language practices are always in contact, or said another way, their languaging always includes features that are said to belong to two or more different languages, although it is the product of what Otheguy et al. (2015) have called their unitary language system. Thus, bilingual students are evaluated in schools as having poor command of their named languages, and are often labeled as “semilingual,” or even “alingual.” This creates great linguistic insecurity among bilingual learners. Because these students’ dynamic bilingualism is not recognized as fitting the acceptable definition of bilingualism in education (i.e., an addition of two or more standardized languages), their language is dismissed as inadequate, poor, deficient. It is important to underscore that the stigmatization of the bilingualism of minoritized learners is produced precisely through the process of schooling. It is the expectations that schools hold about the language of bilinguals that create the educational failure of minoritized bilingual students. The least recognition of bilingualism in education is reserved for those who are deaf. The language capacity of the Deaf relies heavily on bodily gestures, and yet oralism remained largely the only way to educate the Deaf until the 1980s. Even after signed languages were standardized, Deaf learners’ bimodal bilingualism has remained mostly unrecognized in education. With deaf youth increasingly immigrating to contexts with different signed and oral languages, Deaf bilingualism is becoming more complex, and yet continues to be mostly absent in schools (Swanwick et al., 2016; see Tang & Sze, Chapter 24, this volume). It is education, then, through the process of accepting only the balanced and equal use of standardized languages as the only form of bilingualism, that is responsible for the academic failure of many bilingual learners. Bilingual learners and their language practices are not celebrated, but measured against monolingual learners in the majority language. The lens of deficiency is most often cast over them.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

395

2.3 Performances of Bilinguals in Communities Whereas schools continue to demand equal and balanced performances in standardized languages for academic success, bilingual practices go beyond the borders in which schools construct and constrain notions of language. In schools, bilingual learners and their families are subjected to concepts that make little sense in their everyday language performances. When bilingual families register their children in schools, they are stumped when asked: Which language do you speak at home? Which is your dominant language? Which is your mother tongue? Which is your first language? These are all questions that monolingual families can reply to easily, but for many bilingual families, these answers are complicated. In today’s globalized world and complex family structures, families’ language use is much more complex than the single answers expected by these questions. In many educational settings, bilingual speakers are socialized very early to recognize some features of their language performances as belonging to one named language or another. However, in informal unmonitored situations, many bilinguals use their full linguistic and semiotic repertoire to make meaning among themselves. Unmonitored situations include interactions in contexts such as bilingual homes or neighborhoods where there is no external authority watching over how speakers use what are deemed to be appropriate linguistic features and constraining the ways in which bilinguals may deploy their meaning-making resources. Unlike in schools, where only certain language performances are authorized, these more informal contexts may allow the use of a bilinguals’ full language repertoire, especially when interacting with other bilinguals. For example, Ofelia’s daughter recently dropped off her two-year-old son and the stroller. She knew that Ofelia had a hard time operating the stroller, so she said, as she demonstrated: “Mami, atie´ndeme please. Push the palanquita pa’rriba to open the stroller.” In many schools in the United States, she would have been forced to say: “Push the lever up to open the stroller.” In a school in Cuba, she would have been asked to say: “Empuja la palanquita para arriba para abrir el coche.” But Ofelia and her daughter and grandchild live a bilingual life in the United States, so in the life of the bilingual family it is not necessary to use solely features of what is considered English, or what is considered Spanish, to have a meaningful interaction. Ofelia and her daughter are free to use their entire language and semiotic repertoire (see De Houwer, Chapter 17, this volume, for discussion of language choice). Like Ofelia and her daughter, many bilingual speakers are constantly translanguaging, that is, using their unitary language repertoire (Otheguy et al., 2015) purposefully and strategically. In unmonitored situations, they are free to use their own entire language repertoire in agentive ways. But schools, with their monoglossic ideologies about bilingualism, are the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

396

OF ELIA GARCÍA AN D R UAN N I TU PAS

language-monitoring situations par excellence. In monitoring the language use of students, schools turn bilingual speakers to governable subjects (Flores, 2013) whose social and economic opportunities are constrained through the imposition of an artificial, monolingual, monoglossic language norm. The translanguaging of bilingual students is evaluated negatively and simply not recognized, for this would require a heteroglossic bilingual lens of liberation of minoritized speakers, not domination. That is, for translanguaging practices to be recognized as legitimate, language would have to cease to be used as a way to produce governable subjects. Instead, however, schools label the students’ translanguaging as linguistic interference. Based on a monolingual standard, the bilinguals’ translanguaging is subjected to an analysis of language contact phenomena, described as loans, calques, code-switching, and so on (see Aalberse & Muysken, Chapter 26, this volume), introduced by Western scholars of bilingual studies beginning with Wilhelm von Humboldt in the late 18th and early 19th centuries and particularly since the work by Uriel Weinreich and Einar Haugen in the mid-20th century. In education, the traces of such language contact phenomena and code-switching have often been seen as a kind of interference, that is, as deviations from the standard named language, the only one legitimized for education. Bilinguals’ translanguaging is constructed as an aberration from a monolingual standard, and then identified as the cause of academic failure (see Bigelow & Collins, Chapter 2, this volume, for a complementary perspective). We have reviewed the competing understandings of bilingualism in the lives of different bilingual learners, both in and out of school. In the next section, we turn to describing how bilingualism has been promoted or not in schools, for whom, and why.

3 Ways in which Bilingualism and Biliteracy are Learned and Unlearned in Schools Schools have had a most important role in doing, but mostly undoing, bilingualism. As we have said, most educational systems in the world promote and enforce monolingualism in education in order to inculcate the state’s children with certain values, histories, and ways of using the dominant language. In so doing, schools produce governable subjects and serve the interests of the elites. The academic subject called language arts in some educational systems (i.e., classes that focus on reading, writing, and communicating in the language of instruction) is a most important part of the monolingual monoglossic educational enterprise. But as today’s globalized world has started to acknowledge its vast bilingualism, specialized educational programs that focus on either promoting bilingualism or ensuring that it is

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

397

controlled or eradicated have proliferated. Next, we discuss the two different goals of these specialized educational programs – learning bilingualism or unlearning bilingualism – as we discuss whom they serve, how, and why.

3.1 Learning to Do Bilingualism through Schools We begin by identifying how educational programs to learn to be bilingual serve three types of learners and social goals: elite and middle-class learners, minoritized learners, and learners in multilingual states. Elite and middle-class learners, often monolingual to begin with, learn to be bilingual in school by adding a so-called foreign or world language. Throughout the world, and especially in Asia, the dominant language added is increasingly English (Park, 2009; Park & Wee, 2012), although Mandarin Chinese is also competing for attention. To learn to be bilingual, elite and middle-class learners participate in different types of educational programs that are described in depth by Juan-Garau and Lyster (Chapter 11, this volume) and Mun˜oz and Spada (Chapter 12, this volume): • Foreign/world language education programs, the most popular way of learning an additional language. These programs focus on the acquisition of an additional language, and abound in secondary and tertiary education. • Content and Language Integrated Learning (CLIL) programs, where the additional language is used as medium of instruction, going beyond core foreign language programs. • Immersion bilingual education programs, where majority children, usually young, are initially taught through the medium of an additional language, until both languages achieve instructional parity. Especially popular today are English-medium instruction (EMI) programs, where often immersion in English outweighs instruction in the other language. • Developmental bilingual education programs, used to instruct elite and middle-class children who want to become bilingual, and where two languages are used as medium of instruction. Minoritized youth learn to be bilingual through the very same developmental bilingual education programs listed in the last item just given (see also Bigelow & Collins, Chapter 2, this volume). Especially since the ethnic revival and the civil rights and decolonization efforts of the mid-20th century, some minoritized groups have been successful in demanding such programs for their children. This is the case of many indigenous groups where language revitalization projects are under way. Perhaps one of the most successful indigenous groups in this regard has been the Ma¯oris of New Zealand. This has also been the case of many regional minorities, for example, the Welsh in the United Kingdom and the Basque

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

398

OF ELIA GARCÍA AN D R UAN N I TU PAS

in Spain. Although most efforts to make minoritized youth bilingual and reverse language loss take place in well-established developmental bilingual education programs that focus on developing the minority language, heritage language education can also help in this process. These programs, in which instruction is solely in the minority language, are generally found in secondary schools and university programs or in complementary schools (Blackledge & Creese, 2010). Some nation-states that aim to be multilingual also make available some multilingual school options for their youth. The trend, of growing importance in some Asian countries in particular, refers to efforts by many countries to ensure that their entire population becomes multilingual through multilingual education programs (MLE). The reasons for MLE programs vary. In some contexts the goal is to use the pupils’ mother tongues to develop proficiency in two national languages, since developmental bilingual education in these highly multilingual contexts is deemed insufficient. This is what is known as mother tongue-based multilingual education (MTB-MLE) (Tupas, 2015). The Philippines, East Timor, and India are good examples of societies that have introduced MTB-MLE to ensure the inclusion of the languages of minoritized youth in education, while developing other national languages (Curaming & Calidjernih, 2014; Mohanty, 2006; Tupas, 2015). In yet other contexts, multilingual education is fueled by the desire to teach English so as to access the symbolic goods that English supposedly represents (Adamson & Feng, 2009; Hu, 2007), while also supporting national languages. This is the case, for example, of Kazakhstan, where trilingual education in Kazakh, Russian, and English is being promoted. This is also the case of the Basque Autonomous Region of Spain, where bilingual education programs in Spanish and Euskara have been transformed to trilingual programs in Spanish, Euskara, and English (Cenoz, 2009). An unusual case of MLE is that of Luxembourg, a country that promotes trilingual education to ensure that not only Luxembourgish but also German and French are spoken throughout its territory. Implementation of MLE programs for all citizens remains elusive. On the one hand, only schools in the most resourced areas with wealthier students seem to be successful. On the other hand, in most cases, there is a privileging of resources in English and the national language (e.g., Mandarin in China, Thai in Thailand, Khmer in Cambodia). In Luxembourg, despite a trilingual education system, immigrants, especially those of Portuguese descent, do more poorly than others. Thus, despite efforts to make multilingual education equitable across social classes in many contexts, socioeconomic class and ethnicity have huge impact on youth’s access to multilingual education. Tupas (2015), referring to MTB-MLE education in Southeast Asia, warns of the inequalities of multilingualism (see also Lo Bianco, Chapter 8, this volume, for related discussion).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

399

3.2 Unlearning Doing Bilingualism through School Most monolingual education in the world promotes the unlearning of bilingualism, and many specialized educational programs focus on unlearning bilingualism (Skutnabb-Kangas, 2000). Among the most important types of programs to unlearn bilingualism are second language programs in the dominant language, transitional bilingual education, EMI programs, and even some developmental bilingual education and MTB-MLE programs. Second language programs can be of different types. But whether the teacher pulls out some students, supports them within a classroom, or is the sole teacher of the class, the minoritized students in these programs are deemed to be in need of remediation (see also Bigelow & Collins, Chapter 2, this volume). This is also the case of students in transitional bilingual education programs, used throughout Africa, Asia, Latin America, and the United States to unlearn the students’ nondominant language practices, with their home languages used only temporarily. The accelerated growth of EMI programs is also viewed as a threat to bilingualism (Dearden, 2014). Throughout the world, English is being introduced as a medium of instruction because it has been found to attract students and increase the profits of educational institutions that are the product of the neoliberal push for the privatization of education (Park, 2011). Taguchi (2014) presents case studies of EMI education in China, Qatar, Japan, Hong Kong, and South Africa. It is important to underscore that although developmental bilingual education and MTB-MLE programs have been introduced as a way to break the cycle of educational failure among pupils who have been educated in languages they are unfamiliar with, they can often function in the same ways as transitional bilingual education. In some contexts, developmental bilingual education programs pay more attention to the development of the majority language than to the minority language. Most MTB-MLE programs are implemented only in kindergarten and the first three years of education. The goal of these programs is for students to transition slowly into monolingualism in the national language, or into bilingualism in the national language and a foreign language (mostly English nowadays) (Dekker & Young, 2005), thus helping children unlearn their bilingual and multilingual practices (Tupas, 2015).

4 Language Education Policy Issues to Teach (or not) Bilingualism After decisions are made by educational authorities to set up programs to learn or unlearn bilingualism, the real work of educational policy makers and educators starts (see Lo Bianco, Chapter 8, this volume). In this section we discuss issues having to do with the language policy of the programs, focusing specifically on timing decisions. Language education policies

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

400

OF ELIA GARCÍA AN D R UAN N I TU PAS

guide the ways in which the education program and curriculum are organized. The main language policy decisions have to do with time – the time in which the educational program starts, how long it takes, how much time is devoted to it, how language is allocated, as well as whether the languages are introduced simultaneously or sequentially. Programs to unlearn bilingualism start as early as learners arrive in schools. To learn bilingualism, on the other hand, there are different types of programs: some start in early childhood, others when learners enter primary schools, still others when students enter middle school or even secondary schools or universities. There has been much controversy about when it is most beneficial to start these programs. Although there are advantages to starting earlier, especially in phonological benefits, there is research that shows that the same milestones can be achieved if learning for bilingualism starts in middle school or secondary schools, as long as there are adequate resources and ample investment in the learning (see further discussion in Juan-Garau & Lyster, Chapter 11, this volume, and Mun˜oz & Spada, Chapter 12, this volume). There is scholarly consensus, however, that the most success in achieving bilingualism occurs when students are educated bilingually throughout their schooling (Cummins, 1979a; see also Collier & Thomas, 2017). Another language education policy decision that has to be made has to do with how much time should be devoted to instruction in one or the other language. When two languages are used to unlearn bilingualism, the language policy calls for less and less use of the home language of the students. Sometimes the home language is used initially 90% of the time, or even 100% of the time, with that proportion diminished, until it eventually almost disappears, relegated maybe to a language arts class. Scholars seem to agree that the time intensity decision for programs to learn bilingualism has to do with the social standing of the students and the status of the named language. In situations where schools work to revitalize the minoritized language, more time needs to be devoted to that language in order to work against the power of the dominant language in society. This is also the case when a group wants its children to learn a language that is not dominant in society. In many of these cases, immersion bilingual programs start early, with 100% of the time devoted initially to the language that is not dominant in the social context. But bilingual and multilingual education programs and CLIL programs most often use two or more languages in different combinations, sometimes for different grades, and depending on the instructional material and teachers that are available. The scholarly recommendation, however, is that no less than 50% of the instructional time be devoted to instruction through the additional language so that there is parity among languages (Baker, 2011; see Juan-Garau & Lyster, Chapter 11, this volume). In reality, however, we have seen many educational programs where learners become bilingual even though the nondominant language is used for less than half of the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

401

instructional time. In these programs, however, the nondominant language is validated and deemed important, spoken by administrators and teachers, and treasured by the community. Jim Cummins suggested already in 1979 (see 1979a, 1979b) that it takes three to five years for emergent bilingual learners to be able to perform academic tasks in what is considered an additional language. In reality, bilingualism is always emerging, for the features that we use and the linguistic performances in which we engage depend on the interlocutors and the tasks that we are performing. There is no end point to bilingual learning. That is, bilingual learners never “have” bilingualism; they “do” bilingualism; they engage in what we might call bilingual performances. A most important language policy decision has to do with whether the two languages ought to be used separately in instruction. Classroom teachers who work against bilingualism, as well as those who work for bilingualism, often end up establishing and firming up a diglossic relationship between two or more languages that maintains the hegemony of the dominant language and solidifies a linguistic hierarchization based on the power and market value of the language being learned. Foreign language programs, second language programs in a dominant language, and CLIL, immersion, and EMI programs all claim to use only the so-called target language so that students can be allegedly immersed in the language of instruction (for this position, see Juan-Garau & Lyster, Chapter 11, this volume). In reality, however, much use is often made of the “other” language, as instruction has to be scaffolded so that meaning is made (Garcı´a, 2009a). Most types of bilingual and MTB-MLE programs formally advocate for a language separation or isolation approach. Languages are said to be separated according to day or week, time of day, subject, and so on. Often the argument made for the language separation approach in bilingual programs has to do with protecting the minoritized language and ensuring its important place in the curriculum (Juan-Garau & Lyster, Chapter 11, this volume). Once again, this language separation is artificial, not responding to sociolinguistic principles of language use and adaptability (Fishman, 1977) or to the translanguaging practices of bilingual speakers. The complete language separation arrangements in MTB/MLE programs ensure that only standardized named languages are admitted, ignoring the more dynamic language practices of bilingual learners, that is, their translanguaging. In programs to learn bilingualism, another policy decision has to be made as to whether both languages should be introduced at the same time (simultaneously), or whether one language needs to be taught first before an additional language is introduced (sequentially). In most programs to learn bilingualism, the additional language is introduced sequentially. This is the case of all foreign language and CLIL programs (Mun˜oz & Spada, Chapter 12, this volume). It is also the case of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

402

OF ELIA GARCÍA AN D R UAN N I TU PAS

immersion bilingual education (Juan-Garau & Lyster, Chapter 11, this volume), and of most developmental bilingual education and MLE programs. Many transitional bilingual education and MTB-MLE programs also prefer this sequential introduction to additional languages. This is evident particularly in educating indigenous poor children in the African, Latin American, and Asian contexts. When more than one language is used in primary education, the traditional thinking has been that one language ought to be used first to teach literacy, and only then could another language be introduced. Especially when languages have different scripts, the typical understanding is that children need to master the conventions in one language first. However, this conventional wisdom of the sequential introduction of languages for literacy has repeatedly been questioned throughout history (see, for example, Escamilla, Hopewell, Butvilofsky et al., 2014; Fishman, 1977; for more on the acquisition of biliteracy, see Bigelow & Collins, Chapter 2, this volume).

5 Classroom Instruction to Teach Minoritized Learners for Bilingualism Schools, with their monolingual monoglossic ideologies, have been quite successful in teaching for monolingualism, even when on the surface bilingualism is the goal. This has been especially so when the learners are minoritized bilingual learners. In this section we focus simply on teaching for bilingualism, and on teaching those who are most vulnerable – minoritized bilingual learners who are ironically bilingual to start with. We turn to issues of pedagogies and resources that are most important to successfully deliver the classroom instruction to teach for bilingualism. We focus on three questions: What pedagogical practices and principles are needed to engage these students in an education that would support and extend their own bilingual practices? What shifts would be needed in assessment? What type of teachers and instructional material would need to be included? We offer a critical pedagogical approach for bilingual learners, an approach that is based on a dynamic and heteroglossic understanding of bilingualism. We describe a holistic approach to bilingual education that might be a way of working with the complex linguistic practices of bilingual learners, that is, their translanguaging.

5.1 Pedagogy for Bilingualism: Translanguaging Until the mid-20th century, foreign language education programs emphasized grammar and reading. Being bilingual was a mark of high culture, and knowing the grammar and literature of another language group was said to provide one with superior understandings about other cultures and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

403

the world, but also about one’s own. In fact, the study of foreign language was often said to help develop one’s own language. This emphasis began to change in the mid-20th century when the grammar-translation method was abandoned for more communicative approaches. Oracy was then emphasized and the study of traditional grammar receded, as the emphasis on literacy skills gave way to oral communication. The communicative approach to language learning resulted in the development of immersion pedagogies, where the language being learned was solely used as a medium of instruction. The conventional wisdom became that bilingual learners need to perform solely in the language of instruction, always keeping the two languages separate and isolated from each other. But neither the explicit language pedagogy that focused on the structure of the language being learned, nor the immersion approach that focused on communicative interaction in only one language at a time resulted in being appropriate for the schooling of minoritized bilingual learners in the long run. To engage these learners, a pedagogy must focus on leveraging the entire linguistic repertoire of bilinguals and their dynamic language practices. How can the concept of translanguaging be turned into pedagogical practice? A translanguaging pedagogy relies on different epistemological understandings of language and bilingualism that go beyond the concept of two or more named languages (Garcı´a & Li Wei, 2014; Otheguy et al., 2015). Although translanguaging recognizes the material and real effects that the construction of named languages has had in the lives of learners, it does not solely respond to the external constructions of standardized named languages of nation-states that schools reify (see Fuller, Chapter 6, this volume, for an analysis of the language ideologies associated with the project of the nation-state). Rather, a translanguaging pedagogy leverages the unitary full repertoire of linguistic features and semiotic practices that bilingual learners use. Rather than starting with named languages, a translanguaging pedagogy starts with the language practices of bilingual learners. By leveraging their entire language repertoire, translanguaging returns the power of language to speakers and engages their communicative human potential, rather than authorize only the conventions of named languages that have been codified by the nation-state to develop governable subjects. Garcı´a, Johnson, and Seltzer (2017) identify three components of a translanguaging pedagogy: stance, design, and shifts. Teachers who leverage bilingual students’ translanguaging have a stance, a deep belief that their students’ language practices are a resource that transcends the standardized language(s) of schools and can be used to make meaning of academic tasks, learn, and develop bilingual practices. Besides a stance, these teachers also design units, lessons, instruction, and assessment that integrate home and school languaging practices. But beyond the translanguaging design, a translanguaging pedagogy also relies on teacher shifts,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

404

OF ELIA GARCÍA AN D R UAN N I TU PAS

the appropriate moment-by-moment decisions teachers make to respond to learners’ languaging. Translanguaging pedagogy supports students’ bilingual practices, even in classrooms that aim at making them unlearn their bilingualism. It makes space for students’ bilingualism and ways of knowing. It also supports bilingual students’ socioemotional development and bilingual identities. In doing this, a translanguaging pedagogy is a tool for social justice, disrupting the linguistic hierarchizations that schools have created (Garcı´a et al., 2017). A translanguaging pedagogy does not only scaffold the learning of a new language. It transforms the notions that learners have about named languages and legitimizes the linguistic practices of bilingual speakers (Garcı´a & Li Wei, 2014; Otheguy et al., 2015). Furthermore, a translanguaging pedagogy engages students in a critical examination of how named languages have been invented (Makoni & Pennycook, 2007) and used in schools to create inequalities. Only by leveraging the fluid language practices that make up the linguistic repertoire of bilinguals, and by building their critical understandings of how the invention of named languages has worked to produce governable subjects, will bilingual youth be able to engage in their education and become agentive learners. Some literacy programs have begun to emphasize that it is important for students to leverage their translingual practices especially in writing (Escamilla et al., 2014; Horner, Lu, Royster, & Trimbur, 2011). But this recognition is still not in any way tapped by assessments, especially of the standardized type, as we discuss in the next section.

5.2 Assessment for Bilingualism: Translanguaging Assessment of language proficiency is intimately tied to instruction. Language proficiency has been constructed as knowing conventional language features that have been standardized and using them appropriately to speak, read, and write. But the translanguaging performances of bilingual learners many times expose linguistic features that go beyond those of the monolingual standard. Yet the psychometric construction of language assessment requires that language be limited to specific features that can be explicitly and reliably measured. Thus, the use of language for real purposes in social and academic interactions is rarely measured (see also Treffers-Daller, Chapter 15, this volume). Some efforts have begun to appear, but these tend to be limited to classroom assessments (De Backer, van Avermaet, & Slembrouck, 2017). Moreover, assessment of language proficiency is most often used to limit access to educational opportunities. A form of assessment that would measure linguistic performances not for access to educational opportunities, but for success in schools would be more authentic, evaluating the success of the multiple creative interactions of how learners do language in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

405

different circumstances and for various tasks. But for now, the standardized assessment of the ability to language in order to make an argument, tell a joke, find text-based evidence, etc., is hard to find. Technological innovation might help the testing industry come up with these more authentic assessments of language performance. But for this to occur, attention has to be paid to the psychometric measures of consequential validity, that is, to the differential effects of the tests between bilingual minoritized students and other groups, rather than simply on reliability measures. That is, language testing must be explicitly connected to social justice (Deygers, 2017). Most summative assessments are almost never concerned with how learners “do” language, but with whether learners “have” certain features that are considered standard. It is then important to ensure that teachers of bilingual learners remain critical and vigilant about the pitfalls of standardized language assessments in one or another named language. Assessing students in two or more languages is not sufficient if the comparison group consists of monolinguals (Treffers-Daller, Chapter 15, this volume), and if the bilingual context of the lives of bilinguals is not taken into account (Bigelow & Collins, Chapter 2, this volume). Besides holding a critical stance about standardized language assessments, teachers need to develop and use their own formative assessments where bilingual students’ language ability is assessed holistically, without regard for their use of standardized language features of one named language or another. That is, true and fair language assessments must match the way in which bilingual people use language in society, not an artificially constructed monolingual norm. Assessments are important to enable teachers of bilingual learners to differentiate between the student’s language capacities and their language proficiency in one or another named language. These types of language assessments can then be used to shape instruction so that bilingual learners achieve educational success.

5.3

Teachers and Materials for Bilingual Learning: Translanguaging The most important source of success for bilingual learners is the teacher herself. This has to do with the fact that teachers who have the appropriate stance (in the sense outlined by Garcı´a et al., 2017) can go beyond the strict monolingual monoglossic language policies of schools. It is most important to have teachers who not only are bilingual, but who are critical of the ways in which language and bilingualism have been constructed in schools and the reasons for doing so. That is, teachers must hold a translanguaging stance. Only then will teachers be able to assume the role that will lead to success in learning for bilingual learners. But beyond the stance, teachers need to be prepared to design instruction and assessment that will lead to learners’ success (Garcı´a et al., 2017).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

406

OF ELIA GARCÍA AN D R UAN N I TU PAS

Schools that work for bilingualism need to have appropriate instructional material that is rich and culturally appropriate, besides being bilingual. Bilingualism cannot be developed just in oral interaction. The schools’ version of bilingualism requires books that can be read, discussed, and analyzed as well as authors that write using different language practices, including translanguaging ones. These books cannot solely be fiction, but must also include nonfiction books and textbooks. Local publishers in so-called small languages cannot compete with the publication conglomerates of the developed world, but locally produced books must be held in high esteem by educators, students, and families. These books must reflect who the students are, their language practices, and their translanguaging, instead of depicting only a standardized language that is artificial and that cannot be used to communicate meaningfully in the community. The writing that bilingual learners produce must be displayed publicly, with the language produced by students, and not simply the edited standardized language(s). It is unfair to ask educators to think critically about the controlling role that standardized languages have had unless they have been aptly educated to do so. Teacher education programs need to shift away from teaching about bilingualism as strictly additive, and instead encompass the translanguaging practices of bilingual communities. Only by doing careful sociolinguistic ethnographies of bilingual communities will prospective teachers be able to develop a stance that views language differently from the standardized notions learned in school (see Hammer & Edmonds, Chapter 19, this volume, for a similar call for careful sociolinguistic ethnography by speech and language clinicians). Only then will teachers be able to develop and incorporate instructional material that reflects the bilingual community practices and to leverage those practices in a translanguaging pedagogy.

6 Conclusion In this chapter we have critically examined how educators and educational researchers in different societal contexts understand bilingualism, and the ways in which bilingualism is learned and unlearned by youth in various educational settings. Our review of the state of bilingualism in education shows clearly that although most bilingual learners come to schools with diverse language practices, schools continue to build their programs and instruction on monolingual models. This has to do with the role that schools have had in controlling opportunities and resources in society and producing governable subjects. In addition, we would like to note here that the continued hegemony of Western educators and scholars who are not familiar with language use in colonized contexts perpetuates misunderstandings about bilingualism in education. Although Asia, Africa, and Latin America

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

Doing and Undoing Bilingualism in Education

407

are rich in linguistic diversity, scholars from those geographical regions seldom enter the scholarly conversation of the supposed multilingual turn (May, 2014). Perhaps more than in other geographical locations, schools in these richly diverse areas continue to impose the monolingual diglossic ideology that they inherited from their colonial histories. As long as schools constrain their view of language as society’s authorized values, histories, and ways of languaging confined to standard repertoires and to named languages in isolation, they will exclude the complex linguistic practices of bilingual learners, and they will fail minoritized bilingual learners. To truly reverse how bilingualism in education is perceived, we would need to turn our attention to how minoritized bilingual communities speak, interact, and live. The situation, however, is quite hopeless, for it is precisely these subaltern people who are most subjected to schooling that is inadequate for them to learn and become educated. The sati that Spivak describes in her foundational text, “Can the subaltern speak?” (1988), is repeated daily by minoritized learners who fail in schools. But this speech of failure and despair is not visible or audible to most Western scholarship. And so schools continue to construct bilingualism to benefit the elites of the world, and especially to further consolidate the power of English in a market global economy, while minoritized bilinguals continue to be excluded from educational opportunities precisely because of the nature of their bilingualism and their translanguaging practices.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:51:08, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.021

21 Second Language Acquisition as a Road to Bilingualism Lourdes Ortega

1 Introduction Many people around the globe learn new languages at various points in life beyond early childhood and in varied contexts that are located outside the family unit. The new languages can enter people’s lives due to events that are variably self-elected, compulsory, or forced upon, personal or professional, and fulfilling or traumatic. In many cases, the development of each new language occurs through variable mixtures of instruction and naturalistic immersion. The outcomes are not guaranteed and may include language learning but also unlearning, non-learning, and even relearning at any point of the adult life. The language learning biography of famous Russian-French painter Marc Chagall, chronicled by Lvovich (2015), offers a good illustration. Born in present-day Belarus in 1887, Chagall grew up with Yiddish at home and started to learn to read in Hebrew when he reached school age (he never learned to speak it). He added spoken and written Russian at the age of 13 when he began to be schooled in that language. French came into his life when he moved to France after age 24 to pursue his passion for art. Between 1914 and 1921, Chagall spent eight years back in Russia, which soon became the Soviet Union, before moving back to France in 1922. In 1941, when he was 54, he fled from Europe and its raging World War II and sought asylum in the United States. Harriss (2003) writes that Chagall “never got used to the pace of New York life, never learned English” (no page). At age 60, Chagall returned to his beloved France and led a rich and long life, interspersed with a great deal of international travel, using French as well as Russian and perhaps some Yiddish, until his death in 1985 at 98 years of age. Chagall’s example is not singular, nor is fame a prerequisite for people to learn additional languages beyond the ones learned at a tender age in the family unit. In 2017, there were 258 million international migrants

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

409

worldwide (United Nations, 2017). This figure amounts to 3.4% of the total world population and ranks fifth after the four most populous nations in the world (China, India, the United States, and Indonesia). It is impossible to estimate how many international migrants experience learning new languages as a result of their mobility, but the experiences and needs associated with functioning in new languages are likely to be vast for many of them (see Simpson, Chapter 13, this volume). Another substantial population with language learning needs and experiences consists of international students, that is, young adults seeking a university degree outside their country of origin. In 2015, they amounted to 4.5 million or 5.6% of all higher education enrollments (Organisation for Economic Co-operation and Development [OECD], 2017). Equally important are the additional language learning experiences and needs of children and adolescents enrolled in compulsory education around the world, given that many if not most education systems include additional language learning (most often, English) in their compulsory curriculum. In the European Union (EU), for example, almost all children begin learning a foreign language in primary school and almost all continue in secondary education (Eurostat, 2017). Three-fifths of adolescents study two or more foreign languages as part of their regular schooling, which may also help explain why about two in three working-age adults in the EU report knowing a foreign language. Then there are the millions of children in the world who do not speak the school language but who are placed in the sink-orswim situation of having to learn it without the help of specific secondlanguage (L2) instruction. For example, Kuchah (2016) discusses the educational perils for a country like Cameroon, which represents 2% of Africa’s total population and features 13.5% of Africa’s total local languages within its borders, and where children’s only access to the curriculum is increasingly via English, a language that neither they nor their parents understand. And in a country like the United States, where in 2014–2015 an estimated 4.6 million students or about 9.4% of all public school children were classified as English language learners (National Center for Educational Statistics, 2017), many start off not understanding the L2 and are thrown into the mainstream classroom to learn content with no language learning support. The study of how people learn new languages across such variegated contexts beyond the time of primary language socialization and outside the family unit – in the case of Chagall, all his languages but his first language (L1) Yiddish – has given rise to the field known as second language acquisition (SLA). By convention the new language being learned is called additional language, second language, or L2. This label is an abstraction, given that the L2 may be learned as a third language (L3) or any subsequent language (Ln) and that, in many cases, a strict sequential order is difficult to establish. Moreover, people may start off as either monolinguals or bilinguals when they begin to learn the new language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

410

LOURDES ORTEGA

in question. The object of inquiry in SLA is thus sequential, late-timed bilingualism, which is studied from vantage points partly intersecting and partly diverging from the study of bilingualism in other fields represented in this Handbook. This chapter offers a selective characterization of the research conducted in SLA. In keeping with the traditional disciplinary scope, my main focus is on young adult learners and on mature adults, but where relevant I refer to studies of SLA at other life stages as well. I first characterize SLA and compare it to other fields that study bilingualism. Next, I synthesize key findings in two areas of central importance: the role of previously known languages, and the contribution of the surrounding linguistic environment. The remainder of the chapter examines the question of how successful L2 learning can be. I do so by first problematizing the notion of success, and then inspecting SLA research into two important populations of adult L2 learners: immigrants and international students. The coverage is by necessity partial, but I hope to show a selection of what is known in SLA about the trajectories, linguistic and extralinguistic, that individuals take on the path to developing an L2. I argue that the enterprise of L2 learning has a fundamental potential for success and must ultimately be understood as a road to late-timed, sequential bilingualism.

2 SLA as a Discipline The field of SLA dates back to at least the founding of the journal Language Learning in 1948, but it is the publication dates of two seminal articles that are often taken to mark its beginnings: “The Significance of Learners’ Errors,” by Pit Corder in 1967, and “Interlanguage,” by Larry Selinker in 1972. After the creation in 1979 of another flagship journal, Studies in Second Language Acquisition, SLA began to grow steadily. During the 1980s, it saw the establishment of two key conferences (the Second Language Research Forum [SLRF] in 1982, and the conference of the European Second Language Association [EuroSLA] in 1989), as well as the creation of a third flagship journal, Second Language Research, in 1985. Since the 1990s, a large number of handbooks, encyclopedias, and textbooks and a cadre of well-established doctoral programs across continents attest to the vibrancy of this field. A digital repository of instruments and materials for research into second languages, IRIS, has provided a free online platform (www.iris-database.org/iris/app/home/index) where the SLA community shares research tools since 2011 (Marsden, Mackey, & Plonsky, 2016). Newcomers to the field may be surprised by the great number of heated debates that have been published about desirable theoretical allegiances and epistemic goals (e.g., Atkinson, 2011). In the 1980s, the mainstream of SLA research was decidedly cognitive in outlook, but by the late 1990s, many researchers had begun to embrace multiple theories that aided in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

411

approaching L2 development as a phenomenon that is as fundamentally social as it is cognitive. In parallel to this social turn, by the 2000s, the investigation of linguistic development and L2 processing mechanisms had bifurcated into two main strands that continue to burgeon today. Formal generativist SLA (Slabakova, 2016) is motivated by the view that L2 acquisition must be studied as part of the genetic endowment of human language. Functional usage-based SLA (Ellis, Ro¨mer, & O’Donnell, 2016) follows the tenet that, like all language development, L2 development emerges from iterative, meaningful experiences with language and from the abstraction of statistical distributions of form-and-meaning units over thousands of processed instances. While the epistemological debates continue, at present many SLA scholars seem content with theoretical and methodological diversity (Douglas Fir Group, 2016).

3 SLA in the Context of Bilingualism Studies How is SLA to be characterized in relation to bilingualism research? Precise boundaries are not easy to draw, as SLA and bilingualism studies converge and diverge in several respects. Indeed, SLA work is published and presented in venues whose titles prominently display the words bilingual and multilingual (including this Handbook). Nevertheless, some disciplinary preferences distinguish SLA from other disciplines that focus on bilingualism. I highlight four here. First, more so than is the case in much of bilingual research, there is a widespread assumption in SLA that there is a single and clearly identifiable L1, and that this L1 does not need to be empirically inspected in order to answer questions about the development of the L2. Thus, only for specific research purposes, for example, to investigate crosslinguistic influence (Section 5), will SLA researchers collect data from the same participants in their L1 or other previously known languages in addition to the L2 (Ortega, 2016). Bilingualism researchers tend to investigate all the languages of a bilingual, either in the sense of holistic communicative repertoires, as proposed by scholars who champion anti-essentialist views of bilingualism (Fuller, Chapter 6, this volume; see also the notion of translanguaging, Garcı´a & Tupas, Chapter 20, this volume), or by focusing on measuring within the same study the development of the two languages of a bilingual, thus characterizing the linguistic development not of one or another language in isolation, but in terms of same-speaker language pairs (Serratrice, Chapter 1, this volume, and Treffers-Daller, Chapter 15, this volume). Second, the prototypical participants in SLA studies are adults learning a new language during young adulthood and in midlife. Yet a not negligible number of SLA researchers also direct attention to school-aged children who begin to learn an L2 in middle childhood or adolescence, sometimes

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

412

LOURDES ORTEGA

even earlier. In SLA circles, the term for this subfield is child SLA. What seems to distinguish child SLA from other work in bilingualism with children of overlapping age ranges (often, 4–12 years old) is that in SLA: (a) the child participants tend to be language-majority speakers learning a new language, (b) only the L2, rather than the full language repertoires or L1–L2 pairs, is analyzed, and (c) the research questions tend to fall along either a pedagogical interest in determining best classroom practices for teaching a new language in primary and secondary schools (e.g., JuanGarau & Lyster, Chapter 11, this volume; Mun˜oz & Spada, Chapter 12, this volume), or a theoretical interest in asking whether Universal Grammar is available in the process of L2 learning, or whether it somehow becomes inaccessible once L1 learning has taken place (see Haznedar, 2013; Slabokova, 2016). Third, a special staple of SLA is that language teaching and pedagogy have always occupied a central place in the field. The reason is that, even when an L2 comes into someone’s life mostly in the form of naturalistic immersion, at some point many adults will avail themselves of formal instruction. The investigation of pedagogical and instructional concerns has given rise to the subfield known as instructed SLA (Loewen & Sato, 2017). Many other researchers studying bilingualism, by contrast, investigate bilingual development in the absence of language instruction, particularly when the bilinguals are children learning their languages prior to formal schooling (Serratrice, Chapter 1, this volume), or they draw convenience samples from L2-speaking adults who may in fact be enrolled in language classrooms (e.g., as in some of the psycholinguistic processing research synthesized by Paap, Chapter 22, this volume) but do not consider that the phenomena investigated could be due to or confounded by influence from L2 instruction. A fourth and final observation is that SLA is often about emergent or incipient levels of L2 learning. Many studies capture the initial trajectory, over, say, the first two or three years of formal study of the new language, or the first few months of learners being surrounded by the new language. Or they may investigate the initial state of proficiency, knowledge, or processing that obtains after some short-term exposure or training in an L2. Far less research examines the language use and possibly continued language learning by accomplished L2 learners in depth (Section 8). This focus on initial learning trajectories and states, together with the almost exclusive focus on a single language, is aggravated by the problem of constantly comparing adults who embark in language learning to monolingual native speakers, reflecting an ideology of nativespeakerism. This research habitus sets up a situation where the SLA literature may implicitly portray learners as deficient, forever striving to attain perfection in their L2 abilities and/or completeness in their mental grammars, and forever deviating from an idealized native speaker norm. Many scholar have criticized this state of affairs as problematic (Cook, 1991; Ortega, 2013,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

413

2016; see Treffers-Daller, Chapter 15, this volume). The 2010s have been characterized by SLA-internal calls to overcome these problems, which are rooted in a monolingual bias and the deficit orientation it unwittingly reproduces.

4 The Scope of SLA: A Bird’s-Eye View The findings and contexts for L2 learning I survey in this chapter are by necessity only a few slices of a much larger field. Therefore, I would like to use Chagall’s language learning biography as a device that allows me to offer a bird’s-eye view into SLA’s research interests and that helps show where it converges and where it parts ways with other fields that study bilingualism. What range of research questions in Chagall’s language learning history would ignite the imagination of SLA researchers? Most would not expend any attention to Chagall’s Yiddish, since he learned it as an L1 from birth via primary caretaker socialization. In contrast, as shown by Serratrice (Chapter 1, this volume), Ko¨pke (Chapter 18, this volume), and Keijzer & de Bot (Chapter 14, this volume), bilingualism scholars are quite interested in what happens to the L1s of a bilingual. Many scholars in SLA would want to examine age effects and critical periods by comparing the outcomes resulting from Chagall’s different onsets of learning for his different L2s: a start in middle childhood (Hebrew) or around the onset of puberty (Russian), versus in young adulthood (French) and in midlife (English). Bilingualism scholars have also been interested in the role of age of onset of exposure, but for young children especially (Armon-Lotem & Meir, Chapter 10, this volume). Connected with the age of onset issue, SLA scholars would want to explore Chagall’s aptitude or natural talent for languages. By comparison, language aptitude has not been of particular concern to bilingualism scholars, who instead may focus on bilinguals’ enhanced metalinguistic awareness or improved cognitive functioning (see Paap, Chapter 22, this volume). Moreover, in order to investigate age, aptitude, or both, SLA researchers would want to measure the levels of proficiency Chagall reached in each of his L2s. Within these proficiency explorations, many SLA researchers would find it unproblematic and indeed a research imperative to compare Chagall’s proficiencies to the proficiencies of monolingual native speakers, particularly for his strongest L2s (Russian and French), with the purpose to determine whether he ever reached levels of competence that were nativelike. Heeding Grosjean’s (1989) longstanding dictum that bilinguals are not two monolinguals in one, bilingualism scholars tend to be less focused on monolingual comparisons (but see Treffers-Daller, Chapter 15, this volume). As in bilingualism studies (Paap, Chapter 22, this volume), SLA researchers would also be interested in the cognitive and psycholinguistic processes Chagall used in parsing the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

414

LOURDES ORTEGA

surrounding linguistic input and in learning to comprehend and produce each language. As in bilingualism studies (Hernandez, Chapter 23, this volume), a burgeoning but minority group of SLA researchers would likewise want to explore the underlying neural bases of such processes. Many SLA researchers are interested in asking what crosslinguistic influence may be at work as each new language is added along the way. To address this question, they would study any influence stemming from Chagall’s knowledge of his L1 Yiddish on the other languages he learned to speak afterward (L2 Russian and L2 French), asking how his French may have been influenced by his Russian as well. These multilingual connections are also of interest to bilingualism scholars studying the use of more than two languages (Quay & Montanari, Chapter 27, this volume). In addition, some of the SLA researchers studying crosslinguistic influence would also ask how Chagall’s L1 Yiddish may have been subtly changed by his later knowledge of other L2s, an interest that is addressed at the societal rather than individual level by contact linguists in bilingualism (Aalberse & Muysken, Chapter 26, this volume). SLA researchers would also want to explore the role of learning contexts, including the quality of formal language instruction impacting Chagall’s Russian, the contributions of naturalistic immersion nourishing his French, and perhaps his (seemingly voluntary) lack of access to English (see further Section 6). The relative merits of different approaches to schooling and language instruction versus immersion in bilingual learning are also explored by educational researchers, although these focus more on the academic achievement and school adjustment of bilingual children who are speakers of a minority language being schooled in a majority language (Bigelow & Collins, Chapter 2, this volume; Garcı´a & Tupas, Chapter 20, this volume). The precise relation between quantity and quality of input and exposure and bilingual outcomes is of keen interest to bilingualism scholars, but mostly by those who investigate bilingual development in infants and children and the role of the speech directed to them by caretakers and intimate others (Serratrice, Chapter 1, this volume; Armon-Lotem & Meir, Chapter 10, this volume); when bilingualism researchers focus on bilingual interactions among adults, they usually do so with respect to language use rather than learning (De Houwer, Chapter 17, this volume). Yet other SLA scholars might be interested in knowing what Chagall himself thought of his language learning, and what strategies he applied to regulate his own learning in each case. Rather than approach the question via ethnographic methods, as typically done in the many studies in bilingualism that address bilinguals’ emic views (e.g., Bigelow & Collins, Chapter 2, this volume; Fuller, Chapter 6, this volume; Gonc¸alves, Chapter 3, this volume), SLA researchers would most likely use surveys or think-aloud protocols to elicit such insights. SLA scholars would also study Chagall’s motivation to invest effort and to persist in order to learn

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

415

each language, partly depending on his reasons to learn Hebrew vs. Russian vs. French vs. English and partly depending on his attitudes toward the various speaking communities he came into contact with through these languages. Language attitudes and broader language ideologies are widely studied in bilingualism (Fuller, Chapter 6, this volume), but this research has not sought to formalize and measure the construct of motivation to learn a language the way SLA researchers have. Finally, since the turn of the century, socially oriented SLA scholars, much in attunement with work on attitudes and ideologies by bilingualism researchers, would also want to explore a host of forces having to do with the personalaffective, social, ideological, and historical-political dimensions that probably shaped what Chagall was ultimately able to achieve in linguistic and communicative terms by bringing (or not bringing) Hebrew, Russian, French, and English into his life. Having characterized SLA as a field, I now turn to a review of two research areas of central importance: the role of previously known languages (Section 5) and the contributions of the surrounding linguistic environment (Section 6).

5 Crosslinguistic Influence A central question in understanding L2 learning is how previously known languages, and particularly an individual’s L1, influence the process of learning an additional language. This area of SLA, known as crosslinguistic influence (CLI), is rich in findings that have been authoritatively reviewed by Jarvis and Pavlenko (2008) and Odlin (2014). Other terms historically preceding CLI, such as interference and transfer, have fallen into disfavor because they suggest that CLI necessarily hinders L2 learning. In actuality, the accumulated research shows that CLI can have felicitous as well as less desirable consequences, as I discuss here. The consensus among CLI researchers is that L2 learners rely on their already familiar languages to aid the learning process, and that they do so both unknowingly and strategically. For example, Singleton (1987) inspected L2 French speech by Philip, an L1 English professional writer who had picked up French during three short visits to France. He spoke some colloquial Spanish and had studied Irish and Latin in school. In three half-hour interviews with the researcher, Philip produced 154 non-target-like instances in L2 French, many of which could be characterized as Frenchified Spanish lexical items. For a third of them, he issued some overt indication (i.e., a hesitation, apology, interrogative intonation, or laughter) that suggested he might have been conscious of the imperfect solution. When asked about a specific French form retrospectively, Philip noted: “I knew that it probably wasn’t right, but it was the nearest I could get to something that might be right” (p. 335). Like Philip, many L2 learners purposefully and actively use

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

416

LOURDES ORTEGA

other languages they know in order to find solutions and convey their desired meanings in their (as yet) less developed L2. Many other cases of CLI, in addition, are likely below the level of consciousness. Corpus linguistic methodologies have been particularly successful at furnishing evidence of this type of CLI, through the group-level analysis of massive collections of L2 written texts. Focusing on the ungrammatical use of English morphemes in a 4-million-word corpus of English texts written for proficiency exams, Murakami and Alexopoulou (2016) were able to show that L1 German and L1 French examinees of English were less likely to reach 90% levels of accuracy in their use of -ing, even at the highest proficiency levels, than examinees from other L1 backgrounds. They argue the key difference is that German and French do not express continuous events through a morpheme, whereas the other languages included in the corpus as L1s do (e.g., Japanese -tei-; Korean -koiss-, -oˆiss-; Spanish -ando, -iendo; Turkish -iyor). Behavioral experiments and brain imaging studies also demonstrate the extent to which CLI can be pervasive and likely below conscious level. For example, in a brain imaging study, Dowens, Vergara, Barber, and Carreiras (2010) found that L1 English–L2 Spanish late bilinguals showed somewhat different electrophysiological patterns when they judged L2 sentences containing number violations (e.g., los suelo = the+plural floor+*singular) than when the sentences contained violations of gender agreement (e.g., la suelo = the+feminine floor+*masculine). The researchers attributed the results in part to the fact that number morphology exists in both the L1 and L2 (e.g., these floors, in English), whereas gender exists in the morphology of the L2 only. CLI can manifest itself not only in overt non-target-like uses but also in global qualities of language where no errors per se are involved. For example, the SLA scholars gathered in the volume by Jarvis and Crossley (2012) created computational algorithms that detected the L1 backgrounds of English texts written by L2 writers as a function of L1-influenced L2 characteristics of cohesion, lexical sophistication, syntactic complexity, and word length, none of which involved any error tagging. Subtle CLI effects can also be detected at the level of prelinguistic and linguistic conceptualization, which can affect preferred L1 and L2 phrasings. A wellstudied case is that of L2 motion events (Brown & Gullberg, 2008; Stam, 2010), where L1–L2 typological differences create interesting effects resulting in non-idiomatic choices rather than errors. For example, when the event described in the boy ran into the room is verbalized by L2 English speakers from Spanish or Japanese L1 backgrounds, they will tend to deemphasize the manner of motion (i.e., running, e.g., as opposed to walking) by expressing it in a non-prominent satellite phrase: el nin˜o entro´ en la habitacio´n corriendo, literally, the boy entered the room running. This preference is congruent with their L1 but clashes with the typical English choice, which expresses manner in a salient verb position (ran). Interestingly, the synchronization of gestures with motion event

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

417

descriptions differs for speakers of verb-framed and satellite-framed languages as well, and SLA research has shown that these L1 preferences can influence the L2. Specifically, when using a verb-framed language like English as an L2, adult speakers of satellite-framed L1s have been found to draw on their L1 preferences at the level of manner encoding and accompanying gestures (e.g., in L1 Japanese–L2 English, Brown & Gullberg, 2008; in L1 Spanish–L2 English, Stam, 2010). Research shows that CLI cannot be viewed as just a side effect of L2 learning, a sign of lesser competence in the L2 that will slowly but surely disappear from the L2 of those learners that reach extremely advanced levels of proficiency. This insight is based on three sets of empirical findings. First, rather than a hindrance, CLI is also seen to afford adult learners an acceleration of development in many areas; indeed, in acquiring English -ing CLI may slow down some French and German speakers but it will speed up some Spanish speakers of English (Murakami & Alexopoulou, 2016). Second, CLI is not linearly related to proficiency. For example, some CLI effects can surface at incipient levels of L2 proficiency and eventually disappear (e.g., accuracy in the use of English -ing), while others are seen to occur only at advanced levels (e.g., overpassivization errors like he was arrived early, Oshita, 2000). Also, reverse influence from L2 to L1 may occur when the L1’s strength remains intact and the L2 has reached only intermediate levels of L2 proficiency (Brown & Gullberg, 2008). Third, as Odlin has repeatedly cautioned (e.g., 2014), crosslinguistic patterns are always probabilistic rather than deterministic, and what can be predicted at the group level may hold for some individuals but not others. For example, cognates, that is, words with similar forms and meanings across languages (e.g., English–Romanian affluence – afluent¸a˘, English–Japanese ice-cream – aisu kurı¯mu, English–Russian airport – аэропорт) are known to greatly boost comprehension and accelerate learning, particularly but not only when the L1 and L2 are genetically related (Ringbom, 2007). Nevertheless, different individuals vary widely in their cognate guessing skills, which in turn depend on L1 vocabulary knowledge and fluid intelligence (Vanhove & Berthele, 2015). It follows that not all individuals will be able to reap the benefits expected from cognates to the same degree, even when they share the same L1-L2-Ln constellation. Chagall’s illustration is a reminder that many adults venture into learning not just one but several languages at some point in their lives (see the review of multilingual child and adult learners by Quay & Montanari, Chapter 27, this volume). The reality of multilingual learning by adults has become increasingly impossible to ignore, and a burgeoning area within CLI research asks: when an L1 and an L2 are available during L3/Ln learning, which one will be the source of help (or hindrance) for the new language? The evidence has been mixed (e.g., Gonza´lez Alonso, Rothman, Berndt et al., 2017), but it may be unrealistic to expect that there be an ultimate language source that predicts transfer in L3 acquisition across the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

418

LOURDES ORTEGA

board, for all language pairs, all structures, and all L2 speakers; this would counter the well-established insights about CLI presented here. At the most global levels of multilingual learning, however, language proximity does matter: the smaller the typological distance between earlier learned languages and an L3, the better the L3 is learned. Schepens, van der Slik, and van Hout (2016) showed this in a study of hitherto unprecedented scale in SLA, involving 39,300 adults tested for their proficiency in Dutch, which was their L3 or Ln. All participants were newcomers to the Netherlands from 119 countries, and together they spoke 56 different L1s. The L2 of these multilinguals was operationalized as the language they reported to be strongest or learned best outside the family unit (for 25% of these multilinguals, their best L2 was a language other than English). Using mixedeffects regression, the researchers found a reliable association between higher proficiency in L3 Dutch and the proximity of L1 and/or L2, with the facilitative effect being particularly large for lexical (as opposed to morphological) closeness. They also uncovered some evidence that the L1 may occupy a privileged place in late-timed multilingualism, since the distance effect was stronger for L1–L3 than for L2–L3, although both distances made distinct contributions. Interestingly, this same study seems to answer the question of whether multilinguals are better language learners than bilinguals with a resounding yes. After accounting for the relative linguistic distance/proximity factor, and across different comparisons performed, the 39,300 adults who spoke more than two languages tended to exhibit higher L3/Ln Dutch proficiency when compared to a new sample of L2 learners drawn from the same database, for whom Dutch was truly the only additional language learned sequentially, after growing up with a single language. The conclusion is that being multilingual or knowing at least two languages will enhance the odds of success with subsequent languages, although the magnitude of the advantage will be moderated by the relative distances between the specific languages involved. Ultimately, the many decades of CLI research offer a lesson from SLA that can inform the study of bilingualism more broadly. Namely, at a fundamental level, cross-language influence is an unavoidable fact of having more than one language at one’s disposal. Furthermore, the languages of an adult are known to interact not just in the process of L2 learning but also at the level of fluent and proficient processing (Paap, Chapter 22, this volume). Thus, CLI is part of the broader reality of bilingualism (Biedron´ & Birdsong, Chapter 16, this volume; Treffers-Daller, Chapter 15, this volume).

6 The Contribution of the Surrounding Linguistic Environment: Exposure and Agency The relation between the surrounding linguistic environment and L2 learning has been examined by the long-standing SLA research tradition

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

419

of what is known as the interaction approach (Mackey, 2012). Using experimental designs in the laboratory or in the classroom, this approach has furnished ample evidence of L2 learning benefits in the areas of vocabulary, pronunciation, and grammar when learners interact in the L2 with expert speakers, such as teachers or native speakers, and also when they interact with L2 peers at the same or different proficiency level. In such interactions, learners are able to receive linguistic feedback, both positive (in the form of models) and negative (in the form of overt error correction or more subtle signals of comprehension difficulties or reformulations into more target-like utterances), and in this way they push themselves to clarify and modify their own speech (see De Houwer, Chapter 17, this volume, on bilingual interactions). There is also an increasingly large body of SLA research that uses Conversation Analysis to investigate the development of L2 interactional competence as a social phenomenon, and as a calibration of L1 resources for the purpose of L2 communication (Pekarek Doehler, 2013). Nevertheless, SLA investigations probing the precise link between the quantity and/or quality of learners’ exposure and input to a new language and L2 outcomes are scarce. The neglect is surprising, as it stands to reason, for example, that Chagall’s lesser success in English than French should be in great part traceable to frequent and supportive social interactions during his learning of French, which he needed as a lingua franca to interact with other artists in Paris, where he went as a lone individual, versus a low frequency of interactions in English in the United States, where he could rely on his beloved Russian-speaking wife until her death in 1944, and where he continued to surround himself by family members and intimate others who spoke fluent Russian and/or French, while also actively participating in the vibrant Yiddish-speaking community in New York. Some SLA studies have collected self-reported information about the frequency of activities in the L2. The results suggest that many L2 learners have low access to the L2 despite being in a potentially L2-rich environment. For example, in a rare study eliciting information on the frequency of L2 use via a computerized log, Chinese L1 students enrolled in a graduate degree at a Canadian university reported that they used L2 English for interactions with friends on average less than 11 minutes per day (Ranta & Meckelborg, 2013). A rare SLA study conducted by Collins, Trofimovich, White et al. (2009) with 10- and 11-year-olds enrolled in English-intensive classrooms in Francophone Canada suggests that even school-aged children who are acquiring an L2 under apparently ideal conditions – in this study, with intensive L2 instruction from teachers who speak that L2 natively (and who are often imagined to possess a perfect and complete command of the language) – may not always receive the kind of relevant exposure to specific linguistic content that is needed for acquisition to take place. In an analysis of 40 hours of native-

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

420

LOURDES ORTEGA

speaking teacher talk directed to the children, Collins et al. found only 354 tokens of the past tense -ed morpheme out of 15,130 finite verbs. This is a surprisingly low frequency of occurrence. The morpheme -ed is known to be potentially difficult to master for L2 learners (e.g., Murakami & Alexopoulou, 2016), but SLA researchers have presumed that it is amply available to L2 learners in regular communication. However, the findings in Collins et al. (2009) suggest it is possible to be immersed in English and yet not to frequently hear -ed tokens. Thus the little evidence that exists gives reason to posit that L2 learners’ access to the L2 is often less than ideal. When access to the L2 is available and realized, however, exposure and engagement with language in the surrounding environment has been shown to foster L2 learning. For example, youth’s extramural exposure to online media for leisure purposes, such as gaming, brings about L2 learning in specific areas such as vocabulary, idiomaticity, and even grammar – provided that the English L2 users in the samples reported frequent online use, that is, four or more hours per week (see, e.g., Jensen, 2017, with 8- to 10-year-olds in Denmark after two lessons per week for one year, and the studies collected by Sylve´n & Sundqvist, 2017, with a variety of L2 adult learners). A vibrant community of SLA scholars has been able to link actual use of the L2 in online environments to a variety of L2 learning benefits, both when technology is used in the wild and when it is domesticated to complement formal instruction. This work fills the pages of several SLA journals exclusively dedicated to technology-mediated L2 learning topics, such as CALICO Journal, Language Learning & Technology, and ReCALL Journal. The reasons why the potential for rich access to the L2 is often not realized can be complex, involving attitudes and ideologies by all parties involved (De Houwer, Chapter 17, this volume; Gonc¸alves, Chapter 3, this volume). For example, Subtirelu (2014) suggests that the problem is particularly acute for L2 speakers who develop what he called a deficit ideology. In his study, based on interviews with international students in the United States, these were speakers who blamed their not-so-good English for the difficulties they constantly experienced when communicating in the L2 and, as a result, they dreaded and gradually avoided L2 communication. This avoidance naturally lowers the chances at further developing the L2. By comparison, in the same study other international students turned out to be more resilient in the face of very similar negative experiences, and they seemed to embrace a lingua franca ideology, whereby they allocated the blame for failures in communication equally between two interlocutors. Such L2 speakers did not avoid further L2 encounters and thus were able to continue using their L2, which, at least in theory, ought to afford them more of the necessary L2 usage to further their learning (Ellis et al., 2016). In addition, it must not be forgotten that to the task of learning and using the new language adults will bring their preferred ways of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

421

communicating, of learning new information, of being in the world, and of viewing and evaluating the world. Thus, they often exercise their agency in choosing or declining to learn certain aspects of the L2. An increasing number of SLA studies adopting longitudinal and qualitative methodologies has revealed agency to be a factor that complicates the contributions of the linguistic environment to L2 learning. For example, Ohara (2001) found that after spending a year in Japan, female learners of Japanese as an L2 became aware of the fact that a high pitch in female Japanese voices indexes femininity; however, upon this realization, some learners accommodated and began to use a higher pitch in order to sound more Japanese, whereas others refused to change their pitch. One of Ohara’s participants explained her decision: “[T]hose Japanese girls, they were not even girls, some of them were in their late twenties, but they would use those real high voices to try to impress and make themselves look real cute for men. I decided that there was no way I wanted to do that” (p. 244). Likewise, Diao (2017) documented the interesting case of Hasan, a young man born in the United States to a family from the Middle East, who had studied Mandarin for only one semester in college but had learned some Chinese as a child through a nanny from the southwestern Chinese province of Sichuan who spoke little English. During a study abroad semester in Shanghai, he discovered that his pronunciation was considered nonstandard by locals. The culprit, Diao documented, was a dental/retroflex merger replacing retroflex initials (zh, ch, sh) with dental ones (z, c, s), a stigmatized feature associated with southern Chinese dialects, which Hasan had picked up as a child from his nanny. He initially attempted to speak in standard Putonghua pronunciation. But he abandoned these efforts at accent modification once he visited his nanny in Sichuan and reconnected with her. During the rest of his stay abroad, Hasan refused to make his accent more standard, despite continued criticism by his teacher and his roommate. For him, his nanny’s Sichuan pronunciation was a symbol of intimacy and of a transnational diasporic identity that was more desirable than the national Chinese identity conferred by standard Mandarin pronunciation (see De Houwer, Chapter 17, this volume, on the social meaning of language choice). The two examples by Ohara and Diao show that L2 learners are clearly capable of understanding and learning socially meaningful linguistic variation in the varieties of language that they are exposed to (see Chevrot & Ghimenton, Chapter 25, this volume, for discussion of bidialectalism vs. bilingualism). However, the discovery of linguistic diversity by L2 adult learners does not come stress free, and some choose to appropriate certain features of the L2 while others consciously reject them. In the study of bilingual first language acquisition (i.e., bilingual acquisition from birth) and early second language acquisition (i.e., early sequential bilingualism), it is widely accepted that the quality and quantity of the language to which children are exposed in their surrounding linguistic

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

422

LOURDES ORTEGA

environments greatly shape their bilingual outcomes. How exactly this happens continues to be the object of intense research (Armon-Lotem & Meir, Chapter 10, this volume; De Houwer, 2018b). Similar efforts would be desirable in the future in SLA, given that the little research that exists suggests it is neither an immersive L2 environment nor a classroom environment that sustains L2 learning, but the actual quality and quantity of the language that learners can experience across their learning contexts. In the case of adult learners, access to the L2 seems less than ideal and is heavily complicated by considerations of ideology, identity, and the agency many adults exercise in choosing or declining to learn certain aspects of the L2.

7 Contesting Notions of Success in Adult L2 Learning From early on, SLA researchers have asked the question of how likely it is that adult learners will succeed in L2 learning and what factors might explain variable success. The two explanatory factors most frequently examined in SLA have been starting age and aptitude for language learning. These explanations echo two entrenched myths in folk wisdom: that children pick up languages effortlessly, and that some adults have a rare natural gift for language learning but most are bad at it. Singleton and Pfenninger (Chapter 4, this volume) review the relation between success and starting age, cautioning that it would be empirically premature to accept the hypothesis that there is a critical period for L2 learning, and pointing at the need to consider many other factors beyond pure biological age. I refer readers to their exposition, and I expand on the issue of age as a predictor of L2 learning success in Section 8. Aptitude for language learning is reviewed by Biedron´ and Birdsong (Chapter 16, this volume), who also problematize it. The conundrum with aptitude is puzzling. On the one hand, over the years SLA researchers have developed three well-known tests of aptitude: the Modern Language Aptitude Test (Carroll & Sapon, 1959/2002), the LLAMA (Meara, 2005), and the Hi-Lab (Linck, Hughes, Campbell et al., 2013). All three instruments reliably predict how well adults will do if they attempt to study a foreign language, based on the measurement of key cognitive abilities known to differ across individuals: (a) memory for new vocabulary; (b) perceptual ability to discriminate, remember, or mimic new sounds; (c) the ability to figure out structural regularities in the input deductively when consciously trying hard to guess the patterns during exposure to language stimuli (measured only in the Modern Language Aptitude Test and the LLAMA); and (d) the ability to do the same inductively during exposure to nonlinguistic strings that engage statistical learning and without being told that there are any patterns that should be guessed (measured only in the HiLab). On the other hand, good predictions based on these cognitive abilities are insufficient to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

423

illuminate what the construct of aptitude really entails. As Biedron´ and Birdsong (Chapter 16, this volume) discuss, there is reason to argue (but little research so far to know with certainty) that foreign language aptitude may be both more multifaceted and less innate than previously thought, in that the existence of high aptitude for language learning is likely to be a combination of cognitive abilities measured by existing aptitude tests but also of personality traits such as resilience and strong will as well as educational opportunities, including the opportunity to experience success in learning multiple languages. Ironically, the measurement successes that L2 aptitude researchers have achieved seem to have barred them from investigating the deeply complex conceptual issues surrounding aptitude. Stepping back from attributions of success to age, aptitude, or any other variables, it must be acknowledged that success is in the eye of the beholder. What researchers would deem as successful may be quite different from what bilinguals would have to say about themselves, or what would be deemed as success by those bilinguals’ teachers, employers, relatives, partners, or friends. Wes, one of the best-known L2 learners in the history of SLA, was in his early 30s when, looking to expand his sphere of fame as a Japanese artist, he migrated to Honolulu and began learning English naturalistically (Schmidt, 1983). Comparing his grammar at two points that were 2.4 years apart, one can hardly see any changes in several areas of morphology, such as regular past tense, third person singular, possessive, or article usage: from zero levels in his speech, Wes only progressed to no higher than 20% suppliance (p. 146). Yet the analysis of changes in his L2 discourse shows his global communicative competence did develop in the almost three years of study. Moreover, by all interlocutors’ accounts, Wes was a charming conversationalist. He also was able to use English for 75–90% of his daily activities, including the promotion of his artistic career, for example, during exhibits in galleries (pp. 140–141). Thus, although on objective measures such as the ones that SLA researchers might use Wes was a particularly unsuccessful language learner, for Wes himself and his interlocutors, his English was quite successful indeed. Schmidt insightfully reflects on this point, noting that Wes is “quite clearly proud of what he has accomplished and knows that he can communicate much better in English than many nonnative speakers with much greater linguistic knowledge” even though “he knows that he speaks funny English” (p. 168, italics in the original). When it comes to adults, as opposed to children, there is also a tendency to inadvertently underscore failure and downplay success in L2 learning. Klein and Perdue (1997) reported the findings of a seminal investigation they undertook across five European countries in the 1980s, lasting more than 2.5 years and involving 10 different L1–L2 pairs. The researchers analyzed multiple waves of L2 speech data from 40 adult immigrants who were largely acquiring the L2

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

424

LOURDES ORTEGA

naturalistically. This research culminated in the description of a Basic Variety in oral L2 development, a target-language system (attested across all the L2s investigated) in which no morphology or subordination exists and phrases and information are organized by semantic and pragmatic principles instead. By the end of the study, two-thirds of the immigrants had moved beyond the Basic Variety and were developing both morphology and subordination in the L2. Only one-third seemed to have plateaued there. Yet the literature on SLA tends to focus just on the Basic Variety, implying that this is a typical finished line for naturalistic adult L2 learning. L2 success also depends on the criteria researchers choose to measure linguistic development and attained proficiency or ability (Treffers-Daller, Chapter 15, this volume). As mentioned, in SLA the norm against which L2 use is measured is a monolingual native speaker norm. In connection to this, some SLA researchers have proposed that close linguistic scrutiny in the laboratory should be the gold standard for deciding on success (Abrahamsson & Hyltenstam, 2009; Granena & Long, 2013a). Others have objected that when this linguistic scrutiny depends on a comparison to monolingual native speakers, then apples and oranges are being compared, because what is taken as falling short of the L1 norm may simply be signs of being or becoming bilingual, precisely because bilinguals are not double monolinguals (Birdsong, 2013; Cook, 2016; Grosjean, 1989). The acceptance of linguistic scrutiny as a valid basis for proclaiming success in SLA thus depends on the conceptualization of L2 development as an arduous and never-ending process of eliminating errors and other forms of crosslinguistic influence, a ladder toward nativelikeness, or, rather, monolingual-likeness (Ortega, 2013, 2018). It therefore also rests on the privileging of monolingualism as the default of human language and on the idea that languages are bounded and fixed systems. Bilinguals and monolinguals alike exhibit great language variability stemming from multilectal and multilingual repertoires they develop as a function of the variable levels of education different people attain (Da˛browska, 2012) and the variable exposure to different language varieties individuals experience (Chevrot & Ghimenton, Chapter 25,this volume). There is no ladder, no end state, in language learning (Larsen-Freeman, 2017), whatever and however many languages and language varieties the learning may involve. In fact, as already noted by classic bilingualism authors in the 1980s, monolingualism and bilingualism are best thought of as a continuum. As Luk and Bialystok (2013) put it: “[t]he criteria that determine an individual’s designation as monolingual or bilingual are fuzzy at best and minimally involve an interaction of language proficiency and use” (p. 605). In the remainder of the chapter, I examine the degree of success that L2 learners may experience, zooming in and out of two important populations of adult L2 learners that have been studied in SLA, namely immigrants and international students.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

425

8 L2 Learning Success in the Context of Immigration In this section I discuss two SLA studies that examined the success of L2 adult learners who learned the new language in the context of immigration (for a broader discussion of the bilingual lives of young adult immigrants, see Gonc¸alves, Chapter 3, this volume; for a discussion of their educational needs, see Simpson, Chapter 13, this volume). The two studies here are unique by inspecting multiple-year-long evidence, rather than the shorter periods that are more typical in the SLA literature, and by comparing adult learners to one another, rather than to external baselines made up of monolingual native speakers, as has been the norm in SLA. I then discuss their findings in light of two other studies on the same topic conducted in the field of migration studies. The four studies yield a number of key factors beyond age and aptitude that jointly support L2 learning progress and ultimate success. Kozar and Yates (2018) sought to evaluate the progress in L2 English made by 24 immigrants to Australia who had arrived from 11 different countries (the L1s were not reported). All had emigrated between the ages of 40 and 73 and had been residing in Australia for 2 to 14 years at the beginning of the study, which ran over a period of 5 years. Success was operationalized as a combination of self-assessment of L2 proficiency, progress through the free 510-hour curriculum offered by the government, eventual ability to dispense with interpreters, and (for a few learners only) oral proficiency ratings. Five learners were deemed to have made good progress. They all: (1) had at least 12 years of formal education upon arrival, (2) started off with sufficient English to be placed into the last of three levels in the government-sponsored English program, (3) engaged in proactive and self-directed language learning and study strategies, and (4) never made comments about their old age or deteriorating memory being a hindrance to their efforts at learning English. In addition, these most successful L2 learners consistently reported making frequent use of English over the entire span of the study (specifically, 80% of the time or higher), and at some point they held a job at the same level of qualification as pre-immigration, in occupations that required professional use of the L2 in tasks such as written reports or oral presentations. Frequent L2 use and use of the L2 in a fulfilling and demanding job can, of course, be both a drive for L2 development and an outcome of better language skills. At the other extreme, the nine adults who were seen to make little progress tended to exhibit markedly low values in all the areas identified as supporting the successful learners. The middle group of nine adults who were judged to have made reasonable progress exhibited some but not all of the supportive factors of the most successful group. Included in this group, for instance, was the oldest person in the sample, a highly educated Chinese woman who was 73 years old at the time of migration. She was invested in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

426

LOURDES ORTEGA

English language learning and expended much effort at studying it, although she reported speaking it only 10% of the time on average. Eventually, after six years of residence in Australia and having taken advantage of the free English classes offered by the government, she achieved two landmarks of L2 success: being able to dispense with interpreters when interviewing for the study and receiving medical services in the hospital without an interpreter. Kozar and Yates concluded that in this sample of mature adults, age was uninformative in explaining who had made good, reasonable, or little progress over up to five years of investigation. In a very different study, Lahmann, Steinkrauss, and Schmid (2016) analyzed archival oral history testimonies by 102 German-Jewish immigrants who had fled from Germany to Australia, the United Kingdom, or the United States between 1938 and 1939, forced by the two historical tragic events of Kristallnacht (the “Night of Broken Glass”) and the outbreak of World War II. All were between 7 and 17 years of age at the time of the traumatic relocation and were between 57 and 87 at the time of the recordings. Thus, they had lived their adult lives in an English-speaking environment and could be considered “long-term L2 speakers” (p. 378). Educational levels varied but most interviewees had completed high school, and 45% held a college degree or higher. In the oral testimonies, interviewees made comments about their involvement in activities for which they needed German or English. On the whole, the immigrants still used their L1 German relatively frequently, with a small minority (14%) reporting that they used their L1 German at work. The researchers evaluated the interviewees’ L2 English proficiency in the first 30 minutes of speech by applying a number of measures indexing syntactic complexity, use of the passive voice (which is often used in SLA as an indicator of advanced L2 proficiency in English), and lexical diversity and sophistication. Age of L2 learning onset failed to differentiate attained levels of L2 syntactic and lexical complexity in this sample. In a linear mixed-effects regression followed by qualitative inspection of patterns, the three best predictors of higher L2 grammatical and lexical complexity in the testimonies were: being male, having high levels of education, and (for lexical complexity only) using their L1 German for work. Lahmann et al. concluded that L2 learners can attain very advanced levels in the L2 and that factors related to experience, rather than age at the first time of immersion in an L2, jointly help predict L2 outcomes. The findings in the two SLA studies just discussed can be augmented by two large-scale investigations conducted in the field of migration studies, both based on self-reported L2 proficiency. Kristen, Mu¨hlau, and Schacht (2016) examined predictors of L2 progress over the first 1.5 years of settlement in a sample of 2,261 immigrants from Poland and Turkey to four European countries, with age of arrival between 18 and 60. AdamutiTrache (2013) looked into associations between pre- and post-immigration

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

427

variables. Her sample comprised 6,090 immigrants to Canada from a variety of source countries, with age of arrival between 20 and 59. Across all four studies, one clear finding is the primordial importance of frequent use of the L2, that is, the importance of having access to practicing the L2 in the new environment through “formal and informal education and language training, labor market participation, [and] social networking” (Adamuti-Trache, 2013, p. 124). Interestingly, related to L2 frequency of use, employment in a language-intense occupation emerged as a specific factor in all four studies, such that L2 learners or speakers with a job requiring rich use of the L2 tended to improve faster and/or ultimately showed higher L2 proficiency. Kristen et al. (2016) refer to this factor as “the job’s language-intensity” (p. 204) and call for future research on the relation between occupation and better access to (and ultimate better learning of) the L2 (p. 204). Gonc¸alves (Chapter 3, this volume) discusses ethnographic studies capturing how jobs that are menial or done alone offer low opportunities for L2 practice, whereas language-intense jobs, even if low-skilled, such as waitressing, can become opportunities for L2 use and growth. Biedron´ and Birdsong (Chapter 16, this volume) also note that many, though not all, polyglots and hyperpolyglots tend to choose careers and occupations connected to language, and are interpreters, writers, or language teachers. The one study here that included frequency of the L1 use (Lahmann et al., 2016) also supports the importance of language use in the workplace for language growth. That frequent use of L1 German at work in business or academic occupations should be one of only three predictors of higher L2 lexical complexity suggests that if a job requires rich use of the L1, both the L1 and the L2 can be boosted. This finding joins a wealth of accumulated robust evidence showing that the L1 and the L2 do not compete against each other but instead can nourish each other in supportive ways when developed in tandem, for example in the development of academic literacy (Bigelow & Collins, Chapter 2, this volume). Overall, then, the language-related demands of different occupations may emerge in the future as a fruitful factor to investigate when predicting L2 learning, because adults’ jobs likely moderate access to life-important contexts for meaningful and frequent use of the L2 being learned, and/or the L1 being maintained. In addition to frequent L2 use, the importance of level of education found by the two SLA studies is corroborated by Adamuti-Trache (2013). While Kristen et al. (2016) failed to find a relation between education and L2 proficiency, they dismissed this null finding as counterintuitive and attributed it to measurement error (p. 204). It should not come as a surprise that level of education matters for ultimate attainment in L2 learning, as it is known to create individual differences among native speakers’ L1 skills as well (Da˛browska, 2012). Moreover, Andringa (2014) has provided convincing evidence that the uniformly high levels of education associated

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

428

LOURDES ORTEGA

with the native speaker samples against whom SLA researchers compare L2 learners has distorted findings of age effects. Three of the studies – Adamuti-Trache (2013), Kozar and Yates (2018), and Kristen et al. (2016) – converge on finding robust evidence for the importance of (a) initial levels of L2 proficiency prior to emigration, as well as (b) being “active self-learners” (Adamuti-Trache, p. 116) and energetically pursuing formal L2 instruction once in the new country. (Neither factor was investigated by Lahmann et al., 2016.) The two immigration studies, in particular, robustly show a major role for pre-migration L2 skills, or language capital as Adamuti-Trache puts it. The extent to which migrants already have some knowledge of the new country’s language helps predict their growth in proficiency and will ultimately affect their economic and social integration, even after four years of residence (cf. Adamuti-Trache). The findings about the role of biological sex in L2 learning are conflicting. Lahmann et al. (2016) and Adamuti-Trache (2013) found an L2 proficiency advantage of males over females, but the reverse pattern was reported by Kristen et al. (2016), and although Kozar and Yates (2018) did not consider the sex of learners, no strong effects in either direction can be discerned in their data. It is likely, however, that sex as a factor is not informative enough without consideration of the gendered roles the culturally diverse adults in these studies were socialized into, given that both education and employment were strong predictors of L2 learning success in the four studies, and that access to and type of education and employment can greatly vary depending on gendered roles (see also Simpson, Chapter 13, this volume). What about the role of another biological factor, age? At first sight, the two immigration studies proclaim to have found clear evidence that the younger immigrants learned faster and better, whereas the two SLA studies assert that age was predictive neither of measured L2 proficiency gains over up to five years of settlement nor of the proficiency ultimately attained after a lifetime of L2 use, respectively. The conflictive findings cannot be taken at face value, however. First and foremost, any discussion of the role of starting age of exposure to an L2 is limited by the fact that neither the two immigration studies nor Kozar and Yates (2018) report on the exact starting ages for first meaningful exposure to the L2. Second, in both immigration studies, the younger participants, who are seen to make more gains in self-reported proficiency than the older ones, were also often the ones with higher L2 skills at pre-immigration. In fact, the joint findings for 8,351 individuals in these two studies cannot be interpreted in terms of a simple age effect, because the age ranges in both studies covaried with pre-migration L2 knowledge and educational levels (as well as gender and the length of time in the new country). In sum, the available evidence suggests four robust predictors of L2 learning success for immigration contexts: prior levels of L2 knowledge

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

429

before moving to an L2 environment, levels of education, frequency of L2 use, and being active self-learners who deliberately and strategically regulate their own learning processes.

9 Developing Academic Language in an L2: Evidence from International Students The internationalization of higher education worldwide has created a new population of L2 learners: international students who move away from their country of origin seeking a university degree, which they will obtain through the use of a language of instruction that is not one they know best. As mentioned, this population is very large worldwide, amounting to 4.5 million or 5.6% of all higher education enrollments in 2015 (OECD, 2017). English-speaking countries make up a large portion of destinations. In 2015, the United States accounted for 30% of the 4.5 million international students worldwide, the United Kingdom for 14%, and Australia for 10%, closely followed by Canada (OECD, 2017, pp. 286–288). Given these figures, it is no surprise that SLA findings for this population speak mostly to the development of English as an L2 in academic registers and domains. My discussion hence refers to English. In most cases, international students have received many years of formal instruction in the L2 in their home countries and have passed demanding language university entrance exams. But as is true of all students in higher education, the new academic context will make heavy demands on academic language and registers as well as propel significant growth in these. Yet the linguistic development international students experience, at least in terms of L2 writing and L2 vocabulary, appears more modest than one would expect from the intense engagement with the L2 as the medium of advanced academic instruction. To illustrate in the area of vocabulary development, both L2 knowledge of polysemy and derivational morphology have emerged as slow to be learned, despite academic immersion in the L2. For example, Schmitt (1998) tracked three adult learners’ knowledge of 11 academic words (e.g., spur, illuminate, convert) by means of three introspective interviews administered over the graduate students’ years in the United Kingdom. The students arrived with the advanced level of English needed to be admitted to the university, and the study started soon after their arrival. Schmitt found some lexical improvements within the first four months of the study, followed by a plateau. Even after a year there was little overall evidence of improvement in the students’ awareness of the polysemous meanings or derivational morphology of the 11 target academic words. Crossley, Salsbury, and McNamara (2010) investigated the L2 lexical development of six international students in the United States. Their profiles were not dissimilar from those of the students in Schmitt’s

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

430

LOURDES ORTEGA

study, although the US students were likely at a lower starting proficiency level. The students’ conversations with an interviewer were recorded every two weeks for a year. Crossley et al. replicated Schmitt’s previous finding of growth within the first four months and stagnation for the rest of the year. However, Crossley et al. also found that the students increasingly used five highly frequent words in polysemous ways. For example, while think/thinking/thought were initially used only as verbs and only in three senses (consider/judged, expect/suppose, cogitate), in the final-year interviews, these forms were used both as verbs and as nouns, and in 13 additional senses (e.g., imagine/visualize, intend, organized beliefs of a period, decide by pondering, ponder but not decide). Drawing from his own and others’ research in past decades, and extrapolating from frequency profiling in corpora rather than from actual developmental evidence, Nation (2006) estimated that about 6,000 word families (i.e., each family including all word forms that are morphologically related, e.g., think, thinking, thought, thoughtful, etc.) will suffice in order for an adult L2 user to attain 98% coverage of the words likely to come up in informal English conversations or in an average English-speaking movie. For the purposes of reading and understanding a novel or a newspaper in English at 98% comprehension levels, 8,000 or 9,000 word families will be needed. Nation’s estimates for L2 users have acquired somewhat of a golden status in the literature as worthwhile goals for vocabulary size that international students can strive for. Yet they are still far from the average comprehension vocabulary size of a typical well-educated L1 speaker of English in the United States, which comprises an estimated 11,000 word families on average for 20-year-old college students and reaches an average of about 13,400 word families by age 60 (Brysbaert, Stevens, Mandera, & Keuleers, 2016). Piecing together the findings of limited vocabulary development just discussed and the implication that L2 users must work hard to hit the 9,000 word families benchmark, should we conclude that the norm for international students who are L2 users of academic English is to have smaller vocabularies in comparison to L1 English students in higher education, who typically understand 11,000 word families? One plausible guess is that international students can settle for 98% comprehension and thus have little incentive to challenge themselves and increase their L2 vocabularies. Nevertheless, remember that no actual estimates comparable to those of Brysbaert et al. (2016) exist for L2 users. Moreover, as Brysbaert et al. noted, their L1 benchmarks are sensitive to level of education (the less education, the smaller the vocabulary) and to whether individuals are avid readers or people who mostly get their vocabulary through social interactions. It is equally important to consider that, as Treffers-Daller (Chapter 15, this volume) discusses in relation to vocabulary size in bilingual children, in order to avoid deficit conclusions in comparisons with monolingual children, bilinguals’ vocabulary

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

431

knowledge is best measured as a score of their total vocabulary, that is, as all the words known summed across both languages. For example, if an international student knows perro and dog and inculcar and instill, they would know four different words, but if only their English were to be considered, they would only know two words. If all the words that international students know were counted, the total vocabulary sizes of L2 educated speakers would likely be higher than the estimates proposed by Brysbaert et al., which represent total vocabulary knowledge in English monolingual speakers. To the best of my knowledge no SLA researcher has ever attempted to study the total vocabulary size of L1–L2 educated adults. Doing so would be of great relevance. Specifically, international students’ mental lexicons are likely to exhibit complex lexical knowledge that includes unique lexical items, or words that exist only in one of their two languages, English, as is often the case with terminology in a given field, as well as singlets, or words known to a bilingual in one but not the other language; in the case of international students, the unknown equivalent may be either in the L2 or the L1, given that during academic training many words are newly acquired that may not have been acquired in the L1. For example, an international student who is a very experienced L2 English academic writer may know the word instill in their L2 without knowing the translation equivalent in their L1. Moreover, total vocabulary knowledge may be ideal to adequately investigate the depth of vocabulary knowledge that an L1–L2 adult develops through polysemy, cognate status, collocational restrictions, and other lexical relations. Consider, for example, the complex relations potentially crisscrossing the poysemy networks, collocational restrictions, and so on, among instill, inculcar, imbuir, and inculcate. The findings for academic language development and L2 vocabulary produced by SLA researchers are typically interpreted as meaning that international students meet with limited success in the development of the L2. However, this conclusion may be an artefact of the relatively short length of observation typically found in those studies as well as a consequence of considering the L2 in isolation from L1 knowledge. A positive picture of L2 growth may be more clearly revealed if future SLA research also investigates the linguistic academic competencies of former international students who, upon employment in academia either back in their countries or in English-speaking centers of knowledge, join the large international academic community of scholars who regularly negotiate publishing in English for both international and national outlets. SLA studies in this area are unfortunately lacking, although ethnographies have documented the role that L2 learning and knowledge play in international publishing from all corners of the world, from Hungary, Slovakia, and Spain (Curry & Lillis, 2004) to Sri Lanka (Canagarajah, 2006), to Tanzania (Thomas, 2018).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

432

LOURDES ORTEGA

10 Conclusion I hope this chapter has served to illustrate how SLA researchers seek to investigate all and any relevant forces and processes that shape additional language learning. The outcomes of L2 learning are not guaranteed – but neither are the outcomes of bilingual first language development (De Houwer, 2009) – and may include language learning but also unlearning, non-learning, and even relearning at any point of the adult life. However, real-life evidence makes abundantly clear that for all practical purposes adults are perfectly able to achieve success when learning a new language at various points in life beyond early childhood and outside the family unit. Several well-known bilingual figures achieved great prominence in domains of life that make heavy demands on language, for example, in political (German L1–English L2 Henry Kissinger in the United States), literary (Czech L1–German L2 Franz Kafka in the Czech Republic), acting (Afrikaans L1–English L2 Charlize Theron in the United States), and scientific careers (L1 Polish–L2 French Marie Curie in France). The preponderance of empirical evidence surveyed in this chapter too suggests that for all practical purposes, adults are perfectly able to achieve success in an L2. Clearly, a great deal of variation can be found when people learn new languages beyond primary socialization in the family. The starting age of learning onset varies greatly, but so do the preferred ways of communicating, of learning new information, of being in the world, and of viewing and evaluating the world that individuals who reach a certain age will have developed and will inevitably bring to the task of learning and using the new language. An important role will be played also by the preexisting language skills each individual brings to the task, depending on the languages and language varieties already known at the point of starting another one. Cross-language interactions, that is, the well-studied phenomenon of crosslinguistic influence, cannot be viewed as a sign of lesser competence in the L2 that will slowly but surely weed out if sufficiently advanced levels of proficiency are reached. Instead, crosslinguistic influence is an unavoidable fact of having more than one language at one’s disposal, and part and parcel of the broader reality of being bilingual. Moreover, when no or less than expected learning is observed, L2 adult learners may not have engaged in the frequent L2 use presumed available in their contexts. Furthermore, the quality of language directed to them may have been insufficient to foster L2 learning, or the non-learning of particular features may have been motivated by choices adults make in order to fashion themselves as the kind of L2 speaker they desire to be in their social worlds. These plausible explanations must be verified before an L2 learner, or even worse, entire learner populations are pronounced

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

Second Language Acquisition as a Road to Bilingualism

433

incapable of achieving L2 success, and indeed before the entire enterprise of L2 learning is deemed very difficult or even impossible. SLA researchers will continue to investigate age and aptitude, as has traditionally been done in the field (and other important factors not examined in this chapter, such as motivation and effective instruction; see Loewen & Sato, 2017). However, they ought to do so under new prisms that are critical of certain premises in the field, particularly premises that reflect biases privileging monolingualism and casting deficit on bilinguals. Perhaps the hardest premise that must be questioned is: can monolinguals provide the golden benchmark for deciding whether bilinguals, including L2 speakers, have reached the finish line in their linguistic development? Other more abstract questions must also be asked: Is there a finish line in language learning? What constitutes success in L2 learning, and who is to tell? Future SLA research will have to give the center stage to important factors that have been found to jointly support L2 learning success. In a nutshell, these are: frequent use of the L2 in language-rich and language-demanding situations, maintenance of active L1–L2 bilingual use, and high levels of education. Two additional factors play a large role in adults’ pathways to bilingualism: adults’ deliberate and strategic management of their own learning processes, and high-quality language instruction (Loewen & Sato, 2017). To that, one must add language proximity (Schepens et al., 2016). In the context of immigration, in particular, this latter factor means that the more distant a newcomer immigrant’s L1s and best L2s are to the new language, the more time and resources will need to be made available by receiving governments before that individual can be expected to have learned the new language and begin functioning in the new environment. Also in the context of immigration, it is important to keep in mind that pre-migration L2 skills, or language capital, help predict not only how fast newcomers will grow in their L2 proficiency but also their economic and social integration, even after four years of residence (cf. Adamuti-Trache, 2013). Finally, a pending disciplinary challenge is to overcome the monolingual bias and nativespeakerism that inadvertently have seeped into much SLA research. Research portrayals of L2 learners as forever striving to attain perfection in their L2 abilities and/or completeness in their mental grammars, and forever failing to learn and/or deviating from an idealized native speaker norm can be overturned by producing more studies that look into long-term learning rather than mostly initial learning trajectories, by analyzing whenever possible and relevant all of a speaker’s languages, instead of the new language only, so as to make all the languages of an individual count as part of her or his competence, and by ceasing to compare adults who embark in language learning to monolingual native speakers. The 2010s have been characterized by SLA-internal calls to overcome these problems (e.g., Douglas

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

434

LOURDES ORTEGA

Fir Group, 2016; Ortega, 2018), and some progress seems to be on the horizon. Given the very large numbers of L2 learners in the world, the importance of critically calibrating the insights that SLA produces about success in L2 learning and of treating L2 learners as bilinguals in the making cannot be sufficiently underscored. Whether it is immigrants with allegedly ample access to the L2 in the wild, younger and older students who learn a new language in school, or young adults who join the world of higher education through the use of an additional language, SLA researchers have the obligation – and the opportunity – to both elucidate and support the human potential for sequential emergent bilingualism across these different contexts. The disciplinary aspiration is to generate knowledge about L2 development with positive consequences for adults’ lives and well-being. The guide to do so is an explicit orientation toward L2 learning as a road to bilingualism.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:52:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.022

22 Bilingualism in Cognitive Science The Characteristics and Consequences of Bilingual Language Control Kenneth R. Paap 1 Introduction How do bilinguals represent and access words that belong to two lexicons? How do bilinguals coordinate two languages? Does managing two languages enhance general cognitive control? Does speaking two languages afford greater mental flexibility? This chapter describes how cognitive psychologists go about answering these questions, provides reasonably clear-cut answers to some of them, and explains why one is vigorously debated and remains controversial. The chapter also explains how bilinguals may display different styles of moral reasoning and emotionality as a function of which language they are using. On the more equivocal side, seminal studies appearing to show a bilingual advantage in various components of executive functioning are difficult to replicate and the necessary and sufficient conditions for producing them remain elusive. On the intriguing side, there are demonstrations of linguistic relativity within the same person showing that activating one language rather than the other can influence the way a bilingual perceives, understands, evaluates, and remembers life experiences.

2 Bilingual Language Control Acquiring two or more languages will have consequences for the cognitive architecture supporting language processing, and may also have consequences for cognitive processing outside the domain of language. At the lexical level, a speaker of two languages on average knows more words than a monolingual and also knows, for any given word, if it is a word in LX, LY, or both. Consequently, a bilingual can produce sentences in LX, LY,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

436

K E N N ET H R . PA A P

or strategically switch between languages (see De Houwer, Chapter 17, this volume, for what factors help explain such switching). Identifying a word’s language, selecting a target language, and switching languages are unique to bilingual processing. One simple solution for managing two languages would be to have two lexicons that are functionally independent and a control switch that allows selective access to LX or LY as needed. This proposal has intuitive appeal as balanced bilinguals rarely experience an unintended intrusion from the nontarget language and feel as if they usually switch effortlessly between their two languages. But intuitions are imperfect guides and there is considerable evidence that during reading, listening, and speaking, words in the nontarget language become coactivated and create conflict that needs resolution. This section reviews some of the seminal research on language coactivation and selective access to a particular language in bilingual processing.

2.1 Cognate Facilitation and Interlingual Homograph Interference Lexical decision tasks (LDTs) provide an opportunity for detecting coactivation from a printed word in the other language without explicit task demands on accessing the word’s meaning or pronunciation in either language. LDTs as used in bilingual processing research simply require participants to press one response key if a letter string is a word in the designated target language, and the other button if it is a nonword in that same target language. If access is nonselective, then the input letters should activate word representations in proportion to the degree of orthographic overlap with a word from the other language, regardless of targeted language membership. This makes cognates and interlingual homographs fertile candidates for detecting coactivation. Cognates as used in LDTs are translation equivalents that are spelled in highly similar ways (such as Spanish papel and English paper) or identically (e.g., real in both Spanish and English). In LDTs, cognates are responded to faster than non-cognate control words, an effect called the cognate facilitation effect. Interlingual homographs have identical spellings but different meanings: for example, fin is part of a fish in English and signifies “end” in Spanish. Responses to homographs are slower and less accurate than to nonhomograph controls, thus showing an interference effect. Even when sequential bilinguals are engaged in lexical decisions targeting only their first language (L1), there is both cognate facilitation and interlingual homograph interference from their second language (L2) (Dijkstra, Timmermans, & Schriefers, 2000). Thus, words in the L2 are coactivated even when the L1 is the task-relevant one. However, the very presence of cognates and interlingual homographs serves as a periodic exogenous cue to activate the nontarget language. An even stronger test of automatic coactivation would be the demonstration of an effect of words

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

437

from the other language when similar words in the other language are never presented. Van Heuven, Dijkstra, and Grainger (1998) showed that Dutch–English bilinguals were sensitive to the number of Dutch neighbors (i.e., Dutch words that differ from the English target word by only a single letter) in an English LDT that presented no cognates or interlingual homographs, and even though they did not use Dutch, their first language (L1), at any time during the experiment. Effects consistent with the coactivation of both lexicons have also been reported for spoken word recognition and conceptually driven word production. In Dutch lexical decision, homophone interference occurs as it took Dutch L1 speakers with English as a second language (L2) longer to respond “word” when they heard an interlingual homophone like /li:f/ (pronounced similarly to leaf in English, but meaning “sweet” in Dutch) than when they heard control words that were not interlingual homophones (Lagrou, Hartsuiker, & Duyck, 2011). With respect to production, cognate facilitation routinely occurs in picture naming (Costa, Caramazza, & Sebastia´n-Galle´s, 2000), suggesting there are two (nearly) identical lexical nodes (e.g., one Spanish and one Catalan) that provide converging activation to their shared phonological segments. Perhaps more indicative of unintended coactivation, Colome´ (2001) showed that in a phoneme-monitoring task using Catalan names of target pictures, Catalan–Spanish bilinguals needed more time to respond no if the Spanish translation included the target phoneme. The cognate, interlingual homograph, interlingual homophone, and neighbor effects discussed to this point build a strong case for nonselective access and the automatic coactivation of both a bilingual’s lexicons. The magnitude of these effects can be modulated by L2 proficiency, cross-language similarity, recent activity in each language, and sentence context. If interlingual homographs create conflict, then bilingual language control may involve inhibition, that is, the active suppression of nontarget representations. Macizo, Bajo, and Martin (2010) used an ingenious design to uncover evidence for inhibition. Spanish–English bilinguals were instructed to decide if two visually presented English words (e.g., pie-toe) were semantically related. The correct response, of course, is no. However, pie coactivates its orthographic representation in the Spanish lexicon and, in turn, its meaning in Spanish, which is “foot.” Foot is semantically related to toe, and this relation supports the incorrect response yes. If this conflict leads to the active suppression of the “foot” concept, then its current activation will be depressed. The critical test of this inhibition hypothesis occurs on the following trial. Suppose that pie-toe is followed by foot-hand. If the concept “foot” is now in an inhibited state, then it should take longer to correctly respond that foot and hand are semantically related. This is exactly what happened in comparison to a control condition. In order to investigate the possible consequences of bilingualism on nonlinguistic cognition it is necessary to draw subtle distinctions about

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

438

K E N N ET H R . PA A P

Concepts

L1 Lexical Forms CAT

L2 Lexical Forms GATO

Figure 22.1 Visualization of the processing of cat/gato by Spanish–English bilinguals according to the Revised Hierarchical Model (adaptation of figure 3 in Kroll & Stewart, 1994)

the processes underlying bilingual language control. Logically, coactivation is not a sufficient condition for triggering compensatory processing mechanisms – the comprehension and production systems could live with the crosslinguistic competition and accept the small delays in processing time and the occasional error. Setting that aside and granting that a compensatory mechanism is triggered, it could involve either upregulation (also termed over-activation) of the target language, or inhibition of the nontarget language. Among the effects reviewed so far, only the pie-toe interference effect reported by Macizo et al. (2010) requires inhibition as a mechanism, given that over-activation of the “crust-filled” meaning of pie cannot explain the results. As a final complication, inhibition in bilingual language control may be localized within the lexicalsemantic system or be under executive control from a supervisory system that selects task schemas, that is, overall goals that guide smaller actions, and sets decision rules. To sort through these fine distinctions, it is useful to consider functional and computational models of bilingual language control.

2.2 The Revised Hierarchical Model Consider a speaker of English (L1) learning Spanish (L2) in school. A typical technique for teaching vocabulary is for a teacher to say that the Spanish word for cat is gato. This invites an initial stage of L2 acquisition that associates new L2 word forms with their existing equivalents. Thus, access to the meaning (conceptual representation) is primarily mediated by back translation to L1. These ideas were instantiated in a model developed by Kroll and Stewart (1994) referred to as the Revised Hierarchical Model (RHM) (see Figure 22.1, in which all the connections are excitatory). The stronger link from L2 to L1 in the model, compared to the reverse,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

439

accounts for why L2 learners can translate faster from the L2 to the L1 than from the L1 to the L2. A critical problem for the RHM is that the strong excitatory connections from L2 to L1 predict that L2 gato should prime L1 cat (that is, should influence its activation), but, as Grainger, Midgley, and Holcomb (2010) point out, translation priming in proficient bilinguals is “a surprisingly elusive phenomenon” (p. 274) when primes are in the L2 and targets are in the L1 (see also Brysbaert & Duyck, 2010). A reconciliation proposed by Grainger et al. (2010) is that the RHM captures the earliest stage of foreign language acquisition, but that the need to control the growing competition between translation equivalents leads first to a dampening of the excitatory connections and eventually to connections that are mutually inhibitory. This more developed state has been implemented as a computational model called the Bilingual Interactive Activation (BIA) model (Dijkstra & van Heuven, 1998).

2.3 The Bilingual Interactive Activation (BIA) Model Grainger and Dijkstra (1992) formulated the rudimentary framework of what they later developed into a much fuller computational model of visual word recognition in proficient bilinguals called the Bilingual Interactive Activation (BIA) model (Dijkstra & van Heuven, 1998; van Heuven, Dijkstra, & Grainger, 1998). Figure 22.2 visualizes the architecture of the BIA model. It uses solid dots to represent inhibitory connections and arrows for excitatory connections. The curved inhibitory link on the ellipse representing the integrated lexicon represents the assumption that all word forms are mutually inhibitory regardless of their language membership. Activated word forms also start to activate their corresponding language nodes, which, in turn, have inhibitory connections to all word forms from the other language. The dotted lines represent the assumption that selecting a task schema can lead to top-down inhibition of nontarget lexical forms. Thus, in the BIA model non-cognate written translation equivalents are not serious competitors because the sublexical input c-a-t strongly activates cat but only weakly activates gato, and the inhibitory connections between words enable cat to “strangle” gato early in processing. Any small amount of competition created by partial overlap in orthographic forms (e.g., cat and gato) is quickly resolved by the inhibitory links within the integrated lexicon, and without recourse to any cognitive control from outside the lexical processing system. In the BIA model, translation equivalents with dissimilar forms (e.g., table and mesa) are not habitually coactivated during reading. In contrast to non-cognate translation equivalents, consider the much smaller set of interlingual homographs such as Spanish–English fin/fin. The BIA represents homographs as two separate lexical forms with

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

440

K E N N ET H R . PA A P

Language Nodes English

L1 CAT

Spanish

L2 GATO

Integrated Lexicon of Orthographic Forms

Sublexical Orthographic Forms

Visual Input CAT Figure 22.2 Visualization of the processing of cat by Spanish–English bilinguals according to the Bilingual Interactive Activation Model (BIA) (adaptation of figure 6.1 in Dijkstra & van Heuven, 1998)

base-rate activations adjusted for the relative frequency of occurrence in the language the form belongs to. With both lexical forms receiving strong bottom-up activation on each cycle from the letter string f-i-n it is difficult for even the higher-frequency homograph to get the upper hand through mutual inhibition. In order to resolve the competition rapidly, the BIA assumes that situational factors can also affect the activity in the two word nodes. If, for example, a participant was in a “go/nogo” English LDT where the response button is pressed for English words but withheld for non-words or Spanish words, then the selection of the “respond-to-English” task schema would introduce asymmetric interference through top-down control whereby the English language node would exert more inhibition on the Spanish word forms compared to the reverse. Depending on the parameter setting (i.e., the selected numerical strength of the top-down inhibition in a given simulation run), the amount of asymmetric interference can mimic a fine filter or a mere sieve with respect to screening out the competition from the nontarget homograph. Close simulations of the data often require sieve-like parameter values. That is, even when bilinguals are making lexical decisions in their first language, an L2 homograph can slow recognition of the corresponding L1 word and lead to a complete failure to recognize and respond to the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

441

L1 target word in 25% of trials. As Dijkstra and van Heuven (2002) observed, “parallel bottom-up activation of lexical candidates from both languages is so strong that suppressing it seems practically impossible” (p. 189). Viewed from this perspective, it is ironic that the very phenomena often said to demonstrate the need for inhibitory control (i.e., homograph interference effects) in managing two languages can also be seen as evidence that such top-down suppression is ineffective or even absent. As illustrated in the discussion of interlingual homographs, coactivation in the BIA is caused by orthographic overlap between the input letters and the word forms. For example, the target word gato will activate its neighbor gate because three of the letters match and are in the same order. On the other hand, the mismatching letter enables the target word to eventually get the upper hand. This illustrates that in the BIA, selecting a single competitor from among the candidates critically depends on the mutually inhibitory links between the word forms themselves. If a bilingual is reading a series of words in a single language (and if the language node decays more slowly than the individual word nodes), then selection receives an additional helping hand from the cross-language inhibition stemming from the language node. Note that the BIA account of the homograph and neighbor effects discussed earlier only uses the two inhibitory mechanisms nested within the Word Identification System: (1) mutual inhibition between word forms, and (2) cross-language feedback from the language nodes. When inhibition occurs within the Word Identification System, it is not recruiting domainfree executive functioning (EF) and is not providing ubiquitous practice that might enhance general inhibitory control. But the BIA also assumes that inhibition can be driven by task schema or decision rules through topdown activation of the language nodes and consequent inhibition of the other-language word forms. Thus the relative activity of the language nodes might be set to reflect a bilingual’s strategy in the task at hand – for example, increasing the inhibition on L1 words in order to compensate for a perceived weakness in L2, or because the task is to respond only to L2 words and ignore L1 distractors.

2.4 The BIA+ Model Dijkstra and van Heuven (2002) have updated the original BIA model and refer to the extended model as BIA+. As shown in Figure 22.3, the BIA+ adds phonological and semantic representations to the orthographic representations of the original BIA. Thus bilingual word recognition will be affected not only by cross-language orthographic overlap, but also by crosslanguage phonological and semantic overlap. Dijkstra and van Heuven (2002) did not actually add phonological and semantic representations to their computer model, but the functional model provides a framework for

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

442

K E N N ET H R . PA A P

TASK SCHEMA Procedures for target task Sets decision criteria Receives continuous input WORD IDENTIFICATION SYSTEM

Orthographic Forms L1

Language Nodes English

Spanish

L2 /gato/

L2 GATO

CAT

Sublexical Orthographic Forms

L1 /kaet/

Phonological Forms

Semantics

Sublexical Phonological Forms

Visual Input CAT Figure 22.3 Visualization of the processing of cat/gato by Spanish–English bilinguals according to the BIA+ Model (adaptation of figure 2 in Dijkstra & van Heuven, 2002)

extending the BIA principles to auditory word recognition (initiated by sublexical phonological representations) and to word production tasks whereby nodes in the conceptual/semantic system can be activated by, for example, the processing of pictures. The BIA+ makes a clear distinction between the encapsulated Word Identification System and domain-general cognitive control (the Task Schema/Decision System). Because the Word Identification System is encapsulated, the activation of the word form units and the language nodes cannot be affected by top-down control. Not shown in Figure 22.3 is that the Word Identification System is assumed to be part of a larger Language Processing System. The BIA+ further assumes that the syntactic and semantic constraints of this language processor can affect the activation of the lexical nodes. In the BIA+ word form unit, activation is affected only by processing within the Word Identification System. Linguistic information from the visual input (c-a-t) or from the preceding sentence context (e.g., the mouse was chased by the . . .) can and will affect the activation of word form units, but nonlinguistic contexts such as the participant’s expectations and strategies will not. Rather, these nonlinguistic factors influence performance via adjustments to parameter settings in the Task Schema/Decision System. The decision mechanism is part of a task schema and reads out the activation of nodes within the Word Identification System. For example, in general lexical decision tasks, a word response can be made when any word unit crosses an activation threshold. As the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

443

task unfolds, the activation thresholds and temporal deadlines may be adapted based on the frequency of different types of items. These adaptations involve dynamic adjustments of response criteria, but do not affect the activation level of individual word form units or the language nodes. Within the general BIA framework, the case of non-cognate translation equivalents is quite different for word production compared to word recognition. In word recognition, the bottom-up input of a letter string coactivates lexical forms that are homographs and neighbors of the target word, but not translation equivalents. In contrast, in picture-naming tasks, both translation equivalents will be coactivated because the BIA+ assumes that the conceptual representations are shared and that a picture of a cat will coactivate the addressed phonology of both gato/gato/ and cat/kaet/. In the original BIA model, the competition could be resolved at the lexical level by externally adjusting the Spanish language node to inhibit the English pronunciation. In the BIA+, this option is no longer available and a response would need to be withheld until information about language membership arrived from the language nodes.

2.5 More on Translation Priming and Coactivation In both versions of the BIA, non-cognate translation equivalents are automatically coactivated in production (e.g., in picture naming), but not during word recognition (e.g., LDT). Is the prediction that lexical access (during word reading) takes place without competition from the translation equivalent reasonable? Or, does cat have to compete with gato in the recognition of c-a-t? The answer requires a distinction between activations that occur before versus after a word is recognized. Obviously a Spanish–English bilingual, when asked to translate the English word cat, can say gato. Similarly, after recognizing the word cat as a prime, activation may spread to semantically related words, including the translation equivalent gato, and facilitate their recognition. Thus, BIA-type models can predict translation priming, but the priming effect should take time to develop and does not entail early competition between the translation equivalents. As pointed out in the section on the RHM, there may also be differences depending on whether L1 to L2 priming or L2 to L1 priming is involved, with the latter being considered “nearly impossible” (Brysbaert & Duyck, 2010, p. 365). This conclusion likely still stands today when behavioral measures are considered. However, research by Wu and Thierry (2010) involving neurophysiological methods and an implicit priming paradigm shows that bilinguals may constantly coactivate the word being processed in the target language (L2) and the corresponding translation in their other language (L1). Chinese–English bilinguals had to decide whether two sequentially presented words in English, their L2, were semantically related. The intriguing condition consisted of pairs like experience–surprise

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

444

K E N N ET H R . PA A P

whose members were unrelated in English orthography, phonology, and semantics but whose oral translations in their L1, Chinese, shared the same initial syllable: the respective Chinese translation equivalents for experience and surprise are JingYan and JingYa. This was the sound repetition condition. The event-related brain potential N400 amplitude was reduced in this condition compared to control conditions. This is consistent with a fast coactivation of the Chinese translations. Although there were no behavioral differences in speed or accuracy, the N400 results led Wu and Thierry to entitle their article: “Chinese–English Bilinguals Reading English Hear Chinese.” This titular conclusion seems to challenge the BIAtype assumption that activation of non-cognate translations primarily occurs after lexical access. However, if the N400 response is time-locked to the onset of the second word, then it is fair to suggest that it is reflecting activity well past the point in time that the second word has been recognized and reflects the degree to which the second word fits the semantic context of the first. A more far-reaching challenge to the general hypothesis that both a bilingual’s languages are quickly and automatically coactivated stems from an alternative interpretation of the general evidence for coactivation recently advanced by Costa, Pannunzi, Deco, and Pickering (2017). They propose that sequential bilinguals “carry over” remnants of the structure of their L1 to the L2 as a consequence of using the L1 to learn the L2. This learning alternative can explain cognate facilitation and homograph interference, as well as the Wu and Thierry (2010) N400 effects, even in singlelanguage contexts when the other language is totally irrelevant. Using the Wu and Thierry example, native Chinese speakers have a form-based association between JingYan and JingYa based on the overlapping first syllable. As these speakers learn the corresponding translation equivalents, experience will be paired with JingYan and activation will spread to the overlapping form JingYa and then to its meaning, surprise. Based on the Hebbian principle that units that fire together wire together, these connections will strengthen. As a result, the corresponding English translations (experience and surprise) will acquire a rudimentary association in a budding Chinese–English bilingual even though experience and surprise are not related in English in any way. According to these principles, at some point during the learning of the L2, the presentation of the word experience will lead to the activation of surprise in the total absence of any current contribution from their Chinese translations. Using a computational model, Costa et al. (2017) demonstrated that these types of rudimentary associations can in fact be acquired and may facilitate the output even when the “online” connections between translation equivalents are completely severed and can no longer influence the output. They conclude that their simulation provides support that effects usually interpreted as evidence of coactivation of both languages could simply be the product of learning two languages and that there can be

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

445

subtle but important differences between the English lexicon of a native English speaker and the English lexicon of a fluent bilingual who has English as one of her/his languages. Although Costa et al.’s work supports the hypothesis that highly proficient bilinguals could transition to a state where lexical access is selective and translation equivalents are no longer coactivated, it does not necessarily replace the popular view of rampant coactivation, nonselective access, and a ubiquitous need to inhibit the nontarget language (see later in this chapter). All these could be caused by coactivation, learning, or a combination of both.

2.6 Green’s Inhibitory Control Model Perhaps the most influential model of bilingual language control is Green’s (1998) Inhibitory Control Model (ICM). The ICM shares several critical assumptions with the original BIA. Foremost is the need for higherlevel (language) task schemas. For example, when presented with a printed word, a bilingual can be given the task of reading it silently, reading it aloud, translating it, generating an associate, classifying it as animate or inanimate, and so on. Each requires a different task schema. The ICM emphasizes that task schemas compete with each other for controlling action. In the Stroop color-word task (see Note 1), the task schema specifies naming the color of the printed word (without reading the word), but for a skilled reader, the “name-the-color” schema will have to inhibit the more automatic “read-the-word” schema. Both the BIA and the ICM models assume top-down control over the representations of nontarget words. However, the ICM makes an interesting and more specific assumption, namely, that the actual inhibition is not applied until competition for production occurs. Consider a picturenaming task where a pre-cue indicates whether the upcoming picture should be named in English or Spanish. When the pre-cue requires a switch, say from English to Spanish, the new schema (“name-inSpanish”) must be selected and this sets the stage for the global inhibition of all the lemmas with English tags. However, the actual inhibition is not applied until: (1) the picture appears, (2) the participant initiates plans to name the picture, and (3) response competition is detected. For example, if the stimulus is a picture of a cat, then the competition between the correct response gato and the competitor cat will trigger inhibition of all English lemmas in proportion to the degree that they were activated by the picture. Thus, cat would be inhibited more than unrelated English words like table. Although Green (1998) discusses some early evidence that is consistent with the reactive inhibition assumption, a more recent study by Runnqvist, Strijkers, Alario, and Costa (2012) poses an intriguing challenge to the ICM (and, for that matter, to all models of bilingual control discussed here). Participants were instructed to name a long series of pictures that

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

446

K E N N ET H R . PA A P

belong to different semantic categories such as vehicles, clothing, or pets. The categories were not presented in blocks, but nonetheless each additional item in the same category took about 20 ms longer to name. One way of explaining this assumes that naming an object leaves its representation in a more highly activated state (i.e., above baseline) and that consequently it is a more formidable competitor when it becomes reactivated by a picture of an object from the same semantic category. This Cumulative Semantic Interference (CSI) is thought to occur when the effects of each new object from the same category are additive. The intriguing twist here is what happens if bilinguals switch between their two languages. In the critical condition of the Runnqvist et al. (2012) experiments, participants alternated between naming pictures from a particular category in Spanish and other pictures from the same (or another category) in Catalan. The surprising result was that the slope of the CSI effect was identical in the languagealternating and language-invariant (cf. previous paragraph) conditions. Models assuming any type of global inhibition of the nontarget language (whether immediate or reactive) would predict no CSI or at least a substantially reduced CSI. To summarize, both the original BIA and the ICM assume that task schemas can globally inhibit nontarget word units through their language nodes/tags and that this is needed to quickly and consistently filter out the competitors. As described earlier, this assumption is inconsistent with the fact that instances of CSI are not diminished by language switching. If the global inhibition assumption does not hold, then bilingual language control becomes more and more like monolingual lexical processing where competition is resolved simply by the mutual competition between word forms and where language tags simply function like another semantic feature.

2.7 The Role of Inhibition in Language Switching An aspect of language control unique to bilingualism is the need to switch from one language to another. For example, bilinguals switch languages to accommodate the languages spoken by their conversational partners (De Houwer, Chapter 17, this volume). Most theories assume that language switching involves the selection of a new language task schema, for example, a “speak-English” schema. There is a cost when bilinguals switch from one language to the other in tasks like cued picture naming. These switch costs are defined as the difference in mean reaction time (RT) between switch trials and repeat trials and are generally assumed to consist of two components. The first component is the time it takes to select and implement the new language schema on a switch trial. This component reflects the intrinsic time it takes to reset the experimental goal and is assumed to be the same when switching from LX to LY or from LY to LX. The second

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

447

component refers to any additional time it takes to select and prepare the correct word for production due to the switch. For bilinguals who are still in the middle of learning their L2, switch costs are asymmetric: it is more difficult for them to switch back from the L2 to the L1 than to switch from the L1 to the L2 (Costa & Santesteban, 2004). The fact that it is more difficult to get “back into the groove” of one’s L1 may seem counterintuitive, but fits in perfectly with the ICM’s assumption that bilingual language control involves reactive inhibitory control whereby L1 words are fierce competitors and require more inhibition during L2 word naming. Further, assuming that the inhibition carries over to the next trial, a switch to the L1 will lead to a greater delay in lexical access. Thus, the ICM account is that the observed switch cost asymmetry in L2 learners is caused by reactive inhibition in the second component because L1 words have been more severely inhibited and it will therefore take longer to select the correct word compared to when the switch is into the L2. The inhibition account of asymmetrical switching costs is also consistent with the result that highly proficient bilinguals usually show symmetrical switch costs. This follows because words in both languages would require the same amount of reactive inhibition. As an alternative to inhibition, Costa and Santesteban (2004) suggested that proficient bilinguals learn how to selectively access the lexical representations in the target language and ignore the nontarget words regardless of their activation levels. This hypothesis is superior to reactive inhibition in accounting for the pattern of switch costs that Costa and Santesteban (2004) obtained with a third group of participants, i.e., trilinguals who were highly proficient in Spanish (L1) and Catalan (L2), but much less proficient in their L3, English. In a picture-naming task that cued either L1 or L3 (but never L2), the switch costs were symmetrical. This is consistent with the hypothesis that these bilinguals have learned how to selectively access their lexicons. In contrast, a reactive inhibition hypothesis predicts that switch costs should be asymmetric, with greater costs for switching into the L1 than into the L3.

2.8

The Relation between Language Switching and General Task Switching Green (1998) developed the ICM within the context of a general supervisory attentional system (SAS). As such, selecting and switching to a language schema such as “name the object in Spanish” may involve the same supervisory processes as selecting and switching to a nonverbal “make guacamole” schema. Paap, Myuz, Anders et al. (2017) discuss the relation between language switching and general task switching as a contrast between the overlap and independence hypotheses. Consistent with the ICM, the overlap hypothesis assumes that switching between

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

448

K E N N ET H R . PA A P

two nonverbal tasks involves the same control mechanism as switching between two languages. However, there is substantial behavioral evidence that language switching is not subsidiary to general task switching and that the functional overlap between the two types of switching is minimal. The independence hypothesis is consistent with research summarized by Paap et al. (2017) showing that language switching performance does not correlate with performance on nonverbal switching tasks, that nonverbal task switching shows a much greater decline with age compared to language switching (see further Goral, Chapter 5, this volume, on language switching in aging populations), and that general switching abilities can be severely impaired while language switching is not (for more on language switching under impaired conditions, see Hernandez, Chapter 23, this volume). The question of whether bilingual language control relies on general executive control has been explored in the same task for both inhibition and switching. Branzi, Calabria, Boscarino, and Costa (2016) tested 62 Spanish–Catalan bilinguals who had learned English as a foreign language at school under conditions where they had to switch between three different languages (Catalan, Spanish, English) or between three different tasks (color, size, shape). Using A, B, and C as markers of the three languages or three nonverbal tasks, consider the sequence CBA versus CAA. The difference in RT on the last trial (trial n) of these two sequences provides the standard measure of switch costs as CBA is a switch trial and CAA is a repeat trial. Following Branzi et al. (2016), this comparison is called the n-1 shift cost. Next consider the difference between sequences CBA and ABA. Both are switch trials, but ABA requires a return to a task (A) that presumably was inhibited by the switch to B on trial n-1. If the inhibition of A carries over to trial n, then ABA sequences should produce slower RTs compared to CBA. This difference is called the n-2 repetition cost. When nonlinguistic tasks are used, n-2 repetition costs are robust, reliable, and uniformly interpreted as an index of inhibitory control. Branzi et al.’s (2016) design enables multiple tests of whether the relation between bilingual language control and general executive control is independent or overlapping. First, consistent with the results of other experiments, there was no significant correlation between the switch cost (n-1 shift cost) in language switching and in general task switching. This confirms that language switching and general switching rely on different mechanisms. Second, there was no significant correlation between n-2 repetition costs in language switching and nonlinguistic task switching. Given the assumption that switching costs are due to inhibition, it follows that the inhibitory mechanisms are not the same for the two types of switching. Furthermore, Branzi et al. (2016) found that the magnitude of the n-2 repetition costs was much smaller in language switching (13 ms) than in nonlinguistic switching (47 ms). This suggests

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

449

not only that the inhibitory mechanism is different for the two types of switching, but also that it is barely used at all during language switching in this population of proficient bilinguals.

2.9 The Role of Inhibition in Bimodal Language Control Another way of exploring the role of inhibition in bilingual language control is to compare speakers of two languages (e.g., English and Spanish) to fluent signers who are also fluent in a spoken language (e.g., English and American Sign Language [ASL]). Setting literacy skills aside, the first category of bilinguals is unimodal (auditory) while the second is bimodal (visual and auditory). Because hand gestures and articulatory gestures can be produced simultaneously there is no response competition between a bimodal bilingual’s two languages. Perhaps this is why bimodals reveal a strong preference for code-blending over code-switching (see Emmorey, Giezen, & Gollan, 2016, for a review). For example, if a bimodal is speaking English, (s)he will often simultaneously produce the sign for the translation equivalent (or even a word that is never spoken but that adds relevant information; see Tang & Sze, Chapter 24, this volume). Studies of picture naming by bimodal bilinguals suggest that such code-blended production does not incur a lexical processing cost, despite the fact that two lexical items must be retrieved and produced simultaneously. During the production of sentences, one can assess the cost of moving into a blend separately from moving out. Consistent with picture-naming data, sentence production results indicate no processing cost of moving into a blend. In contrast, moving out of a blend does incur a cost. Although generalizing from bimodals to unimodals may be risky, this pattern suggests that turning on an inactive language schema requires little cognitive effort, but that turning off (inhibiting) an active schema takes more time and effort.

3 The Bilingual Advantage in Executive Functioning Hypothesis The architecture of the bilingual lexicon and the nature of its control mechanisms may have consequences in other domains (e.g., Bialystok & Craik, 2010). In particular, if bilingual language control – relative to monolingual control – involves constant competition between cross-language lexical representations and if that competition is managed by top-down inhibitory control, then this ubiquitous practice may enhance general inhibitory control and lead to a bilingual performance advantage in nonverbal interference tasks. That is, the inter-linguistic competition in bilinguals may lead to an advantage in overall executive functioning. More specifically, a bilingual advantage in inhibitory control would be signaled

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

450

K E N N ET H R . PA A P

by smaller interference scores in bilinguals than monolinguals (mean RT on incongruent trials where conflict is present minus mean RT on congruent trials where conflict is absent; see Note 1 for a brief explanation of (in)congruent trials in various experimental tasks). Similarly, if the overlap hypothesis (cf. Section 2.8) has merit, bilinguals should be better general task switchers compared to monolinguals depending on the extent to which they engage in frequent language switching. The bilingual advantage hypothesis is attractive because it is grounded in popular models of bilingual language control (e.g., the BIA and the ICM) and counteracts earlier deficit views that acquiring a second language has negative consequences for cognition (see Hakuta, 1986, for a detailed methodological critique). However, both the underlying logic of why bilingual advantages should accrue and the empirical evidence supporting the hypothesis have been challenged, and the author is one of the skeptics (e.g., Paap & Greenberg, 2013; Paap, Johnson, & Sawi, 2015). These articles examined many aspects of the debate and our conclusion was that bilingual advantages in EF either do not exist, or are restricted to very specific and undetermined circumstances. The main arguments supporting this conclusion are summarized in Sections 3.1 and 3.2. These are followed by a section exploring circumstances under which bilingual advantages might still exist and a final section presenting an alternative.

3.1

Bilingual Advantages are Relatively Small in Number, Effect Size, and Sample Size Paap et al. (2015) tallied the number of significant and nonsignificant bilingual advantages in studies using nonverbal interference tasks. Only 20% of the tests yielded significant bilingual advantages. More importantly, all of the tests based on large sample sizes showed null results. This is the opposite of what would be expected if the null hypothesis was false and bilingual advantages were real, but small in size. With respect to formal meta-analyses focusing primarily on interference effects there is a small effect size of about d = 0.3 (de Bruin, Treccani, & Della Sala, 2015). Furthermore, both Paap and colleagues and the Basque Center for Cognition, Brain, and Language (BCBL) group in Spain have consistently used Bayes Factor analyses in addition to null hypothesis testing. A Bayes Factor is a ratio of the probability of the null hypothesis being true (given the data) over the probability of the alternative being true (given the data). Common guidelines suggest that Bayes Factors in the range of 0 to 3 are not very meaningful, but as they grow in the range from 3 to 10 they provide substantial evidence that the null hypothesis is true. To take some examples, Paap, Johnson, and Sawi (2014) computed Bayes Factors for 12 different measures of EF comparing bilinguals to monolinguals. Nine of the 12 comparisons produced substantial Bayes Factors in the range of 5 to 9. Across the BCBL studies, Bayes Factor analyses showed substantial

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

451

support for the null hypothesis in the flanker task (5.6), the color-word Stroop task (5.6), and the numerical Stroop task (3.6) (for brief explanations of what these tasks involve, see Note 1). Sometimes studies showing no differences with respect to the interference effect show that bilinguals were faster on both congruent and incongruent trials. This global RT effect (averaged across both trial types) is often interpreted as a bilingual advantage in monitoring (Hilchey & Klein, 2011). However, Paap et al.’s (2015) review showed that bilingual advantages in measures of monitoring were also in a distinct minority. Furthermore, as Hilchey, Saint-Aubin, and Klein (2015) reported in their updated review, the consistent advantages in global RT on nonverbal interference tasks observed with schoolchildren, young adults, and the elderly have vanished in the wake of null results reported in newer, often larger, and often better-controlled studies. The seminal study showing a bilingual advantage in switching cost (Prior & MacWhinney, 2010) used a color-shape switching task that appeared to have a medium effect size and adequate power. However, when Paap et al. (2017, Table 1) examined the 17 studies and 53 tests for bilingual advantages that used some version of the color-shape task, significant advantages in switching costs were obtained in only 2 of 26 replications. Thus, the empirical evidence strongly tilts in the direction of no differences between bilinguals and monolinguals when it comes to switching ability. Prior (2012) also contributed a seminal paper testing for bilingual advantages in n-2 repetition costs. In a comparison of 30 Hebrew–English bilinguals in Israel to 30 English-speaking monolinguals in the United States, the bilinguals had significantly larger n-2 repetition costs. Under the prevailing view at the time, Prior predicted this difference under the assumption that bilinguals were better at inhibitory control and that the larger n-2 repetition costs reflected “too much of a good thing” (cf. Prior’s article title). However, this significant group difference was not reproduced by Branzi, Calabria, Gade et al. (2018) in two experiments using Spanish–Catalan bilinguals and Spanish monolinguals, who, admittedly, all had some degree of proficiency in a foreign language (English) that they had learned from primary school onward but did not use in their daily lives (as such, especially the monolinguals were likely different from most of the monolingual participants in Prior, 2012). Branzi et al.’s (2018) second experiment measured both n-1 switch costs and n-2 repetition costs with about 100 participants in each group.

3.2 Other Issues with the Bilingual Advantage Hypothesis This section discusses additional issues that impinge on the validity of the bilingual advantage hypothesis. The first relates to the well-known publication bias against null results in science: editors and reviewers are more

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

452

K E N N ET H R . PA A P

likely to recommend publication of positive results over null or negative ones. This bias often leads researchers to place new studies with null results in the proverbial “file drawer” and not even attempt publication (Rosenthal, 1979). As demonstrated by de Bruin et al. (2015), psychological science and the literature on the bilingual advantage in particular are not immune to these biases. The existence of bias implies that a small average effect size of d = 0.3 might easily shift to zero if the analysis were to include studies that were missing because they were rejected and/or placed in the file drawer. Furthermore, as Paap et al. (2015) demonstrated, the extant literature on an assumed general bilingual cognitive advantage contains some statistical and methodological problems that in combination can lead to a substantial number of false positives. These include the use of questionable research practices such as reporting only the experiment that works, the inappropriate use of the analysis of covariance, the misinterpretation of critical interactions, and the lack of control for confounding factors. The latter is a major challenge to this research question because bilinguals and monolinguals inevitably differ on other dimensions than just the number of languages known, such as culture, socioeconomic status, educational level, immigrant status, age, and general fluid intelligence. A study that controlled for many such factors found no bilingual–monolingual differences in measures of EF (de Bruin, Bak, & Della Sala, 2015). A difficult complication for presenting compelling evidence for a bilingual advantage for any of the specific components of EF such as updating, monitoring, inhibition, and shifting is that constructs like inhibitory control must be captured with impure measures of task performance. Part of the inconsistency across studies is likely due to the use of different measures of the same construct that have inadequate levels of convergent validity (see Paap & Sawi, 2014, for extensive discussion). Finally, another problem with interpreting assumed bilingual advantages is that the relation between language experience and EF might embody a reverse causality, particularly in studies featuring sequential bilingual samples. Although the executive control required to learn a second language could enhance general EF, it is just as likely that individuals with better EF are better able to master a second language.

3.3

Searching for the Necessary and Sufficient Bilingual Experiences Although proponents of the bilingual advantage hypothesis often claim that most bilinguals have better cognitive control than most monolinguals, there appears to be a growing consensus that bilingual advantages may be restricted to specific types of bilinguals. The risk of over-generalizing is illustrated by considering the implications of the meta-analytic effect size

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

453

of d = 0.3 mentioned in Section 3.1. This indicates that 88% of the bilingual–monolingual distributions overlap and that the probability of superiority of a randomly selected bilingual compared to a monolingual is just 58%. One possibility is that age of acquisition modulates the findings for a bilingual advantage. In a study comparing monolinguals and bilinguals with an age of learning onset before or after 10, Luk, de Sa, and Bialystok (2011) reported no group differences in monitoring but an advantage in inhibitory control for (i) early bilinguals over both (ii) late bilinguals and (iii) monolinguals, thus suggesting that there are differences among bilinguals that relate to when they acquired their two languages. As Paap et al. (2015) observed, though, none of the six subsequent studies comparing these three groups resulted in a unique advantage of early bilinguals over either late bilinguals or monolinguals in inhibitory control. Another aspect distinguishing bilinguals beyond age of acquisition is the frequency of language switching. If language switching enhances general task switching ability, then bilingual switching cost in nonverbal switching tasks should be negatively correlated with the frequency of language switching. As a corollary, if language switching involves the top-down inhibition of lexical representations, then frequent language switchers should also have better inhibitory control. Indeed, Verreyt, Woumans, Vandelanotte et al. (2016) found that the frequency with which bilinguals switch languages is the key determinant in facilitating a bilingual advantage in inhibitory control. However, in an analysis using considerably larger samples Paap et al. (2014) reported no differences between high-switch bilinguals, low-switch bilinguals, and monolinguals on interference scores obtained from either Simon or flanker tasks (for brief information on what Simon tasks usually involve, see Note 2). Furthermore, Paap et al. (2017) analyzed a composite database of 320 bilinguals and found no relation between reported frequency of language switching and switching costs. On balance, there is thus very little evidence supporting the hypothesis that frequency of language switching affects either bilinguals’ inhibitory control or their switching ability. Another possibility is that highly proficient bilinguals are most likely to show bilingual advantages because coactivation and competition are always strong, regardless of the target language. Paap et al. (2014) used their composite database to compare bilinguals with high, middle, and low L2 proficiency to one another and to monolinguals. No significant group differences appeared in flanker interference or global RT. Likewise, no group differences were found in switching costs or mixing costs. Thus, differences or similarities in L1 and L2 proficiency levels as a possible factor for bilingual advantages are likely not relevant. Indeed, none of the individual factors mentioned so far may be sufficient to consistently enhance EF, but perhaps in combination they can. The BCBL has taken advantage of their local population of highly

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

454

K E N N ET H R . PA A P

proficient early bilinguals who switch languages frequently to test this hypothesis. In the Spanish Basque country, Spanish and Basque are coofficial languages. Both languages are acquired early, 50% of all academic subjects are taught in both languages, and the local bilingual community provides continuous opportunity for language switching. The BCBL studies (e.g., Anto´n, Dun˜abeitia, Este´vez et al., 2014; Dun˜abeitia, Herna´ndez, Anto´n et al., 2014; Anto´n, Ferna´ndez-Garcı´a, Carreiras, & Dun˜abeitia, 2016) are also distinguished by large sample sizes, multiple measures of EF, excellent matching on potential confounding variables, and testing over a wide range of ages ranging from first graders through the elderly. The results are consistent in showing no bilingual–monolingual differences on measures assumed to reflect monitoring, inhibitory control, or switching ability. The BCBL studies lead to a rejection of the hypothesis that highly proficient bilinguals have the necessary experience to produce bilingual advantages. Perhaps the opposite is true, that is, it may be the case that general EF is recruited only during initial L2 learning when L1 is clearly dominant and requires constant monitoring and inhibition in order to prevent intrusions when the intention is to speak L2. In recent symposia, Paap has incorporated this idea into the formulation of a controlled-dose hypothesis that, building on Chein and Schneider’s (2012) work on skill acquisition, proposes that L2 learners transition from a state that heavily relies on executive control to one that becomes automatic and less reliant on executive control. The controlled-dose hypothesis is predicated on two basic assumptions. The first is that any regimen that requires intense EF generates a boost in EF ability, provided that the required EF is above the individual’s current adaptation level. The second critical assumption is that the boost emerges rapidly, but dissipates with either plain disuse or with disuse caused by the transition from controlled to automatic processing. The controlled-dose hypothesis makes a prediction that initially seemed attractive – namely, that bilingual advantages are most likely to develop in the early stages of L2 learning, especially when the learners are preschoolers or elderly and thus presumably have lower adaptation levels than young or middle-aged adults. However, this prediction failed in two recent studies. In another BCBL study (Ramos, Ferna´ndez Garcia, Anto´n et al., 2017), elderly Spanish monolinguals were given an intense dose of learning Basque for an entire academic year. However, they showed no improvement at all in switch costs derived from a color-shape switching task and no differences compared to a control group. At the other end of the age spectrum, Woumans, Surmont, Struys, and Duyck (2016) carried out a longitudinal study of French-speaking four-year-olds. Half of the children were then immersed in one year of Dutch–French classes while the other half continued in single-language instruction in French. There were significant bilingual advantages on the Ravens’ test of nonverbal intelligence,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

455

but more relevant to the discussion of the controlled-dose hypothesis is that there were no group differences in either Simon or flanker interference effects or in global RT. Thus, the plausible idea that there is a critical L2/L1 ratio that promotes bilingual advantages in EF has not achieved any compelling empirical support. Finally, there is some interesting evidence that for bilinguals, the interactional context may affect switch costs in nonlinguistic tasks. The Adaptive Control Hypothesis (ACH) (Green & Abutalebi, 2013) extends and enriches the ICM by focusing on how the cognitive demands of bilingual language control dynamically adapt to demands imposed by different types of interactional contexts. In a pure single-language context, the context-appropriate language is used throughout the interaction. For example, a student’s classroom, work, and campus interactions may be conducted exclusively in English and conversations at home exclusively in Mandarin. The recurrent interactional contexts for this individual are strongly single-language oriented. In strong dual-language contexts, switching is more frequent and triggered by partners entering or leaving the conversation in the same context and by other sociolinguistic factors (De Houwer, Chapter 17, this volume). Not only are there frequent switches in a dense code-switching context, but many occur within sentences, including the adaptation of words from one language in the context of the other (see also Aalberse & Muysken, Chapter 26, this volume). The ACH considers that each context is best served by a different subset of control processes. Green and Abutalebi (2013) speculated that adaptation to the dual-language context is likely to increase the demand on several control processes: goal maintenance, interference control, salient cue detection, selective response inhibition, and task disengagement and engagement. Dual-language contexts rely on more control processes than single-language contexts, while dense code-switching does not. Furthermore, if an individual predominantly engages in one control process it is assumed that the language control system will adapt to maximize efficiency, capacity, and/or connectivity for that process and context. Consistent with predictions derived from the ACH, Hartanto and Yang (2016) compared bilinguals in a single-language interactional context to bilinguals in a dual-language context on the performance of a color-shape switching task. (Note that a partition based on single- versus dual-language contexts is different from a partition based simply on the overall frequency of language switching.) The dual-language context group had significantly smaller switch costs. Furthermore, in regression analyses, ratings of dual-language use were better predictors of switch costs compared to the simple frequency of intersentential switches. Finally, the frequency of intrasentential switches positively predicted (i.e., exacerbated) switch costs. If intrasentential switches are valid measures of the degree to which bilinguals operate in an open control mode (see Section 3.4), then the results are consistent with the ACH. However, the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

456

K E N N ET H R . PA A P

group differences reported by Hartanto and Yang (2016) do not directly show that dual-language-context bilinguals have better switching ability than monolinguals, as no monolinguals were tested. To date, the search for the necessary and sufficient conditions that will produce bilingual advantages has not paid many dividends in examinations of age-of-acquisition, L2 proficiency relative to the L1, frequency of language switching, and interactional context. Hartsuiker (2015) suggests that this is inevitable until cognitive science can formulate more precise theories of language control, skill generalization, and cognitive control.

3.4 Why Are Bilingual Advantages in EF Rare or Nonexistent? The most likely reason why bilingual advantages in EF rarely occur is that bilingual language control only modestly involves top-down cognitive control. Section 2 of this chapter reviewed the substantial literature consistent with the assumption that there is nonselective access of the bilingual lexicon and that cognates, homographs, and neighbors are coactivated and compete for selection in both comprehension and production. However, the evidence that this competition triggers ubiquitous top-down inhibitory control was not as compelling. Among the alternatives are: (1) bilingual language control can usually operate in an open control mode where cross-language competition is simply tolerated, leading to relatively small delays in lexical access and occasional intrusions, (2) cross-language competition is managed within the languageprocessing module and is not augmented by top-down control external to the module, and (3) learning an L2 by reference to an L1 may leave traces of L1 organization that can mimic the effects commonly taken as strong evidence for coactivation and competition. Bialystok (2015) has been quite explicit in dismissing inhibitory control as being relevant to EF: “The bilingual advantage is not in inhibition; rather it is the failure of bilinguals to inhibit attention to the nontarget language that leads to the involvement of executive function and the eventual consequences for its development and function” (p. 4, original emphasis). Thus, according to this formulation, bilinguals do not inhibit conflicting representations or responses, rather, they manage them. What does this mean? One answer is that bilinguals are better at selectively attending to goal-relevant information and at disengaging from irrelevant information. Friesen, Latman, Calvo, and Bialystok (2015) reported a bilingual advantage in the speed with which respondents identified target shapes in a visual search task with many distractors. In contrast, Ratiu, Hout, Walenchok et al. (2017) found no such advantages in a series of experiments that could separate visual search time from decision time because eye movements were monitored. The view that bilinguals are better at monitoring for important environmental cues and reconfiguring task schema as needed is also challenged by the results of nonverbal

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

457

switching tasks that compare single task-performance (e.g., a block of pure shape or pure color trials) to performance on a block where the two tasks are randomly mixed together: as Paap et al. (2017) reported in their Table 1, only 1 of 20 such comparisons reported bilingual advantages in mixing costs. Summing up, bilinguals who primarily operate in a dual-language mode could in principle increase their general task-switching ability. Yet we know that language switching and general task switching often dissociate and that there is substantial behavioral evidence favoring the independence hypothesis (Section 2.8). The most likely way that language switching becomes independent of general task switching is through a rapid transition from a controlled process to one that is automated within the language-processing module. This would entail the ability of external cues (e.g., a change in conversational partner) and internal states (e.g., a plan to change the topic) to automatically trigger a language switch without the need for higher-level cognitive control.

4 Linguistic Relativity across a Bilingual’s Two Languages Even if there is no bilingual advantage in EF, there may be advantages in mental flexibility. One manifestation of that may be that bilinguals can perceive and understand the world in different ways through the different languages they speak. For this to be true, there must be some form of linguistic relativity, that is, the specific language monolinguals speak must influence perception and cognition in ways that are different from monolingual speakers of other languages. This, of course, is the famous Sapir-Whorf hypothesis that has spawned a vast empirical literature and continued debate. With respect to the effect of language and bilingualism on lower-order perception, the most modest claim would be that language acts as an attention-directing mechanism that influences judgments of perceptual qualities. There is only weak evidence for this in work with bilinguals. Athanasopoulos, Damjanovic, Krajciova, and Sasaki (2011) had participants make similarity judgments between color patches matched on their physical separation that either fell within a color name or crossed a naming boundary between Japanese basic color terms for “light” versus “dark” blue. Participants included monolingual speakers of English or Japanese, and Japanese–English bilinguals. In addition, the physical differences in wavelength could be “near” or “far.” The significant three-way interaction had a surprising pattern. When the colors were physically near, the similarity ratings were the same for all three groups regardless of whether colors crossed the word category boundary. In contrast, when the colors were physically more dissimilar, Japanese monolinguals rated pairs that crossed the boundary as more dissimilar than the pairs on the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

458

K E N N ET H R . PA A P

same side. That is, whether each color was associated with the same color name or a different name only mattered for the pairs that were physically quite clearly different from one another. If having different names actually enhanced the observer’s perceptual sensitivity to a small physical difference in hue, then one would have expected the opposite. The most likely explanation for the obtained interaction is that speaking Japanese did not influence the perceptual processing of color, but rather affected performance post-perceptually. The specific pattern may be related to the fact that people are much better at making fine judgments of similarity than they are of degrees of dissimilarity. This may have rendered those judgments most susceptible to an additional (non-perceptual) source of information. For each bilingual Athanasopoulos et al. (2011) computed a Categorical Perception Index (CPI), which is the mean similarity on the withincategory trials minus the mean on the cross-category trials. The CPI is an index of the degree to which the L1 distinction between Japanese hues of blue affected the similarity judgments. The CPI was correlated with several possible predictors: a measure of L2 proficiency, the proportion of daily activities conducted in the L2, and the experimental setting (i.e., whether participants were given instructions in Japanese or in English). The only significant predictor was a negative correlation with L2 use. Athanasopoulos et al. (2011) characterize this result as displaying “cognitive flexibility” because the bilinguals’ judgments depended on which language they used most frequently in their daily activities. This makes sense if all one means by cognitive flexibility is that the influence of the L1 on color judgments is modifiable. But this has nothing to do with an online flexibility that is under the control of the bilingual or can be selectively tuned by the current language environment. The study actually suggests that the opposite may be true, as the interactional context of the experimental session (Japanese- or English-speaking experimenter) had no influence on the CPI at all. The effects of language and bilingualism on higher-order event conceptualization appear to be more robust. Athanasopoulos, Bylund, Montero-Melis et al. (2015) capitalized on crosslinguistic differences in the expression of how events relate to the flow of time. In aspect languages, an ongoing action is obligatorily marked on the verb (e.g., the progressive -ing in walking). Speakers of aspect languages are therefore more likely to perceive events as ongoing and place less emphasis on the eventual destination. In contrast, languages such as German that lack viewpoint aspect describe events more holistically and hence direct less attention to the ongoing motion. Speakers of German are more likely to include the endpoint as an important component of the overall event. To the extent that language affects event perception, speakers of aspect and non-aspect languages should show differences in how they judge the similarity of events that vary with respect to how salient the end goal is.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

459

Each trial of the task Athanasopoulos et al. (2015) used to assess participants’ event judgments consisted of a sequence of three video clips, each lasting six seconds. The clips depicted different actors, pathways, and destinations, but differed systematically with respect to goal salience: one had low salience as the goal was far off in the distance, another had high goal salience ending with arrival at the destination, and the third was intermediate. The task required participants to decide if the last (intermediate) clip was more similar to the high or low goal salience. German monolinguals, exercising their preference for events with endpoints, selected the high-salience alternative about 41% of the time, compared to about 23% of the time for English monolinguals. This comparison supports the hypothesis that language affects judgments of event similarity. More interesting were Athanasopoulos et al.’s (2015) findings concerning two groups of German–English bilinguals. Bilinguals were randomly assigned to a condition where the experiment was conducted in English or in German. The group in the German-language context performed statistically equivalently to the German monolinguals, while the group in the English-language context performed statistically the same as the English monolinguals. This suggests that German–English bilinguals conceptualize events from different perspectives depending on which of their languages is currently active. They are not stuck in a position somewhere between their L1 and L2. Even more intriguingly, when the bilinguals were instructed to overtly rehearse two-digit numbers while they were viewing the clips, their similarity judgments shifted in the direction usually associated with the inactive language. Apparently, the task-irrelevant digits held and rehearsed in working memory were blocking the influence of that language and allowed the influence of the “unblocked” language. To summarize, online language context can affect the judged similarity of recently observed motion events (Athanasopoulos et al., 2015), but language context does not affect the judged similarity of simultaneously viewed color patches (Athanasopoulos et al., 2011). This set of studies is consistent with the hypothesis that language affects post-perceptual judgments, but that the ability of bilinguals to switch between perspectives may be restricted to more abstract representations.

5 Retrieving Memories in a Bilingual’s Two Languages An interesting question is how bilingualism affects memory of specific life events. Such autobiographical memories typically include the time, place, actors, actions, sensations, emotions, and other aspects of the event’s context. For bilinguals, the language in use is often an important aspect of the context. Given the general principle that reinstating contextual cues facilitates retrieval of autobiographical memories it is not surprising that autobiographical memories are retrieved faster, more accurately, with

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

460

K E N N ET H R . PA A P

more elaboration, and with greater emotional intensity when the language used during retrieval matches the one used during the retrieved event (Schroeder & Marian, 2014). Another interesting research question is whether there are differences between bilinguals and monolinguals in episodic memory that is not autobiographical. On the basis of their extensive literature review, Schroeder and Marian (2014) tentatively suggest that bilinguals may be advantaged in remembering nonverbal information but disadvantaged in remembering verbal information. As discussed in Section 4, language can influence the conceptualization of events. It is thus not surprising that language also affects the memory of events (see review by Schroeder & Marian, 2014). Even grammatical differences between languages may affect bilingual memory. For instance, in a well-controlled study, Boroditsky, Ham, and Ramscar (2002) compared bilinguals when they were using a language that marks tense (English) to when they were using a language that does not convey tense (Indonesian). Participants first viewed a series of pictures depicting actions taking place in the recent past, in the present, or about to occur. At test all three versions of each scenario were presented and participants simply had to indicate the picture they had seen earlier. Performance was poorer when the instructions were given in Indonesian than in English, and this is consistent with the view that the language active during encoding and retrieval affects the salience of different attributes of the event.

6 Moral Judgments and Emotions in Sequential Bilinguals The stakes and the magnitude of any perceptual or cognitive differences associated with using a native versus a foreign language appear to be much greater in the arena of moral reasoning. Whether this “flexibility” is good or bad is arguable: many bilinguals might prefer to choose the same path (the one based on their moral “core”) when arriving at the crossroads of an important decision, regardless of whether the interactional context led them to ponder the problem in their L1 or L2. Would you push a man off a footbridge in order to save five others from being hit by a runaway trolley? The decision itself involves a collision between one’s individual code of behavior and a utilitarian one (i.e., one that gives greatest weight to the alternative that serves the greatest good for the greatest number of people). Costa, Foucart, Hayakawa et al. (2014) reported that an average of 18% of bilinguals would sacrifice the one man in order to save five when using their native language, but up to an average of 44% would do so when they were using their foreign language. Participants were relatively late (with onset mostly around puberty) young adult learners of the foreign language in question who did not grow up speaking it at home and who self-rated as having low-to-medium proficiency in it.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

461

There are different interpretations of the shift in preferences toward the utilitarian choice when operating in a foreign language. One possibility is that native language processing is more automatic and relies on quick gut-level feelings whereas foreign language processing is more controlled and deliberate. Although this may fit the description of the trolley dilemma – contemplating physically pushing a man to his death gives a quick rush of aversion that must give way to deliberate consideration of the greater good of saving five – the deliberate thought hypothesis clashes with the results of Geipel, Hadjichristidis, and Surian (2016), who studied participants’ evaluations of scenarios in which good intentions led to bad outcomes or the reverse. Reading these in a foreign language (i.e., English, which these participants knew at relatively good proficiency levels) compared to a native language led to judgments that placed greater weight on outcomes and less weight on intentions. This is inconsistent with the hypothesis that foreign languages promote deeper thinking, given that other research has shown that asking people to be more reflective is an effective way to get them to take intentions into account. Furthermore, when participants are judging the deserved punishment for causing harm, the intention behind the action has a reduced effect on judgment when they are using a foreign language compared to a native one. The data presented by Geipel et al. (2016) appear to show that operating in a foreign language promotes more consideration of the outcome and less consideration of the means or intentions. A related conjecture is that it is easier to keep track of whether outcomes are good or bad than whether the means or intentions were good or bad. If so, sequential bilinguals whose native language proficiency is degrading from disuse might show “foreign language” psychosocial behavior in their native language (for linguistic effects, see Keijzer & de Bot, Chapter 14, and Ko¨pke, Chapter 18, this volume). One aspect of differences in moral reasoning in a native versus foreign language is that the use of native languages seems to elicit more emotional memories and thus increases the salience of the affective components of a moral dilemma. As summarized by Caldwell-Harris (2014), native and foreign languages tend to differ in their emotional impact as measured through a variety of experimental tasks (see Panicacci & Dewaele, 2018, for converging evidence obtained through nonexperimental means). For example, native languages enhance the emotional appeal of advertising slogans and increase the potency of emotional words to block taskrelevant processing and to trigger greater changes in skin conductance. Although there are discrepant findings, there appears to be a trend for these effects to be stronger with earlier ages of acquisition, higher levels of L2 proficiency, and when the L2 is learned through immersion rather than through explicit classroom learning. These findings cohere around a constructivist view that language and emotion are part and parcel of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

462

K E N N ET H R . PA A P

our episodic memories and the conceptual knowledge that is tapped during language comprehension. This emotional boost during native language relative to foreign language processing can be viewed as another form of linguistic relativity.

7 Bilingual Advantages in Other Cognitive Areas? If the notion that bilinguals have enhanced inhibitory control is correct (see discussion earlier), then bilinguals may be more creative in tasks that involve the suppression of the typical or conventional solution for a problem in order to discover a more creative one. Following earlier studies demonstrating that bilinguals have greater creative capacities than their monolingual counterparts, Kharkhurin and Wei (2015) set out to try to explore the extent to which this bilingual advantage is related to inhibitory control. They proposed a two-step account of these advantages: regular code-switching leads to an enhancement of EF (specifically the shifting and inhibition components), which, in turn, facilitates creativity. To focus on the code-switching part of their account a large group of bilinguals were partitioned into those who habitually code-switched and those who did not. To focus on the inhibitory control part all participants completed a nonverbal flanker task. The flanker interference effect (mean RT on incongruent trials minus mean RT on congruent trials) is usually assumed to reflect selective attention and inhibitory control. While, as expected, the bilinguals who frequently code-switched showed increased innovative capacity compared to those who did not, the surprising result was that the flanker effects were the same for both groups. If the groups did not differ in this direct measure of interference control, then the positive relation between code-switching and creativity cannot be mediated by the positive effects of code-switching on inhibitory control. Thus, creativity advantages favoring frequent code-switchers over nonhabitual switchers (or bilinguals over monolinguals) are likely to be unrelated to the inhibitory control component of EF. Another research strategy that has been underutilized is to test for creativity in both languages of a bilingual. Kharkhurin and Altarriba (2016) reported both a main effect of language (an English advantage) and a remarkable cross-over interaction between induced mood (positive or negative) and test language (a single one, English or Arabic) on a measure of creativity used with Arabic–English bilinguals. The interaction is difficult to interpret because the participants rated themselves as more fluent in their L1 Arabic while being better at picture naming in their L2 English. It would also be interesting to see if creativity is sparked by a language context that has both languages in an active and ready state compared to a single-language context. This would enable bilinguals to take advantage of linguistic relativity because verbal labels from different languages can

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

463

trigger different ways of seeing or thinking about the same object. Creativity may be facilitated by toggling between languages versus staying locked in one. Another topic that has received attention in cognitive approaches to bilingualism is Theory of Mind (ToM). ToM refers to the conceptual ability to understand that others have viewpoints and beliefs different from your own. While the studies discussed so far used mainly adult participants, the study of ToM in bilinguals has mainly focused on children. In one widely used test of ToM, children are asked to predict the behavior of a character holding a belief that they know to be false. For example, in the Sally-Anne task, Sally puts a marble in a box before going out to play. When Sally is outside, Anne moves the marble to a basket. In the standard version, children are prompted with the question: “When Sally comes back, where will she look for her marble?” At some point around the age of four, children transition from failing to predict the answer (instead, they incorrectly point to the basket, thereby expressing their true belief about where the marble actually is) to succeeding (thereby taking Sally’s false belief into account). Competence accounts attribute true belief responses to an inability to represent other people’s beliefs, that is, to an immature ToM. In contrast, performance accounts assume that ToM is in place but attribute true belief responses to an EF failure to inhibit them on the basis of the more salient knowledge that the marble is now in fact in the basket. The transition from true to false belief responses occurs consistently earlier for bilingual children in comparison to monolinguals, and some evidence has been found for bilingual adults’ better performance in false belief tasks compared to monolingual adults (see review in Rubio-Ferna´ndez, 2017). The question is why this should be the case. Bilingualism could confer an advantage in conceptual representation, EF, or a combination of both. One could rule out the contribution of EF to the bilingual advantage in ToM if there were no correlation between EF and ToM, but a meta-analysis by Devine and Hughes (2014) of 100 studies with monolingual participants does show a medium effect size. Furthermore, in longitudinal studies, individual differences in EF predict the later development of ToM, but not the reverse. However, finding a consistent relationship between EF and ToM is not sufficient toward establishing that EF differences are the cause of the early bilingual advantage in ToM. One also needs to show that young bilinguals have enhanced EF. More specifically, if the bilingual advantage in ToM is due to a bilingual advantage in inhibitory control, then young bilingual children (in the age range of three to six, at a time when ToM is supposed to mature) should show smaller interference scores in nonverbal interference tasks than monolingual peers. This, however, is not unequivocally the case (see, e.g., De Cat, Gusnanto, & Serratrice, 2018; Gathercole, Thomas, Kennedy et al., 2014; Tran, Arrendondo, & Yoshida,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

464

K E N N ET H R . PA A P

2015). Thus, performance explanations for ToM based on superior bilingual inhibitory control lack direct support. Although there are no bilingual advantages for nonverbal interference tasks, many of the false belief studies with preschoolers do show bilingual advantages on overall accuracy or latency. Consistent with this pattern of results, Rubio-Ferna´ndez (2017) proposes that bilingual children are not better at inhibiting the true-belief response, but rather that they are better at staying focused on the protagonist’s (i.e., Sally’s) perspective and resisting distraction. This suggests a bilingual advantage in attention management and monitoring. Although Rubio-Ferna´ndez (2017) favors this more specific performance explanation rather than a more global EF inhibition performance explanation, she concedes that ToM competence likely plays a supporting role and that, for example, as Tare and Gelman (2010) have suggested earlier, bilinguals’ sociolinguistic awareness of their interlocutor’s language abilities allows young children to appreciate that other people’s knowledge and perspectives can be different from their own at an earlier age than is the case for monolingual children.

8 Conclusion General inhibitory control has played a lead role in theories of bilingual language control such as the Bilingual Interactive Activation and Inhibitory Control Models (Section 2). However, challenges are mounting. Some of the effects taken as evidence of the continuous coactivation of two languages may be due to the rudimentary association of forms across languages during initial L2 learning (Costa et al., 2017; see Section 2.5). Second, inhibition often does not work, as shown by the frequent failures in recognition of L1 words with L2 homographs (see Section 2.3). Furthermore, in contrast to what the Inhibitory Control Model predicts, language switching and the reactive inhibition it presumably triggers do not diminish the magnitude of the CSI effect whereby instances of previously shown semantic categories exert a greater effect on naming than the language in which they are to be named (Section 2.6). In addition, n-repetition costs in nonverbal tasks (as a measure of general inhibition) do not correlate with n-2 repetition costs in language switching, suggesting that the inhibitory mechanism in language control is a horse of a different color (Branzi et al., 2016). There are also challenges to the assumption that language switching recruits general task switching, given that the two types of switching dissociate with respect to switching costs, mixing costs, and the effects of aging. If this trend continues, then models like the BIA+ will gain favor over models assuming strong topdown cognitive control. Dong and Li’s (2015) review offers a somewhat different perspective on this continuing debate.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

Bilingualism in Cognitive Science

465

If bilingual language control does not consistently rely on general cognitive control, then there is no reason for bilingual advantages in executive functioning. This may be why the seminal studies showing advantages on nonverbal interference tasks and nonverbal switching tasks do not replicate. The steady drip of new studies reporting bilingual advantages, against the stream of null results, may be due either to small effect sizes that occur only under specific circumstances or to methodological problems and questionable research practices that can sustain belief in a nonexistent phenomenon. Even if bilingualism does not enhance executive functioning, there are many other benefits to being bilingual, such as communicating directly with more people, expanding educational and career opportunities, enhancing understanding and appreciation of other cultures, perceiving and understanding the world from additional perspectives, gains in creativity, and a head start on acquiring a theory of mind.

Notes 1. Very broadly, a flanker task in cognitive psychology generally aims to test respondents’ ability to control interference from nontarget stimuli when target stimuli are “flanked,” that is, simultaneously presented with nontarget stimuli that are either similar to the target stimuli (congruent trials) or not (incongruent trials) (see more via, e.g., www.sciencedirect .com/topics/neuroscience/eriksen-flanker-task). In the classic color-word Stroop task, respondents have to name or otherwise select the color in which visually presented words (often names of colors) are presented to them. In congruent trials, the printed color and the semantics of the words match. In incongruent trials, there is a mismatch. Numerical Stroop tasks are variations on the color-word Stroop task that use digits instead of colors as stimuli. The numerical Stroop task in the BCBL studies referenced in this chapter involved competition between numerical magnitude as indicated by number digits and the font size in which these digits were presented. 2. The Simon effect refers to the fact that in an experiment, participants find it easier to respond to a stimulus that matches the response/action asked of them (congruent condition, e.g., pressing a key on the right when a stimulus appears on the right of a screen) than to a stimulus that is a positional mismatch with the required action (incongruent condition, e.g., pressing a key on the right when a stimulus appears on the left of a screen) (see, e.g., www.psytoolkit.org/experiment-library/simon.html). Usually, color and position are the two dimensions manipulated in Simon tasks.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:25:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.023

23 Bilingualism in Neurolinguistics From Static to Dynamic Approaches Arturo E. Hernandez

1 Introduction The bilingual brain has been a source of interest in the published literature for close to 150 years. Before anyone could capture the brain in action in living people, neuropsychologists were making inferences about the neural representation of two languages by looking at patients who had suffered some kind of brain injury. Here is just one example from the late 19th century (Bianchi, 1886): Arturo A. (A.A.) was a 24-year-old man in good health who was short and muscular but well proportioned. He was a successful international businessman who traveled abroad a lot to make sales. He was well suited for this type of work thanks to his ability to speak several languages. He had first learned Italian, his L1, and also spoke French and English, which he had learned a bit later, but early in life. One day after having severe convulsions he was promptly taken to the doctor for treatment. A.A. suffered from paralysis to the right side of his body and had difficulty with spoken language. Further tests revealed that he could say some words in Italian but could not understand the connection between what he said and the corresponding things in the real world. As the time post-insult increased, the link between spoken words and their real-world referents became stronger. A.A. was never able to fully recover all the languages he had previously spoken. Whereas he showed a considerable amount of recovery in Italian, he was never able to say a single word in French or English again. In the case of A.A., the first language was the last one left standing. Based on such cases of the selective recovery of languages that patients had used prior to brain insult, neuropsychologists began to develop hypotheses about bilingual neural representation that would help to account for the greater loss in production and comprehension in one language relative to the other after brain damage. As elucidated in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

467

Section 2, early hypotheses mainly focused on how factors relating to language learning and use history such as familiarity with a language and assumed concomitant increases in language proficiency and age of acquisition, as well as cognitive control, led to differences in the loss of one or the other language. A.A.’s case presents evidence for all three factors. This patient was able to recover the language he spoke the best, which just happened to be the first language he had learned. It is also possible that his losing languages was due to difficulties in language control, which may play a greater role for chronologically second and third (etc.) languages than for the first. The ensuing sections explore how the pioneering ideas about bilingual neural representation have been approached in more recent research (I focus mainly on adults who engage with oral language). However, as explained in Sections 6 and 7, both computational simulations of language learning and novel approaches to neural functioning have opened the door for a more dynamic view of the neuropsychological basis of language in bilinguals.

2 Classic Hypotheses Regarding the Neuropsychological Basis of Language in Bilinguals 2.1 Familiarity and Its Neural Instantiation As explained by Lebrun (1995), French medical doctor and researcher JeanAlbert Pitres was the first to suggest that familiarity, which may result in greater knowledge of a particular language, might render a language more resistant to damage from brain insult. Already in 1895 Pitres noted that the impairment of one specific language in bilinguals or multilinguals was quite common. In reviewing a number of cases, Pitres (1895) described the loss of language in multilingual aphasics. Some patients retained access to only a single language. For example, there is the case of a 56-year-old Frenchspeaking male who lived in France but learned German as a child and Basque, English, Spanish, and Italian as an adult. After losing consciousness due to a stroke, the right side of his body was paralyzed. Over time he recovered his ability to use French but not his other five languages. Pitres reviewed many such cases, which always revealed a differential amount of recovery in some languages relative to others. Based on these cases, Pitres formulated the generalization that the most familiar language would remain resistant to damage. This has come to be termed “Pitres’s rule” (e.g., Lebrun, 1995). In addition, Pitres noted that many cases showed what he called temporary “inertia” and that once it was overcome there was no need to think that aphasics were growing new language centers for each language. The fact that familiarity with a language is important for the recovery of language in aphasics has played an important role both in the child and in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

468

ARTURO E. HERNANDEZ

the adult bilingual literature. Although Pitres was referring to intensity and frequency of use, researchers have more recently considered the product of these processes through which a language becomes familiar by taking up the notion of language proficiency, a likely result of familiarity (see Treffers-Daller, Chapter 15, this volume, for discussion of language proficiency in bilinguals). Language proficiency has been found to play a central role in the neural representation of bilingualism (Abutalebi, Cappa, & Perani, 2001; Mechelli, Crinion, Noppeney et al., 2004) and in the errors or speed of behavioral responses in the cognitive literature (Birdsong, 2006; Costa, Santesteban, & Ivanova, 2006). So far, the research literature accepts the idea that greater proficiency in a language leads to different patterns of brain activity.

2.2 Age of Acquisition In addition to language familiarity and concomitant proficiency, the age at which a language was learned, that is, age of acquisition (AoA), can also be seen as playing a role in the recovery of language observed with patient A.A. On the surface, A.A.’s case supports Pitres’s rule since Italian, his most familiar language, was preserved relative to the other two. However, like for many bilinguals (henceforth this term also relates to multilinguals), familiarity is confounded with AoA since Italian was the language that A.A. had learned first (L1) (see Armon-Lotem & Meir, Chapter 10, this volume, for discussion of similar difficulties in distinguishing between AoA and other exposure effects in child bilinguals; see also Serratrice, Chapter 1, this volume). French and English, which A.A. learned in early childhood as second and third languages, were also spoken well but did not recover as well after brain damage. One could interpret this as showing that the L1 takes precedence. The first to explicitly propose that the age at which a language was learned mattered was The´odule-Armand Ribot (1881, English version of 1887). Many consider Ribot the father of French psychology. In his view, the importance of age of acquisition was a natural extension of the general laws that rule the organization of memory. For people suffering from what he called “l’amne´sie progressive des signes” (Ribot, 1881, p. 136), that is, progressive memory loss, he observed the following pattern of loss in the ability to produce words (see Goral, Chapter 5, this volume, for discussion of age-related language changes in nonimpaired populations). The first types of words lost are proper names, followed by common nouns, then adjectives, verbs, interjections, and, finally, gestures. The loss of words within each category starts with the ones most recently learned and progresses to those learned earliest, that is, it proceeds from items that were new, complex, voluntary, and least organized to items that were old, simple, automatic, and best organized.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

469

In short, Ribot proposed that “age of acquisition” was a central part of memory formation (and retention). At the time that Ribot was developing his theories, there was little in the way of theoretical or empirical psychological work with children. However, there was a literature on second language acquisition, which built on existing ideas of the time that distinguished between speech as a motoric memory and comprehension, which was considered a component of the mind. In fact, Ribot was hinting at the distinction between speech and other forms of memory (as conceptualized at the time). Ribot also considered the case of the recovery of lost languages in bilinguals. Of particular interest to him were patients who under anesthesia or on their deathbeds showed some recovery of a language even after protracted periods of little use. Ribot interprets such findings as indicating that a second language is built using the conscious mind and the first language is built using organic and motor memories. The distinction between conscious memory and motor memory is the basis for Ribot’s claim that AoA plays a crucial role in the neural representation of both languages in a bilingual (see also Keijzer & de Bot, Chapter 14, this volume, on representation and memory).

2.3 Beyond Language: The Role of Language Control Ribot’s view that first and second languages could be distinguished based on the use of different memory systems provides a good segue to this section. Pitres and Ribot proposed the two mechanisms of familiarity (assumed to index proficiency) and AoA, respectively, as playing a crucial role in the neural instantiation of two languages. However, language loss could be due to the general loss of capacity. Specifically, neural damage could affect brain areas involved in helping to select or switch to the correct language, that is, in language control. Rather than showing language loss per se, not being able to speak a particular language may in fact be due to the loss of language control and the ability to switch between languages. The first to propose the idea of a switching mechanism was Otto Po¨tzl, an Austrian neurologist born in 1877 in Vienna, where he completed his medical training. In 1922, after he took a position as head of psychiatry in Prague, Czech Republic, he encountered a bilingual patient who had a strange syndrome (Herschmann & Po¨tzl, 1983 [translated from the German 1920 original]). The patient was a 60-year-old Austrian businessman, a speaker of Czech as his L1, who had learned his L2, German, at age 14. As an adult he had moved to Vienna and used German exclusively in his family and daily life. The patient suffered two strokes that had caused him to lose the ability to see the space directly to his right. Furthermore, he suffered from a peculiar language impairment: he could understand both Czech and German but could speak only Czech. A subsequent patient with

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

470

ARTURO E. HERNANDEZ

a very similar problem led Po¨tzl to conclude that some bilinguals suffered from a problem with language fixation. Although these patients understood both languages perfectly, they remained fixated on the language they had used during their first few days after a stroke. More recently, Franco Fabbro encountered a similarly odd case of a 56-year-old Friulian–Italian bilingual (Fabbro, Skrap, & Aglioti, 2000). Patient S.I. was a land surveyor who learned Friulian, a dialect spoken in northern Italy, as his L1 but was educated in his L2, Italian, for 13 years. The patient noticed that he began to intersperse Friulian sentences with his Italian friend who was monolingual. He kept apologizing and translating for his friend. At the same time, he began to notice problems moving the right side of his body, which made walking and driving difficult. After the patient met with Fabbro, it was quickly determined that the patient had a tumor, which was then resected. After it was removed, the patient would alternate between his two languages across sentences. He would consistently say one sentence in Friulian and one in Italian even when he knew the person listening only understood Italian (see De Houwer, Chapter 17, this volume, on factors normally guiding language choice in bilinguals). When tested with a set of standardized neuropsychological instruments, S.I. showed no real language deficits. He understood each language very well. He could repeat words in each language almost perfectly. He could even translate between his languages. The problem seemed to be restricted to spontaneous speech. The cases presented by Po¨tzl and Fabbro provide us with an intriguing deficit. Patients understood both languages very well. Hence, according to both these researchers, patients such as these are not exhibiting language loss. Rather, they have a problem with accessing each language. Researchers have come to think of this as a problem with language switching. The metaphor of a language switch, originally proposed by Wilder Penfield in the 1950s (Penfield & Roberts, 1959), has gained popularity in the literature in recent years. The idea is that bilinguals have to use general cognitive control mechanisms in order to access each of their two languages at the appropriate time. The importance of cognitive control has also been explored in bilinguals relative to monolinguals. Relevant studies have suggested certain advantages for bilinguals on tasks that, at least on the surface, have nothing to do with language (see Paap, Chapter 22, this volume, for critical review).

3 Views of Bilingualism in Processing Approaches As discussed in the previous section, Ribot’s approach to language in bilinguals suggested that memories can be split into conscious and unconscious memories. Pitres suggested that less familiar languages involve different systems than more familiar languages. Finally, Po¨tzl suggested

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

471

the presence of a language switch. Recent theories of bilingual language representation have built on the foundation established by these classic authors. In particular, recent approaches to the bilingual brain have sought to characterize whether the processing of each language might rely on similar or different brain mechanisms. In general, current theories of bilingual language processing have built on a dual-systems view of language processing. This dual-systems view seeks to distinguish languages according to whether they involve different memory systems. This view is best exemplified in the declarative/ procedural (DP) model (Paradis, 2009; Ullman, 2005). In some ways, this model is a direct descendant of Ribot’s (1881) view that there are conscious vs. unconscious memories. The DP model proposes that first and second languages involve different brain systems. This is particularly true with respect to grammatical aspects of each language. The DP model proposes that the L1 grammar is learned in an implicit or procedural manner and associated with left frontal (i.e., anterior) neural regions and the basal ganglia, which is highly interconnected with sensory and motor systems. Second and subsequent language grammars are assumed to be learned in a more explicit way. In the DP model, this explicit learning places greater emphasis on declarative memory and will involve temporo-parietal (i.e., posterior) regions (Ullman, 2005). The DP model proposes that in processing spoken sentences bilingual speakers show different brain activity patterns depending on the language presented (Dehaene, Dupoux, Mehler et al., 1997). The model is also supported by neuropsychological case reports on patients suffering brain damage. Cases reporting selectively impaired recovery in the native language have seen damage to basal ganglia structures (Fabbro & Paradis, 1995; Garcı´a-Caballero, Garcı´a-Lado, Gonza´lez-Hermida et al., 2007). Theories have also begun to consider whether bilingualism and monolingualism might be a product of similar or different neural substrates. Whereas the DP model proposes that the L2 uses different circuits from those used by the L1, other accounts do not posit such a difference. One example of this approach is the neural convergence theory, which argues that identical neural regions mediate language in bilinguals and monolinguals (Green, 2008) during both lexical (Parker Jones, Green, Grogan et al., 2012) and grammatical processing (Abutalebi, 2008). This does not imply that brain activity is identical between languages, or that bilinguals are identical to monolinguals. In fact, a main difference between bilinguals and monolinguals can be seen in the functional demand on language-mediating regions, which is higher for bilingual speakers (Abutalebi & Green, 2007; Parker Jones et al., 2012). One source of bilingual–monolingual differences is the presumed reduced language-processing skill in speakers of two languages (Bialystok, Craik, Green, & Gollan, 2009; Portocarrero, Burright, & Donovick, 2007) and the simultaneous activation of both languages even under single language-

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

472

ARTURO E. HERNANDEZ

processing conditions (Hoshino & Thierry, 2011; Kroll, Bobb, & Wodniecka, 2006; see Paap, Chapter 22, this volume, for discussion). In order to overcome interference from the other language as well as the effect of assumed lower proficiency in each language compared to monolinguals, the convergence hypothesis proposes that bilinguals must use control mechanisms. This view follows in the theoretical footsteps of Po¨tzl by proposing that selective recovery of language post-stroke might be due to a loss of language control (Abutalebi & Green, 2007). Beyond interest from scholars working on language processing, the question of bilingual–monolingual differences has also been of interest in functional and structural neuroimaging (e.g., Grant, Dennis, & Li, 2014).

4 Back to Bilingual Aphasics Also in the aphasia literature the question of bilingual–monolingual differences has been addressed. For instance, a study comparing 33 bilingual and 174 monolingual patients with aphasia found intriguing group differences in the link between the locus of brain damage and language impairment (Hope, Parker Jones, Grogan et al., 2015). In addition to noncognitive tasks, groups of bilingual and monolingual patients were given a battery of tests measuring a variety of language skills including comprehension, repetition, reading, confrontation, naming, and writing. The scores on each of these tasks were then mapped onto brain areas to see how specific symptoms might be associated with brain damage in particular areas. The findings revealed an interesting group difference between monolinguals and bilinguals in their L2, which was the same as the single language monolinguals knew, viz., English: the breakdown bilinguals experienced in their L2 was much larger than that observed in monolinguals for their only language. The same size and location of a lesion was associated with a larger language impairment in the language tested (English) in bilinguals relative to monolinguals. Hope et al. (2015) refer to the quantitative sample size differences between bilinguals and monolinguals, the relatively small bilingual sample size, and the diversity of the bilinguals’ L1s as some limitations of their study. In spite of these, the main possible reason that they give for their finding of larger breakdown in English by bilinguals than monolinguals lies in a combination of lower pre-insult proficiency differences in the bilinguals (which, however, they did not measure) and an “enhanced loading (and thus enhanced sensitivity to damage) in a common network of regions recruited to implement language in both patient groups” (pp. 1079–1080). Thus, bilingual and monolingual groups may have already been different before the brain damage occurred, and the damage may have only served to exacerbate these differences. Where possible, Hope et al. (2015) also partially tested bilinguals in their L1; interestingly, patients performed better in their L2 than in their L1. The

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

473

authors speculate that this is a result of patients using their L2 more than their L1. This explanation is reminiscent of Pitres’s view that familiarity plays a role in the neural representation of language. Hope et al. (2015) emphasize that having much more information on bilingual patients’ language use history, both pre- and post-insult, is essential to help explain bilingual–monolingual differences.

5 Language Familiarity and Associated Proficiency, AoA, and Control Interact As mentioned earlier, Po¨tzl already noted the importance of cognitive control in language processing (although he used different terms). Along with cognitive control, the question of how factors identified by other pioneers working on bilingual aphasia, viz., language familiarity (and resulting proficiency) (Pitres) and age of acquisition (Ribot), influence language processing was taken up in a series of studies by Archila-Suerte and colleagues. In a behavioral study, Archila-Suerte, Zevin, Bunta, and Hernandez (2012) investigated phonological processing. Adult Spanish L1 speakers who had started learning their L2 English at different ages were presented with pairs of English-like nonsense words and asked to rate whether they were similar or different (there also was a monolingual English-speaking control group). Vowels in the stimulus words were either similar to Spanish vowel sounds (saf and sef), while others were much less so (sof and suf). Each of the 160 stimulus tokens was unique and differed from any other token in intonation, timbre, and duration. Participants were presented with 780 pairs of words and asked to decide how similar they were on a four-point Likert scale going from 1, very similar, to 4, very different, that was presented while participants heard a pair. The data from each participant can be visualized as having four different clusters, one for each vowel sound in saf, sof, suf, and sef. The blurriness of each cluster indicates how well a person heard the similarities across the sounds. On one extreme is the notion that each cluster is a single point, in which case a person recognizes all variations of suf as a single item. Reality, of course, is not nearly so clean. Natural speech varies across speakers and instances of speech. In the Archila-Suerte et al. (2012) study, the size of the clusters was correlated with AoA of the L2. Age of acquisition played a role when participants had to decide that two items from the same category (saf–saf) were the same. Despite the small variations in the pronunciation of the four vowel sounds, native English speakers recognized each instance as coming from a single speech category. Like these speakers, early bilinguals who had learned English before age five revealed four tight clusters for each sound, showing that they were able to ignore slight variations in a sound in order to recognize them as a single phoneme. On the other hand,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

474

ARTURO E. HERNANDEZ

intermediate learners, who had started to acquire English between ages six and nine, showed clusters that were distinct, but a bit more blurry. Late learners, who had acquired English after the age of 10, showed even more blurry clusters. Late L2 learners used a very different strategy from the early bilinguals, viz., they utilized a relative comparison strategy to compare different sounds. These results show that the age at which a person learns English plays a role in how well they can categorize two English sounds as being from the same speech sound category. The question is how these data fit in with existing theories of bilingual language processing. In some ways, they are resonant with the DP model described earlier: specifically, early learners may be depending on more procedural memory systems whereas late learners are relying on declarative memory systems. They are also resonant with views first proposed by Ribot, viz., that memories formed in early life are more automatic, whereas those formed later in life are not. In another experiment, Archila-Suerte, Zevin, and Hernandez (2015) set out to see how the earlier behavioral findings translated into brain activity. To do this the behavioral task used in Archila-Suerte et al. (2012) was adapted for a functional magnetic resonance imaging scanning experiment measuring brain activity. Adult participants were asked to focus on a movie while being scanned. During scanning, participants heard the English-like syllables saf, sof, and suf repeatedly presented in pairs in which either both members contained the same vowel phoneme in English (e.g., saf–saf), or in which they did not (e.g., saf–suf). Participants listened to these sounds without being asked for any response (a research design known as pre-attentive listening) while viewing a silent film with nature scenery in order to make the time pass more easily and, perhaps more importantly, to keep their attention fresh. A total of 82 adult participants were tested (66 Spanish–English sequential bilinguals and 16 English-speaking monolinguals). Like in Archila-Suerte et al. (2012), the bilinguals were divided into two groups. One had first started to learn English prior to age nine (n = 34), and one had done so after age nine. The monolinguals had all been speaking English since birth and spoke no other language fluently. The results were consistent with the dual-systems view of language processing. Specifically, both late bilinguals and monolinguals showed more intense bilateral activity in the superior temporal gyrus (STG) relative to early bilinguals. This suggests that AoA has an effect on the neural activity observed in the STG, an area of the brain that has been linked to the processing of speech sounds. Studies like the ones by Archila-Suerte et al. (2012, 2015) that use a retrospective approach in which developmental factors are explored by looking at adult populations with different language learning histories are informative but leave out valuable information. Specifically, studies with adults tend to obscure the type of change that occurs across childhood, through adolescence, and into adulthood. In order to better

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

475

understand the role of developmental factors, Archila-Suerte, Zevin, Ramos, and Hernandez (2013) used the same research design as in Archila-Suerte et al. (2015), but now with 38 bilingual and 13 monolingual children between the ages of 6 and 10. In addition, this study compared child data to similarly obtained data from adults (19 bilinguals and 17 monolinguals). The adult age range was between 24 and 35. All bilinguals had learned Spanish as an L1 at home and had begun to learn English as an L2 at school; most of them were around five years of age when they first started to acquire English. All monolinguals had only learned English. As Archila-Suerte et al. (2013) found, at the ages of six to eight bilingual children showed similar but weaker neural activity in the bilateral STG compared to monolingual peers. The slightly older bilingual children aged 9 and 10, however, showed a pattern of activity that was quite different from that of the younger bilinguals. Many regions were active while perceiving L2 sounds. These included the bilateral STG (that was the same as for the younger bilinguals) but also the superior and inferior parietal lobules, the inferior frontal gyrus, the parahippocampal gyrus, and the right superior and middle frontal gyrus. The STG has been found to be involved in processing speech in monolinguals. However, the additional areas of brain activity seen in the 9–10-year-old bilinguals are typically involved in motor planning (inferior frontal gyrus) and executive function (middle frontal gyrus), as well as attention (superior and inferior parietal cortices). These results show that in older child second language learners, who have more experience in a language than younger ones, a larger set of brain areas is engaged for the processing of second language speech sounds. An interesting aspect of Archila-Suerte et al.’s (2013) results was that bilingual children were showing greater neural activity in areas involved in cognitive control as a function of increased experience with the L2. Compared to the younger bilingual children, the older ones revealed more activity in the inferior parietal lobule and the middle frontal gyrus. The inferior parietal lobule and the middle frontal gyrus are known to be involved in cognitive control. Another interesting aspect of these data was the need for cognitive control during the listening to speech sounds, even in the absence of specific instructions to do so. Brain areas that are known to be involved in memory and speech processing were also active. This pattern of speech processing in bilingual children might be due to the formation of new speech categories and placing them into memory. Thus the formation of new speech categories during child second language learning may lead to the active use of cognitive control. Finally, the group of early adult bilinguals no longer showed the extensive neural activity seen in the 9–10-year-old bilinguals. (Note that AoA was held constant for all bilinguals.) This finding fits in with the convergence hypothesis by showing that longer experience with the L2, which

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

476

ARTURO E. HERNANDEZ

assumedly goes together with greater proficiency, involves a reduction in the use of cognitive control and an increase in automaticity. The earlier Archila-Suerte et al. (2012) study with adults found that AoA and concomitant language experience differences were reflected in different behavioral indices such that early learners had more compact clusters for different phonemes. Archila-Suerte et al. (2013) interpreted language experience differences among bilingual children and adults as yielding more of a cognitive effect whereas early AoA involved only processing sounds. Results from the neural activity are consistent with this interpretation. These types of studies show how the changes in language proficiency for a second language acquired in childhood reveal activity in brain areas involved in the processing of speech sounds as well as in areas involved in cognitive control. As such they show that age of acquisition and language proficiency interact and thus lead to both greater recruitment of areas involved in the processing of speech sounds as well as those involved in cognitive control. Recent work in the bilingual literature suggests that young bilinguals may have cognitive advantages over monolingual peers (Bialystok, 2017). The advantage may be related to areas of the brain that develop earlier in life and to the need to attend to two different speech sound systems and manage two labels for the same thing. Although compared to young children older children and adults naturally differ in overall brain development, Archila-Suerte et al.’s (2013) results suggest that bilinguals show neural differences in cognitive control in function of the length of bilingual experience. The need for control at so many different levels of language may be the reason that such a pervasive use of executive systems in the brain is seen relative to monolinguals, especially in childhood (Krizman & Marian, 2015).

6 Rethinking Current Models of Bilingual Brain Representation Theoretical models of neural representation such as the DP model and the convergence hypothesis have taken a rather static view of the bilingual brain. These models consider different factors separately. As shown earlier, language proficiency, age of acquisition, and cognitive control are thought to exert different types of effects on the bilingual brain. In contrast, connectionist models and their ability to capture the nuances of change present an alternative way to think about bilingualism on the neuropsychological level. Their flexibility can best be seen in the multiple ways in which a system adapts to learning two languages. As noted in the following paragraphs, work by Ping Li, Brian MacWhinney, and colleagues reveals the power of connectionist networks in bilinguals through the use of language learning simulations based on artificial neural networks.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

477

In a first simulation, Li, Farkas, and MacWhinney (2004) designed a model that was charged with learning vocabulary items in a single language. This neural network model had no structure built into it. The network was presented with 500 early acquired words that child language researchers know are learned by certain ages (Dale & Fenson, 1996). There were words from different word classes, viz., nouns, verbs, adjectives, and closed-class items such as prepositions. Aside from some basic semantic feature information, there was no explicit information about what kinds of words they were (just as in real life children are not explicitly told, either). At first the network was shown 50 words that are typically learned first. Then, along with the original set of words, the network was shown 50 new words that are typically acquired in the next growth stage, and so on. The network showed effects of time of entry. The first words presented showed a strong influence on the network, leading to very sudden shifts in organization. Over time, the network began to self-organize, forming different types of word categories. Nouns were closer to each other than verbs and vice versa. The network also showed AoA effects in that early learned words were much more influential than late learned words. The important thing to note is that the network organized on its own, simply as a result of the competitive process of encountering more and more lexical items. The fact that this self-organizing network could learn a single language led Zhao and Li (2010) to consider the nature of bilingual language acquisition. To do this they designed a network that mapped word sounds to word meanings. They then introduced two languages to the network either simultaneously or in succession. It became clear that learning each language at different times had a very different effect on the ultimate structure of the network. Simultaneous presentation of both languages led to relatively independent maps. Sequential presentation of languages led to the second language being parasitic on the first. These networks give us a way of visualizing many of the principles that were first proposed by Ribot and Pitres. Structure is built up across time and space. As structure emerges, each language starts to occupy different parts of the network. When learning occurs early, the structure of each set of words can exist relatively independently. When learning occurs later in life, a second language becomes attached to the first. The neural network approach is part of a larger movement that has taken a dynamic systems approach to language and cognition. The dynamic systems approach is characterized by the view that different variables are interconnected. Rather than thinking of language as a single thing, it becomes a combination of things. In this view, there is great interconnectedness between different variables (Smith & Thelen, 2003). Unlike very discrete systems in which there is a clear distinction between two different levels of representation, dynamic systems are difficult to predict because the outcome of one variable can affect the outcome of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

478

ARTURO E. HERNANDEZ

other variables: dynamic systems constantly change through interactions with their environments and go through constant self-reorganization.

7 Conclusion Transitioning from neural studies to those considering cognitive processes can be fraught with difficulty. However, brain science has also begun to consider the dynamic nature of cognition. This can be seen in research investigating individual differences in verbal working memory, that is, in the ability to keep language information in mind temporarily. As reviewed by Prat (2011), individual differences in language comprehension abilities show three different neural signatures when one compares individuals with greater vs. lower working memory and/or vocabulary capacity. Specifically, relative to those who do not perform as well, individuals who perform a task better differ in efficiency, neural adaptability, and functional synchronization among activated brain regions. As far as efficiency is concerned, Prat (2011) reports that during language comprehension tasks through reading, participants with larger lexicons than others and thus greater language expertise show less neural activity in brain areas in the left frontal and parietal lobes. This relatively lower activity has been taken to show that “expert brains” are more efficient, that is, they use fewer mental resources. Comparisons of neural efficiency can only be conducted when task performance across groups is similar. When tasks are complex, participants’ neural activity may in addition involve the recruitment of a particular network of brain areas in order to accomplish the task successfully. This adaptability, or “dynamic configuration of neural networks on an ‘as needed’ basis with changing task demands” (Prat, 2011, p. 638), appears to be related to working memory capacity. As Pratt (2011) reports, compared to study participants with lower working memory capacity, participants with higher working memory capacity performed better on a behavioral word recognition task after reading sentences with either simple or complex grammar. The latter also showed increased activity in control areas in the frontal cortex. This increased activity was correlated with increased syntactic complexity, thus showing higher levels of adaptability. Those with lower working memory capacity showed less activity in the frontal cortex and other control related areas. In short, working memory capacity and neural adaptability appear to go hand in hand. Finally, Prat (2011) explains that individual differences in processing also have to do with the synchronization of brain areas. In the same simple-complex sentence comprehension task described earlier, study participants with higher memory capacity showed a stronger coupling of activity in left hemisphere Broca’s and Wernicke’s areas compared to those with lower memory capacity. This tight coupling persisted even

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Bilingualism in Neurolinguistics

479

when an additional distracting task was introduced or sentences were more difficult to process. This suggests that individuals with higher memory capacity have a tighter coupling between brain areas. The interesting part of this framework is that it emphasizes neural connectivity across many loci to explain language processing rather than focusing on particular brain areas, as has been the approach historically. The notions of efficiency, adaptability, and synchronization can be applied to a number of different types of skills. Each of these skills could involve a number of different brain networks. Hence, rather than thinking about the brain as involving mental storage, neuroscientists are beginning to think about it as a set of networks that can be applied to different skills (see also Keijzer & de Bot, Chapter 14, this volume). The notions of efficiency, adaptability, and synchronization offer us a more general framework to think about how two languages are coded and processed in one brain. The interesting aspect with bilingual language development is that there is considerable change happening over time. This is where we can consider the notion of adaptability within a broader context. Finally, Prat’s work and that of others also reveals an interesting variation in the instantiation of skills in the brain by emphasizing the dynamic aspects of neural processing. Rather than skilled and unskilled individuals only showing simple differences in efficiency, there appear to be clear differences in the way that this efficiency plays out as tasks become more complex. Those with better skills are able to adjust or adapt to circumstances better. This dynamic interplay between an individual’s brain and the challenges of the environment provides an interesting foundation for thinking about bilingualism, the brain, and the nature of skilled performance. The current chapter considered the nature of bilingual language processing from the perspective of different theoretical frameworks. While frameworks adopted by researchers in recent years include the view that the bilingual brain involves the use of different memory systems or that bilinguals use more cognitive control, other models have begun to consider the adaptability of the bilingual brain. A neurocomputational approach that involves dynamic and emergentist aspects in which various factors combine in complex ways across time is one that is likely to bear fruit in the future.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:55:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.024

24 Bilingualism and Sign Language Research Gladys Tang and Felix Sze

1 Introduction Compared to the linguistic study of spoken languages, sign language research is a young discipline that emerged only during the 1960s. Despite sign language’s recent history, many researchers are intrigued by what constitutes the crucial properties of sign language, whether theories of linguistics and linguistic universals generated based on spoken languages are attested in sign languages, and how children and adults acquire and use sign language. In this chapter, we focus on the last area: sign language acquisition and use. These have been subsumed under the general framework of bimodal bilingualism, a term adopted recently in the sign linguistics literature that refers to the acquisition and use of a sign language and a spoken language by Deaf or hearing individuals. In keeping with disciplinary conventions, throughout the chapter we use lowercase deaf to refer to the audiological state of not hearing or being hard-ofhearing (henceforth deaf, for simplicity), and capital-case Deaf to refer to individuals who self-identify (or in the case of young children are identified by others) as signing members of the Deaf community. Focusing on how Deaf children acquire sign language, initial studies adopted a first language acquisition methodology and recruited as participants primarily Deaf children. There has since been a shift from a monolingual to a bimodal bilingual approach, with data drawn more often from children of Deaf adults who are hearing and born to Deaf parents (CODAs) and sometimes implanted Deaf children who can produce intelligible speech. One reason for such a shift is that CODAs are likely to acquire both a sign language and a spoken language bilingually at an early age, if given the necessary inputs (Petitto, Katerelos, Levy et al., 2001). Another is that universal newborn hearing screening has become quite common and advancement in assistive hearing technology (i.e., digital hearing aids or cochlear implantation) is in place to support speech and language rehabilitation. This means that even with

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

484

G L A D Y S TA N G A N D F E L I X S Z E

Deaf children born to Deaf parents who acquire a sign language as a first language (L1), there is currently a constant push for their early second language (L2) acquisition of a spoken language. Especially after two to three years of age, these children are generally brought up within a larger hearing community beyond the family and are bound to be exposed to spoken language input at some point during childhood, albeit with varying degrees of speech perception abilities and ultimately reaching different degrees of oral language abilities. Therefore, the issue of quantity and quality of bilingual input to both Deaf individuals and CODAs, and of both signed and spoken input, is rather complex (see Armon-Lotem & Meir, Chapter 10, this volume, for a review of exposure and input to young bilinguals). The nature of input rests upon the hearing status and language use of caregivers, siblings, and peers and the age and effectiveness of clinical intervention, as well as the education setting and educational philosophy experienced by each bimodal bilingual. All these considerations can be seen in recent debates on cochlear implantation and the push for oral monolingualism as threats against sign language and early bimodal bilingualism, with the latter being seen as supporting child development (Humphries, Kushalnagar, Mathur et al., 2014). At any rate, language acquisition research involving sign language nowadays cannot ignore the fact that within individual children, Deaf and hearing, there may exist knowledge of two developing linguistic systems, each of which stems from a different modality: auditory-oral versus visual-spatial. In this chapter, we first discuss two issues behind research on bimodal bilingualism. They are important to understand in order for the research findings to be appropriately interpreted in the context of the different modes of language acquisition. Next, we survey some crucial findings from three major strands of research on bimodal bilingualism: code-blending, bilingual acquisition, and metalinguistic awareness. Our selection of these three areas helps us address two fundamental questions: (1) What evidence is there to suggest language separation in bimodal bilingualism so far, in which two linguistic systems are said to develop autonomously and are differentiated in both development and use? (2) How can evidence of crosslinguistic influence, or the constant interaction between the two systems, be accounted for within a theory of bimodal bilingualism? We conclude by reflecting on the current state of research developments, pointing out the gaps that need to be filled in order to arrive at a better characterization of bimodal bilingualism and to show how it bears on some educational implications.

2 Bimodal Bilingualism: Two Background Issues Recent years have seen attempts by researchers to anchor bimodal bilingualism to bilingual first language acquisition, in particular, to formal

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

485

linguistic approaches in the study of simultaneous bilingual acquisition (Meisel, 2004). In this section, we discuss two important issues behind contemporary research on bimodal bilingualism that provide necessary context for the rest of the chapter. The first issue deals with the context of language acquisition. The complexities of the contexts in which CODAs and Deaf children undergo bimodal bilingual acquisition cannot be sufficiently underscored. We highlight here complexities of a linguistic nature, but attitudes toward communication in an oral language or a sign language are no less complex, for example, among CODAs and their parents (Pizer, Walters, & Meier, 2013). As already mentioned, the backgrounds of CODAs and Deaf children are highly diverse. For some CODA children, spoken language input may not be so readily available to them if both parents are Deaf. For Deaf children, factors like hearing levels (i.e., mild to profound hearing loss), nature of hearing loss (i.e., conductive or sensori-neural), use and effectiveness of assistive hearing devices (i.e., cochlear implants or hearing aids), hearing status of caregivers (i.e., Deaf or hearing families), and educational contexts (i.e., deaf schools or regular schools) all have a direct or indirect impact on their language development, potentially leading to different levels or qualities of their state of knowledge of sign language and/or spoken language. Turning to modes of language acquisition, where 5% of Deaf children are born to Deaf families (i.e., those in which at least one parent is Deaf), these children usually acquire sign language as their L1. CODAs are also born to Deaf families and may or may not acquire sign language as one of their L1s, but, being hearing, they do not fall into this 5%. Nowadays many Deaf children born to Deaf families are also exposed to spoken language at a young age, with support from assistive hearing devices, hence undergoing sequential bilingual acquisition. However, it is unrealistic to assume that these children can process oral language input with a quality similar to that of hearing children because assistive hearing technology cannot restore normal hearing to Deaf children. Therefore, much contemporary research on bimodal bilingualism understandably focuses on child or adult CODAs, who have no hearing loss issues and who have full access to both a sign language and a spoken language, theoretically speaking. In terms of research methodology, studies with CODA participants offer more controlled evidence when the research goal is to demonstrate language interaction effects between a sign language and a spoken language from a developmental or a processing perspective (Emmorey, Borinstein, Thompson, & Gollan, 2008; Lillo-Martin, Koulidobrova, de Quadros, & Chen Pichler, 2012). A sizeable proportion of deaf children (about 95%) are born into hearing families. They are usually prescribed with assistive hearing devices during the first few months of their lives to bolster acquisition of a spoken

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

486

G L A D Y S TA N G A N D F E L I X S Z E

language as their L1. The chance of exposure to a sign language as an early L2 is subject to whether their parents are receptive to the argument that early sign language acquisition is beneficial for child development (e.g., Humphries et al., 2014). Otherwise, exposure to sign language is off-limits to this cluster of the deaf children population. However, schooling is an important opportunity for sign language input at a subsequent stage of their language development. Previously, schools for the deaf that endorse sign language would mean input from Deaf peers, Deaf teachers, and hearing teachers who can sign. However, in recent decades, promotion of inclusive education for the deaf (usually without sign language support) in the mainstream school context has drastically minimized the opportunity for sign language exposure. Recent years have also seen the emergence of a set of pedagogical arrangements to develop sign bilingual and co-enrollment programming in mainstreamed settings, to safeguard early and consistent sign language input for Deaf children (Tang, 2017). Therefore, depending on school policy, the onset of L2 acquisition in a sign language with deaf children born to hearing parents may range from early childhood (i.e., sequential child bimodal bilingual acquisition) to adulthood (i.e., adult second language acquisition). In most cases, onset of a signed L2 is set against a state of knowledge of a spoken language as L1 that is either developing or incomplete. In sum, the diverse backgrounds of D/deaf learners of a sign language as observed in L1, L2, or bilingual acquisition all point to the complexity of research on bimodal bilingualism. The second background issue involves the terminological distinction between bimodal bilingualism and sign bilingualism. Basically, both concepts promote the early exposure to natural sign language as a means to support deaf children’s language and literacy, cognitive, as well as sociopsychological development. Also, both may be seen as attempts to counteract the misconception that arises from perceiving any forms of signing as belonging to a sign language proper, when they do not. For instance, alongside the natural sign language of the community, there usually exists another form of signing that codes a spoken language manually, such as Signed Exact English in English-speaking contexts and signed Chinese in the context of Mainland China. Such manually coded spoken language does not usually adhere to the principles of linguistic organization of natural sign languages but follows those of the spoken language with modifications. Also, it is not uncommon for a manually coded spoken language to combine with mouthing or vocalized speech (sometimes called simultaneous communication [SIMCOM]). In some deaf education settings, SIMCOM is adopted, especially when teachers for the Deaf are biased toward this mode of signing for the promotion of deaf literacy in a spoken language and they keep the natural sign language used by Deaf students off-limits in the classroom. From a bilingual acquisition perspective, however, input from artificial signing varieties does not necessarily

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

487

contain positive evidence pertaining to natural sign languages used by members of the Deaf community. Hence artificial signing is not a target language relevant for research on bimodal bilingualism. Despite the commonalities between bimodal bilingualism and sign bilingualism, however, a difference exists in the philosophical orientation of these two paradigms of research. Sign bilingualism reflects a strong pedagogical undertone. In the late 1970s and early 1980s, this paradigm laid the foundation for attempts to develop sign bilingual or bilingualbicultural education for the Deaf in the Deaf school context in the Scandinavian countries, the United States, the United Kingdom, and Australia, backed by the sign language research findings that had been gradually emerging (Swanwick, Hendar, Dammeyer et al., 2014). In some typical sign bilingual classrooms, pedagogical practices are tailored on the assumption that natural sign language is deaf children’s L1, whether or not their parents are Deaf, and that the sign language should be used in classroom communication in order to support deaf literacy development. When the concept of sign bilingualism started to take root in many Deaf education contexts, various renditions emerged, however, and more often than not they encompassed the adoption of different forms of signing, including those artificial registers that stem from spoken language. In fact, during this period of development of sign bilingualism, the methodology for research on sign language acquisition primarily adopted a monolingual paradigm and little attention was paid to how Deaf children acquire speech. Also, arguments for crosslinguistic influence between L1 and L2 did not stem from language acquisition theories, but largely from theories of bilingual education that contemplated facilitative and supportive unidirectional and subsequently bidirectional transfer of linguistic as well as conceptual knowledge between the two language systems (Cummins, 2006; see also Treffers-Daller, Chapter 15, this volume). But there was silence or even skepticism about the possibility of morphosyntactic transfer if two languages are structurally dissimilar at the surface level, as is typically the case with sign languages and the counterpart oral languages with which they coexist in the wider hearing community. In the absence of linguistic and language acquisition theories to back the development of sign bilingualism, it is easy for some educators to wrongly assume that the lack of a print form in sign language and the superficial structural differences between a spoken and a sign language due to differences in modality mean that there is no facilitative transfer. Under such faulty assumptions, the learning of a natural sign language can be wrongly thought to offer little support to Deaf children’s literacy development in the spoken language initially (see Mayor & Leigh, 2010, for some detailed arguments against the sign bilingualism paradigm). The concept of bimodal bilingualism is gradually gaining currency in sign language research, precisely to fill this gap and to dispel the misconception that early exposure to sign language does not have a positive

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

488

G L A D Y S TA N G A N D F E L I X S Z E

effect, or an effect at all, on Deaf children’s spoken language development. Contrary to sign bilingualism, bimodal bilingualism has a strong linguistic undertone in both its philosophical orientation and research methodology. It takes the study of sign language linguistics as a point of reference, and it attempts to examine the linguistic phenomena that occur in bimodal interactions, embracing the assumption that the kind of knowledge of language posited by linguistic principles of Universal Grammar is amodal, and thus bimodal crosslinguistic influence, which underpins code-mixing (i.e., code-switching and blending), is part and parcel of bimodal bilingual acquisition and bilingual processing. Having discussed these two fundamental background issues about how bilingualism is currently conceptualized in the Deaf and sign language contexts, and the complex learner variables involved, in the bulk of the chapter we focus on research on code-blending, defined as the integration of linguistic materials from both a sign and a spoken language simultaneously during language production. This is a prevalent phenomenon among Deaf and hearing bimodal bilinguals and it sheds light on the question of whether bimodal bilinguals are subject to the same language inhibition processes as unimodal bilinguals are thought to be (see discussion in Paap, Chapter 22, this volume), or whether instead they have simultaneous dual access to two linguistic systems online. We in turn examine the evidence for code-blending in adult CODAs’ language use (Sections 3 through 5) and in Deaf children or CODAs acquiring language (Sections 6 and 7).

3 Code-Blending by Adult Bimodal Bilinguals On the question of whether bimodal bilinguals also code-mix, just as unimodal bilinguals do, findings from the many relevant studies converge in suggesting that bimodal bilinguals prefer code-blending to codeswitching, meaning that while they are capable of switching between a sign language and a spoken language, they show a much stronger preference to simultaneously “blend” into the same utterance the vocabulary or grammatical components of a sign language and a spoken language. Earlier studies involved adult CODAs who were proficient in language pairs like American Sign Language (ASL) and English (e.g., Emmorey, Borinstein, & Thompson, 2005) and Lingua dei Segni Italiana (Italian sign language [LIS]) and Italian (e.g., Bishop, Hicks, Bertone, & Sala, 2006). The CODA data reveal that (a) nouns and verbs take part in code-blending more often than adjectives, and (b) semantically congruent blends on words and phrases outnumber semantically incongruent blends to a great extent. Example (1) shows a semantically congruent code-blend where even sign in English and SIGN in ASL (i.e., [even sign|SIGN]) are articulated more or less co-temporally (the words and signs involved in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

489

code-blending are underlined). Example (2) involves a semantically incongruent code-blend, where the two languages contribute independent grammatical components to the proposition of the utterance. The classifier verb CL:Vwgl [walk down stairs] comes from ASL while the subject my mom’s comes from English. In all examples underlining indicates the extent of code-blending. (1)

ASL–English, Emmorey et al. (2008, Example 1a) English she didn’t even sign until fourteen ASL SIGN

(2)

ASL–English, Emmorey et al. (2008, Example 3b) English and my mom’s you know walking down ASL CL:Vwgl [walk downstairs]

Following Myers-Scotton’s (2002) Matrix Language Frame Model, Emmorey et al. (2008) proposed that for adult ASL–English CODAs, English is usually the matrix language that sets up a morphosyntactic frame for code-blending. The adoption of English may be due to the CODAs’ hearing status and a more dominant spoken language. However, Emmorey et al. also noted the occurrence of ASL as the matrix language, which they referred to as coda talk, as in example (3). In other words, adult CODAs can retrieve the syntax of both languages in production, leading to different forms of code-blending. According to Emmorey et al., such a language choice is subject to language dominance and the discourse context. (3)

ASL–English, Emmorey et al. (2008, Example 4a) English wonder what do ASL WONDER DO DO translation “(Sylvester) wonders what to do”

Instead of CODAs, Fung, Tang, and Lam (2008) recruited native and nonor near-native Deaf signers to analyze their Hong Kong Sign Language (HKSL) and Cantonese code-blends. While Emmorey et al. (2008) claim that coda talk is infrequent among adult CODAs, it is quite frequent among the Deaf bimodal bilinguals Fung et al. studied. Specifically, in example (4), the morphosyntactic frame set up for codeblending comes from HKSL, as judged by the occurrence of an initial adjunct clause with the verb SLEEP, to be followed by two phrases involving a negator NOT-HAVE (i.e., ITCHY NOT-HAVE “not itchy” and FEEL NOT-HAVE “not felt”), reflecting the head-final property of HKSL grammar. In fact, the native Deaf signer of HKSL in (4) chose to complement the HKSL signing with Cantonese mouthing, which this signer seldom does in other contexts. We can speculate that he did so here because he perceived the addressee as a nonnative Deaf signer.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

490

G L A D Y S TA N G A N D F E L I X S Z E

(4)

HKSL–Cantonese, Fung et al. (2008) Cantonese fan3gaau3 han4 HKSL

mou5

sleep

itchy

SLEEP

ITCHY NOTHAVE

translation

gam2gok3 mou5

not-have feeling FEEL

not-have NOT-

NOT-

HAVE

RIGHT

FINISH

“I didn’t feel itchy when I slept, it’s not right (strange)”

Example (5), also from Fung et al. (2008), represents a form of codeblending in which the proposition is contributed by the linguistic components of two languages simultaneously. It was produced by a near-native Deaf signer who had higher fluency in Cantonese than in HKSL. The utterance follows a Cantonese morphosyntactic structure. Although the pronoun [ngo5|IX1], the noun [jat1man1|ONE-DOLLAR], and the verb [bei2|GIVE3] all demonstrate semantically congruent root blends, [bei2|GIVE3] is actually morphosyntactically incongruent in the sense that GIVE3 in HKSL, which carries a verb agreement feature, is morphosyntactically more complex than Cantonese bei2, which is bare, since Cantonese is a language poor in inflection. Also, as Cantonese is a null argument language, the recovery of the indirect object is actually through GIVE3 in HKSL. Modulating the directionality of movement of the sign to locusa in space yields a third-person agreement feature of the indirect object. Additionally, the perfective viewpoint aspect of the event is realized by the Cantonese perfective marker zo2 that occurs post verbally in Cantonese. In HKSL, when the movement of give ends at locusa in space, it marks the endpoint of the event (i.e., it is telic). (5)

HKSL–Cantonese, Fung et al. (2008) Cantonese ngo5 bei2 zo2 jat1man1 pro1 give ASP one-dollar HKSL IX1 GIVE3a ONE-DOLLAR translation “I gave them one dollar”

Examples (4) and (5) show that Deaf bimodal bilinguals may use codeblending as an accommodation strategy in native and nonnative Deaf communication, corroborating an earlier claim that the diverse backgrounds of Deaf signers may influence their language choice and codeblending strategies. It also shows that their possession of two independent linguistic systems enables them to retrieve the relevant grammatical components from both systems to compose a coherent proposition in a grammatically licit, code-blended utterance. Thus, it gives further evidence that code-blended utterances are qualitatively different from those in SIMCOM. Code-blended utterances are based on two grammars but SIMCOM is based on the spoken language grammar and vocabulary from the sign language, plus artificial signs usually created to compensate for the absence of certain grammatical properties of spoken language in signing.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

491

4 Experimentally Induced Cross-Modal Coactivation in Adult CODAs Emmorey et al.’s (2008) observation that cross-modal coactivation occurs among bimodal bilinguals sparked off a series of psycholinguistic and neurolinguistic experiments on such processes, involving CODAs primarily, and adult Deaf signers occasionally, as participants (see Emmorey, Giezen, & Gollan, 2016, for summary). Due to space limitations, we review here only the psycholinguistic studies. Using a picture-naming task, Emmorey, Petrich, and Gollan (2012) found no difference among ASL–English CODAs in their response times for naming in ASL alone versus in ASL in code-blends. Also, code-blending had a facilitation effect on the retrieval of low-frequency ASL signs. Using a processing cost task traditionally employed to measure whether a slowdown is incurred in code-switching because of the need to inhibit one language, Emmorey et al. compared the response time for target ASL and English interspersed with a code-blend in a picture-naming task. They found a time cost for turning off but not turning on the nontarget language (i.e., switching from a code-blend to either language alone vs. switching into a code-blend from either language). In other psycholinguistic studies, similar findings have been reported. For example, the Deaf ASL–English bilinguals in Morford, Kroll, Pin˜ar, and Wilkinson (2014) judged semantically related written word pairs faster when the ASL translation equivalents overlapped in sign phonology. Additionally, they judged semantically unrelated words more slowly under the same condition. Likewise, controlling for ASL phonology in the design of images of the target English words and competitors, Shook and Marian (2012) found that adult ASL–English CODAs took a longer processing time in auditory word recognition than English monolinguals when the target English word and the competitor shared a similar ASL phonology; in other words, they found translation mediated phonological effects. Taken as a whole, the psycholinguistic experiments on code-blending by adult bimodal bilinguals suggest cross-language coactivation at different linguistic levels, similar to unimodal bilinguals, both in production and comprehension. However, bimodal bilinguals tend not to code-switch (i.e., inhibit one language) but instead integrate information through codeblending, leading to a facilitation effect in processing. This is unlike unimodal bilinguals, who tend to select one language and inhibit the other, resulting in a processing cost (see Paap, Chapter 22, this volume, for review and discussion). As Emmorey and colleagues have argued, instead of competing with information from the nontarget language as in unimodal bilingual processing, code-blending facilitates lexical recognition and semantic processing by integrating lexical materials from the two lexicons not only at

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

492

G L A D Y S TA N G A N D F E L I X S Z E

the phonological but also at the semantic-conceptual level. Moreover, there is coactivation despite non-overlapping phonological systems between a signed language and a spoken language, as evinced by the effect on the judgment of semantic relatedness in spoken English words as a function of the presence of ASL phonology. Finally, perceptual overlap between the phonological systems is not required for language coactivation to occur.

5 Theoretical Account of Code-Blending in Adult Processing How can we account for the production of simultaneous ASL–English codeblends? Emmorey et al. (2008) developed a processing model that encompasses speech and gestural production. In this model, the ASL Formulator and English Formulator are distinct, each driven by a lexicon and all the grammatical properties for the respective language. Code-blends are linguistic outcomes when a proposition is channeled into the Message Generator that connects to the ASL as well as English Formulators. The choice of a formulator determines which language forms the matrix language in some forms of code-blending, suggesting that the selection of a matrix language occurs early in the process, whereas lexical representations are activated within both formulators and remain available online for insertion at a later stage. However, to account for code-blends that generate two apparently different or shared syntactic structures, as in example (3) earlier, the model allows both syntactic structures to be activated simultaneously in order to accommodate the insertion of grammatical materials from either formulator (see Paap, Chapter 22, this volume, for discussion of theoretical models for the processing of languages in unimodal bilinguals). In sum, cross-modal coactivation of a signed language and a spoken language during real-time processing must be accounted for in a model that stipulates that lexical and grammatical constituents can be inserted into the process at various stages, thus vitiating the competition and subsequent inhibition of language materials that commonly occur in codeswitching by unimodal bilinguals.

6 Code-Blending by Child Bimodal Bilinguals Earlier research involving child bimodal bilinguals reported similar developmental milestones to child unimodal bilinguals and sensitivity to language choice as early as age one to two (Petitto et al., 2001). On the other hand, bimodal bilingual children tend to code-blend rather than code-switch, and this fact has caught many researchers’ attention. With the exception of a study by van den Bogaerde (2000), which included some Deaf children,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

493

and Fung and Tang’s (2016) study, which investigated a Deaf child, most child studies recruit CODAs as participants. In this section, we review some research findings contributed by research representing six different language pairs. Prinz and Prinz (1981) reported on the early code-mixing of ASL and English by a child CODA at age 1;3 (observation period: ages 0;6–1;9), suggesting the existence of two developing lexicons at an early age. The one-year longitudinal study by Petitto et al. (2001) on Langues des Signes Que´be´coise (LSQ) and French bimodal children also shows similar results. The researchers investigated the bilingual development of three unimodal bilinguals (i.e., French–English) and three child CODAs (i.e., LSQ–French), each representing three age periods (Group I: 0;10–2;10, Group II: 2;26–3;6 and Group III: 3;6–4;6) and compared it to the language input from individuals who regularly interacted with them. Petitto et al. found a correlation between the caregivers’ input and the bilingual children’s output. When parents of the unimodal bilinguals did not mix the two codes in their input, their children barely code-mixed in their output (no more than 2% to any caregivers). In the same vein, when LSQ–French mixing was pervasive in the input to the child CODA of Group I (i.e., 91% for the Deaf mother and 33% for the family friend), it attracted a high rate of intra-utterance language mixing from the child (i.e., 33% to the mother and 20% to the male family friend). The researchers argued that bimodal bilingual children, just as unimodal children, show early language differentiation, based on evidence of their language choice in the discourse context (see Serratrice, Chapter 1, this volume, for further discussion). They also used the code-blending data to disconfirm the myth of early language confusion in bilingual acquisition. Based on a small proportion of simultaneous incongruent mixes, 35/320 of the total congruent and incongruent mixes, as in (6a, 29 tokens) and (6b, 6 tokens), they argued that bimodal bilinguals have access not to a fused system but to two independent grammatical systems. (6a)

LSQ–French, Petitto et al. (2001, Table 6A) French tiens puis du jus here and some juice LSQ BOIS JUS drink juice

(6b)

LSQ–French, Petitto et al. (2001, Table 6B) French mon chien my dog LSQ CHIEN MON dog my

In (6a), while the LSQ noun JUS overlaps with French du jus, the verb BOIS “drink” is supplied by LSQ only, in other words, the simultaneously produced signs and words may express semantically incongruent meanings

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

494

G L A D Y S TA N G A N D F E L I X S Z E

and may come from different grammatical categories. In (6b), the codeblend is represented by two different word orders; hence it is akin to two different utterances, suggesting that child bimodal bilinguals can preserve the word order properties of each language in their output. A study by Baker and van den Bogaerde (2008) on Nederlandse Gebarentaal (Sign Language of the Netherlands) (NGT) and Dutch codemixing by four Deaf mothers speaking to three Deaf and three hearing child bimodal bilinguals between one and three years of age also shows interesting code-blending phenomena. Methodologically, the researchers used semantic congruence between words and phrases instead of morphosyntactic structure as a criterion for classifying these children’s linguistic outputs. The researchers found a relation between hearing status and quantity of three types of utterances: (a) Dutch-based only, (b) NGT-based only, and (c) code-blended NGT–Dutch. Although the input from the Deaf mothers consistently showed a significant proportion of code-blending in parent–child interactions, mothers tended to code-blend less often when interacting with the Deaf children (65–85%) than with the hearing children (75–90%). In turn, the Deaf children produced markedly less code-blending (i.e., 9–22%) than the child CODAs (i.e., 52–67%) of the same age. While the mothers’ code-blending input can be quite complex linguistically, such as the code-blend Dutch–NGT in example (7), they tended to adopt more NGTbased code-blends when interacting with their Deaf children but more mixed code-blends when interacting with their hearing children. The Deaf children, on the other hand, preferred NGT or NGT-based codeblends to other types at all ages and produced hardly any speech. In contrast, two of the three child CODAs code-blended for about onethird of their utterances and the third child simply preferred speech to code-blending during a similar period of development. Baker and van den Bogaerde suggested that the perceived levels of competence in NGT and Dutch in the Deaf children interacted with the types of code-blends adopted by interlocutors. That is, factors such as the perceived language dominance of bimodal bilinguals may have been at play. (7)

NGT–Dutch, Baker and van den Bogaerde (2008; maternal utterance to child aged three years) head nod head nod NGT WIJ-TWEEE¨N GAAN we-two go Dutch mee ik with I translation “yes, together, I will go with you”

Donati and Branchini (2013) studied six older LIS–Italian child CODAs between six and eight and their two Deaf parents. The typology of code-blends that they developed draws heavily on word order syntax as a criterion for classification and comprised: (a) dominant blending,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

495

(b) independent blending, and (c) blended blending. Dominant blending reveals “an autonomous and complete utterance in just one modality, supported by few words or signs in the other modality” (p. 104), as example (8) shows. (8)

LIS–Italian, Donati and Branchini (2013, Example 13) Italian la regina dice the queen say.PRES.3SG LIS SAY SAY translation “the queen says”

The dominant utterance in (8) is in Italian, but the verbs [dice|SAY] are blended, albeit not co-temporally. According to the researchers, this kind of lexical insertion in code-blending is redundant from the point of view of code-mixing in conventional bilingualism research, where under normal circumstances it would be analyzed as a compensation strategy. Namely, in the absence of a word in the lexicon of the stronger, preferred language, a word from the lexicon of the other language is mixed in. Obviously, compensation is not a strategy with semantically congruent code-blends; rather, it reveals language differentiation as the word and the sign are thought to be retrievable from two independent lexicons. A challenge to the theory of word order syntax and linearization ensues, according to Donati and Branchini (2013), when the code-blended utterances follow different word orders (i.e., independent blending in their typology) or when the utterance is not complete without the simultaneous contribution of grammatical constituents from the two languages (i.e., blended blending in their typology). The two cases are illustrated in examples (9) and (10). (9)

LIS–Italian, Donati and Branchini (2013, Example 22) Italian eh? non ho capito uh? NEG have.PRES.1SG understand (past participle form) LIS I UNDERSTAND NOT translation “uh? I haven’t understood”

(10)

LIS–Italian, Donati and Branchini (2013, Example 27) Italian parla con Biancaneve talk.PRES.3SG with Snow White LIS TALK HUNTER translation “the hunter talks to Snow White”

In example (9), the negator precedes the verb in Italian but follows it in LIS. That there are two independent word orders in one code-blended utterance is hypothesized to be computationally costly, whether or not one assumes two syntactic structures, each being projected from the two lexicons, or an abstract structure to accommodate both. In (10), linearization is not an issue because the proposition is built upon supplementary

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

496

G L A D Y S TA N G A N D F E L I X S Z E

grammatical constituents contributed simultaneously by both languages, as in the subject hunter in LIS and the object Snow White in Italian. This example is similar to the Dutch–NGT code-blend in example (7) produced by one of the Deaf mothers in Baker and van den Bogaerde (2008), which was seldom produced by the child bimodal bilinguals in their study, due to their Dutch proficiency. In contrast to the few tokens of code-blends observed among the NGT–Dutch Deaf bimodal bilinguals by Baker and van den Bogaerde (2008), Fung et al. (2008) reported that CC, a HKSL–Cantonese Deaf child, produced a high frequency of code-blends, almost in more than 56% of the data collected at ages 4;9, 5;8 and 6;11. The Deaf child was born to Deaf parents. The father was raised in a Deaf school but the mother learned HKSL in her teens. Most of the code-blends are congruent in terms of morphological roots but some are incongruent in terms of morphosyntactic structures. Example (11) is similar to the LIS–Italian code-blend in (10). (11)

HKSL–Cantonese, Fung et al. (2008, CC at 5;8.24) Cantonese luk@nv green HKSL PEN pen translation “green pen” Note. @nv refers to non-vocalized speech.

In (11), the syntactic constituent of the NP green pen is compositionally derived from an adjective luk “green” in Cantonese and a head noun PEN “pen” in HKSL, hence involving no word order. Another study by Fung and Tang (2016) further examined if Deaf children are sensitive to the head directionality of functional elements. In HKSL, modals, negators, and the auxiliary HAVE are head final, whereas in Cantonese they are head initial. For code-switching in spoken languages, Chan (2008) hypothesized that it is the inherent properties of the functional, not lexical, heads of the language pair that determine the directionality of complements. Accordingly, Fung and Tang (2016) hypothesized that if a functional head is non-blended (i.e., in either HKSL only or in Cantonese only), the directionality of the complement will follow the inherent property of the head that the language entails, and so will the blended elements. When a functional head itself is blended, usually with semantically congruent roots, the complement may occur in either direction. Under this condition, this will invoke a language choice and the head directionality follows. Example (12) was produced by CC at age 5;0.8. The utterance starts with a repeated exclamative wai3 “hey” and ends with a modal CAN, suggesting a head-final syntactic structure. The code-blended complement (i.e., verb [tai2|SEEa], where a refers to a spatial locus, SP) precedes the modal can, hence a complement-head order, reflecting the word order syntax of HKSL.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research (12)

497

HKSL–Cantonese, Fung and Tang (2016, Example 26, CC at 5;0.8) Cantonese wai@nv(x3), tai@nv haa@nv ? hey see ASP HKSL gest.attn SEEa CAN? hey see.SP can Translation “hey, (let me) take a look (at this)?” Note. gest.attn = gesture calling for attention; @nv refers to non-vocalized speech; SP = Singular Person.

Example (13) shows a blended verb [sung3|GIVE] preceding a blended negator [m4-hai6|NOT]), yielding a complement-head order, reflecting the HKSL word order. Note also that the agreeing verb GIVE in HKSL is uninflected for spatial or person agreement, a developmental characteristic of the HKSL–Cantonese Deaf child. (13)

HKSL–Cantonese, Fung and Tang (2016, Example 29, CC at 6:6.26) Cantonese . . .., sung@nv m-hai@nv give (as a gift) NEG.be HKSLL . . .., GIVE NOT give NEG Translation “ . . ., (Mr. Sam) did not give (it) as a gift (to the students)” Note. @nv refers to non-vocalized speech; NEG = negator.

While Fung and Tang’s (2016) hypotheses were largely confirmed by the HKSL–Cantonese code-blending data, there were three tokens that constitute violations to their predictions. In (14), the modal CAN in HKSL is head initial rather than head final as required by adult HKSL, thus the complement follows rather than precedes the head, resulting in a headcomplement order, an illicit word order for modals in adult HKSL. (14)

HKSL–Cantonese, Fung and Tang (2016, Example 27, CC at 5;0.8) Cantonese tai@nv haa@nv, tai@nv haa@nv ? see ASP see ASP HKSL CAN SEE , SEE ? can See see Translation “may I take a look?” Note. @nv refers to non-vocalized speech.

Being a bimodal bilingual acquirer of HKSL and Cantonese, CC basically assumed the head directionality of Cantonese in his production of the code-blended utterance in (14). This is coupled with morphological root blending on the VP complement [tai|SEE] and an independent aspect marker -haa in Cantonese. Fung and Tang (2016) argued that crosslinguistic influence is at play during CC’s bilingual acquisition of sign language and spoken language. In fact, one can assume that when bilinguals have more than one linguistic system at their disposal, drawing resources from both in bilingual acquisition and bilingual performance is by default a naturalistic outcome. Note that CC did produce HKSL utterances with a head-final modal or negator in other contexts, justifying the existence of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

498

G L A D Y S TA N G A N D F E L I X S Z E

a separate linguistic system of HKSL that is distinct from Cantonese. In sum, the HKSL–Cantonese code-blending study potentially offers a window through which one can examine the existence of two linguistic systems and at the same time the conditions under which the resources from both are retrieved during linguistic performance. It is possible that the two developing grammars at different stages of bilingual acquisition impose constraints on crosslinguistic influence, leading to differences in the retrieval of linguistic resources from both languages. Future research is necessary to verify this hypothesis, and to ascertain the factors underlying crosslinguistic influence in both simultaneous and sequential bimodal bilingual acquisition. Such factors may be language dominance, language choice, and even the perceived hearing status and level of linguistic competence by the interlocutors involved in the interaction. In a rare binational project, Lillo-Martin, de Quadros, Koulidobrova, and Chen-Pichler (2010) conducted a particularly systematic bimodal bilingual child acquisition study that involved two groups of bimodal bilinguals: ASL–English and Lı´ngua de Sinais Brasileira, Brazilian Sign Language (LIBRAS)–Brazilian Portuguese. Adopting a generative approach, the team attempted to elucidate central phenomena like language separation and crosslinguistic influence. The reported findings showed a larger proportion of bimodal congruent than incongruent code-blends around age two for two ASL–English child CODAs and around age 2;09–2;10 for a LIBRAS–Brazilian Portuguese child CODA. When analyzing the nonadult utterances that erred in word order, the researchers claimed that most are sanctioned by the grammar of the sign languages concerned, as shown in the OV order in the code-blending data in (15a) and doubling in English in (15b). (15a)

ASL–English, Lillo-Martin et al. (2010, Example 5c, Ben at 2;01) English chocolate eat ASL HOT CHOCOLATE IX EAT Note. IX = pointing.

(15b)

ASL–English, unilingual utterance in English, Lillo-Martin et al. (2010, Example 6, Ben at 2;01) English sleeping mouse sleeping

Lillo-Martin and colleagues have published subsequent findings from this binational project adopting the research framework of bilingual acquisition, and viewing code-blending as an epiphenomenon that arises from crosslinguistic influence. To illustrate, Lillo-Martin et al. (2012, Study 1) investigated the acquisition of wh-questions in speech and sign by the same two ASL–English child CODAs (Ben, 1;11–3;03; Tom, 1;11–4;05) and one LIBRAS–Brazilian Portuguese child CODA (Igor, 2;10–3;02) as in the 2010 study. In English and Brazilian Portuguese, wh-questions are fronted to [Spec CP] or left in situ in some specific contexts such as echo questions. However, ASL and LIBRAS allow more options: fronted, in situ, and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

499

doubled (i.e., in initial and final position). The findings showed a predominance of wh-initial questions produced by both groups of bimodal bilinguals, who also produced more wh-in situ questions and earlier than has been observed in monolingual children acquiring either English or Brazilian Portuguese. In addition, Ben, Tom, and Igor also produced whdoubles in English that are reminiscent of sign language structures, as in the example (16) from Ben. (16)

ASL–English, Lillo-Martin et al. (2012, Example 9, Ben at 2;02) English Where balloon where?

Interestingly, after 2;11, Ben and Tom stopped producing wh-doubles in English while they continued to produce wh-in situ questions, suggesting the child CODAs were aware that wh-doubles pertained to the grammar of ASL or LIBRAS only, respectively, and not English or Brazilian Portuguese. Lillo-Martin et al. (2012, Study 2) administered a follow-up elicited production task targeting wh-questions in English and Brazilian Portuguese to two groups of child CODAs respectively (five ASL–English bilinguals aged between five and six, including Ben and Tom, see earlier; and two LIBRAS–Brazilian Portuguese bilinguals aged 4;9 and 7;4). The whquestions produced by the ASL–English CODAs were overwhelmingly whinitial while those produced by the LIBRAS–Brazilian Portuguese CODAs were consistently so, suggesting that at such ages, these children had mastered the English or Brazilian Portuguese grammars. Using the same stimuli, in Study 3, Lillo-Martin et al. (2012) examined the CODAs’ production of wh-questions in ASL and LIBRAS (six ASL–English bilinguals aged 5;3 to 7;9, and three LIBRAS–Brazilian Portuguese bilinguals aged 4;9 to 7;4). The signing productions were overwhelmingly wh-initial. Although some CODAs also produced other variants (e.g., wh-doubles and wh-finals), both in the contexts of argument and adjunct wh-questions, they did so to a much lesser extent when compared with the monolingual ASL data reported in Lillo-Martin (2000). In this early study, the Deaf child acquirers (ages 4;1–6;9) and native signers of ASL produced a much higher proportion of wh-doubles and whfinals, in addition to wh-initial questions. In other words, the CODAs’ production of wh-questions in ASL is less varied than that in the Deaf child acquirers of ASL.

7 Theoretical Accounts of Code-Blending by Child Bimodal Bilinguals Except for Baker and van den Bogaerde (2008), studies focusing on child bimodal bilingualism have opted for a generative approach in their analysis. At the outset, all followed MacSwan’s (2005) proposal for spoken languages that nothing constrains code-switching other than the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

500

G L A D Y S TA N G A N D F E L I X S Z E

requirements of the mixed grammars; by the same token, invoking a third grammar for constraining code-blending is untenable in this theory. Similar to Emmorey et al.’s (2008) processing account, all scholars agree that bimodal bilinguals have one computation system but two independent lexicons, thus permitting cross-modal insertion of linguistic elements. Nevertheless, they differ in their postulations about how the input system works. We review here three proposals, which correspond to our discussion of research and the data illustrations we showed in (6) through (16). The head directionality account put forward by Fung and Tang (2016) focuses only on the conditions under which linguistic materials from the two lexicons function to constrain code-blending. It assumes that the inherent properties of functional heads determine the directionality of complements in the syntactic derivation. The results based on HKSL–Cantonese modals and negators largely confirm this hypothesis, but future verification through the analysis of other functional or even lexical heads is necessary. Also, the proposal needs to account for those blended utterances in which the proposition is composed simultaneously by supplementary grammatical constituents from both languages, such as the HKSL utterance shown in (11). The proposal by Donati and Branchini (2013) assumes a broader perspective and addresses the question of whether word order is part of the syntax subsumed under the principles and parameters of Universal Grammar or a post-cyclic and post-syntax phenomenon due to a linearization algorithm at the interface level (i.e., the Phonetic Form [PF]). Using the LIS–Italian code-blending data as a litmus test, they opted for the latter answer, positing a single abstract syntactic structure but two PF linearizations, resulting in the possibility of two different word orders as outputs of the single computation system and two independent lexicons. Also, to account for simultaneous composition of a single proposition with linguistic elements distributed over two modalities (i.e., blended blending in their typology), they argued that linearization, being a late phenomenon, can be suspended and no word order difference results. The third theoretical account is the Language Synthesis Model. This model was put forward by Lillo-Martin, de Quadros, and Chen Pichler (2016) to account for crosslinguistic influence in the bilingual acquisition of CODAs or KODAs (i.e., kids). Lillo-Martin et al. consider these to be heritage signers, by association with the wider classification of unimodal heritage language learners, on the grounds that they develop highly variable proficiency in their sign language, just as heritage speakers do in their home language (for discussion of [un]learning and relearning in this population of bilinguals, see Singleton & Pfenninger, Chapter 4, this volume, and Keijzer & de Bot, Chapter 14, this volume). Note that in Lillo-Martin et al.’s account, crosslinguistic influence covers acceleration and delay effects, as well as transfer effects. Drawing from insights in

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

501

distributed morphology, their model assumes one computation system and two independent lexicons, with two loci in the model that allow for what they call synthesis, that is, crosslinguistic influence or interactions between the two languages, to occur as is apparent in the case of codeblending. The first locus for language interaction to occur is at the point of selection of roots and morphemes from either of the bimodal children’s languages. The second locus is what they call vocabulary insertion, which is a late process after syntactic derivation where the vocabulary may come from either language or both. In sum, LilloMartin et al. (2016) propose that language synthesis is not a developmental phenomenon per se; evidence like CODA talk by adult bimodal bilinguals actually suggests bidirectionality of linguistic transfer, a characteristic of bimodal bilinguals in general. If bilinguals have at their disposal two independent but interactive lexicons, bidirectional transfer of linguistic elements is a natural outcome of acquisition (rather than linguistic confusion as wrongly assumed previously, see also Serratrice, Chapter 1, this volume). To conclude, although different accounts have been advanced to explain code-blending in the context of crosslinguistic influence in bilingual acquisition, all agree that bimodal bilinguals can access two linguistic systems simultaneously. What is at issue, then, is how bimodal bilinguals select or combine linguistic elements from one or the other language in the syntactic derivation. One such factor can be language choice. Lillo-Martin, de Quadros, Chen Pichler, and Fieldsteel (2014) have also argued that child bimodal bilinguals are sensitive to the discourse contexts such as the language input of the interlocutors, leading to language choice issues (for a review of the relation between parental input and child development in bilingual child acquisition, see Armon-Lotem & Meir, Chapter 10, this volume). Clearly input is an important factor in language acquisition; however, Lillo-Martin et al.’s argument to some extent rests upon the assumption that these CODAs, child or adult, have attained a sufficient level of competence in both languages for switching or language choice in general to occur. What is at issue here is not just that code-blending occurs, but how it occurs and, specifically, how the linguistic elements are blended into an utterance at different developmental stages in the context of two developing grammars, before the children attain the adult-like properties of their bilingual grammars. There can be two levels of variability. At one level, variability may be conceptualized as adult bilinguals’ sociolinguistic rules of language use, allowing them to code-switch between two languages, or to code-blend in various forms as permitted by the grammar of the two languages, as we observe in both child and adult bimodal bilinguals. At the second level, there may be developmental variability, meaning that the developing grammars allow two options that may be variable at some stages of acquisition before child acquirers assign one option to one language and the other to the other language. In fact, given the observations

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

502

G L A D Y S TA N G A N D F E L I X S Z E

that code-blends occur in adult utterances produced by both CODAs and Deaf parents, some of which may be spoken-language based, the question of how bimodal bilingual children figure out the properties distinct to each of the languages remains. This issue becomes even more critical when child bimodal bilinguals are Deaf and may have problems with auditory perception (even when wearing assistive hearing devices), which potentially leads to differences in acquisition outcomes. Hence, more research is necessary using Deaf bimodal bilinguals as participants. Such research will help shed light on the prevalent phenomenon of code-blending by bimodal bilinguals.

8 Metalinguistic Awareness in Bimodal Bilinguals: Insights for Linguistic Separation and Differentiation with Deaf children Educated in a Sign Bilingualism and Co-Enrollment Context While most research on bimodal bilingualism has attempted to characterize the properties of code-blends as a form of linguistic output and to examine the psycholinguistic processes involved, a small body of research has tackled the issue from the different research angle of metalinguistic awareness, defined here as the extent to which Deaf bimodal bilingual children can tell whether the signing they watch pertains to the grammar of a natural sign language or a manually coded spoken language, the latter of which is an artificially created variety of signing. In the remainder of the chapter, we examine what little is known about the metalinguistic awareness of bimodal bilinguals, as a way to further address the issue of linguistic separation and differentiation. Singleton, Morgan, DiGello et al. (2004) argued that Deaf bimodal bilinguals have a heightened metalinguistic awareness due to their constant mapping of linguistic units of a natural sign language onto those of a spoken language. Analyzing Deaf children’s narrative skills, Rathmann, Mann, and Morgan (2007) suggested that the metalinguistic awareness of bimodal bilingual children develops as they learn to combine narrative skills with linguistic skills in the construction of a coherent narrative structure. In a study that looked into Deaf children’s metalinguistic awareness in sign language, Smith, Andrews, Ausbrooks et al. (2013) argued that such awareness is reflected by Deaf children’ ability to identify the signing errors of hearing teachers and to determine whether the signing is produced based on a spoken language grammar or a natural sign language (i.e., English or ASL in their study). Three studies conducted by our laboratory have investigated to what extent Deaf bimodal bilingual children can differentiate the grammars of HKSL and Cantonese through probing into their metalinguistic awareness. These children, implanted or with hearing aids, and mostly with severe-to-profound hearing loss, are enrolled in a sign bilingualism and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

503

co-enrollment program (henceforth, SLCO) in a regular school setting, spanning Hong Kong’s education system of six years of primary school (from age 6–12) and six years of secondary school (from age 12–17). In this designated SLCO environment, Deaf children are studying with 15 to 25 hearing children, co-taught by a Deaf signing teacher and a hearing teacher. The latter basically uses speech but can also sign to facilitate communication in class. All children, Deaf and hearing, are exposed to three spoken languages: Cantonese, which is the primary language of communication, as well as English and Putonghua, both of which are school subjects. As for written languages, these children are exposed to written Chinese, which is based on Mandarin Chinese grammar, and written English. Last, HKSL is added to the plethora of languages used in the school community, with input usually provided by a Deaf signing teacher and Deaf peers. The SLCO program aims to support sign bilingual education in mainstream education, with a set of pedagogical arrangements to boost early exposure to HKSL alongside spoken language for Deaf children (see Tang, 2017). In the first study, Tang, Lam, and Yiu (2014) compared the language development of 20 SLCO children in HKSL, Cantonese, and written Chinese, sampled from children in primary grades and ranging in age from 7;7 to 13;5. They found a positive correlation in proficiency scores among and between the languages, suggesting no adverse effects when Deaf children acquire sign language alongside speech. To investigate this language interaction effect further, a second study by Tang, Yiu, and Lam (2015) recruited 18 Deaf children from the primary school grade four to the first year of secondary school (aged 9;8 to 15;0), and examined if they could differentiate HKSL from manually coded Chinese, the artificial signing variety mentioned earlier. Note that manually coded Chinese here involves primarily the grammar of Cantonese; however, it sometimes incorporates the grammar of written Chinese. Even if manually coded Chinese is not encouraged at all, the very fact that all hearing teachers and students are L2 learners of HKSL in the SLCO school setting would mean that their signing outputs consequentially reflect properties of L1 Cantonese to some extent. Sometimes, manually coded Chinese is deliberately adopted by teachers, Deaf and hearing alike, to illustrate specific word order rules of sentences or idioms in Chinese. Given such situations, the signing in the classroom can be ambiguous between the grammar of HKSL and of Cantonese (or sometimes written Chinese), leading to the question of how Deaf children weed out manually coded Chinese from the signing input and focus on the positive evidence of HKSL. A corollary issue is what linguistic cues they adopt in this process. The results in Tang et al. (2015) are as follows. On a language differentiation task that included five types of grammatical constructions demonstrating a different linguistic organization in HKSL and manually coded Chinese, with the stimuli randomized and presented by a native

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

504

G L A D Y S TA N G A N D F E L I X S Z E

Deaf signer in a video mode, two-thirds of the Deaf children were able to separate the stimuli presented in HKSL only (condition 1) from those presented in manually coded Chinese with Cantonese mouthing (condition 2). This was in contrast to a third condition in the study that involved manually coded Chinese without mouthing, which resulted in processing difficulty and a much lower score. Additionally, such metalinguistic awareness is correlated with the Deaf children’s proficiency in HKSL but not in spoken Cantonese or in written Chinese, which is understandable because on this task they have to detect clues from the signing to differentiate the varieties. The findings suggest that the Deaf children’s perception of Cantonese mouthing could be a useful cue to differentiate HKSL from manually coded Chinese. Tang et al. (2015) argued that early and consistent exposure to HKSL input in the learning environment has a positive impact on bimodal bilingual children’s ability to identify the relevant linguistic input for acquisition in HKSL. In fact, being able to differentiate mouthing in Cantonese and mouth gestures in HKSL, the latter of which are a linguistic property of natural sign languages, depends to a great extent on Deaf children’s knowledge of HKSL. On a questionnaire survey in which Deaf children were asked to reflect upon their own ability to differentiate the two types of signing and to report on their language choice in daily communication with Deaf and hearing teachers and peers, it was found that their sensitivity toward the signing varieties was quite high. Out of 18 children, 15 claimed they knew the grammar of HKSL and manually coded Chinese are different. Also, 15 claimed they could discern the signing when it is HKSL based, while only 10 claimed they could discern the signing when people “sign and speak” at the same time (i.e., knowing that they were signing in manually coded Chinese). This result suggests that the artificial mode of signing may compromise comprehension initially until the Deaf children’s knowledge of Cantonese has reached a sufficient level for them to be able to differentiate the two signing varieties and to use their knowledge of Chinese, Cantonese, or written Chinese to comprehend manually coded Chinese. In a focus group discussion about the differences that came to their mind regarding the two signing varieties, comments suggested some variability in these children’s degree of explicitness in articulating what aspects of grammatical differences can serve as clues for linguistic differentiation. Children who performed better in the differentiation task were more articulate in their explanations during focus group interactions about the linguistic differences in terms of word order, morphological patterns, non-manuals, and prosody in signing. Arguably, they possessed a higher level of metalinguistic knowledge, awareness, and abilities. The third study we review here is by Sze and Tang (2016), who carried out a follow-up investigation on a group of 15 young Deaf children (i.e., from primary school grades two, four, and six, ages between 7;9 and 12;6).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

505

In this experiment, they only focused on locative sentences in Chinese and HKSL. Locatives in HKSL typically involve simultaneous constructions like classifier predicates, shown in example (17a), and/or pointing signs, shown in (17b), and both require spatial modulations unique to the visual gestural modality. (17a)

HKSL

BALL

gloss

subject

translation (17b)

HKSL

BOXa

be-located DOG be-located +CLround +CLanimate objecta entityb classifier subject classifier predicate predicate “a ball is located here – a dog is located next to the ball” IXa

DOG

HAVE,

be-located +CLanimate entitya gloss object pointing subject verb classifier predicate translation “a box is here, there is a dog inside it” Note. a/b = spatial loci of entities.

In Chinese, to encode the locative relations between entities in sequential representation of word orders, locatives make use of existential verbs, for example, jau5 “have” in (18a), localizers as in soeng6min6 “on top of/ above” in (18b), or locative verbs such as hai2 “at/co-verb” in (18b). 上面 有 三 個 波 soeng6min6 jau5 saam1 go3 bo1 gloss on-top-of/ above have three classifier ball translation “on top of/above the table there exist three balls” Note. The second tier is in Cantonese pinyin.

(18a)

Chinese

(18b)

Chinese gloss

檯 toi2 table

有

三

jau5

saam1 go3

have three

個

波

喺

檯

上面

bo1

hai2

toi1

soeng6min6

classifier ball at/co-verb table on top of/ above (localizer)

translation

“there are three balls on top of/above the table”

Note. The second tier is in Cantonese pinyin.

The experiment in Sze and Tang (2016) involved two main tasks. Task 1 was an elicited production task in written Chinese and HKSL based on pictures and narratives. The HKSL data revealed that all 15 Deaf children demonstrated some use of spatial devices in representing locative relations, though some opted for a version of manually coded Chinese in signing. The use of manually coded Chinese in the Deaf children’s production, we believe, reflects the existence of this signing variety in the co-enrollment primary school, for reasons discussed earlier. Task 2 was comprised of two subtasks. The first involved syntactic and semantic judgments of locatives in written Chinese and HKSL, and the second was a language differentiation

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

506

G L A D Y S TA N G A N D F E L I X S Z E

task consisting of a set of signed stimuli designed to represent manually coded Chinese and HKSL. The results of the first subtask revealed that, as the SLCO children’s years of bilingual exposure increased from primary school grade two all the way to primary school grade six (roughly ages 7–12), there was a corresponding increase in their accuracy in rejecting the semantic and syntactic errors that the stimuli presented, be they in written Chinese or in HKSL. The second subtask targeted metalinguistic differentiation of manually coded Chinese vs. HKSL. In it, the Deaf children scored a set of signed utterances according to how “Chinese” and “HKSL” they “looked like” on a Likert scale from 1 to 5. The stimuli contained a number of locative relations, each represented in four different ways of signing, ranging from manually coded Chinese to HKSL. Interestingly, even the youngest group (i.e., primary school grade two) was able to differentiate manually coded Chinese from HKSL. By primary school grade four, such awareness of linguistic differentiation within the signing modality was well in place. In fact, it is also reflected in the children’s improvement in the perception as well as the production of manually coded Chinese. There have been informal reports that these SLCO Deaf children can switch to manually coded Chinese when interacting with their hearing peers and teachers, probably as an accommodation strategy. In sum, as far as the current methodologies go, the findings here offer some preliminary evidence that by age 7;9, these bimodal bilingual children have developed a high degree of cross-modal metalinguistic awareness especially in terms of linguistic separation and differentiation, despite the timing of auditory and speech intervention as well as sequential exposure to HKSL. As mentioned, medical intervention usually takes up the first year if not longer of their language acquisition lifespan, and, except for the three or four Deaf children who were born to Deaf parents, these children started to receive exposure to HKSL around kindergarten grades two or three onward (i.e., at ages four to five). Therefore, strictly speaking, they should be categorized as sequential bimodal bilinguals (Meisel, 2004), with delayed spoken language input. To what extent this mode of bimodal bilingual acquisition shares similar acquisition processes or differs from other conventional models of bilingual acquisition, or even simultaneous bimodal bilingualism, both in terms of acquisition processes and outcomes, requires further research. Nonetheless, from an applied perspective, if 95% of deaf and hard-of-hearing children are born to hearing parents and almost all deaf children are mainstreamed as far as parent consent goes, the SLCO learning environment, which emphasizes dual language input through Deaf and hearing teachers and students who can sign, and through some who use speech, may offer a better learning environment to bolster development in spoken language, written and speech, and signed language, and at the same time the opportunity to access a full curriculum in a mainstream setting.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

507

9 Conclusion In this chapter we have synthesized contemporary sign language research conducted with a bilingual lens that aims to characterize the phenomenon of bimodal bilingualism from the perspectives of bilingual acquisition and code-blending, psycholinguistic processing, and metalinguistic awareness. The review is not exhaustive, as recent years have seen the burgeoning of research that examines the neurological correlates of bimodal bilinguals (for review, see Emmorey et al., 2016). Taken as a whole, bimodal bilinguals are observed to behave similarly to unimodal bilinguals in terms of dual real-time access to two linguistic systems. Yet bimodal bilinguals differ from unimodal bilinguals in terms of language inhibition in language production because the presence of two sets of articulators potentially results in a preference for code-blending to code-switching among bimodal bilinguals. In other words, the data on code-blending suggest that it is possible to integrate linguistic materials from both languages simultaneously. This observation has led to interesting theoretical consequences. First, from the perspective of linguistic theory and language acquisition theories, the code-blending data lead to a reconsideration of the role of word order in syntax when two simultaneous linearization algorithms at the interface level of phonetic form are possible. Nevertheless, one still needs to address what causes two linearizations to occur, especially when word order outputs go in different directions. The Language Synthesis model as espoused by Lillo-Martin and colleagues (e.g., Lillo-Martin et al., 2016) encourages a reconceptualization of crosslinguistic influence as a manifestation of bilingual acquisition effects, be it in simultaneous or sequential bimodal bilingual acquisition, because bimodal bilinguals have the facility to capitalize on the linguistic resources stemming from two languages. Seen in this light, factors contributing to crosslinguistic influence need to be identified and studied empirically in the context of both simultaneous and sequential bimodal bilingual acquisition, given the complex language backgrounds of CODAs and Deaf children. This brings us to a second important issue: the hearing status of participants in bimodal bilingualism research. The recruitment of Deaf participants for the investigation of bimodal bilingualism is of crucial importance. As discussed, most studies currently recruit CODAs because spoken language data are not confounded by hearing loss issues for this population. However, given that the original goal of sign linguistic research is to uncover how Deaf children acquire sign language, or language in general, future research will need to incorporate the population of Deaf bimodal bilinguals into the pool of participants. This will probably require adjustments in our expectations toward the mode of bilingual acquisition they will be undergoing, so that appropriate predictions can be derived.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

508

G L A D Y S TA N G A N D F E L I X S Z E

The third issue we submit for future research is whether and to what extent manually coded spoken language should be taken as one form of code-blending. Theoretically speaking, manually coded spoken language is a coerced language choice that primarily adopts the syntactic derivation of a spoken language as the base language and, as such, we argue it should be seen as distinct from code-blending. Nevertheless, from the point of view of child bimodal bilinguals, it is sometimes difficult to draw a line between manually coded spoken language as a coerced language choice of an acquirer, on the one hand, and a form of code-blending as part of the natural acquisition process, on the other. In the typologies of code-blends established by different researchers, there is bound to be one category of code-blends that characterizes the predominant contribution from the spoken language, such as the Dutch-based code-blends discussed by Baker and van den Bogaerde (2008), or the syntactic calques considered by Donati and Branchini (2013). Further research is necessary to appreciate under what circumstances knowledge of spoken language is utilized to satisfy the linguistic demands of code-blends and manually coded spoken language. Sociolinguistic factors need to be examined carefully, such as demands of the discourse context and Deaf signers’ perception of the sign language competence of the addressee (see De Houwer, Chapter 17, this volume, for discussion of such factors in oral communication). Alternatively, incorporation of spoken language elements in signing may be seen as an outcome of the bilingual acquisition process, thus providing a fertile ground for the investigation into the spoken language development of Deaf children, especially of those who may not be able to develop intelligible speech. A last issue, which we, however, did not address in this chapter, is whether bimodal bilinguals demonstrate a so-called bilingual advantage, just as unimodal bilinguals have been claimed to do (see Paap, Chapter 22, this volume). This topic is presently subject to heated debate because the findings are rather mixed. Emmorey et al. (2016) caution against a premature conclusion about better executive control with bimodal bilinguals, since such an advantage is actually not observed in some of their studies. As mentioned, bimodal bilinguals do not need to control for the inhibition of the other language to the same extent as unimodal bilinguals do in bilingual processing. One possible future venue in the search for cognitive advantages, if any, is to uncover sign-language-induced cognitive processes – such as the transformation of mental images and spatial memory, as well as the recognition and discrimination of facial expressions – and examine the extent to which these abilities might interact with bimodal bilingual processing abilities. Another possibility could be to investigate the bimodal bilingual ability to monitor attention and integrate information cross-modally because of the availability of linguistic information from two language systems online.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

Bilingualism and Sign Language Research

509

To conclude, bimodal bilingualism opens new research arenas, lending insights that can enhance researchers’ understanding of how linguistic processes work in language acquisition and bilingual acquisition, as well as production and comprehension. Its study also holds great potential for pointing to new directions for research on bilingualism as a whole.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:57:31, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.025

25 Bilingualism and Bidialectalism Jean-Pierre Chevrot and Anna Ghimenton

1 Introduction While linguists have long debated the conceptual and ideological difficulties of distinguishing between a language and a variety of a language, and they find the distinction bewildering, children seem not to have any problems acquiring languages (in bilingual acquisition) and acquiring dialects (in bidialectal acquisition). In fact, very early in development, they are capable of distinguishing between typologically very close linguistic varieties as much as between more distant languages in a bilingual acquisition setting (see Serratrice, Chapter 1, this volume). Infants as young as five or six months prefer to look at native speakers rather than speakers with a foreign accent (Kinzler, Dupoux, & Spelke, 2007). Likewise, babies as young as five months of age are sensitive to the differences separating their home dialect from other dialects (Butler, Floccia, Goslin, & Panneton, 2011). Children’s capacity to discriminate between languages as much as between language varieties suggests that subtle differences in the way others speak are critical cues for organizing the social world. Moreover, their precocious sensitivity to minute differences between varieties of a language legitimizes the research agenda of studying bidialectal acquisition and comparing it with bilingual acquisition. However, as we explain in more detail in Section 2, such an agenda cannot circumvent the difficulty involved in contrasting two objects – language and language variety – with fuzzy borders. The question of whether bidialectalism is achieved through the same process as bilingualism has seldom been addressed, mainly because the two have tended to be viewed as separate objects of research. A way of estimating the similarity of these two acquisitional processes is to ask whether bilingual acquisition and dialect acquisition are shaped by the same attitudes and influenced by the same factors that have been of basic concern in sociolinguistics (e.g., regional background, gender, and socioeconomic status). The present chapter tackles these questions. More

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

511

specifically, it asks how learners come to perceive the social value of different varieties, that is, how they develop language attitudes toward them. It also addresses the question of how individuals learn to produce dialectal varieties. We focus on language attitudes and production within two main life stages: (1) young childhood to preadolescence, and (2) adolescence to adulthood. This grouping in two broad periods is necessary given the paucity of studies focusing on dialect acquisition, in particular when it involves mature and aging adults. Moreover, we focus on the acquisition of spoken varieties, leaving aside the analysis of written realizations of these dialectal varieties as they appear in social media, for example. To our knowledge there has been no investigation of their acquisition in this context, although there has been attention to the actual use of different varieties in written discourse (see, e.g., Jaffe, Androutsopoulos, Sebba, & Johnson, 2012). But first we turn to a brief but necessary discussion of the bewildering difficulties involved in trying to distinguish between a language and a variety of a language (in Section 2) and the settings and contexts in which learners may be exposed to different varieties (in Section 3).

2 On the Difficulty of Defining Dialect and Language Tackling issues focused on the acquisition of several languages and on the acquisition of several dialects means that both the terms dialect and language are clearly definable. Yet finding two separate and clear-cut definitions for the two terms without defining these in terms of each other is an arduous task. The conceptualization of what is a dialect and what is a language is the result of a unique diachronic process combining global differentiation and local convergence between linguistic uses. Moreover, the academic specialization and disciplinary realm in which each of the notions is conceptualized – sociolinguistics and dialectology for dialect; typology, general linguistics, bilingualism, or first and second language acquisition for language – add to the confusion. Consider, for instance, the challenges involved in conceptualizing the status of Alsatian (Alsacien) spoken in the northeastern part of France (the Alsace) and of Venetian (Veneto) spoken in the Veneto, one of Italy’s northeastern regions. Alsatian and Venetian are often considered dialects, including by their speakers, while historical and linguistic arguments converge to a categorization of each of them as languages. The difficulty that sociolinguists find in objectively underscoring dialects’ internal coherence by looking at the covariation of sociolinguistic variables (e.g., Guy & Hinskens, 2016) makes one wonder whether the notions of dialect and language are mere ideological and political constructs. Speakers from stigmatized groups may fight to ensure the recognition of their

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

512

J E A N - P I E R R E C H E V R OT A N D A N N A G H I M E N TO N

culture and this activism may promote the variety they speak every day to the status of language. Considering a variety as a language or a dialect thus is also a matter of ideology and power relationships between groups. A common definition is that dialects are different varieties of the same language. This definition, however, cannot be satisfactory because it defines dialect in relation to language, without defining what the latter refers to. We take the perspective that languages are variable and evolving systems made of several social or regional dialects, overlapping or nested into each other. This heterogeneity of languages is due to their internal dynamics, the contacts between them, and their links with social organization, which is in itself evolving, composite, and multilayered (Laks, 2013). In the same vein, we insist on the fact that dialects have a gradient quality, and we replace discrete categories with a continuous distance between two varieties. Such a distance would be (at least) tridimensional and would include geographical and linguistic distance (operationalized by a comparative analysis of systems and mutual intelligibility), as well as distance in the sphere of identity and politics. Building on this perspective, comparing bilingual and bidialectal acquisition becomes a little clearer. Indeed, examining the extent to which bidialectal acquisition differs from bilingual acquisition entails questioning how the acquisition of two varieties depends on the geographical and linguistic distance separating them, as well as on their identity value and their sociopolitical definition. Consequently, adult production of varieties of a language is influenced by numerous social factors such as the speaker’s regional background, gender, ethnicity, socioeconomic status (SES), and social network structure, as well as factors pertaining to the context of speech, such as the topics of exchange or the degree of formality of the interaction. Moreover, speakers assign social values to the varieties involved and adopt positive or negative attitudes toward them (Niedzielski & Preston, 2000). As a consequence, the linguistic forms that make up different varieties carry socio-indexical information that is nonlinguistic in nature, in that it refers to groups of individuals distinguished on the basis of categories such as gender, SES, regional origin, or ethnicity (e.g., Foulkes, 2010). This understanding of both language and dialect as objects that are internally heterogeneous, gradient in nature, and with fuzzy borders equips us well to shift our attention from acquiring languages (in bilingual acquisition) to acquiring dialects (in bidialectal acquisition), as we do in this chapter.

3 Settings and Contexts in which People Acquire Different Oral Varieties There are numerous settings in which two or more varieties are acquired. Relative timing and contexts of acquisition are two important dimensions that shape this diversity. Depending on the age at which one comes into contact with two language varieties, speakers’ repertoires will be formed

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

513

either through a simultaneous or sequential acquisition process. For languages, the resulting situation is either bilingual first language acquisition (BFLA), early second language acquisition (ESLA), or second language acquisition (SLA). The cut-off ages separating one period from another have been much debated. De Houwer (2009) argues that BFLA differs from ESLA in that, in the former, the exposure to a second language is present from birth while in the latter children are exposed to a second language (L2) after they have heard only a single language (often up to the time they start daycare or preschool) but before any formal literacy instruction. SLA refers to the acquisition of a second language after the age of six, or after formal literacy instruction has begun (see Ortega, Chapter 21, this volume). These distinctions can be transferred to the acquisition of different dialect varieties. Thus, we refer to first bidialectal acquisition when two or more dialect varieties are acquired from birth, that is, in the context of socialization within the family unit. Should one variety be acquired at a later stage, within a socialization context other than the family unit (e.g., extended family, school, through relocation, etc.), we speak of (early) second dialect acquisition. There are four typical living contexts in which children and adults learn more than one variety of what is commonly seen as one sole language. The first concerns children born and growing up in communities where numerous varieties of a single language are used. From early on (but not necessarily simultaneously) children acquire traits of the varieties in contact. The rate of acquisition will depend on the degree of exposure to each of the varieties and their social and identity values (see Fuller, Chapter 6, this volume, for related discussion of the role of language ideologies; see Armon-Lotem & Meir, Chapter 10, this volume, for the role of exposure in early bilingualism). This will be true, for example, of children who are in contact with mainstream American English and a sociolect, African American English. It will also be true for children who live in a country where the regional varieties of one language are used simultaneously along with the standard variety, as, for instance, in many parts of Germany (although here most children will additionally be confronted with a foreign language through school). There is also the widespread situation of young children who are in contact with speakers using other kinds of varieties (e.g., teenage dialects or dialects of different socioeconomic groups) than those spoken by their parents. The second type of context concerns children living in bilingual communities where they are simultaneously or sequentially confronted with varieties of what are considered to be different languages. In the case of bilingualism involving heritage languages, children may learn a regional variety of a minority language at home, additional varieties of the same language in the community, and different varieties of the dominant or majority language. For instance, a child born in France to a Tunisian-origin family that has recently immigrated will be in contact with the variety of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

514

J E A N - P I E R R E C H E V R OT A N D A N N A G H I M E N TO N

Arabic spoken in Tunisia within the family, with other varieties of Arabic spoken in the community, and with both regional and standard varieties of French. If we add to this that children may use nonnative features of the dominant language as a discourse strategy (for example, Arabic-accented English to signal convergence with or divergence from the interlocutor, Khattab, 2013; see De Houwer, Chapter 17, this volume, for discussion of convergence in bilingual interaction), this second type of situation highlights the fact that dialectal variation may be pervasive in children’s language environments and production. The third typical context concerns teenagers and adult learners who are in contact with different varieties of an L2. These varieties may include social and regional dialects spoken by native speakers of the L2. This configuration occurs when teenagers and adults move to a country where they start or continue to learn an L2. Migration and study abroad are two examples of this situation. However, even when learning an L2 in the classroom in their native country, learners are in contact with different varieties of this language, including the teacher’s nonnative variety and social and regional varieties encountered through media. The fourth typical context results from the lifelong possibility of learning a second or a third dialect of the same language. The sequential acquisition of several dialects may occur in the case of geographical (moving to another region) or social (change in profession or in the course of study) mobility, or in more specific settings. For instance, in periods of major political change, as was the case with the fall of the Berlin Wall, the opening up of borders that earlier could not be traversed has allowed speakers to be exposed to new dialects (Auer, Barden, & Grosskopf, 2000). The diversity manifested by these four typical dialect learning contexts suggests that exposure to and acquisition of several dialects is common. While the first two bidialectal learning contexts involve children, the third one mainly concerns adolescents and adults. The fourth context encompasses all life stages following the primary socialization stage.

4 Young Childhood to Preadolescence 4.1 Language Attitudes in Children and Preadolescents In this section we focus on the ways children perceive and socially categorize varieties: we consider how children develop language attitudes and to what extent adults’ attitudes and ideologies exert an influence on children’s attitudes and categorization of dialects. Categorization is a cognitive process that is intrinsically related to perceptions, attitudes, and language ideologies. Both language attitudes and ideologies draw links between language (and dialect) structure or usage and groups as well as

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

515

personal identity (Woolard & Schieffelin, 1994; see Fuller, Chapter 6, this volume, for related discussion). When children are confronted with a variable language environment they will develop language attitudes and ideologies concerning the varieties to which they are exposed. Numerous studies have been designed to tap into children’s attitudes toward dialects of the same language (Kinzler & DeJesus, 2013), foreign-accented realizations of their home language (Floccia, Butler, Girard, & Goslin, 2009), or attitudes toward the languages used in the children’s environment that differ in prestige (Cremona & Bates, 1977). The adults’ attitudes in these contexts provide children with cues regarding the types of uses that are perceived as appropriate. Family language practices and policies often express attitudes favoring the most prestigious language or dialect (e.g., in the European context, see Ghimenton, 2015; see Gonc¸alves, Chapter 3, this volume, for discussion of family language practices and policies). The attitudes that children develop are tightly linked to their ability to distinguish the varieties involved and to exploit indexical information that these carry. Although metacognitive skills emerge before children receive formal education (De Houwer, 2017a), the appearance of meta-pragmatic awareness of sociolinguistic variables in children who are exposed to very similar dialects does not seem to occur before the age of five to six (Barbu, Nardy, Chevrot, & Juhel, 2013). It has been suggested that five- to six-yearolds have a gradient representation of dialect variation as they are able to distinguish dialects when their home variety is contrasted with another variety and not, for example, when they are presented with two unfamiliar dialect varieties (Wagner, Clopper, & Pate, 2013). Kinzler and DeJesus (2013) compared attitudes toward regional accents in the United States among two age groups. The first comprised children aged 5 to 6; the second group consisted of children aged 9 to 10. In each age group, children either came from Illinois (where a northern, prestigious variety of American English is spoken) or Tennessee (where a southern, less prestigious variety of American English is spoken). They found that the younger children showed either a preference for their own dialect or no preference toward any particular dialect. For the older children, however, both groups (northern and southern) viewed the northern and more prestigious variety in a more favorable light. These results suggest that children progressively align their attitudes toward dialects with the mainstream adult preferences (Niedzielski & Preston, 2000) and build categories in function of the perceived prestige of dialectal variation (see De Vogelaer & Toye, 2017, for further discussion). From a developmental perspective, children as young as five are thus able to express sociolinguistic preferences for specific dialectal varieties. As children grow older, their attitudes are aligned with the adults’ categorizations based on dialects’ socio-indexical values, in particular when it comes to perceived prestige of one variety as opposed to another.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

516

J E A N - P I E R R E C H E V R OT A N D A N N A G H I M E N TO N

4.2 Production in Children and Preadolescents In this section, we review some factors that influence and shape children’s and preadolescents’ production of dialectal varieties. Starting from two years of age, children produce the regional variants to which they are exposed through their caregivers’ speech (Ghimenton, Chevrot, & Billiez, 2013). The use of regional variants seems to peak in preadolescence and when children speak to their local friends from the same region (Barbu, Martin, & Chevrot, 2014). Studies that have explored the impact of family socioeconomic background on children’s production of sociolinguistic variables tend toward the same result: the higher the family background’s position on the social scale, the more standard variants children produce. This effect is both early and robust. It appears already at age three and seems to increase along with development. With regards to style, that is, the ability to adapt the usage of dialectal varieties according to the social context, the results of the available studies also converge. Similarly to what has been found in adults, children select more standard variants when participating in a formal interactional context. This skill occurs at an early stage, as children are capable of selecting variants according to the type of interaction with caregivers from the age of three (Smith, Durham, & Richards, 2013; see De Houwer, Chapter 17, this volume, for discussion of similar interlocutor-determined language use choices in bilingual children). Labov (1990) showed that in adult populations there is a gender-based difference with regard to language variation, in that women generally use more standard variants than men. For children, this gender effect is not so clear. Some studies found no gender effect for the use of variants in children aged 2 to 9 (e.g., Ingram, Pittam, & Newman, 1985). Others revealed that between the ages of 3 and 10, girls produce more standard variants than boys (Romaine, 1984), but then there are studies showing that it is boys rather than girls who use more standard variants (Roberts, 1997). In the rural Haute-Savoie region of France, 10- and 11-year-old boys born in the area used more regional variants than girls, but only when they were talking to long-term male friends from the same area (Barbu et al., 2014). Studies linking ethnicity and dialect acquisition have focused primarily on the acquisition of African American Vernacular English (AAVE) (e.g., Green, 2010). The acquisition of AAVE is an acquisition process per se rather than an imperfect attempt to acquire standard English, as some may think. Like children everywhere, children exposed to AAVE learn the specific variable cues of the language they are hearing, in this case, AAVE, from the specific input they are exposed to in their communities (see ArmonLotem & Meir, Chapter 10, this volume, for similar findings in early bilingualism). Although the trajectories of AAVE usage from childhood to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

517

adolescence are diverse, just as most language development trajectories are, AAVE usage generally decreases during the early years of schooling, at which time children’s usage of mainstream English increases. This crossing pattern has been attested for other standard, mainstream varieties that are acquired along with nonstandard varieties (Nardy, Chevrot, & Barbu, 2013), and it is reminiscent of what happens in bilingual children when a minority language is acquired alongside a majority language (see Armon-Lotem & Meir, Chapter 10, this volume). Few studies have examined the impact of children’s social network on their usage of standard and nonstandard varieties. However, the results from the existing research point in the same direction. Like in adults and adolescents, children who are well integrated in their peer group and who have a dense social network are the most likely to use more nonstandard variants. This effect has been found from ages four to five (Nardy, Chevrot, & Barbu, 2014). The foregoing brief review highlights the fact that children acquire dialectal variants from early on. Moreover, sensitivity to the same factors that influence adult production of varieties appears at an early stage in children’s life. The process of dialectal acquisition is not parallel to or lagged behind the acquisition of one or more languages. Rather, it is an inherent part of it. However, the diverging results in studies focused on the influence of child gender on the use of dialectal varieties leads one to reconsider certain aspects in this conclusion. It is likely that the age of acquisition of sociolinguistic patterns depends on the perceptual salience of sociolinguistic variables and on their social salience within a given community. The usage of certain dialectal traits may be transmitted at an early stage in development within the family language environment, whereas others may, in contrast, be subject to a longer socialization or identity construction process.

5 Adolescence to Adulthood 5.1 Language Attitudes in Adolescents and Adults Adolescents’ and adults’ attitudes toward language variation have been studied mainly in the fields of folk linguistics and L2 acquisition. This large body of research reveals that adolescence generally marks a gradual alignment toward adult language attitudes. Folk linguistics studies have tapped language attitudes by looking at how speakers index dialects and their speakers. These studies reveal that language attitudes vary along different dimensions. Niedzielski and Preston (2000) provide a comprehensive account of this type of research conducted in the United States. For example, among the numerous attitudes that emerge in social psychological perception studies, mainstream varieties of English (represented by the northern American English

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

518

J E A N - P I E R R E C H E V R OT A N D A N N A G H I M E N TO N

varieties) index social mobility, whereas the southern varieties index friendliness. Language attitudes have also been investigated through anthropologically oriented ethnographic approaches. For example, fieldwork from Cavanaugh’s (2013) ethnographic studies in Bergamo (northern Italy) revealed that local speakers perceive their regional dialect as carrying traits of masculinity and low education. Similarly to what was found in Preston’s work on language attitudes in the United States (e.g., Niedzielski & Preston, 2000), however, regional varieties are not systematically perceived along a negative dimension. They are also viewed as linguistic resources that can best reflect a speaker’s emotional states (sentiments of affect, for example) as opposed to the stylistic options the standard offers. This echoes findings regarding the role of emotions as related to specific languages in bilinguals (e.g., Pavlenko, 2005). Some L2 acquisition studies have found that L2 learners have a distinct preference for the standard, mainstream variety. For example, Fox and McGory (2007) showed that Japanese learners of English as an L2 exhibited negative attitudes toward southern varieties of American English and preferred the northern (mainstream) American English accent. In addition, L2 variation impacts L2 learners’ attitudes and also their comprehension abilities (Gold, 2015). The perceived negative influence of dialects on comprehension notwithstanding, it has also been shown that exposure to greater dialectal variety in L2 acquisition increases learners’ capacities in perceiving and producing vowels in German and French (Smith & Baker, 2011), as well as in Spanish (Schmidt, 2009). Language attitudes reveal fragments of speaker ideologies that seem to exert a certain influence on both perception and production. The same language ideologies exist in different age groups and shape speakers’ attitudes. Thus, exposure to a language means gaining an attitude toward that specific language and its different varieties. In this perspective-taking process, speakers will develop implicit measures of desirability where some languages or varieties are viewed in a more favorable light than others. It is striking that findings on language attitudes in children, adolescents, adults, and L2 learners point to the same direction: levels of desirability are generally higher toward mainstream varieties. This suggests that language ideologies impact L2 learners in the same way that they do in L1 acquisition. Although expressions of language attitudes do not predict language choices in situations of language contact (Ghimenton, 2015), they allow the tapping of ideologies and categorizations that are collectively constructed (see Fuller, Chapter 6, this volume).

5.2 Production in Adolescents and Adults Dialectal practices learned at an early stage in life can be modified at a later stage, during adolescence or adulthood. For example, some abilities acquired at an early age such as the ability to select variants according to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

519

the interaction setting with caregivers (e.g., teaching and play versus routine and discipline, Smith et al., 2013) could be linked to new indexical values later on, when adolescent or young adults discover the link between variants and social groups (Labov, 2013). But adolescents and adults may also learn new dialectal varieties. This appears generally in two contexts, namely the third and fourth typical contexts outlined in Section 3. First, teen and adult learners may come into contact with different varieties of an L2, including different registers and stylistic variants. They thus have to learn the patterns of sociolinguistic variation implemented by speakers of the L2. Numerous studies of learners of French as an L2 have tried to shed light on the influence of learners’ gender on their usage of French. For example, for some variables but not all, Regan, Howard, and Leme´e (2009) found a tendency on the part of their female participants toward more standard, normative usage. Other studies did not find any gender difference (Dewaele, 2004). Dewaele (2002b) explains the lack of gender-related differences in L2 French usage between male and female learners as resulting from learners’ limited knowledge of socio-indexical values. It should be highlighted that the same difficulty in observing a clear pattern between dialectal usage and gender has been found in children acquiring their first language (see Section 4.2 earlier). This convergence suggests that the construction of gendered speech is a long and winding road, in infancy as well as in adulthood. Studies addressing the impact of learners’ SES on L2 sociolinguistic usage have found that higher-SES adult learners tend to use standard, normative variants of the L2 more frequently compared to lower-SES speakers and to speakers who speak the L2 as a first language (Regan et al., 2009; Rehner & Mougeon, 1999). As Rehner and Mougeon (1999) have shown, L2 learners may not be able (or willing) to switch to less normatively standard, more informal variants in response to changes in the formality of the speech situation (higher-SES learners even less so than lower-SES learners). However, as the authors suggest, opportunities for natural interactions with L1 speakers may play a role in increasing learners’ ability to flexibly adjust stylistic variants to the speech situation (see also Dewaele, 2004). Indeed, everyday interactions with native speakers are likely the key for learners to gain access to more flexible production of L2 varieties. Based on evidence collected during an interview with the same researcher, L2 French learners in a study abroad context whose social network included native L1 speakers of the host country (France) increased the frequency of use of nonstandard variants common in informal interactions during their stay (such as the deletion of ne in French negatives, and the omission of standard liaison), whereas learners of French whose networks were made up of speakers of their own L1 increased the use of standard variants (Gautier & Chevrot, 2015). As Howard, Mougeon, and Dewaele (2013) emphasize, the best acquisition

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

520

J E A N - P I E R R E C H E V R OT A N D A N N A G H I M E N TO N

context for informal varieties is naturalistic immersion during primary socialization, followed by naturalistic immersion in study abroad contexts. Immersion classroom contexts are less conducive to acquiring informal varieties, and least conducive are classroom contexts under drip-feed, limited-input conditions. (For discussion of the latter two contexts, see Juan-Garau & Lyster, Chapter 11, this volume, and Mun˜oz & Spada, Chapter 12, this volume, respectively.) There is more and more research about the processes of acquisition of L2 sociolinguistic patterns. As reviewed in Ortega (Chapter 21, this volume), the evidence across several languages suggests that study abroad fosters quite a bit of awareness of the indexical meanings of different L2 varieties, the learning of L2 variable features, and the development of either negative or positive social attitudes. However, this body of research is still smaller than the research on the acquisition of L1 sociolinguistic variation. There also is relevant research on entirely different populations, such as migrants who are in everyday contact with dialectal varieties of the language of the host country that they did not know prior to immigration. Extensive L2 exposure and the diversity of social settings experienced by Polish-born migrant teenagers in London would explain them acquiring the linguistic constraints of a phonetic variable such as /t/ glottaling after only two years in England, and being able to use the variable as a stylistic resource within the same period (Schleef, 2017). But even after decades of residence in Switzerland, Canadian and English migrants still exhibit different patterns of use of standard German and nonstandard Swiss Alemannic dialects that can be explained through their social experiences, their social expectations about the host community, and the linguistic complexity of the variants (Ender, 2017). The diversity of contexts of acquisition, the description of input, and issues related to identity are among the numerous areas to be explored to better understand the acquisition of sociolinguistic variation in an L2. This exploration should also consider how learners transfer sociolinguistic patterns acquired in their first language. Finally, an important issue concerns the extent to which the categories of sociolinguistic perception and their socio-indexical values (e.g., masculinity, educated, sociability) are the same across different languages and cultures. Finally, we turn to second dialect acquisition, that is, on the acquisition of new dialects of the same language. The field of second dialect acquisition is developing rapidly and is becoming an independent area of research that contributes to major theoretical debates in the language sciences (Nycz, 2013). Our discussion here focuses on adults and adolescents, even if second dialect acquisition concerns children too and has important educational implications. Children who master nonstandard varieties find themselves at a disadvantage in schools where the standard variety is promoted (for related discussion of children

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

521

acquiring different languages, see Garcı´a & Tupas, Chapter 20, this volume). Based on comparisons of the acquisition of different languages and the acquisition of different dialects of the same language, Siegel (2010) set out clear milestones for second dialect acquisition. Noting that second dialect learners did not have much success, Siegel explores the reasons why the acquisition of a second dialect is difficult compared to the acquisition of an L2. The specificity of second dialect acquisition is the great systemic proximity between dialect 1 (D1) and dialect 2 (D2). When they start to learn a second dialect, learners already know the linguistic units and constructions that form the basis of the variety. Given this structural proximity, they struggle to distinguish the two systems and to notice the differences separating them. Consequently, transfers from one dialect to another are numerous. They are even more frequent if the two varieties are perceived as similar (regardless of their structural distance as determined by linguists). Given that often D1 speakers understand D2 speakers, one may wonder why it is necessary to acquire a D2. A D2 is often acquired due to “the desire to be a part of the new community and be viewed as a local” (Siegel, 2010, p. 138). Thus D2 acquirers target social goals that go beyond the ability to communicate. Since dialects are bound with specific identities (e.g., region, class, ethnicity, gender), trying to speak another dialect than our own may trigger negative reactions from speakers of the D2 as well as from speakers of the D1. Some reasons for those negative attitudes include misinterpretation of a learner’s attempt to speak the D2 (e.g., it could be interpreted as making fun of it), folk essentializations of identity and authentic self, and feelings that a dialect belongs to a specific community. Siegel sheds light on the different factors that influence the acquisition of an L2 and the acquisition of a second dialect, such as age of acquisition and age of arrival in the area where the new variety is spoken. A close examination of the available studies reveals two trends. First, the optimal age for acquiring a second dialect is the mid-teens or younger for the lexicon and morphology and seven years and under for phonology. Second, critical factors accounting for differences in second dialect achievement are age, social interaction, and how the learner identifies with the host community.

6 Conclusion Languages are not stable systems: variation is an inherent property of their structure and their usage. A conception of the language environment inspired by sociolinguistic research would consider that, from the very beginning of and throughout their lives, children and adults are confronted with variation, whether they grow up in a monolingual, a

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

522

J E A N - P I E R R E C H E V R OT A N D A N N A G H I M E N TO N

bidialectal, or a bilingual environment. From an early age, learners encounter various linguistic forms that compete for the same referential meaning but different socio-indexical values. Models of bilingual development that have attempted to show the gradual strengthening of form and function pairings within each language, that is, their internal resonance (e.g., Hernandez, Li, & MacWhinney, 2005) are thus applicable to bidialectal contexts too. Modeling the emergence of bidialectal abilities is possible on the basis of the similarities in the processes involved in learning two languages or two dialects. Hence, if two varieties are acquired early, learners have every opportunity to rapidly develop processes of internal resonance linking together units and constructions specific to each language variety. If one variety is learned earlier than another, then its forms will be more entrenched than those of a second variety learned at a later stage. Consequently, the forms of the variety learned first may inhibit the forms of the second variety. Should two varieties be learned at a later stage, as in the case of the acquisition of dialects of a second language, the variety learned first may compromise the process of internal resonance of these varieties. Processes involved in acquiring dialectal variation in a second language and acquiring a second dialect of the same language are two faces of the same coin: both processes involve achieving social goals and adapting social identity to a new context. Interaction with native speakers of the L2 or the D2 is a strong factor that helps to achieve these goals. The interdisciplinary encounter between dialectal acquisition and the broader field of language acquisition has an important heuristic value (Ortega, 2013), fertilizing via sociolinguistic issues a field that is predominantly explored within a psycholinguistic tradition. The study of dialect acquisition shows a major gap in traditional psycholinguistic approaches to language acquisition. That is, there is a general lack of consideration of socio-indexical information in models of language learning and processing. Yet, as has been reported in this chapter, for children, adolescents, and adults alike, such information plays an essential role in both attitude formation and language production. Socio-indexical information can either reinforce or constrain internal resonance of a dialect (or a language). Working toward an integrated model including bilingual and bidialectal contexts will only be possible if both sociolinguistic research and studies concerning bilingual first and second language acquisition are brought together. Taking an even more global and encompassing view of language learning and processing, one could see monolingualism, bidialectalism, and bilingualism as being on the same continuum but with differing levels of differentiation in the linguistic usages speakers are exposed to. As mentioned earlier, both children and adults are confronted with heterogeneous language environments. Although there is a consensus regarding the inherent heterogeneity of language, a more complex issue is to decide whether for each specific sociolinguistic setting this heterogeneity results

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

Bilingualism and Bidialectalism

523

from the contact between several languages, from the contact between several dialects or varieties of the same language, or from variations within a single linguistic system. The solution to this problematic issue does not only depend on linguistic criteria concerning systemic distance or inter-comprehension, but also on different ideological stances. Gaining the status of variety, dialect, or language for a set of more or less convergent linguistic usages depends on their social recognition. If one were to reserve the term dialect or variety for the configurations of usages in relation to specific groups of speakers, for instance, those from a particular region or from a particular ethnic group, then bidialectal acquisition would only concern certain groups of children and adults. If, on the contrary, one agrees to designate as dialect or variety all convergences in linguistic uses noticeable within a social group (e.g., upper-class dialect) or within a context (e.g., informal variety), then bidialectal acquisition would be the norm and not the exception. Children or adults who acquire a language receive exposure to concrete samples of speech, produced in diverse, socially situated contexts by actual speakers, representing diverse social groups based on gender, age, SES, etc. If we accept the broad definition of dialect or variety, the acquisition of one or more languages is systematically achieved through the contact with different varieties of these languages. No matter what definition is adopted, an adult or a child learner should acquire the ability to move from one style to another, from one variety to another, from one language to another, according to similar social constraints. Because this ability to handle variation is profoundly rooted both in language and social cognition, the study of its acquisition throughout life is particularly promising for revealing the depth of the relations between knowledge of language and knowledge of the social world.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.026

26 Bilingualism and Language Contact Suzanne Aalberse and Pieter Muysken

1 Introduction This chapter discusses bilingualism from the perspective of language contact studies. We could consider language contact the umbrella term for all situations where languages come into contact, either through interaction between groups of speakers who have diverse backgrounds or because elements of more than one language come together in an individual speaker. That latter case would refer to bilinguals. Although in principle all parts of bilingualism fall under language contact studies, we focus on work that is more explicitly placed in the domain of language contact, that is, the study of the consequences of bilingualism for language use and structure, rather than on individual bilingualism itself. The term language contact is thus used here to refer to the effects that speakers’ bilingual practices have on the languages they use. These effects may involve any part of the language system: phonetics, intonation, and phonology; the words themselves; morphology; syntax; semantics, discourse, and pragmatics. The effects may be temporary or more permanent, entrenched in the language system as it were. Thus, language contact is a rather vague term for a vast area, particularly since bilingual practices are extremely diverse and all aspects of language are involved. While there has been a steady stream of more incidental observations about language contact in earlier periods, starting at least in the 19th century, Weinreich (1953) and Haugen (1953) gave the field a major impulse both conceptually and empirically. Weinreich had multilingualism in Switzerland (French, Italian, Swiss, German) as his empirical base; he discussed several types of bilinguals and different types of results of what he called interference (i.e., mutual influence of the languages) in their language production. Haugen focused on Norwegian as a heritage language in the Norwegian immigrant community in the United States and brought in historical considerations as well. He can be credited as the founder of the field of heritage language studies. He is also known for his

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

525

focus on the ecology of languages, the basic tenet of which is that the environment in which a language is spoken greatly influences its use and subsequent development. It is impossible to even begin to summarize all the work here in the broad field of language contact studies, which has been approached from very different angles. These include psycholinguistics and neurolinguistics, sociolinguistics and anthropological linguistics, as well as historical and comparative linguistics. Winford (2003) offers a broad but detailed introduction to the field, with chapters providing both summaries and detailed commentaries on lexical and structural borrowing, social and structural aspects of code-switching, bilingual mixed languages, second language learning, pidgins, and creoles. In this chapter, we limit ourselves rather drastically in two ways. First, we discuss just a few types of bilingual practices. They are four: code-switching and lexical borrowing; contact-induced language change; mixed languages, pidgins, and creoles; and heritage languages. Second, we limit ourselves to the lexicon and to syntax. The four fields we cover (and then only in a very superficial way) have received considerable scholarly attention, and the study of the lexicon and syntax has been very important for them. We focus on the more system-oriented domains of bilingual practice, and on convergence in bilingual speech, that is, the process by which two languages become structurally more alike, while disregarding important fields such as bilingual acquisition or second language acquisition. Although the focus of the subdisciplines we discuss has been primarily scholarly, there are important societal implications of the research reported on here. One concerns issues of language purity as linked to ethnic identity: in many communities, the influence of other languages, perceived or real, is a matter of concern both to politicians and to ordinary citizens, and the study of language contact can help explain what is actually happening (see Fuller, Chapter 6, this volume, and Lo Bianco, Chapter 8, this volume). Another one concerns the language portrayed in educational materials for migrant communities in schools (see Bigelow & Collins, Chapter 2, this volume; and Garcı´a & Tupas, Chapter 20, this volume) or adult education (see Simpson, Chapter 13, this volume) or groups of traditional heritage speakers. Is it the traditional varieties of language that are used here, or the evolved varieties bilingual speakers use in their daily lives?

2 Code-Switching and Lexical Borrowing Since Poplack’s (1980) study, code-switching (defined as the use of different languages in the same conversation) and lexical borrowing (that is, the bringing of a word from one language into another one) have been the focus of considerable attention across a wide range of linguistic subdisciplines. In her study, Poplack examined code-switching between Spanish

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

526

SUZANNE AALBERSE AND PIETER MUYSKEN

and English in the Puerto Rican community in New York using audio recordings of spontaneous conversations. The patterns of switching she found were relatively symmetrical, that is to say, speakers switched back and forth evenly between the two languages. The principal constraints on the switching that Poplack discovered were that switches generally occurred where the word order of the two languages was the same and did not occur within a word. An additional finding was that only highly proficient bilinguals switched frequently within a clause. There is an ongoing discussion on the status of asymmetrical code-switching in Poplack’s model, specifically as regards switching in which there is a clear base language into which elements are inserted. Poplack argues that these cases are all instances of borrowing or sometimes nonce borrowing (when a word from another language is only introduced once), while other researchers assume there to be insertional code-switching (i.e., putting a fragment from another language into a sentence while codeswitching). Myers-Scotton (1993) based her Matrix Language Frame (MLF) model primarily on earlier work in which she recorded and analyzed Swahili–English code-switching in Kenya. Most of her data show a clear asymmetrical pattern in which one language, often Swahili, functions as the base or matrix language, and elements from another language, generally English, are inserted into this matrix language. The word order, grammatical structure, and functional elements such as articles and verb endings (which Myers-Scotton labels system morphemes) come from the matrix language. The grammatical model that Myers-Scotton developed is part of a broader pragmatic model in which a particular language functions as the unmarked (i.e., expected) code in specific conversations and the inserted elements come from a marked (i.e., unexpected) language. There has been extensive discussion regarding the possibility of capturing all cases of code-switching in asymmetrical models, as the strict version of Myers-Scotton’s MLF model would require. Many researchers question the possibility of assigning a single matrix language to all cases in which codeswitching occurs. The work by Poplack and Myers-Scotton has been the basis for a series of intense debates on the nature of, and possible grammatical constraints on, code-switching within the clause. The term code-switching refers to the juxtaposition of elements that can be seen to belong to different languages, but sometimes code-mixing may seem to be the better term since, when code-switching occurs within a clause, the result appears to be a new mixed language, subject to specific constraints on mixing. The controversy regarding models and conceptualizations involves several theoretical, grammatical frameworks, from Universal Grammar–based minimalism to cognitive grammar. Muysken (2000) attempted to summarize some of the findings in the debate in order to arrive at some sort of consensus on the types of code-switching

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

527

encountered. The basic distinction he introduced is between asymmetrical insertion, in which there is a clear basic or matrix language, and more symmetrical alternation, in which there is a shift or juxtaposition from one basic language to the other. A third type of code-switching he distinguished is congruent lexicalization, where the languages involved are so similar in structure or lexicon or both that there is neither insertion nor alternation, but rather a rapid back-and-forth between the two languages. While some researchers have recognized the validity of this three-way distinction, it has proven difficult to unambiguously classify all instances of code-switching in terms of these three types. Several studies have tried to put the process of code-switching in a more dynamic perspective, focusing on changes across different generations. Based on Li Wei’s fieldwork on cross-generational change in language choice in families in the Chinese Tyneside (Newcastle, United Kingdom) community, Milroy and Li Wei (1995) proposed a dynamic network model to explain cross-generational shifts in code-switching strategies. In a similar vein, Backus (1996) explored patterns of code-switching in the Turkish-Dutch immigrant community in the Netherlands. Backus shows that first-generation speakers who arrived as adults produce simple insertions of basic culture-specific content words, while the insertions are more complex in the group that arrived from Turkey as a child; the second and third generations show more alternation patterns with a clearer separation of the two languages. A typical insertional pattern from Backus’s work (1998, p. 9) can be seen in example (1) (for explanation of the abbreviations in the glosses, see the Appendix): (1)

iki gu¨n once is¸te bioscoop-a, vragen yap-tιydι-m do-P L U P F - 1 S G two day before interj cinema-D A T ask “but two days before I had asked her along to the movies” (Dutch elements are in italics, Turkish elements are not)

All work on code-switching mentioned so far has taken a sociolinguistic perspective involving recordings in specific bilingual communities and taking variation and speaker characteristics into account. A few efforts have been made to interpret code-switching patterns from a more formal perspective, such as minimalism in generative grammar. A typical example is MacSwan’s (1999) approach, which was originally based on Nahuatl–Spanish code-switching data from central Mexico but takes a much broader perspective. In the minimalist approach, the goal is not to posit grammatical constraints specific to code-switching (such as “do not switch between a pronoun and a verb”) but rather to let all constraints on the process of code-switching inside the clause follow from universal properties of our grammatical competence. Undoubtedly, this is a valuable research strategy. The main problem with it so far is that it postulates a priori that the underlying categorical makeup of different languages is identical (e.g., gender features in one language correspond to gender

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

528

SUZANNE AALBERSE AND PIETER MUYSKEN

features in another language). This assumption is put in doubt in much typological research. In recent years, there has been an increase in psycholinguistic and computational modeling work on code-switching, and it is likely this will increase as new methods and more ecologically valid experimental designs to elicit code-switching become available.

3 Contact Languages The field of contact linguistics looks at contact languages, that is, languages that emerge out of the meeting of different languages in a bilingual population. The field is too vast to cover here in its entirety. The main topics under discussion in contact linguistics are the role of first and second language acquisition for language contact, the role of language typology and language universals, processes of language simplification, complexification, and grammaticalization (i.e., the emergence of new grammatical categories such as “tense” and “number”), and language change. It is difficult to point to a single debate in this area. An overview of the whole field is the introduction by Velupillai (2015). Contact linguists distinguish between new languages that are spoken as second languages for communication between speakers of different language groups, known as pidgins, languages spoken as community languages that have emerged out of earlier contact, known as creoles, and new languages that contain words or morphemes that originate from different language sources, known as mixed languages. There is much debate about the precise definitions and delineation of these terms, however. One point of disagreement, for instance, is whether it is feasible to define creole languages in terms of their typological features, or whether we need to define them solely in terms of the circumstances of their genesis. In the former approach, one would say, for example, that creole languages usually have basic subject-verb-object word-order, have a consonant-vowel syllable structure, etcetera. By contrast, an approach focusing on origins makes no such claims but simply defines creoles as contact languages that became stabilized in their grammar due to acquisition by children and their codification as community languages (see Chevrot & Ghimenton, Chapter 25, this volume, for related discussion of the difficulties involved in distinguishing between “language” and “dialect”). A related question is whether a well-defined class of creole languages exists, or whether the very notion of creole fractures into several dimensions that may or may not be jointly present in individual languages. Whatever the answer, diverse types of language-mixing phenomena can be found, some of which are briefly illustrated here.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

529

Muysken (1981) introduced the notion of relexification, defined as the whole replacement of the lexical shapes of words by those of another language (e.g., Spanish salt-ar “jump” becoming jump-ar, incorporating now an English word form). He based it on rather unusual data from the mixed language of a community in Ecuador often referred to as Media Lengua. In this language, the traditional Amerindian language Quichua (Ecuadorian Quechua) is spoken alongside Spanish. However, for several generations the in-group vernacular was a form of Quichua in which the phonological shape of all roots was replaced by a Spanish shape while most of the Quichua lexical semantics, grammar, morphology, and phonetics was maintained. All word roots are thus Spanish (although pronounced as in Quichua), but the suffixes are almost all Quichua. A typical example is (2) from Muysken’s unpublished fieldwork data, where the underlined elements are of Spanish lexical origin – no “no,” undi (< (d)onde) “where,” i- (< ir) “go,” and kiri- (< querer) “want”: (2)

[[no [undi-mun i-na] kiri-k]-kuna-lla] neg where-dir go-nmlz want-ag-pl-del “those that do not want to go anywhere”

Notice that the sentence in (2) is a far from simple structure, containing a double embedding, following Quichua grammar. Mixed languages are not necessarily simple languages. Bakker (1997) wrote a detailed description of another mixed language, Michif, which involves the Algonquian language Cree as well as French. Michif probably emerged in the 19th century during the fur trade era in eastern Canada, but is currently spoken in some communities further to the west in Canada and in adjacent parts of the United States. In contrast to Media Lengua, where mixing mostly involves the phonological part of the lexical entry, in Michif, the basic split is between the verbal and the clausal systems, which are Cree in origin, and the nominal system, which is French. As shown in example (3) from Bakker (1997, p. 6), sentences are thus split between clearly French (nominal) parts (in italics here) and clearly Cree parts: (3)

eˆkwa paˆstin-am and open-he.it

eˆ-wıˆ-otin-aˆt C O M P -wanttake-he.him “and he opened his mouth and the wolf wanted to take him” sa his.F

bouche mouth

oˆhi this-O B V

le the.M

loup wolf

Every word in Michif is either Cree or French, and the full noun phrase is ordinarily French in origin. The Michif pattern is reminiscent of situations of intra-sentential insertional code-switching in which entire French noun phrases may be inserted, but the origin of Michif as the result of bilingual practices and insertional code-switching is controversial. Thomason’s (1997) book describes a number of the most important contact languages, including Hiri Motu from Papua New Guinea, Pidgin Delaware from the East Coast of the United States, Ndyuka-Trio pidgin

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

530

SUZANNE AALBERSE AND PIETER MUYSKEN

from Surinam, pidgin and creole languages with Arabic lexicons, Sango from the Central African Republic, forms of Swahili that have undergone restructuring as they were learned by populations of second language learners, Michif, Media Lengua, Callahuaya from Bolivia, Medniy Aleut from Siberia, and Ma’a (Mbugu) from Tanzania. Thomason presents many different typological patterns found among these languages, as well as differences in the processes that led to their genesis. Lefebvre (1998) took up the notion of relexification introduced for Media Lengua and tried to account for the genesis of Haitian Creole. She postulated that the primary mechanism involved in its emergence was the relexification of Fongbe grammatical structures with French word forms. Speakers of the West African Fongbe language used relexification as a type of special early second language learning strategy in unusual circumstances. Since there is no inflectional morphology in Fongbe, relexification in this case would lead to structures that only involve French word material. However, the underlying grammatical organization would be Fongbe. An example is (4), from Lefebvre (1998, p. 197): (4)

Mari te prepare pat HA I T I A N CR E O L E ` Mari ko` Da` wO FO N G B E Mary ANT prepare dough “Mary prepared dough” (in the Haitian Creole sentence, the French word is italicized)

Although many grammatical patterns in Haitian Creole are similar to Fongbe patterns, not all researchers agree that the primary mechanism leading to the genesis of Haitian Creole and related languages was indeed relexification. The collective volume edited by Meakins and O’Shannessy (2016) contains a synthesis of the exciting work that has been done on the way speakers of Aboriginal communities in Australia have reacted to colonization and the spread of English in the continent. Contact between Aboriginal languages and the colonial language of English has led to some very interesting contact languages in Australia, the best known of which are Gurindji Kriol and Light Warlpiri. These languages share the main division that Bakker (1997) encountered for Michif. The nominal frame (including case markers and derivational morphology) is from one language, in this case from one of the traditional languages, Gurindji or Warlpiri (in contrast with Michif, where the nominal frame was from the “new” language, French) and the verbal tense-mood-aspect system from another language, in this case English or Kriol. A third mixed language from Australia is New Tiwi, with again the main division between nominal and verbal frames, but in this case New Tiwi sides with Michif. However, these mixed languages are only part of the story of language contact in Australia, which involves mutual influences between the Aboriginal languages themselves, the emergence of contact languages such as Kriol,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

531

and the interaction between English and Kriol and various Aboriginal languages. Speaking broadly, the overviews of mixed languages in the literature show a considerable range of mixing phenomena in different language pairs, often involving opposition or hierarchy. Different oppositions include those between (1) nominal versus verbal frames, as in the case of Michif, where nominal material is from French, and (2) lexical versus grammatical frames, as in the case of Media Lengua, where lexical material is from Spanish. Mixed languages may emerge as languages of ethnically mixed communities, in communities undergoing rapid cultural changes, or sometimes as languages of interethnic communication. Several factors having to do with power and prestige of the languages involved, the degrees of speakers’ proficiency in the different languages, and structural properties of the languages involved determine which of these oppositions comes into play in a specific mixed language.

4 Contact-Induced Language Change The role of language contact in language change has always been controversial and continues to be so. One of the main reasons is that different time frames produce different results: long-term historical developments often seem to lead to more profound changes than short-term contacts in contemporary bilingual communities. Contact-induced language change is by no means omnipresent, it seems, at least on the structural level. Lexical change, as in borrowing, does seem to be virtually always there. The controversy about language change involves three partly related questions asked and discussed in the sections that follow.

4.1

Can Language Change Be the Result of Contact with Another Language? For this question, the consensus among most researchers is yes. There are well-documented cases of contact-induced language change. However, for many individual cases, this remains a controversial issue. While in earlier periods, resistance to the idea that contact-induced change existed came mostly from structuralist and generative quarters, surprisingly, current criticisms of the notion of contact-induced change come mostly from within the variationist paradigm, which posits that variation within spoken language is central. Numerous examples can be given, such as Poplack and Tagliamonte (2001). The reasoning in this paradigm is generally the following. In the variety that has undergone the change, the variable constraints (i.e., the probabilities for a particular change to occur in a specific grammatical context) on the phenomenon in question resemble those in an ancestral variety. Hence, attributing them to an external factor

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

532

SUZANNE AALBERSE AND PIETER MUYSKEN

such as a different language is not necessary and possibly counterintuitive. A typical example was Labov’s original demonstration (1969) that the variable constraints on the presence of the copula (she strong versus she is strong) in African American Vernacular English (AAVE) resembled those on the contraction of the copula in Standard English (she’s strong versus she is strong). This resemblance made postulating the influence of an earlier creole in the genesis of AAVE less attractive as an explanation. This type of argumentation is particularly strong if the variable constraints involved do not follow from universal properties of human processing (an example would be that the sound s is often deleted after a t or k, since pronouncing the sequence of these consonants is not easy), in which case the resemblance could be accidental.

4.2

What Conditions the Processes of Contact-Induced Language Change? The fuller and more specific version of this question is: are the processes of contact-induced language change, where present, mostly conditioned by structural features of the languages involved, or by sociolinguistic conditions? Thomason and Kaufman (1988) in particular have addressed this issue. They argue that different types of language contact processes or scenarios lead to different outcomes, and they introduced the principal distinction between maintenance and shift contexts. In maintenance contexts, a socially subordinate (minority) language is maintained, but undergoes (predominantly lexical) influence from a dominant (majority) language (for discussion of power relations between languages on a societal level, see Fuller, Chapter 6, this volume, and Lo Bianco, Chapter 8, this volume). In shift contexts, languages are abandoned in favor of the dominant language and leave primarily non-lexical traces in the resulting variety (e.g., Italian intonation patterns in Argentinian Spanish). This type of influence is labeled substrate influence. The different contexts are subject to different hierarchies of what elements are taken over from the other language. Thus, in maintenance contexts, there is a hierarchy of which elements are most easily borrowed, starting with nouns, and ending with elements like verb endings. Van Coetsem (1988) introduced the fundamental distinction between recipient language agentivity (i.e., cases where the language undergoing change is strongest, as with borrowing) and source language agentivity (i.e., where the source language is strongest, which van Coetsem labels interference). This roughly corresponds to Thomason and Kaufman’s (1988) distinction formulated in sociolinguistic terms between maintenance-related and shift-related processes. The central idea in van Coetsem’s work is that in borrowing processes, the principles operating in the recipient language determine the shape and structure of a borrowed item (e.g., a word), while in interference, the principles operating in the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

533

source language determine these properties. An example of interference is the case of substrate influence in the Andes, where semantic and phonetic patterns of the substrate language, Quechua, are transferred to the language to which the population of speakers has shifted, viz., to their version of Andean Spanish. When Quechua speakers adopt a word from Spanish (borrowing), the word is adapted to Quechua phonology and morphology. In van Coetsem’s (1988) model, this is a case of recipient language agentivity. Notice that the notion of agency here is couched in terms of the languages involved, rather than the speakers of those languages, but it could be formulated in terms of speaker agency as well.

4.3

Are There Any Restrictions on Which Parts of a Language Can Be Influenced by Another? The question here, stated more precisely, is: if there can be contactinduced change, are there perhaps certain parts or components of language that cannot be influenced by another language? Several scholars have addressed this issue. Most approaches to language contact make a general distinction between lexical and structural transfer. This distinction is based on the type of linguistic material that is transferred from one language into another (Matras, 2009; Thomason & Kaufman, 1988; Winford, 2003). Matras (2009) distinguishes matter replication, which involves the transfer of a word’s phonological form, from pattern replication, which involves “the organization, distribution and mapping of the grammatical or the semantic meaning while the form itself is not borrowed” (p. 146). Matter replication is often lexical, such as the borrowing of the English word awkward into Dutch, but it can also include the borrowing of stock phrases such as the use of sentences like forget it from English into another language. An example of pattern replication is the use of the combination of words as in bij de weg (“by the way”), which is literally translated from English into Dutch. Although all the words are Dutch, the combination of the words to form the meaning “incidentally” is inspired by English (the literal Dutch meaning here would normally be “near the road”). Johanson (2002) studied developments in the Turkic languages and their interactions with surrounding languages, particularly with those belonging to the group of Indo-Iranian languages. His distinction between transfer of form and transfer of structure resembles the distinction put forward by Matras (2009), but rather than speaking of transfer or replication, Johanson speaks of copying and copies. One characteristic of copies is that they are like the original, but almost never exactly the same. The main distinction Johanson makes is the one between global and selective copies. Global copies are code-switches where all components from one language are copied into an utterance in the other language (thus involving both form and meaning). Selective copies import

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

534

SUZANNE AALBERSE AND PIETER MUYSKEN

parts of one language into the other, but not the complete form-meaning package. Johanson then distinguishes four types of selective copies, namely (1) material, (2) semantic, (3) combinatorial, and (4) frequential copies. A material copy means that the external shape of a language element, the word form, is copied, but not necessarily with the same meaning or combinatorial properties as in the original language. For example, the word beamer is borrowed from English into Dutch and German, but with a very different meaning than its use in English: it is only used to refer to video projectors, whereas in English, the word is mainly used to refer to a kind of car. Semantic copies refer to the use of words in a set of meanings from another language. For example, in Russian, xorosˇo means “well,” as in he did well on his exam. It cannot be used in contexts like this was . . . well . . . embarrassing in Russian, but this use is possible in American Russian as the result of a semantic copy of the English word well. The meanings of well have been copied into American Russian. Combinatorial copies refer to the incorporation of collocations in one language into the other. The Spanish word rascacielos, meaning “skyscraper,” is an example of a combinatorial copy, where the combination of the verb rasca, “scrape(s),” and the noun cielos, “skies,” is copied into Spanish, modeled on the English form skyscraper. Last, a frequential copy refers to the process of a change in frequency of a certain structure as the result of language contact. In Section 5, we discuss the decrease of the use of subjunctive in heritage Spanish, making the language more like English. We also discuss the increase of demonstratives in heritage Chinese in the Netherlands. Both changes in frequency are examples of frequential copies, which make the number of uses more like the dominant language. The Australian linguist Ross, who worked on many languages of the Pacific, both in Australia and in New Guinea, coined the concept of metatypy (Ross, 1996). This refers to a change in grammatical typological profile under the influence of a dominant language, hence the term metatypy, typological metamorphosis. A language that undergoes metatypy is usually a smaller community language used primarily within the community itself, while many speakers are also speakers of a language of wider use, which often has more prestige. Some members of the smaller community use the second language more frequently than their community language, and this leads to changes both in the semantic and pragmatic organization of their first language and in its grammar, at the sentence, clause, phrase, and word levels. The case that triggered Ross’s interest was the Austronesian language Takia in New Guinea, which had become more Papuan-like under the influence of the more widely spoken Papuan language Waskia, adopting some of the word order features and semantic patterns of that language. The influence of the other language must be substantial for the outcome to be labeled metatypy, but no exact criteria

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

535

for calling this type of change metatypy have been proposed (and may indeed not be possible). Thus, historical linguists like Ross are willing to assume profound contact-induced language change, often involving long-term processes of adaptation. In contrast, sociolinguists have not found this type of change in communities with a shorter history of contact. Torres Cacoullos and Travis (2015) worked on New Mexican Spanish, including code-switched speech in the New Mexican Spanish–English bilingual community. They worked on processes of grammaticalization and the linguistic outcomes of language contact in a variationist framework. In contrast with the type of results Ross reported for New Guinea, they have argued that it is very difficult to find convincing evidence for structural influence of English on New Mexican Spanish, even though English is also the dominant language in this case.

5 Heritage Languages from a Contact Perspective This section discusses heritage languages from a contact perspective. We use the term heritage language as defined by the editors, that is, to denote a language that someone learns as a child in the family and within the additional context of a community of speakers (e.g., neighborhood, places of worship), often in competition with the dominant language of the majority society, in which the heritage language speaker may end up being more comfortable and/or proficient. Note that all heritage speakers we discuss are proficient (to some extent) both in the home/community language as well as in the official language(s) of the country and thus are always bilinguals. Heritage languages have already been discussed from an acquisition perspective in other chapters in this volume (see in particular Singleton & Pfenninger, Chapter 4, and Keijzer & de Bot, Chapter 14). This chapter focuses on heritage languages from a contact perspective. How would one recognize a language contact framework as opposed to, for example, an acquisition framework? Although the boundaries between the two approaches are not fixed, we identify some factors that play a role in differentiating the two discussed in what follows. We illustrate our points with examples from two edited volumes on Germanic heritage languages in North America that are closely connected to the contact perspective (Johannessen & Salmons, 2015, and Page & Putnam, 2015). One factor that is characteristic of the contact approach to heritage languages is time depth. Language contact studies can involve very long periods of language contact, even as long as a few thousand years. Heritage language studies from a contact perspective could (but do not have to) involve the study of language use in multiple generations of speakers across time. The volume edited by Page and Putnam (2015), for example,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

536

SUZANNE AALBERSE AND PIETER MUYSKEN

includes a study on Volga German in Ellis County, Kansas, in the United States. Volga German is the kind of German spoken by German Catholics who settled along the Volga River in Russia in 1760 and who bought land from the Kansas Pacific Railroad in 1875 to resettle in Ellis and neighboring counties in west central Kansas. During their periods both in Russia and in Kansas, the community kept speaking German. Another study included in the same volume compares heritage Norwegian at various points in time, namely in 1940, 1990, and 2010. Effects of language contact over such longer periods of time and comparisons of the contact languages in different periods are typical of the contact perspective. The time depth issue has an effect on the type of linguistic domains that are studied. Whereas acquisition studies tend to focus more on what is not acquired or lost, language contact studies could also look at what additions language contact has brought to the new variant. For example, Volga German as described earlier has developed a new prepositional case. The volume edited by Johannessen and Salmons (2015) includes a study on the use of English modals that connects this use with hierarchies that can be borrowed from contact languages, showing that transfer of modals is common in contact situations. The same volume includes a study showing that heritage German has lost rounded vowels and that connects loss of rounding to more general patterns of change in historical linguistics. Yet another study shows that Pennsylvania Dutch, a variant of German in the United States, has extended the use of the progressive. All these forms of variation are described from a language change perspective, connecting them to other forms of language change, rather than from the perspective of the lack of acquisition. The time depth issue also influences the methods applied. Information on the language use of eight generations ago, for example, cannot be based on experiments, since the speakers are no longer alive. Data might come from diaries, grammatical descriptions, recordings of speech from earlier time periods, or letters. Speakers who are still alive and know how to speak the language are usually quite old. The implications of working with older speakers are also described in Keijzer and de Bot (Chapter 14, this volume). Another possible difference between the contact perspective and the acquisition perspective is attention to variation and stylization. Rather than focusing on learnability issues, authors with a contact perspective look at the strategic use of different languages. One could look at the role of identity marking in the selection of a particular language and one could study why and in what context the other language is used or why languages are mixed (see De Houwer, Chapter 17, this volume, for discussion of some of these issues). In Section 4, we discussed the use of literal translations from one language into the other. These are known as calques. Speakers can use calques subconsciously or more consciously as a creative form of language play that can mark group identity and creativity. For example, Dutch writer and comedian Arjen Lubach uses the Dutch

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

537

expression wees daar of wees vierkant (“be there or be square”), which is a literal translation from English, but the word square is translated into just one of its English meanings (not the one intended in “be there or be square”). The Dutch translation does not rhyme and sounds odd (comparable to “be there or be rectangular”), but is funny to English–Dutch bilingual speakers who realize it is a direct translation from English. Purposely deviating from the norm to create an ideal self is also discussed by Singleton and Pfenninger (Chapter 4, this volume) in connection with research by Nance (2015) showing how “new speakers” of Scottish Gaelic do not make the traditional velarized/palatalized rhotic distinction, because the preferred ideal self is oriented toward a new speaker mode and the traditional native speaker target of older speakers is considered inauthentic. The time depth issue also affects how linguistic variation is regarded. Studies from a contact perspective will consider variations in forms as the result of language change and as natural reactions to the changing linguistic ecology, rather than consider them errors. Relevant to the linguistic ecology is the fact that in many heritage communities, speakers with different dialect backgrounds come together (see Chevrot & Ghimenton, Chapter 25, this volume, for discussion of dialectal variation). This yields the formation of koines, that is, new variants of a language that arise when mutually intelligible variants of one language come together. Characteristic of koine formation is that features (in the lexicon, morphology, etc.) that are specific to one dialect are lost or replaced. The volume edited by Johannessen and Salmons (2015) includes a study on the effects of koineization in heritage Swedish and argues that the use of some English words in American Swedish is motivated by the need to avoid dialect-specific lexical forms. Language contact studies focus on contact. This implies that both languages under consideration are studied. Whereas second and heritage language acquisition studies may just focus on what parts of language Y are difficult for speakers independently of their other language(s), language contact studies always look at both languages. Looking at both languages is not unique to the contact perspective, however: many acquisition studies do it too, but a contact perspective must involve both languages, whereas this is not obligatory for acquisition studies (although see Treffers-Daller, Chapter 15, this volume, and Hammer & Edmonds, Chapter 19, this volume, for arguments that bilingual abilities must always be measured and tested bilingually). Because language contact can imply the birth of a new contact language, the community perspective is also more popular. How do multiple speakers with the same bilingual background make use of elements of different languages? Having discussed relevant characteristics of the contact perspective to heritage languages, we now move to two more detailed case studies of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

538

SUZANNE AALBERSE AND PIETER MUYSKEN

contact-induced change in heritage speakers. A pioneer in this domain is Silva-Corvala´n (e.g., 1994), who has published extensively on heritage languages from a contact perspective, basing herself on Spanish in Los Angeles as a form of Spanish–English language contact. An important notion she expressed in her work is the notion of indirect change or change in frequency. The fact that she stressed the notion of change is typical of a contact perspective. The idea behind the notion of indirect change is that language contact does not bring about totally new structures that were alien to the language before, but that it can cause an increase or a decrease in the use of a certain structure, just as discussed in Section 4 with regard to the notion of frequential copying. Silva-Corvala´n (1994, p. 133) argued that the existence of superficially parallel structures in the languages in contact makes those structures of the language permeable to change. Here we discuss two types of change in frequency from her work that are due to superficial overlap between structures. The first type concerns the extension of the discoursepragmatic functions of an element to parallel the functions of the translational equivalent in the other language. For example, English can use the proximal demonstratives this and these to introduce indefinite specific referents in order to foreground that referent, as in yesterday I met this guy . . . Non-heritage Spanish does not use deictics to refer to new entities in the discourse position in subject position, but example (5) from SilvaCorvala´n (1994, p. 136) shows how some heritage speakers do make use of deictics this way: (5)

pero esta mula tenı´a la la idea que si no comı´a but this mule holds the the idea that if not eat algo en la casa . . . something in the house “but this mule had the idea that if he did not eat something at the house . . . ”

The use of esta (“this”) in (5) is grammatical in non-heritage Spanish too, but in non-heritage Spanish it would only be used for already mentioned referents, while in heritage Spanish the possibilities of its use are extended to refer to new referents in the discourse. A second type of indirect change that Silva-Corvala´n (1994) mentions is the increased use of a structure that parallels a structure in the other language. For example, both non-heritage Spanish and English allow subordinate clauses with an overt complementizer (such as in the professor thinks that she . . .) as well as with null complementizers (such as the professor thinks Ø she . . .). Null complementizers are much more frequent in English than in Spanish. Silva-Corvala´n shows that the earlier Spanish speakers come into contact with English, the more likely they are to use null complementizers in a conversation. She compared the use of null complementizers in her corpus of audio-recorded conversations across three

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

539

groups of speakers. The first group, speakers who arrived in the United States after the age of 12, showed 4 participants out of 13 using null complementizers. The second group, sequential bilinguals, had 6 out of 11 speakers using null complementizers, and the third group of speakers, simultaneous bilinguals, had 9 out of 11 speakers who used null complementizers. While the use of null complementizers increased per group, no speaker used null complementizers in relative clauses like thank you for the present that you gave me. Although null complementizers are frequently used in this context in English, its use is ungrammatical in Spanish. The absence of null complementizers in relative clauses illustrates the role of superficial parallel structure in permeability to change. Since the use of null complementizers is ungrammatical in Spanish, this context is not permeable to change. Now that we have looked at two forms of indirect change in SilvaCorvala´n’s work, we move to a very special kind of shift in frequency, namely a shift in frequency that can imply the incipient birth of a new grammatical category. The contact-induced rise of new grammatical categories is closely connected to the study of linguistic areas. Linguistic areas are geographical areas where genetically unrelated or distantly related languages are spoken that have become structurally alike due to language contact. Note that in linguistic areas languages become structurally more similar, without speakers thinking of these variants as mixed variants, whereas in mixed languages the role of both languages is clear, because matter of both languages is used (e.g., words and morphology from both languages). Convergence is connected only to pattern replication (see Section 4.3), which is more subconscious and therefore less noticeable to speakers. Convergence is the process whereby languages become structurally more alike. The process of convergence can lead to the emergence of a linguistic area. A form of convergence that has received quite some attention in the study of linguistic areas is the contact-induced rise of definite markers such as articles. Linguistic area studies suggest that a language without articles such as the can develop them under influence of a language with articles, thereby making one of the two languages structurally more similar. Often articles derive historically from demonstrative pronouns. For example, the English article the is derived from the distal demonstrative pronoun that. While demonstrative pronouns indicate deictic and definiteness information, articles indicate definiteness only. The rise of articles tends to go hand in hand with encoding definiteness obligatorily. For example, languages without articles can say I read book but languages with articles have to encode information about definiteness, that is, they can state I read a book or I read the book, whereas bare nouns are only allowed in very restricted contexts. Aalberse, Zou, and Andringa (2017) explored whether in a Mandarin– Dutch contact situation there was any evidence of increased

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

540

SUZANNE AALBERSE AND PIETER MUYSKEN

demonstrative use in Mandarin, possibly as a result of the contact with Dutch, a language that, unlike Mandarin, has articles (demonstratives are part of Mandarin). In order to test this, they compared different speakers of Mandarin with each other: first- and second-generation Chinese immigrants in the Netherlands, where Dutch is the dominant language, and nonimmigrant Mandarin speakers in China. The immigrants were all in contact with Dutch, and Mandarin was a heritage language for them. The second-generation immigrants used Dutch more frequently than Mandarin. The first-generation immigrants varied considerably in their use of Dutch. Based on a narrative elicitation task in Mandarin Chinese, Aalberse et al. (2017) found that both first- and second-generation speakers in the Netherlands used more demonstratives than Mandarin speakers in China when they were referring to persons or objects that had already been mentioned. The use of bare nouns, which is grammatical in Mandarin but ungrammatical in Dutch, decreased. One could interpret the decrease in the use of bare nouns as a form of convergence with Dutch. In addition, Aalberse et al. (2017) found that the second-generation immigrants showed a preference for the Mandarin distal demonstrative nage (“that”) when referring to already mentioned referents or objects. The distal demonstrative in Dutch is most closely connected to the definite article in Dutch both in meaning and contexts of use. Aalberse et al. therefore suggested that second-generation speakers make an extra step in the convergence process by not just avoiding constructions that are ungrammatical in Dutch (bare nouns) but in addition by selecting a translational equivalent nage (“that”) which is closest to the Dutch context. In addition to quantitative differences Aalberse et al. (2017) analyzed qualitative uses of demonstratives in immigrant speakers. They described two contexts where immigrants used demonstratives and nonimmigrant speakers did not use them. One was the use of demonstratives for multiple arguments in one utterance as shown in example (6), from Aalberse et al. (p. 41), and the other was the use of demonstratives in reference maintenance contexts as shown in example (7), from Aalberse et al.’s unpublished data: (6)

他

就

把

那

个

石

头放

在

那个

凳子上，

he

then

BA

DEM

CL

rock

put

on

DEM CL

chair,

另外

那个

朋友

就

站上去。

other

DEM CL

friend

then

stand on

“he then puts the rock on the chair, and the other kid stands on top” (This refers to a scene with two children where a child is trying to reach something in a tree. To make himself taller, he stands on a rock that another child had placed on a chair.)

(7)

他们 they

想 want to

用 use

那个 球 D E M - C L ball

扔 throw

到 at

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

那

个

DEM

CL

541

衣服 上, 希望 那 个 衣服 会 掉下来。 clothes up, hope D E M C L clothes will fall down “they want to throw the ball at the clothes, hoping that the clothes will fall off”

In contexts like (7), the referent (“clothes”) is a maintained argument in the second half of the sentence. Nonimmigrant speakers tend to use bare forms in this case, whereas Mandarin speakers in the Netherlands use demonstratives. Note that the use of demonstratives is grammatical, but feels redundant to nonimmigrant speakers. A first step in the potential use of articles in Mandarin in the Netherlands could be that heritage speakers use more demonstrative markers than non-heritage speakers. Such an increase in frequency relates to the notions of frequential copying (Johanson, 2002) and indirect change (Silva-Corvala´n, 1994). The characteristic that sets this type of frequency change apart from changes in frequency discussed in relation to the increased use of null complementizers in Spanish is that in the long run this change in frequency might end up causing a change in the grammatical function of the element: the change in frequency might be an indication that the demonstrative is becoming an article. The observation that second-generation heritage speakers in the Aalberse et al. (2017) study preferred the distal demonstrative makes the structural difference between Dutch and Mandarin smaller and can thus be seen as the next step in the convergence process.

6 Conclusion In this chapter we have sketched concepts and results from language contact studies in four different domains: code-switching and lexical borrowing; contact-induced language change; mixed languages, pidgins, and creoles; and heritage languages. However, we have mostly limited ourselves to the lexicon and syntax. The systematic study of phonetics, intonation, and phonology is just beginning in contact linguistics, as is the study of semantics, discourse, and pragmatics. Useful starting points for these domains are studies by Queen (2012), Nance (2015), and Yao and Chang (2016) for phonology, and by Babel (2014) and Andersen, Furiassi, and Misˇic´ Ilic´ (2017) for semantics and pragmatics. The volume on Germanic heritage languages in North America (Johannessen & Salmons, 2015) also includes chapters on these less researched topics and thus forms a good starting point for these domains as well. One crucial point in our overview was that the outcome of language contact varies per situation. This notion of variable outcomes is captured in Haugen’s (1953) notion of language ecology, which assumes that the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

542

SUZANNE AALBERSE AND PIETER MUYSKEN

setting in which languages are spoken affects the type of changes we observe in them. The same holds for language development in bilingual situations. Whether and how people code-switch, whether creoles show influence of speakers’ L1 or of universal principles, whether and in which way languages are mixed and how heritage languages change – the answers here will differ given the specific context and given the specific set of languages involved. A crucial distinction introduced in Section 4 is the opposition between what Matras (2009) has called matter versus pattern replication. Matter replication, the use of word forms from one language into the other such as in the case of borrowing words like email or complete phrases like the American dream from English into another language is a very common phenomenon. The specific identity of the donor language is usually also clear. The status of pattern replication is more debatable since the issue to what extent languages may undergo structural influence from each other through bilingual usage is unresolved. A first question that needs to be asked when one observes change in language contact situations is whether a change could have also taken place without language contact. If language contact does seem a probable trigger for change, the next question is whether that change is a general effect of bilingualism or an effect of crosslinguistic influence. Although some issues remain unresolved, we would like to underline that a necessary condition for structural (as opposed to lexical) influence is sustained bilingual interaction in the daily lives of speakers, preferably over a longer period of time. Sustained daily bilingual interaction does not necessarily lead to structural influence, as far as we know, but the conditions under which potential changes are realized are not yet clear. There are some more general observations to be made, which are summarized in the following. Some components of the grammar of a language may undergo change more rapidly than others, but we do not know exactly which components these are. Likely candidates for sensitivity to change may be the tensemood-aspect system of a language as evidenced by the hyperextension of progressives in Pennsylvania German as discussed in Section 5. Also those components of language directly involved in interpersonal interaction such as emphatic marking, politeness marking, and the encoding of information structure may be more sensitive to change, although these may be part of pragmatics rather than grammar proper. Research has also shown that some words are easier to borrow than others. Extra-clausal elements like discourse markers and content words like nouns, for example, are more likely to be borrowed than prepositions or determiners. As far as mixed languages goes, the types of hierarchies noted in Section 3 as regards the division of labor between the two languages generally hold. Furthermore, bilingual practices such as the use of mixed languages and code-mixing cannot be studied without attention to all

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

Bilingualism and Language Contact

543

languages involved. In fact, the role and the effect of all languages involved should be studied in any contact situation. In language change involving language contact there is a crossgenerational scaffolding effect, leading to more contact-induced change over a longer period. However, some speech communities appear to show much less change than others. The reasons for this are not understood so far, but possibly they involve communities’ notions of purism, whether these are below or above levels of awareness. The challenge lies in making the general observations here more concrete. One promising avenue for future research is to bring together insights from experimental language processing studies with formal and computational approaches and observational paradigms. An example of innovative research is the study on code-switching by Goldrick, Putnam, and Schwartz (2016), and the fact that it had commentaries from different research communities points to the way forward. Only interdisciplinary approaches will be able to increase our understanding of the connections between individual bilingualism and the very shape of the languages they use.

Acknowledgments Aalberse’s work on heritage languages was supported by the Traces of Contact project (ERC #230310). Muysken’s work on this chapter was supported by this same project and by the NWO Language in Interaction Consortium.

Appendix: Abbreviations and conventions used in examples

1SG AG AN T BA

CL COMP DA T DE L DE M

first person singular agentive anterior preposition or light verb preceding affected objects classifier complementizer dative delimitative demonstrative

DIR F M NEG NML Z OB V PL PLUPF X.Y

directional feminine masculine negation nominalizer obviation marker plural plusquamperfect original form corresponds to combination of two glosses

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 15:59:13, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.027

27 Bilingualism and Multilingualism Suzanne Quay and Simona Montanari

1 Introduction While this Handbook sees the term bilingualism as referring to learning and using more than a single language, this chapter refers to bilingualism in the narrow sense of learning and using just two languages, and to multilingualism as learning and using three or more languages (Kemp, 2009). Multilingualism is a typical aspect of everyday life for much of the world’s population. While no statistics exist as to the number of people who function with multiple languages on a daily basis, the fact that the world’s almost 7,000 languages are spoken in approximately 200 countries suggests that bilingualism or multilingualism is a characteristic of practically every country (Romaine, 2013). Linguistic diversity within communities has likely existed since the beginning of humanity, indicating that it is perhaps as intrinsic to human nature as cognition and emotion. Indeed, multilingualism is present among individuals of all backgrounds, from the educated to the poor; it also exists in all geographical areas, from along national borders to urban metropolises to rural regions. In Northern Europe, for instance, educated individuals are expected to have knowledge of their home language(s), English, as well as another foreign language, and one out of three Europeans reports speaking three or more languages (European Commission, 2012a). At the same time, multilingualism is part of everyday life for both urban and rural populations in areas as different as South Asia, Africa, and the Middle East. This chapter starts by discussing circumstantial reasons for different languages to coexist in society and continues by looking at multilingual language use in individuals – how they become multilingual as they learn, unlearn, and relearn languages across their lifespan. The latter half of the chapter deals with differences between bilingualism and multilingualism and sketches research areas for further exploration.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

545

2 Societal Multilingualism: Historical, Political, Economic, and Educational Forces The circumstances – historical, political, economic, or due to migration, education, or religion – in the past and today that contribute to the existence of heterogeneous linguistic situations in the world are countless and usually intertwined. The European colonization of the African continent, for instance, imposed a European language on populations who were already bilingual in their local and another indigenous language, with the consequence that multilingualism rather than monolingualism is the norm there today. In South Africa, for example, virtually all speakers master multiple local/regional languages as well as English, the lingua franca used for government, business, politics, and education. The several indigenous languages fall into two typologically related language groups: one, the Nguni, comprising Zulu, Xhosa, Ndebele, and Swazi, and the second, the Sotho group, comprising Sepedi, Sesotho, and Setswana. Speakers of one of these languages are often also proficient in other related languages within the same group. In addition, geographical prominence of specific languages in certain regions (such as Zulu in the KwaZulu-Natal province, Sesotho in the Free State, or Afrikaans in the Western and Northern Cape) makes the knowledge of these languages the order of the day for speakers of other languages. English, however, remains the most common medium of instruction across the country, despite the introduction of indigenous South African languages in education policy in 1997 (Makoe & McKinney, 2014). While the countries in the African continent vary greatly in terms of history and language policies, they all inherited the use of exogenous or European languages (in tandem with indigenous languages) as a consequence of colonialism. Multilingualism can also result from networking and communication rather than from exposure to an external power or originating from government planning (Bhatia & Ritchie, 2013). This is the case of India, where many languages, races, cultures, and religions have coexisted for centuries, and literature, popular media, and trade have been major forces of linguistic diversity and pluralism (the term linguistic pluralism is used in this chapter as a synonym of multilingualism; see Grin, Chapter 9, this volume, for other approaches). Like literature in the past, popular media today is spreading both local and official languages, promoting mutual intelligibility and bridging the gap between urban and rural varieties. Likewise, globalization, the information technology revolution, and improved infrastructure have allowed vernacular and regional languages to permeate old and new media forms, further increasing multilingualism. Trade in local markets and fairs, festivals, and pilgrimages are also contributing to India’s diverse linguistic landscape, promoting pan-Indian multilingualism at the local level. At the same time, education and

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

546

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

government are favorable toward linguistic diversity and pluralismoriented policies such as the “Three Language Formula” in education, which by prescribing trilingualism in schools undoubtedly makes India “the linguistic laboratory of the world” (Bhatia & Ritchie, 2013, p. 843). Indeed, education and communication are major forces behind linguistic pluralism at a global level. First, the spread of personal information technologies has increased the range, diversity, rapidity, and flexibility of instant communication promoting the use of different languages and, therefore, of multilingualism. Competence in multiple languages is also the result of educational policies in many different regions of the world (see also Lo Bianco, Chapter 8, this volume). For instance, nearly 90% of European Union students study English as well as an additional language at school (besides their first language[s]) because the importance of plurilingualism is recognized in a geographical area characterized by cultural and linguistic diversity (European Commission, 2012a; see Fuller, Chapter 6, this volume, on ideologies related to linguistic diversity). Children in most regions of Africa, Asia, and the Middle East also gain competence in different languages because of multilingual schooling. For example, in Northern Africa and the Middle East, virtually all educated speakers are minimally bilingual in the regional, vernacular Arabic variety (the language of daily speech) and Modern Standard Arabic (the high variety learned at school). More often than not, these speakers also know a local minority language, such as Kurdish, Coptic, or Berber, or a European language (i.e., French or English), which they inherited from the period of their colonization. Similarly, in an Asian nation like Vietnam, children today are educated in Vietnamese and English but may also come from 54 ethnic groups with more than 100 different home languages. Due to the legacy of colonial rule and political alliances, French and Russian may also be spoken among the educated older generation while the younger generation has been more motivated to study Japanese and Chinese as Vietnam’s links with other East Asian countries have strengthened (Nguyen, 2012). In these areas, therefore, multilingualism is frequent, usual, and a normal state, not only for immigrants or foreign workers but also for native inhabitants. Migration is also another vehicle of linguistic pluralism in regions that are typically considered monolingual. As millions of immigrants from the world’s poor regions move to wealthier Europe and North America in search of a better life, linguistic and cultural diversity are increased, especially in urban areas where immigrants tend to settle (see Gonc¸alves, Chapter 3, this volume, for immigrants’ experiences in such settings). For instance, in Toronto, Canada, 40% of the population is foreign born and 175 languages are spoken (Mackey, 2013). Similarly, despite its reputation as a monolingual English-only country, at least 364 languages are used in the United States (Mackey, 2013), and cities like New York City, Los Angeles, and Chicago are largely multilingual. The influx of asylum

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

547

seekers and economic migrants to Europe from depressed and war-torn regions of the Middle East that began after the Arab Spring are also further diversifying Europe’s linguistic and cultural landscape. For many of these immigrants – who might already be bilingual to varying degrees – learning the language of the host country is an economic necessity and multilingualism will also become the norm.

3 Language Use in Multilingual Communities Language use in multilingual communities as surveyed earlier is affected by a variety of linguistic, sociolinguistic, and sociocultural factors. Certain areas might be characterized by diglossia; for instance, in some areas of Southern Africa, an indigenous language may be spoken at home, the major indigenous language of the region may be used for the primary years of education, and the rest of schooling occurs in a European language (Kamwangamalu, 2013; see also Fuller, Chapter 6, this volume). Similarly, in the Middle East and Northern Africa, three different languages typically coexist, each fulfilling a different role: in Morocco, Tunisia, and Algeria, for example, the regional, vernacular Arabic variety is typically spoken in informal situations, Modern Standard Arabic is used at school and in formal situations, and French is further acquired for educational and professional reasons (Rosenhouse, 2013). The situation is not very different in other regions of the world: the Sa´mi in Finland, for instance, use Sa´mi at home; they must learn Finnish at school as an economic necessity and further acquire English to communicate beyond national borders (Romaine, 2013). In all these cases, speakers are not expected to be equally proficient in each language; rather, they have conversational competence in their home language, academic competence in the language of instruction, but also functional knowledge of another language for education or professional purposes. In other words, language use is functionally compartmentalized according to specific sociolinguistic domains. Language abilities are expected to be appropriate for and consonant with each of these domains. Language use in multilingual communities might not only be characterized by diglossia but also by code-switching – the use of multiple languages within the same utterance or conversation – especially in informal contexts (see also in this volume De Houwer, Chapter 17, and Aalberse & Muysken, Chapter 26). This is the case in Morocco, for instance, where mixing Moroccan Arabic, Modern Standard Arabic, and French within the same exchange is the usual, everyday means of communication among most literate individuals (Rosenhouse, 2013). Code-switching is also a natural phenomenon and thus pervasive in India, where speakers might use multiple languages within the same interaction for stylistic reasons, for example to quote, interject, specify addressee, and reiterate

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

548

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

or explain a message, but also to express a variety of “socio psychological affiliations” (Bhatia & Ritchie, 2013, p. 861). Code-switching is indeed a hallmark of most multilingual communities because the simultaneous use of multiple languages – in a context characterized by linguistic and cultural diversity – can convey a range of meanings and styles that a single language cannot. In these communities, speakers might have good mastery of each language; yet, code-switching becomes part of their informal linguistic repertoire following a lifetime habit of mixing languages in interactions with multilingual interlocutors. At the other extreme, speakers in multilingual communities where mutually intelligible languages are spoken might also communicate with each other without a lingua franca or without one speaker using the language of the other. This phenomenon – called receptive multilingualism (ten Thije & Zeevaert, 2007) – typically involves languages that are typologically related and share many grammatical, vocabulary, and phonological features. Receptive multilingualism has been a characteristic of many societies for as long as people from closely related ethno-linguistic communities have been in contact. For instance, in the Middle Ages, speakers of different varieties descended from Vulgar Latin would interact each in their first language for purpose-based and face-to-face interactions such as trading and negotiations (Braunmu¨ller, 2007). Receptive multilingualism was also documented among speakers of different American Indian languages with a close genetic relationship (Voegelin & Harris, 1951). Receptive multilingualism, however, has been particularly studied – at least recently – in the context of the closely related Scandinavian languages: Danish, Norwegian, and Swedish. In these communities, speakers can communicate with each other by each using their first language because of the high degree of mutual intelligibility among the three languages (Schu¨ppert & Gooskens, 2012). Interestingly, however, studies show that intelligibility is not only related to linguistic overlap but also to extralinguistic factors such as speakers’ attitude, level of contact with the other language(s), and literacy (Schu¨ppert & Gooskens, 2012). For instance, Norwegian speakers are the best in Scandinavia at understanding languages within the same language group. This is likely because they have more contact with different languages than do Danes or Swedes (Gooskens & Heeringa, 2014). Likewise, in India, speakers of Hindi and Urdu, which tend to be classified as different registers of the same language, Hindustani, report lack of mutual intelligibility due to the complex religious, cultural, and ethnic differences that characterize each community (Bhatia & Ritchie, 2013). Hindi is indeed spoken by the Hindu community; it is historically associated with the Sanskrit tradition indigenous to India and therefore written in the Devanagari script; and it replaced Urdu as the official language of India along with English during the institution of the Indian constitution in 1950. Urdu, on the other hand, is spoken by the Muslim community; it is

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

549

tied to Persian and Arabic literary traditions and thus written in the PersoArabic script, and it is also the national language and lingua franca of neighboring Pakistan. Since speakers of each community identify themselves as ethnically and culturally different from speakers of the other community, linguistic mutual intelligibility is compromised. These findings point to the complexity of language use in multilingual communities and the host of factors – both linguistic and extralinguistic – that affect the use of these languages. While diglossia, code-switching, and receptive multilingualism are also typical modes of communication in bilingual communities, multilingual speakers will understand, manage, and codeswitch in three or more languages, which widens the range and possibilities of their communicative intents.

4 Multilingual Language Use in Individuals: From Development to Loss The occurrence of multilingualism in individuals and their degree of language proficiency in each language depend to a large part on their personal circumstances – their patterns of social and linguistic interactions and their social and geographical mobility, as well as their motivation. In many countries, intercultural marriages, work-related moves, immigration, or just parental aspirations result in the cultivation of multilingualism within families through educational means or other efforts to provide language contact with target language(s) not readily available in the home, community, or school environments – for example, employing au pairs/nannies who can speak the target language(s) with the children, or sending children to summer immersion camps or to supplementary afterschool/weekend language programs. These cases exemplify elective multilingualism as opposed to circumstantial multilingualism, where immigrants and refugees may have no choice but to “pick up” the language of their new linguistic environment in order to survive, that is, to find work and integrate into their new community. This is the case, for instance, for the many Syrian refugees who since 2012 have increasingly sought asylum in neighboring countries like Turkey, Lebanon, and Jordan and in European countries like Germany and Sweden, as well as further afield in English- and French-speaking parts of Canada. Their children have to adapt immediately to educational systems in a language they do not yet fully understand with expectations for them to learn the official language(s) of the public school systems in those countries. Multilingual development across the lifespan often starts with family interactions where two or more languages are used in the home (as is discussed in the next section), but may also be a result of early and later education in a language or languages outside of home life. When children are very young, parents may choose these languages by sending children

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

550

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

to international or bilingual schools, or the choice is left to school or government policies. Later, these same children can decide to study other languages or may experience naturalistic immersion and exposure to other languages when opportunities arise to travel, work, or live elsewhere (see Bigelow & Collins, Chapter 2, this volume, and Gonc¸alves, Chapter 3, this volume). Still others become multilingual proactively through personal motivation and self-study because of interests in particular languages or in the culture or religion to which such languages pertain, as revealed, for instance, in the narratives of the multilinguals featured in Todeva and Cenoz (2009).

4.1 Interactional Strategies in Multilingual Families Parents have more control over languages used in the home with infants, toddlers, and preschool-aged children. Once children enter school, the educational system and peers have a stronger influence over whether multilingual children will retain home languages (see also Bigelow & Collins, Chapter 2, this volume). Depending on interactional situations, frequency of exposure to input, the stage in acquisition, and their preferences, trilingual children exposed to three languages from birth always have dominant or stronger as well as weaker languages, given that they never need to use their languages equally (see Stavans & Hoffmann, 2015, for a review). Trilingual families have generally found success using the one-person one-language interactional strategy, but only when parents had different language backgrounds. Braun and Cline (2014) interviewed 35 trilingual families in England and 35 in Germany and found three main types of parents: Type I parents both had a different first language and neither spoke the societal language natively; Type II parents consisted of one or both being bilingual, with the societal language possibly being one of their languages; and Type III parents consisted of one or both being trilingual, again with the societal language possibly being one of their languages. The parents with different first languages that were not the same as their societal language (Type I) tended to be more successful at raising trilingual children than the bilingual and trilingual parents with the societal language in their repertoire (Types II and III). Type I families also had monolingual grandparents who helped to pass on their first languages and cultures to the children (Braun, 2012). Bilingual and trilingual parents (Types II and III), on the other hand, tended to use only one native language with their children (especially when they also spoke the societal language) and had grandparents who did not provide the minority language support expected because the bilingual grandparents themselves valued the more prestigious societal language (particularly English), which they preferred to use with their grandchildren (Braun, 2012).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

551

The need to avoid using the societal language in the home had already been recommended by De Houwer (2004), who focused on the home language use of 244 trilingual families living in officially monolingual Dutch-speaking Flanders. She found that although all the children in these families received trilingual input, “only about two fifths of the children in the survey who could have been speaking three languages actually do” (p. 124). The parental input pattern found to be the best at supporting children’s active use of three languages was when neither parent used the societal and school language, Dutch, at home and/or both parents spoke two minority languages at home. But even this strategy does not guarantee success: the discourse styles of parents can encourage or discourage the use of home languages (Lanza, 2007), as was also found by Chevalier (2015), who studied the trilingual development of two children from ages two to four in Switzerland. While one child, Elliot, was an active trilingual from the start, the other, Lina, understood the two minority languages, French and English, but mainly spoke the societal language, Swiss German. Although all adults followed the one-person onelanguage strategy, Lina’s father, who provided one of the minority languages, French, did not insist on his daughter’s use of French. He also did not have the didactic conversational style of Elliot’s father, who made much use of the marked adult repetition strategy after language mixing (where the adult repeats the content of the child’s utterance using the target language) and who focused on teaching and eliciting vocabulary (see Lanza, 2007, for a description of parental discourse strategies in reaction to child language mixing; see also De Houwer, Chapter 17, this volume). Also, Lina spent the most time with her mother, who spoke the societal language. Thus, Lina had a much more limited exposure to French and English than Elliot, who not only had equal input but also contact with these two languages from his parents’ bilingual friends and other English speakers in his neighborhood (see Armon-Lotem & Meir, Chapter 10, this volume, for the role of the input in bilingual development). Both children were also aware of the status of their languages as reflected by their greater proficiency in the societal language. Elliot was also equally proficient in English, the language used between his own parents as well as Lina’s. Chevalier’s (2015) study thus highlights the contextual factors along with the discourse styles of caregivers that affect the outcome of raising multilingual children. Although we know that family language policies (conscious planning of language use in the family based on parental language goals, attitudes, or intentions) play a large part in ensuring success in raising multilingual children, these policies may not be enough to encourage minority or heritage languages to flourish in the adolescent and later years (Wang, 2016). Oriyama (2016) points out that just speaking minority/heritage languages at home is not enough to sustain early multilingualism for families living in predominantly monolingual societies. In order to

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

552

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

maintain minority/heritage languages, heritage language (HL) communities of practice are needed where children can have contact with other minority speakers through schools and other activities in the minority language. Oriyama (2016) states unequivocally that “it is participation in HL communities of practice that significantly promotes children’s longitudinal home HL maintenance, not parental HL use” (p. 295). Without joint efforts from parents, extended family, institutional and societal support, and HL communities of practice, maintaining two or more nonsocietal languages in monolingual societies is a challenge when children enter school and become part of the larger society outside the home.

4.2 Language Attrition in Multilinguals Because of the difficulties in maintaining multiple languages, language attrition or the “non-pathological decrease in proficiency in a language that had previously been acquired by an individual” (Ko¨pke & Schmid, 2004, p. 5; see also Ko¨pke, Chapter 18, this volume) often occurs in multilingual individuals. What do we know about language attrition in multilinguals? Unfortunately, not much work, at least not in terms of corpus-based quantitative analyses, has been done on the development, maintenance, and attrition processes in multilingual speakers. Most of the research has concentrated on bilinguals, in particular, the loss of an L1 in an L2 environment as when immigrants lose their mother tongue, or the loss of an L2 or foreign language in an L1 environment as in the case of returnees or foreign language students. However, many so-called bilinguals and L2 learners may have learned an L3 or L4 in classroom settings. The interesting volume edited by Todeva and Cenoz (2009) gives us an unusual insight into the factors supporting the (un)learning and (dis)use of three or more languages through 12 personal narratives of multilinguals with seven or more languages in their linguistic repertoire. These narratives showcase the perspectives of multilingual individuals from diverse countries of origin, diverse family backgrounds, and diverse languages acquired at different ages who have had the experience of developing and maintaining multiple languages with varying levels of proficiency. These personal reflections on multilingual development and attrition converge with findings in studies of L1 and L2 attrition in bilinguals (see Keijzer & de Bot, Chapter 14, this volume, and Ko¨pke, Chapter 18, this volume). Many multilinguals experience language attrition over time with ongoing changes in proficiency in each language, including L1(s), as their multiple language (sub)systems compete for both time and cognitive resources. When multiple languages are learned, dominance in terms of relative proficiency or frequency of use does not necessarily correspond to chronological order of acquisition. For instance, an L2 may be given up for a while due to changes in needs or motivation, and learning an L3 or

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

553

relearning other languages may counteract the maintenance of an L2 or L1 (Jessner, 2015; see also Treffers-Daller, Chapter 15, this volume). This process can be nonlinear and reversible, that is, languages in development, be they the L2, L3, or Ln (up to an indefinite number of languages), can be learned, unlearned, and relearned. Parts of a multilingual system can become fossilized or dormant but may still exert an influence on other parts of the system. Children exposed to two or more languages in early childhood may experience a decrease in proficiency in their dominant home language(s) when they engage in school activities in yet another language that is different from the ones they hear at home. This was the case for Wang’s (2016) two simultaneous trilingual sons whose stronger home languages, French and Chinese, began to show signs of attrition immediately upon entering school in the societal language, English, in the United States. Eventually, trilingual children’s educational or societal language (these often are the same) will become their stronger language and their home languages will continue to weaken unless special efforts are made to maintain them. Oftentimes, languages taught in schools/universities have more prestige than home languages, which may also contribute to the attrition of L1s. Language switching and mixing can perhaps also result in language diminution and loss. Kamanga (2009), an African speaker of 21 languages, saw code-switching as disadvantageous. Not only did he feel that it limited his development of additional languages, but he also claimed that his code-switching practices managed to “kill” some of his languages (p. 123). In addition, several other multilinguals in Todeva and Cenoz (2009) described how the languages they learned the most quickly were the ones they also lost the most quickly. Some observed that grammar learned more slowly is retained even when vocabulary is diminished through lack of experience or use later in life. Interference, that is, unhelpful influence from one language on the other, can occur from a weaker language to a more dominant one as well as from a non-L1 language to a later learned language. The consensus seems to be that languages with the longest exposure thrive, while those with short exposures decline or are completely lost. As in studies of attrition in bilinguals, the multilinguals in Todeva and Cenoz (2009) indicated that vocabulary is more affected than syntax in attrition. They also found that the rate of attrition was slower for languages acquired in naturalistic settings in real-life situations than for languages studied for academic or scholarly purposes in school settings. Just as attrition studies of bilinguals have described, complete attrition can occur when input to a language acquired from birth decreases or stops when a multilingual child enters school. Retention appears to be better if languages are acquired over a sustained period before puberty.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

554

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

In sum, the need and opportunities to use a language in authentic contexts is the strongest predictor for maintaining and increasing proficiency in one or more languages for multilinguals.

5 Differences between Bilingualism and Multilingualism The acquisition and use of multiple languages present more complexity than the learning and use of two languages. Thus, we have made a distinction between bilingualism as referring to just two languages and multilingualism as referring to more than two languages, as does Kemp (2009) who further writes: “there is some research evidence that there are qualitative and quantitative differences between individuals who use two languages and individuals who use three languages” (p. 15). First of all, as described in the previous sections, multilingualism results from a wider variety of social, cultural, and economic reasons than bilingualism, and accounts of this heterogeneity are even more difficult than in the case of bilingual populations. Second, when it comes to development, there are only two possibilities with two languages: they are acquired either simultaneously or sequentially. In the case of three languages, there is greater diversity in the acquisition order, as an individual can have multiple L1s and/or L2s depending on the timing of acquisition of each (Cenoz, 2001), with possibly different outcomes in each situation. For example, a child who learns three languages simultaneously from birth may develop different proficiency levels than an individual who acquires an L2 and L3 later in life in a school setting. Similarly, a child who acquires an L1 and an L2 simultaneously before learning an L3 may behave in linguistically different ways from one whose L2 and L3 are learned simultaneously after the acquisition of the L1 has already started. As shown by the 32 possibilities in Todeva and Cenoz’s (2009, p. 7) overview of possible orders of acquisition for seven languages, the complexity increases exponentially with each additional language. The acquisition order can become even more complicated when we take into account the dynamic nature of multilingualism and how common attrition and the relearning of languages (as mentioned earlier) are across a multilingual individual’s lifespan. Thus, learning several languages is quite different from the sequence “L1 followed by L2” commonly found in second language acquisition research. In addition, the order of acquisition of languages in multilinguals does not necessarily align with a hierarchy of proficiency (see also Treffers-Daller, Chapter 15, this volume). Even when languages are acquired in infancy, development differs for bilinguals and multilinguals as input is considerably more distributed over several languages in multilingual situations than in bilingual ones, which adds to the complexity to describe and explain the outcomes of such development.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

555

When it comes to language use and proficiency, there are further distinctions between bilingualism and multilingualism. McArthur (1992) defines multilingualism as “the ability to use three or more languages, either separately or in various degrees of code-mixing [. . .] for different purposes” (p. 673) and goes on to state that proficiency in each language varies depending on factors such as register, occupation, and education. While this is also partially true for bilingualism, languages typically always have different roles and functions for multilingual individuals; they may be used separately or mixed, and speakers’ proficiency in each language is likely to differ and fluctuate over time in many individuals. Furthermore, societal attitudes toward multilingualism and language proficiency may be less rigorous than in the case of bilingualism. For instance, generally, there are lower expectations of balanced proficiency and/or literacy among multilinguals than among bilinguals. Many multilinguals might actually have passive knowledge of one of their languages or they might be literate in only one. In sum, unlike in the case of bilinguals, the concept of “three monolinguals in one” is rarely expected. This is especially true in regions where multilingualism is common and it is well understood that high abilities in all of a multilingual’s languages are unattainable. These uneven societal expectations surrounding bilingual versus multilingual individuals have far-reaching implications in the realms of work and education. Compare, for instance, the level of English proficiency required of bilingual immigrant students in the United States, who are held to monolingual academic standards beginning from one year after their arrival to the country, to the level of English proficiency expected among multilingual students in Africa or South Asia. Bilingual students in countries traditionally regarded as monolingual are likely to face prejudice and encounter academic challenges; multilingual students in regions of the world that are characterized by linguistic pluralism are encouraged to develop multilingual abilities and are considered successful when they master multiple languages, even though they may not be “perfect” in each of them. Besides setting multilingualism apart from bilingualism, these unbalanced perceptions have serious implications for everyday life, affecting societal interactions, educational policies, professional opportunities, and overall well-being. A further characteristic that sets multilingualism aside from bilingualism is that becoming multilingual is a less effortful process than becoming bilingual. Many studies have documented facilitation of L3 acquisition in speakers who already know two languages as compared to L2 learning in monolingual speakers (see Cenoz, 2003, for a review). L3 learners, it has been argued, are “expert” language learners who not only employ a wider range of linguistic and mnemonic strategies but also are more efficient and flexible in their language use than “novice” L2 learners (McLaughlin & Nayak, 1989). They are also less anxious when learning an additional language than monolinguals (Dewaele, 2002a). Furthermore, L3 learners

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

556

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

also appear to use two languages as base languages to acquire an L3, therefore accelerating the process of L3 acquisition (Cenoz, 2003). Such facilitation has been documented at different ages and with different language combinations. Accelerated L3 learning appears to occur irrespective of typological relation among the languages being acquired, indicating that becoming multilingual once you are bilingual is simply easier than becoming bilingual after you have been monolingual. Abu-Rabia and Sanitsky (2010), for instance, found that Russian–Hebrew bilingual/biliterate sixth graders were better learners of L3 English – as assessed by phonological, vocabulary, morphosyntactic, and orthographic measures – than monolingual Hebrew-speaking peers, even if both Hebrew and Russian are written in orthographies that differ from each other and from English. The authors speculated that knowledge of several different orthographies enhances the acquisition of additional languages, and that language and literacy skills transfer among L1, L2, and L3 even in the case of different alphabets and distinct patterns. Moreover, in a study of 130 multilinguals with 3 to 12 languages, Kemp (2007) found that multilinguals become faster at learning grammar the more languages they know, even when these are not typologically related. In sum, typological or orthographic relation is not a prerequisite for accelerated Ln learning since the cognitive benefits experienced by proficient multilinguals as well as their previous experience with language learning might suffice to make the Ln learning process more efficient. Knowledge of earlier acquired language systems other than one’s L1 appears to affect later language systems as a result of crosslinguistic influence (CLI), called transfer when the effects are positive and interference when the effects are negative. Unlike in bilinguals, where CLI may be from L1 to L2 or vice versa, in multilinguals transfer can occur from the L2 to the L3 or from multiple languages simultaneously, such as L1 and L2 jointly influencing L3. CLI can also be multidirectional in complicated ways: for example, L3 acquisition might also influence L1 and/or L2, or L1 and/or L3 may influence an L2 (Jarvis, 2013). This can be illustrated by a case study of phonological development in two simultaneous English-, Italian-, and Spanish-speaking school-aged sisters in Los Angeles. Mayr and Montanari (2015) showed that while the girls’ Italian voice onset time patterns were influenced by English, their Spanish productions were largely unaffected by the other two languages. Thus, their Italian and Spanish realizations were fundamentally different, even though the Spanish and Italian stops produced by their monolingual Italian- and Spanish-speaking input providers were virtually identical. They conclude that these patterns could only be discovered by investigating trilingual development, because an analysis of just bilingual stop consonant systems could not have identified them (p. 1031).

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

557

Candidates for transfer are based on factors such as frequency, recency, and salience of forms and structures, and most importantly, typological similarity (e.g., De Angelis, 2005). For instance, German L1 learners of French who also spoke Romansch showed not only facilitation but also acquisition patterns that were different compared to German L1 learners of French who did not know Romansch (Brohy, 2001). Similarly, English L1 learners of Italian displayed structures and errors that differed depending on whether the learners knew Spanish or French (De Angelis, 2005). Hammarberg (2009) describes, moreover, the general tendency to activate the L2 rather than the L1 in L3 performance. Hammarberg’s subject, a native English speaker fluent in her L2, German, with advanced knowledge of French and elementary knowledge of Italian, drew from German for lexical constructions in her L3 Swedish but also resorted to English for metalinguistic comments, asides, and requests for help. In addition, in learning an L3, bilinguals will show more evidence of transfer from a language (Spanish) they already know that is related to the L3 (English) than from a language they already know (Basque) that is not related to the L3, regardless of whether this language is speakers’ L1 or L2 (Cenoz, 2001). Jessner (2006) discusses studies suggesting that multilinguals display higher crosslinguistic and metalinguistic awareness than bilinguals. Both L2 status and metalinguistic awareness are of a unique nature in the context of L3 acquisition, as multilinguals appear to develop cognitive abilities not apparent in bilinguals. Literacy development in two languages has also been found to increase metalinguistic awareness and working memory capacity for additional language learning (see Sanz, 2013, for a review). A more developed metalinguistic awareness on the part of the learners, which can be linked not only to their previous experience as language learners but also to their knowledge of different linguistic systems and how these interact, may be advantageous for slowing language attrition or for helping language users to compensate for the effects of attrition more easily and effectively. Nevertheless, there are also costs in language processing, at least with regard to lexical retrieval, that are associated with a larger number of languages spoken. Gonza´lez Alonso, Villegas, and Garcı´a Mayo (2016) compared the response time and accuracy on a lexical decision task with English compound words (e.g., taxi-driver) of Spanish bilinguals with English as an L2 to that of early Spanish–Basque trilinguals with English as an L3. The trilinguals were slower in their responses than the bilinguals. Thus, in terms of speed of lexical access and retrieval, bilingual speakers may have an advantage over multilingual ones.

6 Further Research Needed and Other Areas to Explore Much more research has investigated bilinguals than multilinguals. Research on L3 acquisition or multilingualism has intensified only from

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

558

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

the beginning of the 21st century (Jessner, Megens, & Grauss, 2016). The work of Jessner and her colleagues as well as that of Aronin and Hufeisen (2009) and Todeva and Cenoz (2009) has stressed that the far greater complexity of crosslinguistic interaction when more than two languages are involved makes it impossible to consider multilingualism as part of or similar to bilingualism. We also need to address the problem of assessing multilingual proficiencies in individuals in ways that go beyond linguistic measures to consider also their dynamic multilingual system and metalinguistic awareness (see also Treffers-Daller, Chapter 15, this volume). As already discussed, dominance or greater proficiency in a specific language does not necessarily correspond to chronological order of acquisition, as the multilingual system is in constant change (as when one language falls into disuse due to changes in needs or motivation while other languages are relearned). While bilingualism can be considered a form of multilingualism, the same cannot be said of the reverse. Since it is apparent that multilingual development is qualitatively different from bilingual development, we cannot depend on the results from bilingual studies to inform our understanding of multilingual acquisition and processing (Jessner et al., 2016). However, de Bot and Jaensch (2015, p. 142), in reviewing evidence from acquisition to attrition and from aphasia to neurolinguistics as well as from functional brain mapping, suggest that it is not yet possible to definitively decide that “cognitive mechanisms differ between multilinguals and bilinguals” without much more research on the mechanisms underlying L2 and L3 processing. Therefore, more studies are needed that look specifically at the neural architecture that makes the learning, storing, and processing of multiple languages unique and possibly different from the case of bilingualism. More investigations are also needed on different groups of multilingual learners, ranging from those with two or three languages from birth to those with early versus late additional languages added to their repertoire. How do children exposed to three or more languages in the home from an early age differ from those who acquire additional languages at a later age? And what about deaf children who have plural and diverse language practices at home and at school with the use of sign and spoken languages? A preliminary study by Swanwick, Wright, and Salter (2016) based on five case studies in the United Kingdom found that multilingual and bimodal interactions are common among deaf children from ethnically diverse families. The children they studied used translanguaging practices where they blended “sign, spoken and written language to make themselves understood, accommodate mixed deaf and hearing audiences and respond to the differing language fluencies and preferences of others” (p. 139). Research on multilinguals should thus move beyond users of spoken languages and examine language acquisition and practices in users of multiple signed and spoken

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

Bilingualism and Multilingualism

559

modalities. In addition, for both hearing and deaf individuals, what triggers a monolingual mode, bilingual mode, trilingual mode, etc. in interactions? That is, how do multilinguals decide which language to use predominantly and how much of the other language(s) to keep (in) active? Also, what is the minimum input needed and until what age for any critical threshold of proficiency to develop that would prevent total attrition should exposure to a particular language cease? It would be interesting to investigate more precisely the link between proficiency levels and rate of attrition in multilinguals’ language repertoires. What processes help multilingual children and adults overcome language loss or decline? What is the impact of instruction versus immersion in revitalizing dormant languages? How can educational systems encourage multi-literacies to help children maintain home languages? Since Sanz (2013) has suggested that biliteracy in typologically similar languages can help enhance metalinguistic awareness for additional language learning, can biliteracy in two distant languages be even more advantageous? Research focus should move beyond children to adolescents and their continued development of home and school languages in terms of CLI, multidirectional transfer, and metalinguistic awareness. Finally, what about elderly multilinguals? Which of their languages (L1, L2, Ln) are affected more with cognitive decline? And how can participation in language-specific communities of practice at different ages encourage lifelong multilingualism in monolingual societies? More studies are of course needed to determine the extent to which multilinguals do benefit from their prior experience of knowing more than one language in learning and maintaining additional languages (Aronin & Hufeisen, 2009). We still have only a very limited understanding of the complicated mechanisms underlying multilingual transfer and need to build up more research into multilinguals with three or more languages from typologically different language families beyond the Indo-European language context. We know that learners’ L1, L2, and Ln interact with each other, but we do not know to what extent language distance influences learners’ types and processes of learning and their ultimate attainment. Hammarberg (2009) has claimed that it is the learner’s perceived distance or psychotypology rather than the actual linguistic distance that is a more powerful trigger for linguistic transfer. And to make matters even more complex, the perceived distance for learners of multiple languages changes over time. What similarities do learners perceive and how do these similarities aid or hinder additional language learning? Research on multilingualism in terms of acquisition/(re)learning and attrition/ maintenance and the accompanying effects of crosslinguistic interaction is much needed to advance the field.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

560

S U Z A N N E Q U AY A N D S I M O N A M O N TA N A R I

7 Conclusion Altogether, the findings reviewed in this chapter and the questions they raise indicate that the acquisition of multiple languages – and hence multilingualism – is a complex phenomenon that is influenced by a host of linguistic and extralinguistic forces. It is true that similar variables also impact bilingualism. It is also true that, unlike monolinguals, both bilinguals and multilinguals have to learn, store, process, and use more than one language system that can interact with each other. In this respect, it can be argued that bilingualism and multilingualism share more similarities with one another than bilingualism and monolingualism. After all, only bilingual and multilingual individuals, but not monolinguals, can experience phenomena such as code-switching and transfer. Yet, the myriad of ways that linguistic and nonlinguistic forces can combine and operate when more than two languages are at play brings about a level of complexity and diversity that is unique to the learning and use of multiple languages. Sociolinguistic ties in multiple language contexts, including the home, work, and social relations, as well as connections to ethnic communities, in the case of migration, are all factors that increase the dynamic nature of multilingual identity and language use. Therefore, multilingualism should not be seen as a variant of bilingualism but should rather be studied in its own right as further evidence of the human potential and capacity for language. The research done so far on multilingualism, both as individual and social phenomena, has shown that stimulating insights into the acquisition and attrition processes can arise from a focus on learning three or more languages from multidisciplinary areas such as psycholinguistics and linguistics, applied linguistics, second language acquisition, sociolinguistics, neurolinguistics, and education. As discussed, research on many of the issues raised in this chapter is limited and many more studies are needed to understand what makes multilingualism unique or rather similar with respect to bilingualism.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:00:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.028

References

Aalberse, S., Zou, Y., & Andringa, S. (2017). Extended use of demonstrative pronouns in two generations of Mandarin Chinese speakers in the Netherlands. Evidence of convergence? In E. Blom, L. Cornips, & J. Schaeffer (Eds.), Cross-linguistic influence in bilingualism (pp. 25–48). Amsterdam: John Benjamins. Abrahamsson, N. & Hyltenstam, K. (2009). Age of onset and nativelikeness in a second language: Listener perception versus linguistic scrutiny. Language Learning, 59(2), 249–306. Abu-Rabia, S. & Sanitsky, E. (2010). Advantages of bilinguals over monolinguals in learning a third language. Bilingual Research Journal, 33(2), 173–199. Abutalebi, J. (2008). Neural aspects of second language representation and language control. Acta Psychologica, 128(3), 466–478. Abutalebi, J., Cappa, S. F., & Perani, D. (2001). The bilingual brain as revealed by functional neuroimaging. Bilingualism: Language and Cognition, 4(2), 179–190. Abutalebi, J. & Green, D. (2007). Bilingual language production: The neurocognition of language representation and control. Journal of Neurolinguistics, 20(3), 242–275. Abutalebi, J., Miozzo, A., & Cappa, S. F. (2000). Do subcortical structures control “language selection” in polyglots? Evidence from pathological language mixing. Neurocase, 6(1), 51–56. Adamson, B. & Feng, A. (2009). A comparison of trilingual education policies for ethnic minorities in China. Compare, 39(3), 321–333. Adamuti-Trache, M. (2013). Language acquisition among adult immigrants in Canada: The effect of premigration language capital. Adult Education Quarterly, 63(2), 103–126. Admiraal, W., Westhoff, G., & de Bot, K. (2006). Evaluation of bilingual secondary education in the Netherlands: Students’ language proficiency in English. Educational Research and Evaluation, 12(1), 75–93.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

562

References

Ag, A. & Jørgensen, J. N. (2013). Ideologies, norms, and practices in youth poly-languaging. International Journal of Bilingualism, 17(4), 525–539. Ag˘ırdag˘, O., Jordens, K., & Van Houtte, M. (2014). Speaking Turkish in Belgian schools: Teacher beliefs versus effective consequences. Bilig, 70(3), 7–28. Ag˘ırdag˘, O. & Vanlaar, G. (2018). Does more exposure to the language of instruction lead to higher academic achievement? A cross-national examination. International Journal of Bilingualism, 22(1), 123–137. Alamillo, R. & Collingwood, L. (2017). Chameleon politics: Social identity and racial cross-over appeals. Politics, Groups, and Identities, 5(4), 533–560. Albirini, A. (2014). Toward understanding the variability in the language proficiencies of Arabic heritage speakers. International Journal of Bilingualism, 18(6), 730–765. Alderson, J. C. (1980). Native and nonnative speaker performance on cloze tests. Language Learning, 30(1), 59–76. Alladi, S., Bak, T. H., Mekala, S., Rajan, A., Chaudhuri, J. R., Mioshi, E., Krovvidi, R., Surampudi, B., & Kaul, S. (2016). Impact of bilingualism on cognitive outcome after stroke. Stroke, 47(1), 258–261. ALTE (2016). Language tests for access, integration and citizenship: An outline for policy makers. Cambridge: Association of Language Testers in Europe. Altman, C., Armon-Lotem, S., Fichman, S., & Walter, J. (2016). Macrostructure, microstructure, and mental state terms in the narratives of English–Hebrew bilingual preschool children with and without specific language impairment. Applied Psycholinguistics, 37, 165–193. Altman, C., Burstein Feldman, Z., Yitzhaki, D., Armon-Lotem, S., & Walters, J. (2014). Family language policies, reported language use and proficiency in Russian–Hebrew bilingual children in Israel. Journal of Multilingual & Multicultural Development, 35(3), 216–234. Altman, C., Schrauf, R., & Walters, J. (2013). Crossovers and codeswitching in the investigation of immigrant autobiographical memory. In L. Isurin & J. Altarriba (Eds.), Memory, language, and bilingualism: Theoretical and applied approaches (pp. 211–235). Cambridge: Cambridge University Press. Ambridge, B., Kidd, E., Rowland, C., & Theakston, A. (2015). The ubiquity of frequency effects in first language acquisition. Journal of Child Language, 42(2), 239–273. Amengual-Pizarro, M. & Prieto-Arranz, J. I. (2015). Exploring affective factors in L3 learning: CLIL vs. non-CLIL. In M. Juan-Garau & J. SalazarNoguera (Eds.), Content-based language learning in multilingual educational environments (pp. 197–220). Berlin: Springer. American Council on the Teaching of Foreign Languages (ACTFL). (no date). Assigning CEFR ratings to ACTFL assessments. Retrieved from www.actfl.org/

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

563

sites/default/files/reports/Assigning_CEFR_Ratings_To_ACTFL_Assessme nts.pdf, last accessed February 1, 2018. Ammerlaan, T. (1996). “You get a bit wobbly . . . ” – Exploring lexical retrieval processes in the context of first language attrition. Nijmegen: Katholieke Universiteit Nijmegen. Andersen, G., Furiassi, C., & Misˇic´ Ilic´, B. (2017). The pragmatic turn in studies of linguistic borrowing. Journal of Pragmatics, 113, 71–76. Andersen, S. L., Thompson, A. T., Rutstein, M., Hostetter, J. C., & Teicher, M. H. (2000). Dopamine receptor pruning in prefrontal cortex during the periadolescent period in rats. Synapse, 37(2), 167–169. Anderson, M. C., Bjorn, R. A., & Bjorn, E. L. (1994). Remembering can cause forgetting: Retrieval dynamics in long-term memory. Journal of Experimental Psychology: Learning, Memory, & Cognition, 20(5), 1063–1087. Anderson, M. C. & Spellman, B. A. (1995). On the status of inhibitory mechanisms in cognition: Memory retrieval as a model case. Psychological Review, 102(1), 68–100. Andrew, P. (2012). The social construction of age: Adult foreign language learners. Bristol: Multilingual Matters. Andringa, S. (2014). The use of native speaker norms in critical period hypothesis research. Studies in Second Language Acquisition, 36(3), 565–596. Androutsopoulos, J. (2007). Language choice and code-switching in German-based diasporic web forums. In B. Danet & S. C. Herring (Eds.), The multilingual Internet: Language, culture, and communication online (pp. 340–361). Oxford: Oxford University Press. Angelelli, C. V. (2016). Looking back: A study of (ad-hoc) family interpreters. European Journal of Applied Linguistics, 4(1), 5–31. Angermeyer, P. S. (2015). Speak English or what?: Codeswitching and interpreter use in New York City courts. Oxford: Oxford University Press. Ansaldo, A. I. & Ghazi-Saidi, L. (2014). Aphasia therapy in the age of globalization: Cross-linguistic therapy effects in bilingual aphasia. Behavioural Neurology, Article ID 603085. Ansaldo, A. I., Ghazi-Saidi, L., & Ruiz, A. (2010). Model-driven intervention in bilingual aphasia: Evidence from a case of pathological language mixing. Aphasiology, 24(2), 309–324. Anto´n, E., Dun˜abeitia, J. A., Este´vez, A., Herna´ndez, J. A., Castillo, A., Fuentes, L. J., Davidson, D. J., & Carreiras, M. (2014). Is there a bilingual advantage in the ANT task? Evidence from children. Frontiers in Psychology: Language Sciences, 5, Article 398. Anto´n, E., Ferna´ndez-Garcı´a, Y., Carreiras, M., & Dun˜abeitia, J. A. (2016). Does bilingualism shape inhibitory control in the elderly? Journal of Memory & Language, 90, 147–160. Anto´n, E., Thierry, G., Goborov, A., Anasagasti, J., & Dun˜abeitia, J. A. (2016). Testing bilingual educational methods: A plea to end the language-mixing taboo. Language Learning, 66(S2), 29–50.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

564

References

Antoniou, M., Gunasekera, G. M., & Wong, P. C. M. (2013). Foreign language training as cognitive therapy for age-related cognitive decline: A hypothesis for future research. Neuroscience & Biobehavioral Reviews, 37(10), 2689–2698. Aram, D., Bates, E., Eisele, J., Fenson, J., Nass, R., Thal, D., & Trauner, D. (1997). From first words to grammar in children with focal brain injury. Developmental Neuropsychology, 13(3), 275–343. Arcand, J.-L. & Grin, F. (2013). Language in economic development: Is English special and is linguistic fragmentation bad? In E. Erling & P. Seargeant (Eds.), English and development: Policy, pedagogy and globalization (pp. 243–266). Bristol: Multilingual Matters. Archila-Suerte, P., Zevin, J., Bunta, F., & Hernandez, A. E. (2012). Age of acquisition and proficiency in a second language independently influence the perception of non-native speech. Bilingualism: Language and Cognition, 15(1), 190–201. Archila-Suerte, P., Zevin, J., & Hernandez, A. E. (2015). The effect of age of acquisition, socioeducational status, and proficiency on the neural processing of second language speech sounds. Brain and Language, 141, 35–49. Archila-Suerte, P., Zevin, J., Ramos, A. I., & Hernandez, A. E. (2013). The neural basis of non-native speech perception in bilingual children. Neuroimage, 67, 51–63. Ardila, A., & Ramos, E. (2008). Normal and abnormal aging in bilinguals. Dementia and Neuropsychologia, 2(4), 242–247. Armon-Lotem, S., de Jong, J., & Meir, N. (2015). Assessing multilingual children: Disentangling bilingualism from language impairment. Bristol: Multilingual Matters. Armon-Lotem, S., Joffe, S., Abutbul-Oz, H., Altman, C., & Walters, J. (2014). Language exposure, ethnolinguistic identity and attitudes in the acquisition of Hebrew as a second language among bilingual preschool children from Russian and English-speaking backgrounds. In T. Gru¨ter & J. Paradis (Eds.), Input and experience in bilingual development (pp. 77–98). Amsterdam: John Benjamins. Armon-Lotem, S., Walters, J., & Gagarina, N. (2011). The impact of internal and external factors on linguistic performance in the home language and in L2 among Russian–Hebrew and Russian–German preschool children. Linguistic Approaches to Bilingualism, 1(3), 291–317. Aronin, L. & Hufeisen, B. (2009). The exploration of multilingualism. Amsterdam: John Benjamins. Aronin, L. & Singleton, D. (2012). Multilingualism. Amsterdam: John Benjamins. Ashaie, S. & Obler, L. K. (2014). Effect of age, education, and bilingualism on confrontation naming in older illiterate and low-educated populations. Behavioural Neurology, Article ID 970520.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

565

Athanasopoulos, P., Bylund, E., Montero-Melis, G., Damjanovic, L., Schartner, A., Kibbe, A., Riches, N., & Thierry, G. (2015). Two languages, two minds: Flexible cognitive processing driven by language of operation. Psychological Science, 26(4), 518–526. Athanasopoulos, P., Damjanovic, L., Krajciova, A., & Sasaki, M. (2011). Representation of colour concepts in bilingual cognition: The case of Japanese blues. Bilingualism: Language and Cognition, 14(1), 9–17. Atkinson, D. (Ed.). (2011). Alternative approaches to second language acquisition. London: Routledge. Au, T. K., Knightly, L. M., Jun, S. A., & Oh, J. S. (2002). Overhearing a language during childhood. Psychological Science, 13(3), 238–243. Auer, P. (Ed.). (1998). Code-switching in conversation: Language, interaction and identity. London: Routledge. Auer, P., Barden, B., & Grosskopf, B. (2000). Long-term linguistic accommodation and its sociolinguistic interpretation: Evidence from the inner-German migration after the Wende. In K. Mattheier (Ed.), Dialect and migration in a changing Europe (pp. 79–98). Bern: Peter Lang. Auerbach, E. R. (1993). Reexamining English only in the ESL classroom. TESOL Quarterly, 27(1), 9–32. August, D. & Shanahan, T. (2006). Developing literacy in second-language learners: Report of the National Literacy Panel on Language-Minority Children and Youth. Mahwah, NJ: Lawrence Erlbaum. Babel, A. (2014). Time and reminiscence in contact: Dynamism and stasis in contact-induced change. Spanish in Context, 11(3), 311–334. Bacchi, C. (2009). Analysing policy: What’s the problem represented to be? Sydney: Pearson Education. Bachman, L. F. & Palmer, A. (2010). Language assessment in practice. Oxford: Oxford University Press. Backus, A. (1996). Two in one: Bilingual speech of Turkish immigrants in the Netherlands. Tilburg: Tilburg University Press. Backus, A. (1998). Turks–Nederlandse codewisseling. Universele en taalspecifieke aspecten van taalcontact [Turkish–Dutch code-switching: Universal and language specific aspects of language contact]. Nederlandse Taalkunde, 3, 2–17. Baddeley, A. D. (2015). Working memory in second language learning. In Z. Wen, M. B. Mota, & A. McNeill (Eds.), Working memory in second language acquisition and processing (pp. 17–28). Bristol: Multilingual Matters. Baetens Beardsmore, H. (1982). Bilingualism: Basic principles. Clevedon: Tieto Ltd. Bahrick, H. P., Hall, L. K., Goggin, J. P., Bahrick, L. E., & Berger, S. A. (1994). Fifty years of language maintenance and language dominance in bilingual Hispanic immigrants. Journal of Experimental Psychology: General, 123(3), 264.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

566

References

Bahtina, D. & ten Thije, J. D. (2013). Receptive multilingualism. In C. A. Chapelle (Ed.), The encyclopedia of applied linguistics. Blackwell Publishing Ltd. doi: 10.1002/9781405198431.wbeal1001 Baker, A. & van den Bogaerde, B. (2008). Code-mixing in signs and words in input to and output from children. In C. Plaza-Pust & E. Morales-Lo´pez (Eds.), Sign bilingualism: Language development, interaction, and maintenance in sign language contact situations (pp. 1–28). Amsterdam: John Benjamins. Baker, C. (2011). Foundations of bilingual education and bilingualism (5th edn.). Bristol: Multilingual Matters. Bakhtin, M. (1981). Dialogic imagination: Four essays. Austin: University of Texas Press. Bakker, P. (1997). A language of our own: The genesis of Michif, the mixed Cree– French language of the Canadian Me´tis. Oxford: Oxford University Press. Ballantyne, P. J., Yang, M., & Boon, H. (2013). Interpretation in cross-language research: Tongues tied in the health care interview? Journal of Cross Cultural Gerontology, 28(4), 391–405. Ballinger, S., & Lyster, R. (2011). Student and teacher language use in a two-way Spanish/English immersion school. Language Teaching Research, 15(3), 289–306. Ballinger, S., Lyster, R., Sterzuk, A., & Genesee, F. (2017). Context-appropriate crosslinguistic pedagogy: Considering the role of language status in immersion. Journal of Immersion and Content-Based Language Education, 5(1), 30–57. Banegas, D. L. (2014). An investigation into CLIL-related sections of EFL coursebooks: Issues of CLIL inclusion in the publishing market. International Journal of Bilingual Education and Bilingualism, 17(3), 345–359. Barbu, S., Martin, N., & Chevrot, J.-P. (2014). The maintenance of regional dialects: A matter of gender? Boys, but not girls, use local varieties in relation to their friends’ nativeness and local identity. Frontiers in Psychology, 5, Article 1251. Barbu, S., Nardy, A., Chevrot, J.-P., & Juhel, J. (2013). Language evaluation and use during early childhood: Adhesion to social norms or integration of environmental regularities? Linguistics, 51(2), 281–411. Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in Cognitive Sciences, 6(6), 248–254. Bates, E., Dale, P., & Thal, D. (1995). Individual differences and their implications for theories of language development. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 96–151). Oxford: Blackwell. Bates, E. & Goodman, J. C. (1999). On the emergence of grammar from the lexicon. In B. MacWhinney (Ed.), The emergence of language (pp. 29–80). Mahwah, NJ: Lawrence Erlbaum.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

567

Baynham, M., Roberts, C., Cooke, M., Simpson, J., Ananiadou, K., Callaghan, J., McGoldrick, J., & Wallace, C. (2007). Effective teaching and learning: ESOL. London: National Research and Development Centre. Beacco, J.-C., Krumm, H.-J., & Little, D. (Eds.). (2017). The linguistic integration of adult migrants: Some lessons from research. Berlin: De Gruyter/ Strasbourg: Council of Europe. Beacco, J.-C., Little, D., & Hedges, C. (2014). Linguistic integration of adult migrants. Guide to policy development and implementation. Strasbourg: Council of Europe Publishing. Bedore, L. M., & Pen˜a, E. D. (2008). Assessment of bilingual children for identification of language impairment: Current findings and implications for practice. International Journal of Bilingualism, 11(1), 1–29. Bedore, L. M., Pen˜a, E. D., Griffin, Z. M., & Hixon, J. G. (2016). Effects of age of English exposure, current input/output, and grade on bilingual language performance. Journal of Child Language, 43(3), 687–706. Behrens, H. (2000). Lexical representation and acquisition theory. Psychology of Language and Communication, 4(2), 33–44. Bell, J. S. (1995.) The relationship between L1 and L2 literacy: Some complicating factors. TESOL Quarterly, 29(4), 687–704. Benditt, T. M. (1975). The concept of interest in political theory. Political Theory, 3(3), 245–258. Benmamoun, E., Montrul, S., & Polinsky, M. (2013). Defining an “ideal” heritage speaker: Theoretical and methodological challenges. Reply to peer commentaries. Theoretical Linguistics, 39(3–4), 259–294. Benson, M. & O’Reilly, K. (2009). Migration and the search for a better way of life: A critical exploration of lifestyle migration. Sociological Review, 57(4), 608–625. Bergmann, C., Meulman, N., Stowe, L. A., Sprenger, S. A., & Schmid, M. S. (2015). Prolonged L2 immersion engenders little change in morphosyntactic processing of bilingual natives. NeuroReport, 26(17), 1065–1070. Bergmann, C., Sprenger, S. A., & Schmid, M. S. (2015). The impact of language co-activation on L1 and L2 speech fluency. Acta Psychologica, 161, 25–35. Berk-Seligson, S. (1990). The bilingual courtroom: Court interpreters in the judicial process. Chicago: University of Chicago Press. Berk-Seligson, S. (2009). Coerced confessions: The discourse of bilingual police interrogations. Berlin: Mouton de Gruyter. Bernstein, R. (1994). Dictatorship of virtue: Multiculturalism and the battle for America’s future. New York: Knopf. Berthele, R. (2012). The influence of code-mixing and speaker information on perception and assessment of foreign language proficiency: An experimental study. International Journal of Bilingualism, 16(4), 453–466. Bhatia, T. K. & Ritchie, C. R. (2013). Bilingualism and multilingualism in South Asia. In T. K. Bhatia & W. C. Ritchie (Eds.), The handbook of

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

568

References

bilingualism and multilingualism (2nd ed., pp. 843–870). Malden, MA: Wiley-Blackwell. Bialystok, E. (2015). Bilingualism and the development of executive function: The role of attention. Child Development Perspectives, 9(2), 117–121. Bialystok, E. (2017). The bilingual adaptation: How minds accommodate experience. Psychological Bulletin, 143(3), 233–262. Bialystok, E. & Craik, F. I. (2010). Cognitive and linguistic processing in the bilingual mind. Current Directions in Psychological Science, 19(1), 19–23. Bialystok, E., Craik, F. I., Green, D. W., & Gollan, T. H. (2009). Bilingual minds. Psychological Science in the Public Interest, 10(3), 89–129. Bialystok, E., Craik, F. I., & Luk, G. (2008). Lexical access in bilinguals: Effects of vocabulary size and executive control. Journal of Neurolinguistics, 21(6), 522–538. Bialystok, E. & Luk, G. (2012). Receptive vocabulary differences in monolingual and bilingual adults. Bilingualism: Language and Cognition, 15(2), 397–401. Bialystok, E., Luk, G., Peets, K. F., & Yang, S. (2010). Receptive vocabulary differences in monolingual and bilingual children. Bilingualism: Language and Cognition, 13(4), 525–531. Bianchi, L. (1886). Un caso di sordita verbale. Rivista Sperimentale di Freniatria, 12, 57–71. Bice, K. & Kroll, J. F. (2015). Native language change during early stages of second language learning. NeuroReport, 26(16), 966–971. Biedron´, A. (2012). Cognitive-affective profile of gifted adult foreign language learners. Słupsk: Wydawnictwo Akademii Pomorskiej w Słupsku. Biedron´, A. & Pawlak, M. (2016). New conceptualizations of linguistic giftedness. Language Teaching, 49(2), 151–185. Biedron´, A., & Szczepaniak, A. (2012). Working-memory and short-term memory abilities in accomplished multilinguals. Modern Language Journal, 96(2), 290–306. Bigelow, M. (2010). Mogadishu on the Mississippi: Language, racialized identity, and education in a new land. Malden, MA: Wiley-Blackwell. Bigelow, M. & King, K. A. (2015). Somali immigrant youths and the power of print literacy. Writing Systems Research, 7(1), 4–19. Birdsong, D. (1996). An interesting subject indeed. Review of N. Smith and M.-I. Tsimpli, The mind of a savant: Language learning and modularity. Contemporary Psychology, 41, 837–838. Birdsong, D. (2006). Age and second language acquisition and processing: A selective overview. Language Learning, 56(1), 9–49. Birdsong, D. (2007). Nativelike pronunciation among late learners of French as a second language. In O. S. Bohn & M. J. Munro (Eds.), Language experience in second language speech learning (pp. 99–116). Amsterdam: John Benjamins. Birdsong, D. (2013). The Critical Period Hypothesis for second language acquisition: Tailoring the coat of many colours. In M. Pawlak &

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

569

L. Aronin (Eds.), Essential topics in applied linguistics and multilingualism: Studies in honor of David Singleton (pp. 43–50). Berlin: Springer. Birdsong, D. (2016). Dominance in bilingualism: Foundations of measurement, with insights from the study of handedness. In C. Silva-Corvala´n & J. Treffers-Daller (Eds.), Language dominance in bilinguals: Issues of measurement and operationalization (pp. 85–105). Cambridge: Cambridge University Press. Bishop, M., Hicks, S., Bertone, A., & Sala, R. (2006). Capitalizing on simultaneity: Features of bimodal bilingualism in hearing Italian native signers. In C. Lucas (Ed.), Multilingualism and sign languages: From the Great Plains to Australia (pp. 79–118). Washington, DC: Gallaudet University Press. Black, R. W. (2006). Language, culture, and identity in online fanfiction. E-Learning and Digital Media, 3(2), 170–184. Blackledge, A. & Creese, A. (2010). Multilingualism: A critical perspective. London: Continuum. Block, D. (2007). Second language identities. London: Continuum. Block, D. & Corona, V. (2016). Intersectionality in language and identity research. In S. Preece (Ed.), The Routledge handbook of language and identity (pp. 507–522). London: Routledge. Blommaert, J. (2009). Language, asylum, and the national order. Current Anthropology, 50(4), 415–441. Blommaert, J. (2010). The sociolinguistics of globalization. Cambridge: Cambridge University Press. Blommaert, J. (2015). Commentary: Superdiversity old and new. Language and Communication, 44(1), 82–89. Blommaert, J., Leppa¨nen, S., Pahta, P., & Ra¨isa¨nen, T. (Eds.). (2012). Dangerous multilingualism: Northern perspectives on order, purity and normality. London: Palgrave Macmillan. Boissevain, J. (2013). Factions, friends and feasts: Anthropological perspectives on the Mediterranean. New York: Berghahn Books. Boroditsky, L., Ham, W., & Ramscar, M. (2002). What is universal in event perception? Comparing English & Indonesian speakers. In W. D. Gray & C. D. Schunn (Eds.), Proceedings of the 24th annual conference of the Cognitive Science Society (pp. 136–141). Mahwah, NJ: Lawrence Erlbaum. Bosch, L. & Ramon-Casas, M. (2014). First translation equivalents in bilingual toddlers’ expressive vocabulary: Does form similarity matter? International Journal of Behavioral Development, 38(4), 317–322. Bosch, L. & Sebastia´n-Galle´s, N. (2001). Evidence of early language discrimination abilities in infants from bilingual environments. Infancy, 2(1), 29–49. Bosch, L. & Sebastia´n-Galle´s, N. (2003). Simultaneous bilingualism and the perception of a language-specific vowel contrast in the first year of life. Language and Speech, 46(2–3), 217–243.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

570

References

Bourdieu, P. (1991). Language and symbolic power. Cambridge, MA: Harvard University Press. Bowden, H. W. (2016). Assessing second-language oral proficiency for research: The Spanish Elicited Imitation Task. Studies in Second Language Acquisition, 38(4), 647–675. Bowers, J., Mattys, S., & Gage, S. (2009). Preserved implicit knowledge of a forgotten childhood language. Psychological Science, 20(9), 1064–1069. Branzi, F. M., Calabria, M., Boscarino, M. L., & Costa, A. (2016). On the overlap between bilingual language control and domain-general executive control. Acta Psychologia, 166, 21–30. Branzi, F. M., Calabria, M., Gade, M., Fuentes, L. J., & Costa, A. (2018). On the bilingualism effect in task switching. Bilingualism: Language and Cognition, 21(1),195–208. Braun, A. (2012). Language maintenance in trilingual families: A focus on grandparents. International Journal of Multilingualism, 9(4), 423–436. Braun, A. & Cline, T. (2014). Language strategies for trilingual families: Parents’ perspectives. Bristol: Multilingual Matters. Braunmu¨ller, K. (2007). Receptive multilingualism in Northern Europe in the Middle Ages: A description of a scenario. In J. D. ten Thije & L. Zeevaert (Eds.), Receptive multilingualism: Linguistic analyses, language policies and didactic concepts (pp. 25–47). Amsterdam: John Benjamins. Brenner, P. (2004). New state phases. Oxford: Oxford University Press. Briggs, J. G. (2015). Out-of-class language contact and vocabulary gain in a study abroad context. System, 53, 129–140. Brohy, C. (2001). Generic and/or specific advantages of bilingualism in a dynamic plurilingual situation: The case of French as official L3 in the school of Samedan (Switzerland). International Journal of Bilingual Education and Bilingualism, 4(1), 38–49. Brooks, R., Water, J. & Pimlott-Wilson, H. (2012). International education and the employability of UK students. British Educational Research Journal, 38(2), 281–298. Brown, A. & Gullberg, M. (2008). Bidirectional crosslinguistic influence in L1–L2 encoding of manner in speech and gesture: A study of Japanese speakers of English. Studies in Second Language Acquisition, 30(2), 225–251. Brown, P. & Levinson, S. (1987). Politeness. Some universals in language usage. Cambridge: Cambridge University Press. Brown-Blake, C. N. & Chambers, P. (2007). The Jamaican Creole speaker in the UK criminal justice system. International Journal of Speech Language and the Law, 14(2), 269–294. Brownell, M. A. (2012). Receptive and expressive one-word picture vocabulary tests, Fourth edition. Spanish-bilingual versions. San Antonio Pearson. Brysbaert, M. & Duyck, W. (2010). Is it time to leave behind the Revised Hierarchical Model of bilingual language processing after fifteen years of service? Bilingualism: Language and Cognition, 13(3), 359–371.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

571

Brysbaert, M., Stevens, M., Mandera, P. & Keuleers, E. (2016). How many words do we know? Practical estimates of vocabulary size dependent on word definition, the degree of language input and the participant’s age. Frontiers in Psychology, 7, Article 1116. Burns, T. C., Yoshida, K. A., Hill, K., & Werker, J. F. (2007). The development of phonetic representation in bilingual and monolingual infants. Applied Psycholinguistics, 28(3), 455–474. Butler, J., Floccia, C., Goslin, J., & Panneton, R. (2011). Infants’ discrimination of familiar and unfamiliar accents in speech. Infancy, 16(4), 392–417. Buttelmann, D., Zmyj, N., Daum, M. M., & Carpenter, M. (2013). Selective imitation of in-group over out-group members in 14-month-olds. Child Development, 84(2), 422–428. Byers-Heinlein, K. (2013). Parental language mixing: Its measurement and the relation of mixed input to young bilingual children’s vocabulary size. Bilingualism: Language and Cognition, 16(1), 32–48. Byers-Heinlein, K., Behrend, D., Said, L. M., Giris, H., & Poulin-Dubois, D. (2017). Monolingual and bilingual children’s social preferences for monolingual and bilingual speakers. Developmental Science, 20(4), e12392. Byers-Heinlein, K., Burns, T. C., & Werker, J. F. (2010). The roots of bilingualism in newborns. Psychological Science, 21(3), 343–348. Bylund, E. (2009). Maturational constraints and first language attrition. Language Learning, 59(3), 687–715. Bylund, E., Abrahamsson, N., & Hyltenstam, K. (2012). Does first language maintenance hamper nativelikeness in a second language? Studies in Second Language Acquisition, 34(2), 215–241. Cable, C., Driscoll, P., Mitchell, R., Sing, S., Cremin, T., Earl, J., Eyres, I., Holmes, B., Martin, C., & Heins, B. (2010). Language learning at Key Stage 2: A longitudinal study final report (DCSF Research Reports DCSF-RR198). London: Department for Children, Schools and Families. Caldwell-Harris, C. L. (2014). Emotionality differences between a native and foreign language: Theoretical implications. Frontiers in Psychology, 5, Article 1055. Cameron-Faulkner, T., Theakston, A., Lieven, E., & Tomasello, M. (2015). The relationship between infant holdout and gives, and pointing. Infancy, 20(5), 576–586. Cammarota, M. A. & Porquier, R. (1986). Difficulties in the acquisition of a foreign-language by Latin-American political refugees. Langue Franc¸aise, 71, 101–116. Campbell, L. & Grodona, V. (2010). Who speaks what to whom? Multilingualism and language choice in Misio´n La Paz. Language in Society, 39(5), 617–646. Canagarajah, A. S. (2006). Toward a writing pedagogy of shuttling between languages: Learning from multilingual writers. College English, 68(6), 589–604.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

572

References

Canagarajah, S. (2013). Translingual practice: Global Englishes and cosmopolitan relations. London: Routledge. Canagarajah, S. & Liyanage, I. (2012). Lessons from pre-colonial multilingualism. In M. Martin-Jones, A. Blackledge, & A. Creese (Eds.), The Routledge handbook of multilingualism. London: Routledge. Caps, D. (2008). The decades of life: A guide to human development. Louisville Westminster John Knox Press. Caramazza, A., & Zurif, E. B. (1978). Language acquisition and language breakdown: Parallels and divergencies. Baltimore: Johns Hopkins University Press. Carre`re, C. & Masood, M. (2016). Poids e´conomique de la francophonie: son impact via l’ouverture commerciale. In C. Carre`re (Ed.), L’impact e´conomique des langues (pp. 71–92). Paris: Economica. Carroll, J. B. & Sapon, S. M. (1959/2002). Modern Language Aptitude Test. San Antonio The Psychological Corporation. Cavanaugh, J. R. (2013). Language ideologies and language attitudes: A linguistic anthropological perspective. In P. Auer, J. Caro Reina, & G. Kaufmann (Eds.), Language variation: European perspectives IV: Selected papers from the Sixth International Conference on Language Variation in Europe (pp. 45–56). Amsterdam: John Benjamins. Cenoz, J. (2001). The effect of linguistic distance, L2 status and age on crosslinguistic influence in third language acquisition. In J. Cenoz, B. Hufeisen, & U. Jessner (Eds.), Cross-linguistic influence in third language acquisition: Psycholinguistic perspectives (pp. 8–20). Clevedon: Multilingual Matters. Cenoz, J. (2003). The additive effect of bilingualism on third language acquisition: A review. International Journal of Bilingualism, 7(1), 71–87. Cenoz, J. (2009). Towards multilingual education: Basque educational research from an international perspective. Bristol: Multilingual Matters. Cenoz, J. & Gorter, D. (2006). Linguistic landscape and minority languages. International Journal of Multilingualism, 3(1), 67–80. Chamorro, G., Sorace, A., & Sturt, P. (2015). What is the source of L1 attrition? The effect of recent L1 re-exposure on Spanish speakers under L1 attrition. Bilingualism: Language and Cognition, 19(3), 520–532. Chan, A. (2010). The Cantonese double object construction with bei2 “give” in bilingual children: The role of input. International Journal of Bilingualism, 14(1), 65–85. Chan, B. H.-S. (2008). Code-switching, word order and the lexical/functional category distinction. Lingua, 118(6), 777–809. Chang, C. B. (2012). Rapid and multifaceted effects of second-language learning on first-language speech production. Journal of Phonetics, 40(2), 249–268. Chein, J. M. & Schneider, W. (2012). The brain’s learning and control architecture. Current Directions in Psychological Science, 21(2), 78–84.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

573

Chemero, A. (2013). Radical embodied cognitive science. Review of General Psychology, 17(2), 145–150. Cherciov, M. (2013). Investigating the impact of attitude on first language attrition and second language acquisition from a Dynamic Systems Theory perspective. International Journal of Bilingualism, 17(6), 716–733. Chevalier, S. (2015). Trilingual language acquisition: Contextual factors influencing active trilingualism in early childhood. Amsterdam: John Benjamins. Chiappe, P., & Siegel, L. S. (1999). Phonological awareness and reading acquisition in English- and Punjabi-speaking Canadian children. Journal of Educational Psychology, 91(1), 20–28. Chiat, S., Armon-Lotem, S., Marinis, T., Polisˇenka´, K., Roy, P., & SeeffGabriel, B. (2013). Assessment of language abilities in sequential bilingual children: The potential of sentence imitation tasks. In V. C. M. Gathercole (Ed.), Issues in the assessment of bilinguals (pp. 56–89). Bristol: Multilingual Matters. Chiswick, B. & Miller, P. (Eds.). (2007). The economics of language: International analyses. London: Routledge. Chiswick, B. & Miller, P. (2014). International migration and the economics of language. IZA Discussion Paper 7880. http://repec.iza.org/dp7880.pdf, last accessed January 19, 2018 Choi, J. (2014). Rediscovering a forgotten language. Nijmegen: Max Planck Institute for Psycholinguistics. Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press. Churchill, E. & DuFon, M. A. (2006). Evolving threads in study abroad research. In M. A. DuFon & E. Churchill (Eds.), Language learners in study abroad contexts (pp. 1–27). Clevedon: Multilingual Matters. Clahsen, H. & Felser, C. (2006). Grammatical processing in language learners. Applied Psycholinguistics, 27(1), 3–42. Clyne, M. (1977). Bilingualism in the elderly. Talanya, 4, 45–65. Clyne, M. (2005). Australia’s language potential. Sydney: University of New South Wales Press. Clyne, M. (2011). Bilingualism, code-switching, and aging: A myth of attrition and a tale of collaboration. In M. S. Schmid & W. Lowie (Eds.), Modeling bilingualism: From structure to chaos. In honor of Kees de Bot (pp. 201–220). Amsterdam: John Benjamins. Cogo, A. & Jenkins, J. (2010). English as a lingua franca in Europe: A mismatch between policy and practice. European Journal of Language Policy, 2(2), 271–94. Cohen, A. D. & Li, P. (2013). Learning Mandarin in later life: Can an old dog learn new tricks? Contemporary Foreign Language Studies, 396, 5–14. Cohen, J. (1983). The cost of dichotomization. Applied Psychological Measurement, 7(3), 249–253. Collier, V. (1989). How long? A synthesis of research on academic achievement in a second language. TESOL Quarterly, 23(3), 509–532.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

574

References

Collier, V. & Thomas, W. P. (2017). Validating the power of bilingual schooling: Thirty-two years of large-scale, longitudinal research. Annual Review of Applied Linguistics, 37, 203–217. Collins, L., Trofimovich, P., White, J., Cardoso, W. & Horst, M. (2009). Some input on the easy/difficult grammar question: An empirical study. Modern Language Journal, 93(3), 336–353. Collins, L. & White, J. (2011). An intensive look at intensity and language learning. TESOL Quarterly, 45(1), 106–133. Colome´, A. (2001). Lexical activation in bilinguals’ speech production: Language specific or language independent? Journal of Memory and Language, 45(4), 721–736. Comeau, L. & Genesee, F. (2001). Bilingual children’s repair strategies during dyadic communication. In J. Cenoz & F. Genesee (Eds.), Trends in bilingual acquisition (pp. 231–256). Amsterdam: John Benjamins. Comeau, L., Genesee, F., & Lapaquette, L. (2003). The Modeling Hypothesis and child bilingual codemixing. International Journal of Bilingualism, 7(2), 113–126. Communication of Rights Group (2016). Guidelines for communicating rights to non-native speakers of English in Australia, England and Wales, and the USA. www.aaal.org/?page=CommunicationRights, last accessed January 19, 2018. Conboy, B. T. & Thal, D. J. (2006). Ties between the lexicon and grammar: Cross-sectional and longitudinal studies of bilingual toddlers. Child Development, 77(3), 712–735. Conley, J. M. & O’Barr, W. M. (1990). Rules versus relationships: The ethnography of legal discourse. Chicago University of Chicago Press. Cook, V. (1991). The poverty-of-the-stimulus argument and multi-competence. Second Language Research, 7(2), 103–117. Cook, V. (2016). Premises of multicompetence. In V. Cook & Li Wei (Eds.), The Cambridge handbook of linguistic multi-competence (pp. 1–25). Cambridge: Cambridge University Press. Cook, V. & Li Wei (2016). The Cambridge handbook of linguistic multi-competence. Cambridge: Cambridge University Press. Cooke, M. (2006). “When I wake up I dream of electricity.” The lives, aspirations and “needs” of adult ESOL learners. Linguistics and Education, 17(1), 56–73. Cooke, M., Winstanley, B., & Bryers, D. (2015). Whose integration? A participatory ESOL project in the UK. In J. Simpson & A. Whiteside (Eds.), Adult language education and migration: Challenging agendas in policy and practice (pp. 214–224). London: Routledge. Cooper, A. C. & Nguye˜ˆ n, T. T. T. (2015). Signed language communityresearcher collaboration in Vieˆ t Nam: Challenging language ideolo˙ gies, creating social change. Journal of Linguistic Anthropology, 25(2), 105–127.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

575

Corder, S. P. (1967). The significance of learner’s errors. IRAL: International Review of Applied Linguistics in Language Teaching, 5(1–4), 161–170. Costa, A., Caramazza, A., & Sebastia´n-Galle´s, N. (2000). The cognate facilitation effect: Implications for models of lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26(5), 1283–1296. Costa, A., Foucart, A., Hayakawa, S., Aparici, M., Apesteguia, J., Heafner, J., & Keysar, B. (2014). Your morals depend on language. PLoS ONE, 9(4), e94842. Costa, A., Pannunzi, M., Deco, G., & Pickering, M. (2017). Do bilinguals automatically activate their native language when they are not using it? Cognitive Science, 41(6), 1629–1644. Costa, A. & Santesteban, M. (2004). Lexical access in bilingual speech production: Evidence from language switching in highly proficient bilinguals and L2 learners. Journal of Memory and Language, 50(4), 491–511. Costa, A., Santesteban, M., & Ivanova, I. (2006). How do highly proficient bilinguals control their lexicalization process? Inhibitory and language-specific selection mechanisms are both functional. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32(5), 1057–1074. Coulmas, F. (1987). Why speak English? In K. Knapp, W. Enninger, & A. Knapp-Potthoff (Eds.), Analyzing intercultural communication (pp. 95–107). Berlin: Mouton de Gruyter. Council of Europe (2001). Common European framework of reference for languages: Learning, teaching, assessment. Strasbourg: Language Policy Division/Cambridge: Cambridge University Press. Council of Europe (n.d.). Linguistic Integration of Adult Migrants’ (LIAM) Guiding principles. Retrieved from https://rm.coe.int/168049447e, last accessed January 31, 2018. Coupland, N. (2010). Welsh linguistic landscapes “from above” and “from below.” In A. Jaworski & C. Thurlow (Eds.), Semiotic landscapes: Language, image, space (pp. 77–101). London: Continuum. Cox, J. G. (2017). Explicit instruction, bilingualism, and the older adult learner. Studies in Second Language Acquisition, 39(1), 29–58. Cox, J. G. & Sanz, C. (2015). Deconstructing PI for the ages: Explicit instruction vs. practice in young and older adult bilinguals. International Review of Applied Linguistics, 53(2), 225–248. Cremona, C. & Bates, E. (1977). The development of attitudes toward dialect in Italian children. Journal of Psycholinguistic Research, 6(3), 223–232. Crews, F., He, J., & Hodge, C. (2007). Adolescent cortical development: A critical period of vulnerability for addiction. Pharmacology Biochemistry and Behavior, 86(2), 189–199.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

576

References

Crossley, S., Salsbury, T., & McNamara, D. (2010). The development of polysemy and frequency use in English second language speakers. Language Learning, 60(3), 573–605. Cruz-Ferreira, M. (2006). Three is a crowd? Acquiring Portuguese in a trilingual environment. Clevedon: Multilingual Matters. Cruz-Ferreira, M. (Ed.) (2010). Multilingual norms. Bern: Peter Lang. Csize´r, K., & Do¨rnyei, Z. (2005). The internal structure of language learning motivation and its relationship with language choice and learning effort. Modern Language Journal, 89(1), 19–36. Cummins, J. (1979a). Cognitive/academic language proficiency, linguistic interdependence, the optimum age question, and some other matters. Working Papers on Bilingualism, 19, 121–129. Cummins, J. (1979b). Linguistic interdependence and the educational development of bilingual children. Review of Educational Research, 49 (2), 222–251. Cummins, J. (2006). The relationship between American Sign Language proficiency and English academic development: A review of the research. Unpublished report for the Ontario Association of the Deaf, Toronto, Ontario. Retrieved from http://citeseerx.ist.psu.edu/view doc/download?doi=10.1.1.521.8612&rep=rep1&type=pdf, last accessed January 31, 2018. Cummins, J. (2007). Rethinking monolingual instructional strategies in multilingual classrooms. Canadian Journal of Applied Linguistics/Revue canadienne de linguistique applique´e, 10(2), 221–240. Cummins, J. (2016). Reflections on Cummins (1980), “The cross-lingual dimensions of language proficiency: Implications for bilingual education and the optimal age issue.” TESOL Quarterly, 50(4), 940–944. Curaming, R. & Calidjernih, F. (2014). From sentimentalism to pragmatism? Language-in-education policy-making in East Timor. In P. Sercombe & R. Tupas (Eds.), Language, education and nation-building: Assimilation and shift in Southeast Asia (pp. 68–86). Basingstoke: Palgrave Macmillan. Curdt-Christiansen, X.-L. (2016). Conflicting language ideologies and contradictory language practices in Singaporean multilingual families. Journal of Multilingual & Multicultural Development, 37(7), 694–709. Curdt-Christiansen, X.-L. (2014). Family language policy: Is learning Chinese at odds with learning English? In X.-L. Curdt-Christiansen & A. Lancock (Eds.), Learning Chinese in diasporic communities: Many pathways to being Chinese (pp. 35–55). Amsterdam: John Benjamins. Curry, M. J. & Lillis, T. (2004). Multilingual scholars and the imperative to publish in English: Negotiating interests, demands, and rewards. TESOL Quarterly, 38(4), 663–688. Cushing-Leubner, J. & Eik, J. (2018). Our wild tongues: Language justice and youth research. In J. Ayala, J. Cammarota, M. I. Berta-A´vila, M. Rivera,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

577

L. F. Rodrı´guez, & M. E. Torre (Eds.), PAR EntreMundos: A pedagogy of the Ame´ricas (pp. 117–133). Bern: Peter Lang. Da˛browska, E. (2012). Different speakers, different grammars: Individual differences in native language attainment. Linguistic Approaches to Bilingualism, 2(3), 219–253. Dale, P. S. & Fenson, L. (1996). Lexical development norms for young children. Behavior Research Methods, Instruments, and Computers, 28(1), 125–127. Dallinger, S., Jonkmann, K., Hollm, J., & Fiege, C. (2016). The effect of content and language integrated learning on students’ English and history competences—Killing two birds with one stone? Learning and Instruction, 41, 23–31. Dalton-Puffer, C. (2007). Discourse in content and language integrated learning (CLIL) classrooms. Amsterdam: John Benjamins. Dalton-Puffer, C. (2017). Same but different: Content and language integrated learning and content-based instruction. In M. A. Snow & D. M. Brinton (Eds.), The content-based classroom: New perspectives on integrating language and content (pp. 151–164). Ann Arbor: University of Michigan Press. Dalton-Puffer, C., Nikula, T., & Smit, U. (Eds.). (2010). Language use and language learning in CLIL classrooms. Amsterdam: John Benjamins. Danblon, E., De Clerck, B., & van Noppen, J.-P. (2004). Politeness in Belgium: Face, distance and sincerity in service-exchange rituals. In L. Hickey & M. Stewart (Eds.), Politeness in Europe (pp. 45–57). Clevedon: Multilingual Matters. Davis, C. P. (2012). “Is Jaffna Tamil the best?” Producing “legitimate” language in a multilingual Sri Lankan school. Journal of Linguistic Anthropology, 22(2). E61–E82. Davis, N. (2014). Cochlear ear implants for adults: Experts call for review. The Guardian, March 16, 2014. Retrieved from www.theguardian.com/ society/2014/mar/16/cochlear-ear-implants-adults-review#comments, last accessed February, 1, 2018. De Angelis, G. (2005). Interlanguage transfer of function words. Language Learning, 55(3), 379–414. De Backer, F., Van Avermaet, P., & Slembrouck, S. (2017). Schools as laboratories for exploring multilingual assessment policies and practices. Language and Education, 31(3), 217–230. de Bot, K. (2001). Language use as an interface between sociolinguistic and psycholinguistic processes in language attrition and language shift. In J. Klatter-Folmer & P. van Avermaet (Eds.), Theories on maintenance and loss of minority languages. Towards a more integrated explanatory framework (pp. 65–82). Mu¨nster: Waxmann. de Bot, K. (2007). Dynamic systems theory, lifespan development and language attrition. In B. Ko¨pke, M. S. Schmid, M. Keijzer, & S. Dostert

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

578

References

(Eds.), Language attrition: Theoretical perspectives (pp. 53–68). Amsterdam: John Benjamins. de Bot, K. & Clyne, M. (1989). Language reversion revisited. Studies in Second Language Acquisition, 11(2), 167–177. de Bot, K. & Clyne, M. (1994). A 16-year longitudinal study of language attrition in Dutch immigrants in Australia. Journal of Multilingual & Multicultural Development, 15(1), 17–28. de Bot, K. & Jaensch, C. (2015). What is special about L3 processing? Bilingualism: Language and Cognition, 18(2), 130–144. de Bot, K. & Makoni, S. (2005). Language and aging in multilingual contexts. Clevedon: Multilingual Matters. de Bot, K., Martens, V., & Stoessel, S. (2004). Finding residual lexical knowledge: The “savings” approach to testing vocabulary. International Journal of Bilingualism, 8(3), 373–382. de Bot, K. & Stoessel, S. (2000). In search of yesterday’s words: Reactivating a long forgotten language. Applied Linguistics, 21(3), 364–388. De Bot, K. & Weltens, B. (1991). Recapitulation, regression and language loss. In H. W. Seliger & R. M. Vago (Eds.), First language attrition (pp. 31–51). Cambridge: Cambridge University Press. de Bruin, A., Bak, T. H., & Della Sala, S. (2015). Examining the effects of active versus inactive bilingualism on executive control in a carefully matched non-immigrant sample. Journal of Memory and Language, 85, 15–26. de Bruin, A., Treccani, B., & Della Sala, S. (2015). Cognitive advantage in bilingualism: An example of publication bias. Psychological Science, 26(1), 99–107. De Cat, C., Gusnanto, A., & Serratrice, L. (2018). Identifying a threshold for the executive function advantage in bilingual children. Studies in Second Language Acquisition, 40(1), 119–151. Dehaene, S., Dupoux, E., Mehler, J., Cohen, L., Paulesu, E., Perani, D., van de Moortele, P.-F., Lehe´ricy, S., & Le Bihan, D. (1997). Anatomical variability in the cortical representation of first and second language. NeuroReport, 8(17), 3809–3815. De Houwer, A. (1983). Some aspects of the simultaneous acquisition of Dutch and English by a three-year-old child. Nottingham Linguistic Circular, 12(2), 106–129. De Houwer, A. (1990). The acquisition of two languages from birth: A case study. Cambridge: Cambridge University Press. De Houwer, A. (1999). Environmental factors in early bilingual development: The role of parental beliefs and attitudes. In G. Extra & L. Verhoeven (Eds.), Bilingualism and migration (pp. 75–95). Berlin: Mouton de Gruyter. De Houwer, A. (2004). Trilingual input and children’s language use in trilingual families in Flanders. In C. Hoffmann & J. Ytsma (Eds.),

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

579

Trilingualism in family, school and community (pp. 118–135). Clevedon: Multilingual Matters. De Houwer, A. (2007). Parental language input patterns and children’s bilingual use. Applied Psycholinguistics, 28(3), 411–424. De Houwer, A. (2009). Bilingual first language acquisition. Bristol: Multilingual Matters. De Houwer, A. (2013). Early bilingualism. In C. A. Chapelle (Ed.), The encyclopedia of applied linguistics. Blackwell Publishing Ltd. doi: 10.1002/9781405198431.wbeal0351 De Houwer, A. (2014). The absolute frequency of maternal input to bilingual and monolingual children: A first comparison. In T. Gru¨ter & J. Paradis (Eds.), Input and experience in bilingual development (pp. 37–58). Amsterdam: John Benjamins. De Houwer, A. (2015a). Harmonious bilingual development: Young families’ well-being in language contact situations. International Journal of Bilingualism, 19(2), 169–184. De Houwer, A. (2015b). Integration und Interkulturalita¨t in Kindertagessta¨tten und in Kinderga¨rten: Die Rolle der Nichtumgebungssprache fu¨r das Wohlbefinden von Kleinkindern [Integration and interculturality in childcare centers and preschools: The role of the minority language for children’s wellbeing]. In E. Reichert-Garschhammer, C. Kieferle, M. Wertfein, & F. Becker-Stoll (Eds.), Inklusion und Partizipation. Vielfalt als Chance und Anspruch [Inclusion and participation: Diversity as a chance and a challenge] (pp. 113–125). Go¨ttingen: Vandenhoeck & Ruprecht. De Houwer, A. (2017a). Early multilingualism and language awareness. In J. Cenoz, D. Gorter, & S. May (Eds.), Language awareness and multilingualism (pp. 83–97). Berlin: Springer. De Houwer, A. (2017b). Minority language parenting in Europe and children’s well-being. In N. Cabrera & B. Leyendecker (Eds.), Handbook on positive development of minority children and youth (pp. 231–246). Berlin: Springer. De Houwer, A. (2018a). Input, context and early child bilingualism: Implications for clinical practice. In A. Bar-On & D. Ravid (Eds.), Handbook of communication disorders: Theoretical, empirical, and applied linguistic perspectives (pp. 599–616). Berlin: Walter de Gruyter. De Houwer, A. (2018b). The role of language input environments for language outcomes and language acquisition in young bilingual children. In D. Miller, F. Bayram, J. Rothman, & L. Serratrice (Eds.), Bilingual cognition and language: The state of the science across its subfields (pp. 127–153). Amsterdam: John Benjamins. De Houwer, A. (to appear). Individual bilingualism. In S. Mufwene & A.-M. Escobar (Eds.), The Cambridge handbook of language contact. Cambridge: Cambridge University Press.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

580

References

De Houwer, A. & Bornstein, M. H. (2003). Balancing on the tightrope: Language use patterns in bilingual families with young children. Presentation at the Fourth International Symposium on Bilingualism, Tempe, AZ, United States, April 30–May 3. De Houwer, A. & Bornstein, M. H. (2016). Balance patterns in early bilingual acquisition: A longitudinal study of word comprehension and production. In C. Silva-Corvala´n & J. Treffers-Daller (Eds.), Language dominance in bilinguals: Issues of measurement and operationalization (pp. 134–155). Cambridge: Cambridge University Press. De Houwer, A., Bornstein, M. H., & De Coster, S. (2006). Early understanding of two words for the same thing: A CDI study of lexical comprehension in infant bilinguals. International Journal of Bilingualism, 10(3), 331–347. De Houwer, A., Bornstein, M. H., & Putnick, D. L. (2014). A bilingual– monolingual comparison of young children’s vocabulary size: Evidence from comprehension and production. Applied Psycholinguistics, 35(6), 1189–1211. de Leeuw, E., Mennen, I., & Scobbie, J. M. (2013). Dynamic systems, maturational constraints and L1 phonetic attrition. International Journal of Bilingualism, 17(6), 683–700. De Schryver, J. (2012). Het einde van de tussentaal en de Vlaamse standaardtaaldiscussie [The end of in-between language and the Flemish debate on standardization]. In J. Jaspers, K. Absillis, & S. Van Hoof (Eds.), De manke usurpator (pp. 141–165). Antwerpen: Universiteit Antwerpen. De Vogelaer, G. & Toye, J. (2017). Acquiring attitudes towards varieties of Dutch: A quantitative perspective. In G. De Vogelaer & M. Katerbow (Eds.), Acquiring sociolinguistic variation (pp. 117–154). Amsterdam: John Benjamins. DeAnda, S., Arias-Trejo, N., Poulin-Dubois, D., Zesiger, P., & Friend, M. (2016). Minimal second language exposure, SES, and early word comprehension: New evidence from a direct assessment. Bilingualism: Language and Cognition, 19(1), 162–180. DeAnda, S., Bosch, L., Poulin-Dubois, D., Zesiger, P., & Friend, M. (2016). Language exposure assessment tool: Quantifying language exposure in infants and children. Journal of Speech, Language, and Hearing Research, 59(6), 1346–1356. Dearden, J. (2014). English as a medium of instruction: A growing global phenomenon. London: British Council. DeKeyser, R. M. (2000). The robustness of critical period effects in second language acquisition. Studies in Second Language Acquisition, 22(4), 499–533. DeKeyser, R. M. (2007). Study abroad as foreign language practice. In R. DeKeyser (Ed.), Practising in a second language: Perspectives from applied linguistics and cognitive psychology (pp. 208–226). Cambridge: Cambridge University Press.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

581

DeKeyser, R. M. (2013). Age effects in second language learning: Stepping stones toward better understanding. Language Learning, 63(1), 52–67. Dekker, D. & Young, C. (2005). Bridging the gap: The development of appropriate educational strategies for minority language communities in the Philippines. Current Issues in Language Planning, 6(2), 182–199. DELMI (Delegationen fo¨r Migrationsstudier). Utkast till program och strategi [The Migration Studies Delegation. Program and strategy]. Retrieved from www.delmi.se, last accessed February 1, 2018. Denman, J., Tanner, R., & de Graaff, R. (2013). CLIL in junior vocational secondary education: Challenges and opportunities for teaching and learning. International Journal of Bilingual Education and Bilingualism, 16(3), 285–300. Deuchar, M. & Quay, S. (2000). Bilingual acquisition: Theoretical implications of a case study. Oxford: Oxford University Press. Devine, R. T. & Hughes, C. (2014). Relations between false belief understanding and executive function in early childhood: A meta-analysis. Child Development, 85(5), 1777–1794. Dewaele, J.-M. (2002a). The effect of multilingualism and socio-situational factors on communicative anxiety of mature language learners. In J. Ytsma & M. Hooghiemstra (Eds.), Proceedings of the second international conference on trilingualism (CD-ROM). Dewaele, J.-M. (2002b). Using sociostylistic variants in advanced French interlanguage: The case of nous/on. In S. Foster-Cohen, T. Ruthenberg, & M. L. Poschen (Eds.), EUROSLA Yearbook 2002 (pp. 205–226). Amsterdam: John Benjamins. Dewaele, J.-M. (2004). Retention or omission of the “ne” in advanced French interlanguage: The variable effect of extralinguistic factors. Journal of Sociolinguistics, 8(3), 433–450. Dewaele, J.-M. & Li Wei. (2013). Attitudes towards code-switching among adult mono- and multilingual language users. Journal of Multilingual & Multicultural Development, 35(3), 235–251. Dewaele, J.-M. & Salomidou, L. (2017). Loving a partner in a foreign language. Journal of Pragmatics, 108, 116–130. Deygers, B. (2017). Just testing: Applying theories of justice to high-stakes language tests. ITL: International Journal of Applied Linguistics, 168(2), 143–163. Diao, W. (2017). Between the standard and non-standard: Accent and identity among transnational Mandarin speakers studying abroad in China. System, 71, 87–101. Dicker, S. (1993). The universal second language requirement: An inadequate substitute for bilingual education (A response to Aaron Wildavsky). Journal of Policy Analysis and Management, 12(4), 779–787.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

582

References

Dickinson, J. (2010). Languages for the market, the nation, or the margins: Overlapping ideologies of language and identity in Zakarpattia. International Journal of the Sociology of Language, 201, 53–78. Dijkstra, J., Kuiken, F., Jorna, R. J., & Klinkenberg, E. L. (2016). The role of majority and minority language input in the early development of a bilingual vocabulary. Bilingualism: Language and Cognition, 19(1), 191–205. Dijkstra, T., Timmermans, M., & Schriefers, H. (2000). On being blinded by your other language: Effects of task demands on interlingual homograph recognition. Journal of Memory and Language, 42(4), 445–464. Dijkstra, T. & van Heuven, W. (1998). The BIA-model and bilingual word recognition. In J. Grainger & A. Jacobs (Eds.), Localist connectionist approaches to human cognition (pp. 189–225). Mahwah, NJ: Lawrence Erlbaum. Dijkstra, T. & van Heuven, W. J. B. (2002). The architecture of the bilingual word recognition system: From identification to decision. Bilingualism: Language and Cognition, 5(3), 175–197. Diskin, C. & Regan, V. (2015). Migratory experience and second language acquisition amongst Polish and Chinese migrants in Dublin, Ireland. In F. Forsberg Lundell & I. Bartning (Eds.), Cultural migrants and optimal language acquisition (pp. 178–202). Bristol: Multilingual Matters. Donati, C. & Branchini, C. (2013). Challenging linearization: Simultaneous mixing in the production of bimodal bilinguals. In I. Roberts & T. Biberauer (Eds.), Challenges to linearization (pp. 93–128). Berlin: Mouton de Gruyter. Dong, Y. & Li, P. (2015). The cognitive science of bilingualism. Language and Linguistics Compass, 9, 1–13. Do¨rnyei, Z. & Ushioda, E. (Eds.). (2009). Motivation, language identity and the L2 self. Bristol: Multilingual Matters. Doughty, C. J., Campbell, S. G., Mislevy, M. A., Bunting, M. F., Bowles, A. R., & Koeth, J. T. (2010). Predicting near-native ability: The factor structure and reliability of Hi-LAB. In M. T. Prior, Y. Watanabe, & S.-K. Lee (Eds.), Selected proceedings of the 2008 Second Language Research Forum (pp. 10–31). Somerville Cascadilla Proceedings Project. Douglas Fir Group (2016). A transdisciplinary framework for SLA in a multilingual world. Modern Language Journal, 100 (Supplement 1, Centenary Anniversary), 19–47. Dowens, M. G., Vergara, M., Barber, H. A., & Carreiras, M. (2010). Morphosyntactic processing in late second-language learners. Journal of Cognitive Neuroscience, 22(8), 1870–1887. Downes, S. (2001). Sense of Japanese cultural identity within an English partial immersion programme: Should parents worry? International Journal of Bilingual Education and Bilingualism, 4(3), 165–180. Dressler, W. U. (1991). The sociolinguistic and patholinguistic attrition of Breton phonology, morphology, and morphophonology. In

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

583

H. W. Seliger & R. M. Vago (Eds.), First language attrition (pp. 99–112). Cambridge: Cambridge University Press. Duff, P. A. (2010). Research approaches in applied linguistics. In R. Kaplan (Ed.), The Oxford handbook of applied linguistics (2nd ed., pp. 45–59). Oxford: Oxford University Press. Dulay, K. M., Tong, X. & McBride, C. (2017). The role of foreign domestic helpers in Hong Kong Chinese children’s English and Chinese skills: A longitudinal study. Language Learning, 67(2), 321–347. Dun˜abeitia, J. A., Herna´ndez, J. A., Anto´n, E., Macizo, P., Este´vez, A., Fuentes, L. J., & Carreiras, M. (2014). The inhibitory advantage in bilingual children revisited: Myth or reality? Experimental Psychology, 61(3), 234–251. Dunn, L. M. & Dunn, D. M. (2007). Peabody Picture Vocabulary Test, Fourth edition (PPVT™-4). San Antonio, TX: Pearson. Durgunog˘lu, A. Y. & Verhoeven, L. (Eds.) (1998). Literacy development in a multilingual context: Cross-cultural perspectives. Mahwah, NJ: Lawrence Erlbaum. Dussias, P. E. & Sagarra, N. (2007). The effect of exposure on syntactic parsing in Spanish–English bilinguals. Bilingualism: Language and Cognition, 10(1), 101–116. Eades, D. (2008). Courtroom talk and neocolonial control. Berlin: Mouton de Gruyter. Eades, D. (2012). The social consequences of language ideologies in courtroom cross-examination. Language in Society, 41(4), 471–497. Easterly, W. & Levine, R. (1997). Africa’s growth tragedy: Policies and ethnic divisions. Quarterly Journal of Economics, 112(4), 1203–1250. Ebbinghaus, H. (1885). Uber das Geda¨chtnis: Untersuchungen zur experimentellen Psychologie [About memory: investigations into experimental psychology]. Leipzig: Duncker & Humblot. Eccles, J. S. (2009). Who am I and what am I going to do with my life? Personal and collective identities as motivators of action. Educational Psychologist, 44(2), 78–89. Ecke, P. (2004). Language attrition and theories of forgetting: A cross-disciplinary review. International Journal of Bilingualism, 8(3), 321–354. Edelenbos, P., Johnstone, R., & Kubanek, A. (2006). The main pedagogical principles underlying the teaching of young learners: Languages for the children of Europe. Brussels: European Commission. Edmonds, L. A. & Donovan, N. J. (2012). Item-level psychometrics and predictors of performance for Spanish/English bilingual speakers on an object and action naming battery. Journal of Speech, Language, and Hearing Research, 55(2), 359–381. Edmonds, L. A. & Kiran, S. (2006). Effect of semantic naming treatment on crosslinguistic generalization in bilingual aphasia. Journal of Speech, Language, and Hearing Research, 49(4), 729–748.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

584

References

Edwards, J. (2012). Multilingualism. Understanding linguistic diversity. London: Continuum. Egger, E., Hulk, A., & Tsimpli, I. M. (2018). Crosslinguistic influence in the discovery of gender: The case of Greek–Dutch bilingual children. Bilingualism: Language and Cognition, 21(4), 694–709. Egger, P. & Lassmann, A. (2012). The language effects of international trade: A meta-analysis. Economics Letters, 116(2), 221–224. Egger, P. & Toubal, F. (2016). Common spoken languages and international trade. In V. Ginsburgh & S. Weber (Eds.), The Palgrave handbook of economics and language (pp. 263–289). Basingstoke: Palgrave Macmillan. Eijberts, M. & Roggeband, C. (2016). Stuck with the stigma? How Muslim migrant women in the Netherlands deal – individually and collectively – with negative stereotypes. Ethnicities, 16(1), 130–153. Eimas, P. D., Siqueland, E. R., Juscyk, P., & Vigorito, J. (1971). Speech perception in infants. Science, 171(3968), 303–306. Elias-Bursac´, E. (2012). Shaping international justice: The role of translation and interpreting at the ICTY in The Hague. Translation and Interpreting Studies, 7(1), 34–53. Elliott, A. & Urry, J. (2010). Mobile lives. London: Routledge. Ellis, N. C., Ro¨mer, U., & O’Donnell, M. B. (2016). Usage-based approaches to language acquisition and processing: Cognitive and corpus investigations of construction grammar. Malden, MA: Wiley-Blackwell. Ellis, R. (2006). Current issues in the teaching of grammar: An SLA perspective. TESOL Quarterly, 40, 83–107. Elwert, T. (1959). Das zweisprachige Individuum. Ein Selbstzeugnis. [The bilingual individual. A personal testimony.] Wiesbaden: Verlag der Akademie der Wissenschaften und der Literatur. Emmorey, K., Borinstein, H. B., & Thompson, R. (2005). Bimodal bilingualism: Code-blending between spoken English and American Sign Language. In J. Cohen, K. T. McAlister, K. Rolstad & J. MacSwan (Eds.), Proceedings of the 4th International Symposium on Bilingualism (pp. 663–673). Somerville Cascadilla Press. Emmorey, K., Borinstein, H. B., Thompson, R., & Gollan, T. H. (2008). Bimodal bilingualism. Bilingualism: Language and Cognition, 11(1), 43–61. Emmorey, K., Giezen, M. R., & Gollan, T. H. (2016). Psycholinguistic, cognitive, and neural implications of bimodal bilingualism. Bilingualism: Language and Cognition, 19(2), 223–242. Emmorey, K., Petrich, J. A. F., & Gollan, T. H. (2012). Bilingual processing of ASL–English code blends: The consequences of accessing two lexical representations simultaneously. Journal of Memory and Language, 67(1), 199–210. Ender, A. E. (2017) What is the target variety? The diverse effects of standard-dialect variation in second language acquisition. In G. De

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

585

Vogelaer & M. Katerbow (Eds.), Acquiring sociolinguistic variation (pp. 155–185). Amsterdam: John Benjamins. Ennser-Kananen, J. (2014). The right to be multilingual: How two trilingual students construct their linguistic legitimacy in a German classroom. University of Minnesota. https://conservancy.umn.edu/bitstream/han dle/1 129 9/164 846 /Ennse rKanane n_umn_013 0E_1 4950 .pdf; sequence=1, last accessed January 18, 2017. Epstein, N. (1977). Language, ethnicity and the schools: Policy alternatives for bilingual-bicultural education. Washington, DC: Institute for Educational Leadership, George Washington University. Epstein, R. (2016, December 5). The empty brain. Aeon. Retrieved from: https://aeon.co/essays/your-brain-does-not-process-information-and-itis-not-a-computer, last accessed January 18, 2018. Erard, M. (2012). Babel no more. In search for the world’s most extraordinary language learners. New York: Free Press. Escamilla, K., Hopewell, S., Butvilofsky, S., Sparrow, W., SolteroGonza´lez, L., Ruiz-Figueroa, O., & Escamilla, M. (2014). Biliteracy from the start: Literacy squared in action. Philadelphia Caslon. Esser, H. (2006). Migration, language and integration. AKI Research Review (Wissenschaftszentrum Berlin fu¨r Sozialforschung). European Commission (2011). Language learning at pre-primary school level: Making it efficient and sustainable. A policy handbook. Brussels: European Commission. European Commission (2012a). Europeans and their languages. Special Eurobarometer 386. http://ec.europa.eu/public_opinion/archives/ebs/ ebs_386_en.pdf, last accessed January 19, 2018. European Commission (2012b). First European survey on language competences. Brussels: European Commission. Eurostat. (2017). Foreign language learning statistics, October 2017. http:// ec.europa.eu/eurostat/statistics-explained/index.php/Foreign_language _learning_statistics#Further_Eurostat_information, last accessed January 10, 2018. Extra, G. & Yag˘mur, K. (Eds.). (2012). Language rich Europe: Trends in policies and practices for multilingualism in Europe. Cambridge: British Council/ Cambridge University Press. Fabbro, F. (1999). The neurolinguistics of bilingualism: An introduction. Hove: Psychology Press. Fabbro, F. (2001). The bilingual brain: Cerebral representation of languages. Brain and Language, 79(2), 211–222. Fabbro, F. & Paradis, M. (1995). Differential impairments in four multilingual patients with subcortical lesions. In M. Paradis (Ed.), Aspects of bilingual aphasia (pp. 139–176). Oxford: Pergamon Press.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

586

References

Fabbro, F., Skrap, M., & Aglioti, S. (2000). Pathological switching between languages after frontal lesions in a bilingual patient. Journal of Neurology, Neurosurgery & Psychiatry, 68(5), 650–652. Faroqi-Shah, Y., Frymark, T., Mullen, R., & Wang, B. (2010). Effect of treatment for bilingual individuals with aphasia: A systematic review of the evidence. Journal of Neurolinguistics, 23(4), 319–341. Fenson, L., Dale, P. S., Reznick, J. S., Thal, D., Bates, E., Hartung, J., Pethick, S., & Reilly, J. (1993). The MacArthur communicative development inventories: User’s guide and technical manual. San Diego Singular Publishing Group. Ferjan Ramı´rez, N., Ramı´rez, R. R., Clarke, M., Taulu, S., & Kuhl, P. K. (2017). Speech discrimination in 11-month-old bilingual and monolingual infants: A magnetoencephalography study. Developmental Science, 20(1), e12427. Fernald, A., Perfors, A., & Marchman, V. A. (2006). Picking up speed in understanding: Speech processing efficiency and vocabulary growth across the 2nd year. Developmental Psychology, 42(1), 98–116. Ferris, S. H. & Farlow, M. (2013). Language impairment in Alzheimer’s disease and benefits of acetylcholinesterase. Clinical Interventions in Aging, 8, 1007–1014. Festman, J. (2012). Language control abilities of late bilinguals. Bilingualism: Language & Cognition, 15(3), 580–593. Fishman, J. A. (1977). Bilingual education: Current perspectives. Arlington Center for Applied Linguistics. Fishman, J. A. (1991). Reversing language shift: Theoretical and empirical foundations of assistance to threatened languages. Clevedon: Multilingual Matters. Fishman, J. A. (Ed.). (2001). Can threatened languages be saved? Reversing language shift revisited, a 21st century perspective. Clevedon: Multilingual Matters. Fitzsimmons-Doolan, S. (2014). Using lexical variables to identify language ideologies in a policy corpus. Corpora, 9(1), 57–82. Floccia, C., Butler, J., Girard, F., & Goslin, J. (2009). Categorization of regional and foreign accent in 5- to 7-year-old British children. International Journal of Behavioral Development, 33(4), 366–375. Flores, C. (2010). The effect of age on language attrition: Evidence from bilingual returnees. Bilingualism: Language and Cognition, 13(4), 533–546. Flores, N. & Lewis, M. (2016) From truncated to socio-political emergence: A critique of super-diversity in sociolinguistics. International Journal of the Sociology of Language, 241, 97–124. Flores, N. & Rosa, J. (2015). Undoing appropriateness: Raciolinguistic ideologies and language diversity in education. Harvard Educational Review, 85(2), 149–171. Flores, R. (2013). The unexamined relationship between neoliberalism and plurilingualism: A cautionary tale. TESOL Quarterly, 47(3), 500–520. Fodor, J. A. (1975). The language of thought. Boston Harvard University Press.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

587

Footnick, R. (2007). A hidden language: Recovery of a “lost” language is triggered by hypnosis. In B. Ko¨pke, M. S. Schmid, M. Keijzer, & S. Dostert (Eds.), Language attrition: Theoretical perspectives (pp. 169–187). Amsterdam: John Benjamins. Forsberg Lundell, F. & Bartning, I. (Eds.). (2015a). Cultural migrants and optimal language acquisition. Bristol: Multilingual Matters. Forsberg Lundell, F. & Bartning, I. (2015b). Introduction: Cultural migrants: Introducing a new concept in SLA research. In F. Forsberg Lundell & I. Bartning (Eds.), Cultural migrants and optimal language acquisition (pp. 1–16). Bristol: Multilingual Matters. Forsberg Lundell, F. & Bartning, I. (2015c). What can SLA learn from cultural migrants? In F. Forsberg Lundell & I. Bartning (Eds.), Cultural migrants and optimal language acquisition (pp. 221–227). Bristol: Multilingual Matters. Fortune, T. & Song, W. (2016). Academic achievement and language proficiency in early total Mandarin immersion education. Journal of Immersion and Content-Based Language Education, 4(2), 168–197. Fortune, T. & Tedick, D. (2015). Oral proficiency development of K-8 Spanish immersion students. Modern Language Journal, 99(4), 637–655. Foulkes, P. (2010). Exploring social-indexical knowledge: A long past but a short history. Laboratory Phonology, 1(1), 5–40. Fox, R. A. & Tevis McGory, J. (2007). Second language acquisition of a regional dialect of American English by native Japanese speakers. In O. Schwen Bohn & J. Murray (Eds.), Language experience in second language speech learning: In honor of James Emil Flege (pp. 117–134). Amsterdam: John Benjamins. Freed, B. F. (1982). Language loss: Current thoughts and future directions. In R. D. Lambert & B. F. Freed (Eds.), The loss of language skills (pp. 1–5). Rowley Newbury House. Freeman, R. (2000). Contextual challenges to dual-language education: A case study of a developing middle school program. Anthropology & Education Quarterly, 31(2), 202–229. Freire, P. (1970). Pedagogy of the oppressed. New York: Seabury Books. Freire, P. & Macedo, D. (1987). Literacy: Reading the word and the world. Westport Bergin & Garvey. Freud, S. (1899). Die Traumdeutung [Interpreting dreams.]. Leipzig/Wien: Franz Deuticke. Friedland, D. & Miller, N. (1999). Language mixing in bilingual speakers with Alzheimer’s dementia: A conversation analysis approach. Aphasiology, 13(4–5), 427–444. Friesen, D. C., Jared, D., & Haigh, C. A. (2014). Phonological processing dynamics in bilingual word naming. Canadian Journal of Experimental Psychology, 68(3), 179–193. Friesen, D., Latman, V., Calvo, A., & Bialystok, E. (2015). Attention during visual search: The benefit of bilingualism. International Journal of Bilingualism, 19(6), 693–702.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

588

References

Fuller, J. M. (2012). Bilingual pre-teens: Competing ideologies and multiple identities in the U.S. and Germany. London: Routledge. Fuller, J. M. (2013). Spanish speakers in the USA. Bristol: Multilingual Matters. Fung, C. H.-M. & Tang, G. (2016). Code-blending of functional heads in Hong Kong Sign Language and Cantonese: A case study. Bilingualism: Language and Cognition, 19(4), 754–781. Fung, C. H.-M., Tang, G., & Lam, S. (2008). Code-Switching in Hong Kong Sign Language. Paper presented at the 12th Symposium on Contemporary Linguistics, Hua Zhong Normal University, China. Gafaranga, J. (2010). Medium request: Talking language shift into being. Language in Society, 39(2), 241–270. Gagne´, F. (2004). Transforming gifts into talents: The DMGT as a developmental theory. High Ability Studies, 15(2), 119–147. Gal, S. (1978). Peasant men can’t get wives: Language change and sex roles in a bilingual community. Language in Society, 7(1), 1–16. Gal, S. & Irvine, J. T. (1995). The boundaries of languages and disciplines: How ideologies construct difference. Social Research 62(4), 967–1001. Garcı´a, O. (2009a). Bilingual education in the 21st century: A global perspective. Malden, MA: Wiley-Blackwell. Garcı´a, O. (2009b). Education, multilingualism and translanguaging in the 21st century. In T. Skutnabb-Kangas, R. Phillipson, A. K. Mohanty, & M. Panda (Eds.), Social justice through multilingual education (pp. 140–158). Bristol: Multilingual Matters. Garcı´a, O. & Li Wei. (2014). Translanguaging: Language, bilingualism and education. London: Palgrave Macmillan Pivot. Garcı´a, O., Johnson, S., & Seltzer, K. (2017). The translanguaging classroom. Leveraging student bilingualism for learning. Philadelphia Caslon. Garcı´a-Caballero, A., Garcı´a-Lado, I., Gonza´lez-Hermida, J., Area, R., Recimil, M. J., Juncos Rabada´n, O., Lamas, S., Ozaita, G., & Jorge, F. J. (2007). Paradoxical recovery in a bilingual patient with aphasia after right capsuloputaminal infarction. Journal of Neurology, Neurosurgery & Psychiatry, 78(1), 89–91. Garcia-Sierra, A., Ramı´rez-Esparza, N., & Kuhl, P. K. (2016). Relationships between quantity of language input and brain responses in bilingual and monolingual infants. International Journal of Psychophysiology, 110, 1–17. Garcia-Sierra, A., Rivera-Gaxiola, M., Percaccio, C. R., Conboy, B. T., Romo, H., Klarman, L., & Kuhl, P. K. (2011). Bilingual language learning: An ERP study relating early brain responses to speech, language input, and later word production. Journal of Phonetics, 39(4), 546–557. Gardner-Chloros, P. (2009). Code-switching. Cambridge: Cambridge University Press. Gardner-Chloros, P., Sebba, M., & Moyer, M. (2007). The LIDES Corpus. In J. Beal, K. Corrigan, & H. Moisl (Eds.), Models and methods in the

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

589

handling of unconventional digital corpora. Vol. 1: Synchronic corpora; Vol. 2: Diachronic corpora (pp. 91–121). London: Palgrave. Gardner, R. C. (2001). Integrative motivation and second language acquisition. In Z. Do¨rnyei & R. Schmidt (Eds.), Motivation and second language acquisition (pp. 1–19). Honolulu, HI: National Foreign Language Resource Center. Garrido, M. R. (2018). Circulation and localization of a transnational founding story in a social movement. International Journal of the Sociology of Language, 18(250), 113–135. Garrido, M. R. & Oliva, X. (2015). A multilingual, collaborative and functional approach to nongovernmental Catalan classes. In J. Simpson & A. Whiteside (Eds.), Adult language education and migration: Challenging agendas in policy and practice (pp. 94–106). London: Routledge. Gasiorek, J. (2016). The “dark side” of CAT: Nonaccommodation. In H. Giles (Ed.), Communication Accommodation Theory: Negotiating personal relationships and social identities across contexts (pp. 85–122). Cambridge: Cambridge University Press. Gathercole, S. E. & Alloway, T. P. (2008). Working memory and classroom learning. In K. Thurman & K. Fiorello (Eds.), Cognitive development in K-3 classroom learning: Research applications (pp. 15–38). Mahwah, NJ: Lawrence Erlbaum. Gathercole, V. C. M. (2007). Miami and North Wales, so far and yet so near: A constructivist account of morphosyntactic development in bilingual children. International Journal of Bilingual Education and Bilingualism, 10(3), 224–247. Gathercole, V. C. M. (2013). Assessment of bilinguals: Innovative solutions. In V. C. M. Gathercole (Ed.), Solutions for the assessment of bilinguals (pp. 1–9). Bristol: Multilingual Matters. Gathercole, V. C. M. & Thomas, E. M. (2009). Bilingual first-language development: Dominant language takeover, threatened minority language take-up. Bilingualism: Language and Cognition, 12(02), 213–237. Gathercole, V. C. M., Thomas, E. M., Kennedy, I., Prys, C., Young, N. Vin˜as-Guasch, N., Roberts, E. J., Hughes, E. K., & Jones, L. (2014). Does language dominance affect cognitive performance in bilinguals? Lifespan evidence from preschoolers through older adults on card sorting, Simon, and metalinguistic tasks. Frontiers in Psychology, 5, Article 11. Gathercole, V. C. M., Thomas, E. M., Roberts, E., Hughes, C., & Hughes, E. K. (2013). Why assessment needs to take exposure into account: Vocabulary and grammatical abilities in bilingual children. In V. C. M. Gathercole (Ed.), Issues in the assessment of bilinguals (pp. 20–55). Bristol: Multilingual Matters. Gautier, R. & Chevrot, J.-P. (2015). Social networks and acquisition of sociolinguistic variation in a study abroad context: A preliminary study. In R. Mitchell, N. Tracy-Ventura, & K. McManus (Eds.), Social interaction,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

590

References

identity and language learning during residence abroad (pp. 169–184). Eurosla Monographs series, 4. Gazzola, M. (2014). The evaluation of language regimes: Theory and application to multilingual patent organisations. Amsterdam: John Benjamins. Gazzola, M. (2016). Multilingual communication for whom? Language policy and fairness in the European Union. European Union Politics, 17(4), 546–569. Gazzola, M. & Grin, F. (2017). Comparative language policy and evaluation: Criteria, indicators and implications for translation policy. In R. Meylaerts & G. Gonza´lez Nun˜ez (Eds.), Translation and public policy (pp. 83–116). London: Routledge. Gazzola, M., Grin, F., & Wickstro¨m, B.-A. (2016). A concise bibliography of language economics. In M. Gazzola & B.-A. Wickstro¨m (Eds.), The economics of language policy (pp. 53–92). Boston, MA: MIT Press. Gazzola, M. & Wickstro¨m, B.-A. (Eds.). (2016). The economics of language policy (CESifo Series). Boston, MA: MIT Press. Geipel, J., Hadjichristidis, C., & Surian, L. (2016). Foreign language affects the contribution of intentions and outcomes to moral judgment. Cognition, 154, 34–39. Genesee, F. (1987). Learning through two languages: Studies of immersion and bilingual children. Cambridge, MA: Newbury House. Genesee, F. (1989). Early bilingual development: One language or two? Journal of Child Language, 16(1), 161–179. Genesee, F., Boivin, I., & Nicoladis, E. (1996). Talking with strangers: A study of bilingual children’s communicative competence. Applied Psycholinguistics, 17(4), 427–442. Genesee, F. & Delcenserie, A. (Eds.) (2016). Starting over: The language development in internationally adopted children. Amsterdam: John Benjamins. Genesee, F. & Lindholm-Leary, K. (2013). Two case studies of content-based language education. Journal of Immersion and Content-Based Language Education, 1(1), 3–33. Genevska-Hanke, D. (2017). Interpersonal variation in late L1 attrition and its implications for the competence/performance debate. In N. Levkovych & A. Urdze (Eds.), Linguistik im Nordwesten: Beitra¨ge zum 8. Nordwestdeutschen Linguistischen Kolloquium, 13.–14. November 2015 (pp. 1–31). Bochum: Brockmeyer. Gertken, L. M., Amengual, M., & Birdsong, D. (2014). Assessing language dominance with the bilingual language profile. In P. Leclercq, A. Edmonds, & H. Hilton (Eds.), Measuring L2 proficiency: Perspectives from SLA (pp. 208–225). Bristol: Multilingual Matters. Ghimenton, A. (2015). Reading between the code choices: Discrepancies between expressions of language attitudes and usage in a contact situation. International Journal of Bilingualism, 19(1), 115–136. Ghimenton, A., Chevrot, J.-P., & Billiez, J. (2013). Language choice adjustments in child production during dyadic and multiparty interactions:

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

591

A quantitative approach to multilingual interactions. Linguistics, 51(2), 413–438. Giddens, A. (1991). Modernity and self-identity: Self and society in the late modern age. Cambridge: Polity. Giedd, J. N., Clasen, L. S., Lenroot, R., Greenstein, D., Wallace, G. L., Ordaz, S., Molloy, E. A., Blumenthal, J. D., Tossell J. W., Stayer, C., Samango-Sprouse, C. A., Shen, D., Davatzikos, C., Merke, D., & Chrousos, G. P. (2006). Puberty-related influences on brain development. Molecular Cell Endocrinology, 254–255, 154–162. Gierlinger, E. (2015). “You can speak German, Sir”: On the complexity of teachers’ L1 use in CLIL. Language and Education, 29(4), 347–368. Giles, H. (Ed.). (2016). Communication accommodation theory: Negotiating personal relationships and social identities across contexts. Cambridge: Cambridge University Press. Ginsburgh, V. & Weber, S. (2016a). Linguistic distances and ethnolinguistic fragmentation: Fractionalization and disenfranchisement indices. In V. Ginsburgh & S. Weber (Eds.), The Palgrave handbook of economics and language (pp. 137–173). Basingstoke: Palgrave Macmillan. Ginsburgh, V. & Weber, S. (Eds.). (2016b). The Palgrave handbook of economics and language. Basingstoke: Palgrave Macmillan. Gitterman, M. R., Goral, M., & Obler, L. K. (Eds.). (2012). Aspects of multilingual aphasia. Bristol: Multilingual Matters. Glennen, S. (2015). Internationally adopted children in the early school years: Relative strengths and weaknesses in language abilities. Language, Speech, and Hearing Services in Schools, 46(1), 1–13. Global Migration Data Analysis Centre (2017). Global migration trends factsheet. International Organization for Migration. http://gmdac .iom.int/global-migration-trends-factsheet, last accessed January 20, 2018. Gold, M. A. (2015). Learners’ attitudes toward second language dialectal variations and their effects on listening comprehension. Hawaii Pacific University TESOL Working Paper Series, 13, 18–30. Goldrick, M., Putnam, M., & Schwarz, L. (2016). Coactivation in bilingual grammars: A computational account of code mixing. Bilingualism: Language and Cognition, 19(5), 857–876. Goldstein, T. (1997). Two languages at work: Bilingual life on the production floor. Berlin: Mouton de Gruyter. Gollan, T. H. & Ferreira, V. S. (2009). Should I stay or should I switch? A cost-benefit analysis of voluntary language switching in young and aging bilinguals. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(3), 640–665. Gollan, T. H. & Goldrick, M. (2016). Grammatical constraints on language switching: Language control is not just executive control. Journal of Memory and Language, 90, 177–199.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

592

References

Gollan, T. H., Sandoval, T., & Salmon, D. P. (2011). Cross-language intrusion errors in aging bilinguals reveal the link between executive control and language selection. Psychological Science, 22(9), 1155–1164. Golonka, E. M., Bowles, A. R., Frank, V. M., Richardson, D. L., & Freynik, S. (2014). Technologies for foreign language learning: A review of technology types and their effectiveness. Computer Assisted Language Learning, 27(1), 70–105. Gonc¸alves, K. (2013). Conversations of intercultural couples. Berlin: Mouton de Gruyter. Gonc¸alves, K. (2015). The pedagogical implications of ELF in a domestic migrant workplace. In H. Bowles & A. Cogo (Eds.), International perspectives on English as a lingua franca (pp. 136–158). New York: Palgrave Macmillan. Gonc¸alves, K. & Schluter, A. (2017). “Please do not leave any notes for the cleaning lady, as many do not speak English fluently”: Policy, power and language brokering in a multilingual workplace. Language Policy, 16(3), 241–265. Gonza´lez Alonso, J. G., Rothman, J., Berndt, D., Castro, T., & Westergaard, M. (2017). Broad scope and narrow focus: On the contemporary linguistic and psycholinguistic study of third language acquisition. International Journal of Bilingualism, 21(6), 639–650. Gonza´lez Alonso, J. G., Villegas, J., & Garcı´a Mayo, M. P. (2016). English compound and non-compound processing in bilingual and multilingual speakers: Effects of dominance and sequential multilingualism. Second Language Research, 32(4), 503–535. Gonzo, S. & Saltarelli, M. (1983). Pidginization and linguistic change in emigrant languages. In R. Andersen (Ed.), Pidginization and creolization as language acquisition (pp. 181–197). Rowley, MA: Newbury House. Goodwin, C. (2004). A competent speaker who can’t speak: The social life of aphasia. Linguistic Anthropology, 14(20), 151–170. Gooskens, C. & Heeringa, W. (2014). The role of dialect exposure in receptive multilingualism. Applied Linguistics Review, 5(1), 247–271. Goral, M., Levy, E. S., & Kastl, R. (2010). Cross-language treatment generalization: A case of trilingual aphasia. Aphasiology, 24(2), 170–187. Goral, M., Libben, G., Obler, L. K., Jarema, G., & Ohayon, K. (2008). Lexical attrition in younger and older bilingual adults. Clinical Linguistics & Phonetics, 22(7), 509–522. Gorno-Tempini, M. L., Hillis, A. E., Weintraub, S., Kertesz, A., Mendez, M., Cappa, S. F., Ogar, J. M., Rohrer, J. D., Black, S., Boeve, B. F., Manes, F., Dronkers, N. F., Vandenberghe, R., Rascovsky, K., Patterson, K., Miller, B. L., Knopman, D. S., Hodges, J. R., Mesulam, M. M., & Grossman, M. (2011). Classification of primary progressive aphasia and its variants. Neurology, 76(11), 1006–1014. Gorter, D. (2017). Perspectives on linguistic repertoires in adult multilinguals: An epilogue. International Journal of Multilingualism, 14(1), 86–90.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

593

Graham, S., Courtney, L., Tonkyn, A., & Marinis, T. (2016). Motivational trajectories for early language learning across the primary–secondary school transition. British Educational Research Journal, 42(4), 682–702. Grainger, J. & Dijkstra, A. (1992). On the representation and use of language information in bilinguals. In R. J. Harris (Ed.), Cognitive processing in bilinguals (pp. 207–220). Amsterdam: Elsevier. Grainger, J., Midgley, K., & Holcomb, P. J. (2010). Re-thinking the bilingual interactive-activation model from a developmental perspective (BIAd). In M. Kail & M. Hickmann (Eds.), Language acquisition across linguistic and cognitive systems (pp. 267–283). Amsterdam: John Benjamins. Granena, G. (2013). Reexamining the robustness of aptitude in second language acquisition. In G. Granena & M. H. Long (Eds.), Sensitive periods, language aptitude, and ultimate L2 attainment (pp. 179–204). Amsterdam: John Benjamins. Granena, G. (2015). The role of sociopsychological factors in long-term L2 achievement of L1 Chinese learners of L2 Spanish. In F. Forsberg Lundell & I. Bartning (Eds.), Cultural migrants and optimal language acquisition (pp. 203–220). Bristol: Multilingual Matters. Granena, G. & Long, M. H. (2013a). Age of onset, length of residence, language aptitude, and ultimate L2 attainment in three linguistic domains. Second Language Research, 29(3), 311–343. Granena, G. & Long, M. (Eds.). (2013b). Sensitive periods, language aptitude, and ultimate L2 attainment. Amsterdam: John Benjamins. Grant, A., Dennis, N. A., & Li, P. (2014). Cognitive control, cognitive reserve, and memory in the aging bilingual brain. Frontiers in Psychology, 5, Article 1401. Green, D. W. (1986). Control, activation, and resource: A framework and a model for the control of speech in bilinguals. Brain and Language, 27 (2), 210–223. Green, D. W. (1998). Mental control of the bilingual lexico-semantic system. Bilingualism: Language and Cognition, 1(2), 67–81. Green, D. W. (2008). Bilingual aphasia: Adapted language networks and their control. Annual Review of Applied Linguistics, 28, 25–48. Green, D. W. & Abutalebi, J. (2013). Language control in bilinguals: The adaptive control hypothesis. Journal of Cognitive Psychology, 25(5), 515–530. Green, L. (2010). Language and the African American child. Cambridge: Cambridge University Press. Grenier, G. & Nadeau, S. (2016). English as the lingua franca and the economic value of other languages: The case of the language of work in the Montreal labor market. In M. Gazzola & B.-A. Wickstro¨m (Eds.), The economics of language policy (pp. 267–311). Boston MIT Press. Gries, S. & Kootstra, G. J. (2017). Structural priming within and across languages: A corpus-based perspective. Bilingualism: Language and Cognition, 20(2), 235–250.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

594

References

Grigorenko, E. L., Sternberg, R. J., & Ehrman, M. E. (2000). A theory based approach to the measurement of foreign language learning ability: The CANAL-F theory and test. Modern Language Journal, 84, 390–405. Grimes, B. (1985). Language attitudes: Identity, distinctiveness, survival in the Vaupes. Journal of Multilingual & Multicultural Development, 6, 389–401. Grin, F. (2003). Economics and language planning. Current Issues in Language Planning, 4(1), 1–66. Grin, F. (2017). Translation and language policy in the dynamics of multilingualism. International Journal of the Sociology of Language, 243, 155–181. Grin, F., Sfreddo, C., & Vaillancourt, F. (2010). The economics of the multilingual workplace. London: Routledge. Grin, F. & Vaillancourt, F. (2015). The economics of language policy: An introduction to evaluation work. In F. M. Hult & D. Cassels Johnson (Eds.), Research methods in language policy and planning: A practical guide (pp. 118–129). Malden, MA: Wiley-Blackwell. Grosjean, F. (1982). Life with two languages. An introduction to bilingualism. Cambridge, MA: Harvard University Press. Grosjean, F. (1985). The bilingual as a competent but specific speakerhearer. Journal of Multilingual & Multicultural Development, 6(6), 467–477. Grosjean, F. (1989). Neurolinguists beware! The bilingual is not two monolinguals in one person. Brain and Language, 36(1), 3–15. Grosjean, F. (1997). The bilingual individual. Interpreting: International Journal of Research and Practice in Interpreting, 1(2), 163–187. Grosjean, F. (1998) Studying bilinguals: Methodological and conceptual issues. Bilingualism: Language and Cognition, 1(2), 131–149. Grosjean, F. (2001). The bilingual’s language modes. In J. L. Nicol (Ed.), One mind, two languages: Bilingual language processing (pp. 1–22). Malden, MA: Wiley-Blackwell. Grosjean, F. (2010). Bilingualism, biculturalism, and deafness. International Journal of Bilingual Education and Bilingualism, 13, 133–145. Grosjean, F. (2016). The Complementarity Principle and its impact on processing, acquisition, and dominance. In C. Silva-Corvala´n & J. Treffers-Daller (Eds.), Language dominance in bilinguals: Issues of measurement and operationalization (pp. 36–65). Cambridge: Cambridge University Press. Grundy, J. G., & Anderson, J. A. E. (2017). Commentary: The relationship of bilingualism compared to monolingualism to the risk of cognitive decline or dementia: A systematic review and meta-analysis. Frontiers in Aging Neuroscience, 9, Article 344. Gru¨ter, T., & Paradis, J. (Eds.). (2014). Input and experience in bilingual development. Amsterdam: John Benjamins. Guan, S.-S. A., Nash, A., & Orellana, M. F. (2016). Cultural and social processes of language brokering among Arab, Asian, and Latin

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

595

immigrants. Journal of Multilingual & Multicultural Development, 37(2), 150–166. Gullberg, M., de Bot, K., & Volterra, V. (2008). Gestures and some key issues in the study of language development. Gesture, 8(2), 149–179. Gullberg, M., Indefrey, P., & Muysken, P. (2009). Research techniques for the study of code-switching. In B. Bullock & A. J. Toribio (Eds.), The Cambridge handbook of linguistic code-switching (pp. 21–39). Cambridge: Cambridge University Press. Gumperz, J., Jupp, T., & Roberts, C. (1979). Crosstalk: A study of cross-cultural communication. London: National Centre for Industrial Language Training, in association with the BBC. Guo, S. & Mo¨llering, M. (2016). The implementation of task-based teaching in an online Chinese class through web conferencing. System, 62(1), 26–38. Gu¨rel, A. (2004). Attrition in L1 competence: The case of Turkish. In M. S. Schmid, B. Ko¨pke, M. Keijzer, & L. Weilemar (Eds.), First language attrition: Interdisciplinary perspectives on methodological issues (pp. 225–242). Amsterdam: John Benjamins. Gu¨rel, A. & Yilmaz, G. (2011). Restructuring in the L1 Turkish grammar: Effects of L2 English and L2 Dutch. Language, Interaction, Acquisition, 2(2), 221–250. Gutie´rrez-Clellen, V. (2012). Narrative development and disorders in bilingual children. In B. Goldstein (Ed.), Bilingual language development and disorders in Spanish–English speakers (pp. 233–250). Baltimore Brookes Publishing. Gutie´rrez–Clellen, V. F. & Kreiter, J. (2003). Understanding child bilingual acquisition using parent and teacher reports. Applied Psycholinguistics, 24(2), 267–288. Gutie´rrez-Clellen, V. F. & Pen˜a, E. (2001). Dynamic assessment of diverse children: A tutorial. Language, Speech, and Hearing Services in Schools, 32, 212–224. Gutie´rrez-Clellen, V. F., Simon-Cereijido, G., & Leone, A. (2009). Codeswitching in bilingual children with specific language impairment. International Journal of Bilingualism, 13, 91–109. Guy, G. R., & Hinskens, F. (2016). Linguistic coherence; systems, repertoires and speech communities. Lingua, 172–173, 1–9. Haitana, T., Pitama, S., & Rucklidge, J. J. (2010). Cultural biases in the Peabody Picture Vocabulary Test-III: Testing Tamariki in a New Zealand sample. Journal of Psychology, 39(3), 25–34. Hakuta, K. (1986). Mirror of language: The debate on bilingualism. New York: Basic Books. Hale, S. (2004). The discourse of court interpreting: Discourse practices of the law, the witness and the interpreter. Amsterdam: John Benjamins. Haman, E., Łuniewska, M., & Pomiechowska, B. (2015). Designing crosslinguistic lexical tasks (CLTs) for bilingual preschool children. In S. Armon-Lotem, J. de Jong, & M. Meir (Eds.), Methods for assessing

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

596

References

multilingual children: Disentangling bilingualism from language impairment (pp. 194–238). Bristol: Multilingual Matters. Hambly, C. & Fombonne, E. (2012). The impact of bilingual environments on language development in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 42(7), 1342–1352. Hamilton, J., Lopes, M., McNamara, T., & Sheridan, E. (1993). Rating scales and native speaker performance on a communicatively oriented EAP test. Language Testing, 10(3), 337–353. Hammarberg, B. (2009). Processes in third language acquisition. Edinburgh: Edinburgh University Press. Hammer, C. S. (1998). Toward a “thick description” of families: Using ethnography to overcome the obstacles to providing family-centered early intervention services. American Journal of Speech-Language Pathology, 7(1), 5–22. Hammer, C. S., Davison, M. D., Lawrence, F. R., & Miccio, A. W. (2009). The effect of maternal language on bilingual children’s vocabulary and emergent literacy development during Head Start and kindergarten. Scientific Studies of Reading, 13(2), 99–121. Hammer, K. & Dewaele, J.-M. (2015). Acculturation as the key to the ultimate attainment? In F. Forsberg Lundell & I. Bartning (Eds.), Cultural migrants and optimal language acquisition (pp. 178–202). Bristol: Multilingual Matters. Han, W. (2015). First language bidialectism and associative learning in second language acquisition. Studies in English Language Teaching, 3 (1), 67–72. Hansen, L. (1999). Language attrition in Japanese contexts. Oxford: Oxford University Press. Hansen, L., Umeda, Y., & McKinney, M. (2002). Savings in the relearning of a second language: The effects of time and proficiency. Language Learning, 52(2), 653–678. Harding-Esch, E. & Riley, P. (2003). The bilingual family: A handbook for parents. Cambridge: Cambridge University Press. Harley, B., Cummins, J., Swain, M., & Allen, P. (1990). The nature of language proficiency. In B. Harley, P. Allen, J. Cummins, & M. Swain (Eds.), The development of second language proficiency (pp. 7–25). Cambridge: Cambridge University Press. Harris, C. L., Gleason, J. B., & Ayc¸ic¸egi, A. (2006). When is a first language more emotional? Psychophysiological evidence from bilingual speakers. In A. Pavlenko (Ed.), Bilingual minds: Emotional experience, expression, and representation (pp. 257–283). Clevedon: Multilingual Matters. Harris, J. & O’Duibhir, P. (2011). Effective language teaching: A synthesis of research. (Research Report No. 13). Dublin: National Council for Curriculum and Assessment.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

597

Harriss, J. A. (2003). The elusive Marc Chagall. Smithsonian Magazine. www .smithsonianmag.com/arts-culture/the-elusive-marc-chagall95114921/, last accessed January 8, 2018. Hart, B. & Risley, T. R. (1995). Meaningful differences in the everyday experience of young American children. Baltimore, MD: Paul H. Brookes Publishing. Hartanto, A. & Yang, H. (2016). Disparate bilingual experiences modulate task-switching advantages: A diffusion model analysis of the effects of interactional context on switch costs. Cognition, 150, 10–19. Hartsuiker, R. J. (2015). Why it is pointless to ask under which specific circumstances the bilingual advantage occurs. Cortex, 73, 336–337. Hartsuiker, R. J., Beerts, S., Loncke, M., Desmet, T., & Bernolet, S. (2016). Cross-linguistic structural priming in multilinguals: Further evidence for shared syntax. Journal of Memory and Language, 90, 14–30. Hashim, A. (2009). Not plain sailing: Malaysia’s language choice in policy and education. AILA Review, 22, 36–51. Hassanzadeh, S. (2012). Outcomes of cochlear implantation in deaf children of deaf parents: Comparative study. Journal of Laryngology & Otology, 126(10), 989–994. Haugen, E. (1953). The Norwegian language in America: A study in bilingual behavior, I. The bilingual community. II. The dialects of Norwegian. Philadelphia University of Pennsylvania Press. Haviland, J. B. (2003). Ideologies of language: Reflections on language and U.S. law. American Anthropologist, 105, 764–774. Haznedar, B. (2013). Child second language acquisition from a generative perspective. Linguistic Approaches to Bilingualism, 3(1), 26–47. Heffer, C., Rock, F., & Conley, J. (Eds.) (2013). Legal–lay communication: Textual travels in the law. Oxford: Oxford University Press. Heilmann, J., Miller, J., & Nockerts, A. (2010). Sensitivity of narrative organization measures using narrative retells produced by young school-age children. Language Testing, 27(4), 603–626. Heller, M. (2011). Paths to post-nationalism: A critical ethnography of language and identity. Oxford: Oxford University Press. Hennecke, I. (2013). Self-repair and language selection in bilingual speech processing. Discourse, 12, 3–20. Hermes, M., Bang, M., & Marin, A. (2012). Designing Indigenous language revitalization. Harvard Educational Review, 82(3), 281–402. Hermes, M. & King, K. A. (2013). Ojibwe language revitalization, multimedia technology, and family language learning. Language Learning and Technology, 17(1), 125–144. Hernandez, A. E. & Kohnert, K. J. (2013). Investigations into the locus of language switching costs in older adult bilinguals. Bilingualism: Language and Cognition, 18(1), 51–64. Hernandez, A. E., Li, P., & MacWhinney, B. (2005). The emergence of competing modules in bilingualism. Trends in Cognitive Sciences, 9(5), 220–225.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

598

References

Herna´ndez, A. M. (2015). Language status in two-way bilingual immersion: The dynamics between English and Spanish in peer interaction. Journal of Immersion and Content-Based Language Education, 3(1), 102–126. Herschensohn, J. R. (2007). Language development and age. Cambridge: Cambridge University Press. Herschmann, H. & Po¨tzl, O. (1983). Observations on aphasia in polyglots. In M. Paradis (Ed.), Readings on aphasia in bilinguals and polyglots (pp. 148–154). Montreal: Marcel Didier. [posthumous English translation of Herschmann, H., & Po¨tzl, O. (1920). Bemerkungen u¨ber die Aphasie der Polyglotten. Neurologisches Centralblatt, 39, 114–120]. Herve´, C., Serratrice, L., & Corley, M. (2016). Dislocations in French–English bilingual children: An elicitation study. Bilingualism: Language and Cognition, 19(5), 987–1000. Higby, E., Lerman, A., Korytkowska, M., Malcolm, T., & Obler, L. K. (2019). Aging as a confound in language attrition research. In M. S. Schmid & B. Ko¨pke (Eds.), The Oxford handbook of language attrition. Oxford: Oxford University Press. Hilchey, M. D. & Klein, R. M. (2011). Are there bilingual advantages on nonlinguistic interference tasks? Implications for the plasticity of executive control processes. Psychonomic Bulletin & Review, 18(4), 625–658. Hilchey, M. D., Saint-Aubin, J., & Klein, R. M. (2015). Does bilingual exercise enhance cognitive fitness in non-linguistic executive processing tasks? In J. W. Schwieter (Ed.), The Cambridge handbook of bilingual processing (pp. 586–613). Cambridge: Cambridge University Press. Hirsch, E. D. (1988). Cultural literacy: What every American needs to know. New York: Vintage Books. Hladek, G. A. (2002). Cochlear implants, the deaf culture, and ethics: A study of disability, informed surrogate consent, and ethnocide. Monash Bioethics Review, 21(1), 29–44. Hoey, M. (2005). Lexical priming: A new theory of words and language. London: Routledge. Hoff, E. & Core, C. (2013). Input and language development in bilingually developing children. Seminars in Speech and Language, 34(4), 215–226. Hoff, E., Rumiche, R., Burridge, A., Ribot, K. M., & Welsh, S. N. (2014). Expressive vocabulary development in children from bilingual and monolingual homes: A longitudinal study from two to four years. Early Childhood Research Quarterly, 29(4), 433–444. Holmqvist, J., Van Vaerenbergh, Y., & Gro¨nroos, C. (2014). Consumer willingness to communicate in a second language: Communication in service settings. Management Decision, 52(5), 950–966. Honigsberg, P. J. (2014). Linguistic isolation: A new human rights violation constituting torture, and cruel, inhuman and degrading treatment. Northwestern University Journal of International Human Rights, 12(1), Article 2.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

599

¯ ., Grogan, A., Crinion, J., Rae, J., Ruffle, L., Hope, T. M. H., Parker Jones, O Leff, A. P., Seghier, M. L., Price, C. J., & Green, D. W. (2015). Comparing language outcomes in monolingual and bilingual stroke patients. Brain, 138(4), 1070–1083. Hopp, H. & Schmid, M. S. (2013). Perceived foreign accent in L1 attrition and L2 acquisition: The impact of age of acquisition and bilingualism. Applied Psycholinguistics, 34(2), 361–394. Horner, B., Min-Zhan, L., Royster, J., & Trimbur, J. (2011). Language difference in writing: Toward a translingual approach. College English, 73(3), 303–321. Horner, K. (2015). Discourses on language and citizenship in Europe. Language and Linguistics Compass 9(5), 209–218. Hoshino, N. & Thierry, G. (2011). Language selection in bilingual word production: Electrophysiological evidence for cross-language competition. Brain Research, 1371, 100–109. Housand, A. M. (2009). Adult, gifted. In B. Kerr (Ed.), Encyclopedia of giftedness, creativity and talent (pp. 28–31). Thousand Oaks, CA: Sage. House, J. (2003). English as a lingua franca: A threat to multilingualism? Journal of Sociolinguistics, 7(4), 556–578. Howard, M., Mougeon, R., & Dewaele, J.-M. (2013). Sociolinguistics and second language acquisition. In R. Bayley, R. Cameron, & C. Lucas (Eds.), The Oxford handbook of sociolinguistics (pp. 340–359). Oxford: Oxford University Press. Hu, G. (2007). The juggernaut of Chinese–English bilingual education. In A. Feng (Ed.), Bilingual education in China: Practices, policies and concepts (pp. 94–126). Clevedon: Multilingual Matters. Hua, Z. (2008). Duelling languages, duelling values: Codeswitching in bilingual intergenerational conflict talk in diasporic families. Journal of Pragmatics, 40(10), 1799–1816. Hudson, R. (2012). How many languages can a person learn? In E. M. Rickerson & B. Hilton (Eds.), The five-minute linguist: Bite-sized essays on language and languages (2nd ed., pp. 102–105). London: Equinox. Hulsen, M. (2000). Language loss and language processing. Three generations of Dutch migrants in New Zealand. PhD dissertation (ISBN: 9090139907). Nijmegen: Katholieke Universiteit Nijmegen. Hulstijn, J. (2012). The construct of language proficiency in the study of bilingualism from a cognitive perspective. Bilingualism: Language and Cognition, 15(2), 422–433. Hulstijn, J. (2015). Language proficiency in native and non-native speakers: Theory and research. Amsterdam: John Benjamins. Hult, F. M. (2014). Covert bilingualism and symbolic competence: Analytical reflections on negotiating insider/outsider positionality in Swedish speech situations. Applied Linguistics, 35(1), 63–81.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

600

References

Hult, F. & Hornberger, N. (2016). Revisiting orientations in language planning: Problem, right, and resource as an analytic heuristic. Bilingual Review, 33(3), 30–49. Hultsch, D. F., MacDonald, S. W., & Dixon, R. A. (2002). Variability in reaction time performance of younger and older adults. Journals of Gerontology: Series B, 57(2), 101–115. Humphries, T., Kushalnagar, P., Mathur, G., Napoli, D. J., Padden, C., Rathmann, C., & Smith, S. (2014). Bilingualism: A pearl to overcome certain perils of cochlear implants. Journal of Medical Speech-Language Pathology, 21(2), 107–125. Hurtado, N., Gru¨ter, T., Marchman, V. A., & Fernald, A. (2014). Relative language exposure, processing efficiency and vocabulary in Spanish– English bilingual toddlers. Bilingualism: Language and Cognition, 17(1), 189–202. Hurtado, N., Marchman, V. A., & Fernald, A. (2008). Does input influence uptake? Links between maternal talk, processing speed and vocabulary size in Spanish-learning children. Developmental Science, 11(6), F31–F39. Hyltenstam, K. (Ed.). (2016). Advanced proficiency and exceptional ability in a second language. Berlin: Mouton de Gruyter. Ingram, J., Pittam, J., & Newman, D. (1985). Developmental and sociolinguistic variation in the speech of Brisbane schoolchildren. Australian Journal of Linguistics, 5(2), 233–246. Ingvalson, E. M., Nowicki, C., Zong, A., & Wong, P. C. M. (2017). Non-native speech learning in older adults. Frontiers in Psychology, 8, Article 148. Iqbal, I. (2005). Mother tongue and motherhood: Implications for French language maintenance in Canada. Canadian Modern Language Review, 61(3), 305–323. Isphording, I. E. (2015). What drives the language proficiency of immigrants? IZA World of Labor, 2015, Article 177. Jacquemet, M. (2009). Transcribing refugees: The entextualization of asylum seekers’ hearings in a transidiomatic environment. Text & Talk, 29 (5), 525–46. Jaffe, A. (2003). Talk around text: Literacy practices, cultural identity and authority in a Corsican bilingual classroom. International Journal of Bilingual Education and Bilingualism, 6(3–4), 202–220. Jaffe, A., Androutsopoulos, J., Sebba, M., & Johnson, S. (Eds.) (2012) Orthography as social action: Scripts, spelling, identity and power. Berlin: Mouton de Gruyter. Jakobson, R. (1941). Kindersprache, Aphasie und allgemeine Lautgesetze. [Child language, aphasia, and general laws of sound] Uppsala: Almqvist & Wiksell. Jarvis, S. (2013). Crosslinguistic influence and multilingualism. In C. A. Chapelle (Ed.), The encyclopedia of applied linguistics. Blackwell Publishing Ltd. doi: 10.1002/9781405198431.wbeal0291

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

601

Jarvis, S. & Crossley, S. A. (Eds.). (2012). Approaching language transfer through text classification: Explorations in the detection-based approach. Bristol: Multilingual Matters. Jarvis, S. & Pavlenko, A. (2008). Crosslinguistic influence in language and cognition. London: Routledge. Jensen, S. H. (2017). Gaming as an English language learning resource among young children in Denmark. CALICO Journal, 34(1), 1–19. Jessner, U. (2006). Linguistic awareness in multilinguals. Edinburgh: Edinburgh University Press. Jessner, U. (2015). Multilingualism. In J. Wright (Ed.), International encyclopedia of the social and behavioral sciences (2nd ed.). Amsterdam: Elsevier. Jessner, U., Megens, M., & Graus, S. (2016). Crosslinguistic influence in third language acquisition. In R. Alonso (Ed.), Crosslinguistic influence in second language acquisition (pp. 193–214). Bristol: Multilingual Matters. Jia, F., Gottardo, A., Koh, P. W., Xi, C., & Pasquarella, A. (2014). The role of acculturation in reading a second language: Its relation to English literacy skills in immigrant Chinese adolescents. Reading Research Quarterly, 49(2), 251–261. Johannessen, J. B., & Salmons, J. C. (Eds.) (2015). Germanic heritage languages in North America: Acquisition, attrition and change. Amsterdam: John Benjamins. Johanson, L. (2002). Contact-induced change in a code-copying framework. Contributions to the Sociology of Language, 86, 285–314. Johns, B. T., Sheppard, C. L., Jones, M. N., & Taler, V. (2016). The role of semantic diversity in word recognition across aging and bilingualism. Frontiers in Psychology, 7, Article 703. Jørgensen, J. N. (2008). Polylingual languaging around and among children and adolescents. International Journal of Multilingualism, 5(3), 161–176. Jourdan, C. & Angeli, J. (2014). Pijin and shifting language ideologies in urban Solomon Islands. Language in Society, 43(3), 265–285. Juan-Garau, M. & Pe´rez-Vidal, C. (2001). Mixing and pragmatic parental strategies in early bilingual acquisition. Journal of Child Language, 28(1), 59–86. Juan-Garau, M. & Salazar-Noguera, J. (Eds.). (2015). Content-based language learning in multilingual educational environments. Berlin: Springer. Kadric´, M. (2001). Dolmetschen bei Gericht: Erwartungen, Anforderungen, Kompetenzen [Court interpretation: Expectations, requirements, competencies.]. Wien: WUV Universita¨tsverlag. Kalashnikova, M., Mattock, K., & Monaghan, P. (2015). The effects of linguistic experience on the flexible use of mutual exclusivity in word learning. Bilingualism: Language and Cognition, 18(4), 626–638. Kamanga, C. M. M. (2009). The joys and pitfalls of multiple language acquisition: The workings of the mind of a simultaneous multilingual. In E. Todeva & J. Cenoz, (Eds.), The multiple realities of multilingualism:

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

602

References

Personal narratives and researchers’ perspectives (pp. 115–134). Berlin: Mouton de Gruyter. Kamwangamalu, N. M. (2013). Multilingualism in southern Africa. In T. K. Bhatia & W. C. Ritchie (Eds.), The handbook of bilingualism and multilingualism (2nd ed., pp. 791–812). Malden, MA: Wiley-Blackwell. Kanto, L., Huttunen, K., & Laakso, M.-L. (2013). Relationship between the linguistic environments and early bilingual language development of hearing children in deaf-parented families. Journal of Deaf Studies and Deaf Education, 18(2), 242–261. Kanto, L., Laakso, M-L., & Huttunen, K. (2015). Differentiation in language and gesture use during early bilingual development of hearing children of Deaf parents. Bilingualism: Language and Cognition, 18(4), 769–788. Kasparian, K. & Steinhauer, K. (2016). Confusing similar words: ERP correlates of lexical-semantic processing in first language attrition and late second language acquisition. Neuropsychologia, 93(A), 200–217. Kasparian, K., Vespignani, F., & Steinhauer, K. (2017). First language attrition induces changes in online morphosyntactic processing and re-analysis: An ERP study of number agreement in complex Italian sentences. Cognitive Science, 41(7), 1760–1803. Kasuya, H. (1998). Determinants of language choice in bilingual children: The role of input. International Journal of Bilingualism, 2(3), 327–346. Kave´, G. & Goral, M. (2017). Do age-related word retrieval difficulties appear (or disappear) in connected speech? Aging, Neuropsychology and Cognition, 24(5), 508–527. Kave´, G., Knafo, A., & Gilboa, A. (2010). The rise and fall of word retrieval across the lifespan. Psychology and Aging, 25(3), 719–724. Kay-Raining Bird, E. (2016). Bilingualism and children with Down syndrome. In J. Patterson & B. Rodriguez (Eds.), Multilingual perspectives on child language disorders (pp. 49–73). Bristol: Multilingual Matters. Kay-Raining Bird, E., Genesee, F., & Verhoeven, L. (2016). Bilingualism in children with developmental disorders: A narrative review. Journal of Communication Disorders, 63, 1–14. Keijzer, M. (2007). Last in first out? An investigation of the regression hypothesis in Dutch emigrants in Anglophone Canada. Utrecht: LOT Publications. Keijzer, M. (2010). The regression hypothesis as a theoretical framework for first language attrition. Bilingualism: Language and Cognition, 13 (1), 9–18. Keijzer, M. (2011). Language reversion versus general cognitive decline: Towards a new taxonomy of language change in elderly bilingual immigrants. In M. S. Schmid & W. Lowie (Eds.), Modeling bilingualism: From structure to chaos (pp. 221–232). Amsterdam: John Benjamins. Keijzer, M. C. J. & Schmid, M. S. (2016). Individual differences in cognitive control advantages of elderly late Dutch–English bilinguals. Linguistic Approaches to Bilingualism, 6(1/2), 64–85.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

603

Kemp, C. (2007). Strategic processing in grammar learning: Do multilinguals use more strategies? International Journal of Multilingualism, 4(4), 241–261. Kemp, C. (2009). Defining multilingualism. In L. Aronin & B. Hufeisen (Eds.), The exploration of multilingualism (pp. 11–26). Amsterdam: John Benjamins. Kennedy, K. & Romo, H. (2013). “All colors and hues”: An autoethnography of a multiethnic family’s strategies for bilingualism and multiculturalism. Family Relations, 62(1), 109–124. Kharkhurin, A. V. & Altarriba, J. (2016). The effect of mood induction and language of testing on bilingual creativity. Bilingualism: Language and Cognition, 19(5), 1079–1094. Kharkhurin, A. V. & Li Wei. (2015). The role of code-switching in bilingual creativity. International Journal of Bilingual Education and Bilingualism, 18 (2), 153–169. Khattab, G. (2013). Phonetic convergence and divergence strategies in English–Arabic bilingual children. Linguistics, 51(2), 439–472. Kheirkhah, M. & Cekaite, A. (2015). Language maintenance in a multilingual family: Informal heritage language lessons in parent– child interactions. Multilingua, 34(3), 319–346. Khubchandani, L. M. (1997). Revisualizing boundaries: A plurilingual ethos. New Delhi: Sage. Kidd, E., Chan, A., & Chiu, J. (2014). Cross-linguistic influence in simultaneous Cantonese–English bilingual children’s comprehension of relative clauses. Bilingualism: Language and Cognition, 18(3), 438–452. King, K. A. (2000). Language ideologies and heritage language education. International Journal of Bilingual Education and Bilingualism, 3(3), 167–184. King, K. A., Fogle, L., & Logan-Terry, A. (2008). Family language policy. Language and Linguistics Compass, 2(5), 907–922. King, K. A. & Hermes, M. (2014). Why is this so hard? Ideologies of endangerment, passive language learning approaches, and Ojibwe in the United States. Journal of Language, Identity & Education, 13(4), 268–282. Kinsella, C. & Singleton, D. (2014). Much more than age. Applied Linguistics, 35(4), 441–462. Kinzler, K. D. & DeJesus, J. M. (2013). Children’s sociolinguistic evaluations of nice foreigners and mean Americans. Developmental Psychology, 49(4), 655–664. Kinzler, K. D., Dupoux, E., & Spelke, E. S. (2007). The native language of social cognition. Proceedings of the National Academy of Sciences, 104(30), 12577–12580. Kirby, J. R., Deacon, S. H., Bowers, P. N., Izenberg, L., Wade-Woolley, L., & Parrila, R. (2012). Children’s morphological awareness and reading ability. Reading and Writing, 25(2), 389–410. Klein, W. & Perdue, C. (1997). The Basic Variety (or: Couldn’t natural languages be much simpler?). Second Language Research, 13(4), 301–347.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

604

References

Kliesch, M., Giroud, N., Pfenninger, S. E., & Meyer, M. (2018). Research on second language acquisition in old adulthood: What we have and what we need. In D. Gabrys´-Barker (Ed.), Third age in second language acquisition (pp. 48–76). Bristol: Multilingual Matters. Kohn, K. (2011). English as a lingua franca and the Standard English misunderstanding. In A. De Houwer & A. Wilton (Eds.), English in Europe Today: Sociocultural and educational perspectives (pp. 71–94). Amsterdam: John Benjamins. Kohnert, K. (2013). Language disorders in bilingual children and adults (2nd ed.). San Diego: Plural Publishing. Kohnert, K., Kan, P. F., & Conboy, B. T. (2010). Lexical and grammatical associations in sequential bilingual preschoolers. Journal of Speech, Language, and Hearing Research, 53(3), 684–698. Kolk, H. H. J. (2001). Does agrammatic speech constitute a regression to child language? A three-way comparison between agrammatic, child and normal ellipsis. Brain & Language, 77(3), 340–350. Kong, A. P. H. (2011). The main concept analysis in Cantonese aphasic oral discourse: External validation and monitoring chronic aphasia. Journal of Speech, Language, and Hearing Research, 54(1), 148–159. Ko¨pke, B. (2004). Attrition is not a unitary phenomenon: On different possible outcomes of language contact situations. In A. M. L. Sua´rez, F. Ramallo, & X. P. R. Ya´n˜ez (Eds.), Bilingual socialization and bilingual language acquisition. Proceedings from the Second International Symposium on Bilingualism (pp. 1331–1347). Vigo: Servizo de Publicacions da Universidade de Vigo. Ko¨pke, B. (2007). Language attrition at the crossroads of brain, mind, and society. In M. S. Schmid, B. Ko¨pke, M. Keijzer, & S. Dostert (Eds.), First language attrition: Theoretical perspectives (pp. 9–37). Amsterdam: John Benjamins. Ko¨pke, B. & Schmid, M. S. (2004). First language attrition: The next phase. In M. Schmid, B. Ko¨pke, M. Keijzer, & L. Weilemar (Eds.), First language attrition: Interdisciplinary perspectives on methodological issues (pp. 1–43). Amsterdam: John Benjamins. Koven, M. (2017). Language awareness and emotion. In D. Gorter & J. Cenoz (Eds.), Encyclopedia on language and education, Volume 6: Multilingualism and language awareness (pp. 67–82). Berlin: Springer. Kozar, O. & Yates, L. (2018). Factors in language learning after 40: Insights from a longitudinal study. IRAL: International Review of Applied Linguistics in Language Teaching. doi:10.1515/iral-2015–0113 Kramsch, C. (2014). Teaching foreign languages in an era of globalization: Introduction. Modern Language Journal, 98(1), 296–311. Krashen, S. D., Long, M. H., & Scarcella, R. C. (1979). Age, rate and eventual attainment in second language acquisition. TESOL Quarterly, 13(4), 573–582.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

605

Krausneker, V. (2015). Ideologies and attitudes toward sign languages: An approximation. Sign Language Studies, 15(4), 411–431. Kremmel, B. & Schmitt, N. (2016) Interpreting vocabulary test scores: What do various item formats tell us about learners’ ability to employ words? Language Assessment Quarterly, 13(4), 377–392. Krieger, N. (2014). Discrimination and health inequities. International Journal of Health Services, 44(4), 643–710. Kristen, C., Mu¨hlau, P., & Schacht, D. (2016). Language acquisition of recently arrived immigrants in England, Germany, Ireland, and the Netherlands. Ethnicities, 16(2), 180–212. Krizman, J. & Marian, V. (2015). Neural consequences of bilingualism for cortical and subcortical function. In J. W. Schwieter (Ed.), The Cambridge handbook of bilingual processing (pp. 614–630). Cambridge: Cambridge University Press. Kroll, J. F., Bobb, S. C., & Wodniecka, Z. (2006). Language selectivity is the exception, not the rule: Arguments against a fixed locus of language selection in bilingual speech. Bilingualism: Language and Cognition, 9(2), 119–135. Kroll, J. F., Dussias, P. E., Bice, K. & Perroti, L. (2015). Bilingualism, mind, and brain. Annual Review of Linguistics, 1(1), 377–394. Kroll, J. F. & Stewart, E. (1994). Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language, 33(2), 149–174. Kroskrity, P. V. (Ed.). (2000). Regimes of language: Ideologies, polities, and identities. Santa Fe: School for Advanced Research Press. Kuchah, K. (2016). English-medium instruction in an English–French bilingual setting: Issues of quality and equity in Cameroon. Comparative Education, 52(3), 311–327. Kuhl, P. K. (2004). Early language acquisition: Cracking the speech code. Nature Reviews Neuroscience, 5(11), 831–843. Kulick, D. (1992). Language shift and cultural reproduction: Socialization, self, and syncretism in a Papua New Guinean village. Cambridge: Cambridge University Press. Kupisch, T. & Rothman, J. (2016). Terminology matters! Why difference is not incompleteness and how early child bilinguals are heritage speakers. International Journal of Bilingualism. doi:10.1177/13670069166 54355 Kupisch, T., & van de Weijer, J. (2016). The role of the childhood environment for language dominance: A study of adult simultaneous bilingual speakers of German and French. In C. Silva-Corvala´n & J. Treffers-Daller (Eds.), Language dominance in bilinguals: Issues of operationalization and measurement (pp. 174–194). Cambridge: Cambridge University Press. Labov, W. (1969). Contraction, deletion and inherent variability of the English copula. Language, 45(4), 715–762.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

606

References

Labov, W. (1990). The intersection of sex and social class in the course of linguistic change. Language Variation and Change, 2(2), 205–254. Labov, W. (2013). Preface: The acquisition of sociolinguistic variation. Linguistics, 51(2), 247–250. Lagrou, E., Hartsuiker, R. J., & Duyck, W. (2011). Knowledge of a second language influences auditory word recognition in the native language. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(4), 952–965. Lahmann, C., Steinkrauss, R., & Schmid, M. S. (2016). Factors affecting grammatical and lexical complexity of long-term L2 speakers’ oral proficiency. Language Learning, 66(2), 354–385. Laks, B. (2013). Why is there variation rather than nothing? Language Sciences, 39, 31–53. Lambert, R. D. & Freed, B. F. (1982). The loss of language skills. Rowley, MA: Newbury House. Lambert, W. E. (1974). Culture and language as factors in learning and education. In F. E. Aboud & R. D. Meade (Eds.), Cultural factors in learning and education (pp. 91–122). Bellingham, WA: 5th Western Washington Symposium on Learning. Language and National Origin Group (2004). Guidelines for the use of language analysis in relation to questions of national origin in refugee cases. International Journal of Speech Language and the Law, 11(2), 261–266. Lantolf, J. P. & Poehner, M. E. (2014). Sociocultural theory and the pedagogical imperative in L2 education: Vygotskian praxis and the research/practice divide. London: Routledge. Lanvers, U. (2001). Language alternation in infant bilinguals: A developmental approach to codeswitching. International Journal of Bilingualism, 5(4), 437–464. Lanza, E. (1992). Can bilingual two-year-olds code-switch? Journal of Child Language, 19(3), 633–658. Lanza, E. (1997). Language mixing in infant bilingualism: A sociolinguistic perspective. Oxford: Clarendon Press. Lanza, E. (2007). Multilingualism and the family. In P. Auer & Li Wei (Eds.), Handbook of multilingualism and multilingual communication (pp. 45–67). Berlin: Mouton de Gruyter. Lanza, E., & Li Wei. (Eds.) (2016). Multilingual encounters in transcultural families. Special Issue. Journal of Multilingual & Multicultural Development, 37(7). Lapkin, S., Hart, D., & Harley, B. (1998). Case study of compact Core French models: Attitudes and achievement. In S. Lapkin (Ed.), French second language education in Canada: Empirical studies (pp. 3–30). Toronto: University of Toronto Press. Larsen-Freeman, D. (2015). Complexity theory. In B. VanPatten & J. Williams (Eds.), Theories in second language acquisition (2nd ed., pp. 227–244). London: Routledge.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

607

Larsen-Freeman, D. (2017). Complexity Theory: The lessons continue. In L. Ortega & Z. Han (Eds.), Complexity Theory and language development: In celebration of Diane Larsen-Freeman (pp. 11–50). Amsterdam: John Benjamins. Larsen-Freeman, D., Schmid, M., & Lowie, W. (2011). Introduction: From structure to chaos: Twenty years of modeling bilingualism. In M. Schmid & W. Lowie (Eds.), Modeling bilingualism: From structure to chaos. In honor of Kees de Bot (pp. 1–12). Amsterdam: John Benjamins. Lasagabaster, D. & Doiz, A. (2017). A longitudinal study on the impact of CLIL on affective factors. Applied Linguistics, 38(5), 688–712. Lebrun, Y. (1995). The study of bilingual aphasia: Pitres’ legacy. In M. Paradis (Ed.), Aspects of bilingual aphasia (pp. 11–22). Oxford: Pergamon Press. LeClair, N., Doll, B., Osborn, A., & Jones, K. (2009). English language learners’ and non-English language learners’ perceptions of the classroom environment. Psychology in the Schools, 46(6), 568–577. Lee, A. S. & Baskerville, R. L. (2003). Generalizing generalizability in information systems research. Information Systems Research, 14(3), 221–243. Lee, J. (2010). Interpreting reported speech in witnesses’ evidence. Interpreting, 12(1), 60–82. Leeman, J. & King, K. A. (2015). Heritage language education: Minority language speakers, second language instruction, and monolingual schooling. In M. Bigelow & J. Ennser-Kananen (Eds.), The Routledge handbook of educational linguistics (pp. 210–223). London: Routledge. Leeman, J. & Martinez, G. (2007). From identity to commodity: Ideologies of Spanish in heritage language textbooks. Critical Inquiry in Language Studies, 4(1), 35–65. Lefebvre, C. (1998). Creole genesis and the acquisition of grammar: The case of Haitian Creole. Cambridge: Cambridge University Press. Legacy, J., Zesiger, P., Friend, M., & Poulin-Dubois, D. (2016). Vocabulary size, translation equivalents, and efficiency in word recognition in very young bilinguals. Journal of Child Language, 43(4), 760–783. Lenet, A. E., Sanz, C., Lado, B., Howard, J. H. J., & Howard, D. V. (2011). Aging, pedagogical conditions, and differential success in SLA: An empirical study. In C. Sanz & R. P. Leow (Eds.), Implicit and explicit language learning: Conditions, processes, and knowledge in SLA and bilingualism (pp. 73–84). Washington, DC: Georgetown University Press. Lenneberg, E. H. (1967). Biological foundations of language. New York: John Wiley and Sons. Lesaux, N. K. & Harris, J. R. (2017). An investigation of comprehension processes among adolescent English learners with reading difficulties. Topics in Language Disorders, 37(2), 182–203. Levy, B. J., McVeigh, N. D., Marful, A., & Anderson, M. C. (2007). Inhibiting your native language: The role of retrieval-induced forgetting during second language acquisition. Psychological Science, 18(1), 29–34.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

608

References

Lewis, L. & Gray, L. (2016). Programs and services for high school English learners in public school districts: 2015–16 (NCES 2016–150). US Department of Education. Washington, DC: National Center for Education Statistics. https://nces.ed.gov/pubs2016/2016150.pdf, last accessed July 20, 2018. Li, P., Farkas, I., & MacWhinney, B. (2004). Early lexical development in a self-organizing neural networks. Neural Networks, 17(8–9), 1345–1362. Licoppe, C. & Vernier, M. (2013). Interpreting, video communication and the sequential reshaping of institutional talk in the bilingual and distributed courtroom. International Journal of Speech Language and the Law, 20(2), 247–275. Lieberman, A. M., Borovsky, A., Hatrak, M., & Mayberry, R. I. (2015). Realtime processing of ASL signs: Delayed first language acquisition affects organization of the mental lexicon. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41(4), 1130–1139. Lieberman, A. M. & Mayberry, R. I. (2015). Studying sign language acquisition. In E. Orfanidou, B. Woll, & G. Morgan (Eds.), Research methods in sign language studies: A practical guide (pp. 281–299). Malden, MA: Wiley-Blackwell. Lieven, E. V., Pine, J. M., & Baldwin, G. (1997). Lexically-based learning and early grammatical development. Journal of Child Language, 24(1), 187–219. Lightbown, P. (2014). Focus on content-based language teaching. Oxford: Oxford University Press. Lightbown, P. M., Halter, R., White, J., & Horst, M. (2002). Comprehensionbased learning: The limits of “Do it yourself.” Canadian Modern Language Review, 58(3), 427–464. Lightbown, P. M. & Spada, N. (1994). An innovative program for primary ESL in Quebec. TESOL Quarterly, 28(3), 563–579. Lightbown, P. M. & Spada, N. (2013). How languages are learned (4th ed.). Oxford: Oxford University Press. Lillo-Martin, D. (2000). Aspects of the syntax and acquisition of WH-questions in American Sign Language. In K. Emmorey & H. Lane (Eds.), The signs of language revisited: An anthology in honor of Ursula Bellugi and Edward Klima (pp. 401–413). Mahwah, NJ: Lawrence Erlbaum. Lillo-Martin, D., de Quadros, R. M., & Chen Pichler, D. (2016). The development of bimodal bilingualism. Linguistic Approaches to Bilingualism, 6(6), 719–755. Lillo-Martin, D., de Quadros, R. M., Chen Pichler, D., & Fieldsteel, Z. (2014). Language choice in bimodal bilingual development. Frontiers in Psychology, 5, Article 1163. Lillo-Martin, D., de Quadros, R. M., Koulidobrova, H., & Chen Pichler, D. (2010). Bimodal bilingual cross-language influence in unexpected domains. In J. Costa, A. Castro, M. Lobo, & F. Pratas (Eds.), Language

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

609

acquisition and development: Proceedings of GALA 2009 (pp. 264–275). Newcastle upon Tyne: Cambridge Scholars Publishing. Lillo-Martin, D., Koulidobrova, H., de Quadros, R. M., & Chen Pichler, D. (2012). Bilingual language synthesis: Evidence from WH-questions in bimodal bilinguals. In A. K. Biller, E. Y. Chung, & A. E. Kimball (Eds.), BUCLD 36 Proceedings. Somerville, MA: Cascadilla Press. Linck, J. A., Hughes, M. M., Campbell, S. G., Silbert, N. H., Tare, M., Jackson, S. R., Smith, B. K., Bunting, M. F., & Doughty, C. J. (2013). HiLAB: A new measure of aptitude for high-level language proficiency. Language Learning, 63(3), 530–566. Linck, J. A., Kroll, J. F., & Sunderman, G. (2009). Losing access to the native language while immersed in a second language: Evidence for the role of inhibition in second-language learning. Psychological Science, 20(12), 1507–1515. Linck, J. A., Osthus, P., Koeth, J. T., & Bunting, M. F. (2014). Working memory and second language comprehension and production: A meta-analysis. Psychonomic Bulletin & Review, 21(4), 861–883. Lincoln, P. C. (1979/80). Dual lingualism: Passive bilingualism in action. Te Reo, 22/23, 65–72. Lindholm-Leary, K. (2014). Bilingual and biliteracy skills in young Spanish-speaking low-SES children: Impact of instructional language and primary language proficiency. International Journal of Bilingual Education and Bilingualism, 17(2), 144–159. Lindholm-Leary, K. J. & Block, N. (2010). Achievement in predominantly low-SES Hispanic dual language schools. International Journal of Bilingual Education and Bilingualism, 13(1), 43–60. Lippi-Green, R. (2012). English with an accent: Language, ideology and discrimination in the United States (2nd edn.). London: Routledge. Llinares, A., Morton, T., & Whittaker, R. (2012). The roles of language in CLIL. Cambridge: Cambridge University Press. Llurda, E. (2016). “Native speakers,” English and ELT. In G. Hall (Ed.), The Routledge handbook of English language teaching (pp. 51–63). London: Routledge. Lo, Y. Y. & Lo, E. S. C. (2014). A meta-analysis of the effectiveness of English-medium education in Hong Kong. Review of Educational Research, 84(1), 47–73. Lo Bianco, J. (2010). Language policy and planning. In N. Hornberger & S. McKay (Eds.), New perspectives on language and education (pp. 143–176). Bristol: Multilingual Matters. Lo Bianco, J. (2014). Domesticating the foreign: Globalization’s effects on the place/s of languages. Modern Language Journal, 98(1), 312–325. Lo Bianco, J. & Aliani, R. (2013). Language planning and student experiences. Bristol: Multilingual Matters. Lo Bianco, J. & Slaughter, Y. (2016). The Australian Asia project. In G. Leitner, A. Hashim, & H.-G. Wolf (Eds.), Communicating with Asia:

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

610

References

The future of English as a global language (pp. 296–312). Cambridge: Cambridge University Press. Loewen, S. & Sato, M. (Eds.). (2017). The Routledge handbook of instructed second language acquisition. London: Routledge. Long, M. H. (2013). Maturational constraints on child and adult SLA. In G. Granena & M. H. Long (Eds.), Sensitive periods, language aptitude, and ultimate L2 attainment (pp. 3–41). Amsterdam: John Benjamins. Long, M. H. (2015). Second language acquisition and task-based language teaching. Malden, MA: Wiley-Blackwell. Lopez, F. A. (2012). Moderators of language acquisition models and reading achievement for English language learners: The role of emotional warmth and instructional support. Teachers College Record, 114(1), 1–30. Lo¨vde´n, M., Wenger, E., Maº rtensson, J., Lindenberger, U., & Ba¨ckman, L. (2013). Structural brain plasticity in adult learning and development. Neuroscience & Biobehavioral Reviews, 37(9), 2296–2310. Lu, J., Jones, A., & Morgan, G. (2016). The impact of input quality on early sign development in native and non-native language learners. Journal of Child Language, 43(3), 537–552. Luk, G. (2015). Who are the bilinguals (and monolinguals)? Bilingualism: Language and Cognition, 18(1), 35–36. Luk, G. & Bialystok, E. (2013). Bilingualism is not a categorical variable: Interaction between language proficiency and usage. Journal of Cognitive Psychology, 25(5), 605–621. Luk, G., de Sa, E., & Bialystok, E. (2011). Is there a relation between onset age of bilingualism and enhancement of cognitive control? Bilingualism: Language and Cognition, 14(4), 588–595. Lvovich, N. (2015). Translingual identity and art: Marc Chagall’s stride through the gates of Janus. Critical Multilingualism Studies, 3(1), 112–134. Lyster, R. (2007). Learning and teaching languages through content: A counterbalanced approach. Amsterdam: John Benjamins. Lyster, R., Saito, K., & Sato, M. (2013). Oral corrective feedback in second language classrooms. Language Teaching, 46(1), 1–40. Macdonald, S. (2013). Migration and English language learning in the United Kingdom: Towards a feminist theorising. Power and Education, 5(3), 291–303. Macizo, P., Bajo, T., & Martin, M. (2010). Inhibitory processes in bilingual language comprehension: Evidence from Spanish–English interlexical homographs. Journal of Memory and Language, 63(2), 232–244. Mackenzie, I. (2014). English as alingua francaTheorizing and teaching English. London: Routledge. Mackey, A. (2012). Input, interaction, and corrective feedback in L2 learning. Oxford: Oxford University Press. Mackey, A. & Sachs, R. (2012). Older learners in SLA research: A first look at working memory, feedback, and L2 development. Language Learning, 62(3), 704–740.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

611

Mackey, W. F. (2013). Bilingualism and multilingualism in North America. In T. K. Bhatia & W. C. Ritchie (Eds.), The handbook of bilingualism and multilingualism (2nd ed., pp. 707–724). Malden, MA: Wiley-Blackwell. MacLeod, C. (1976). Bilingual episodic memory: Acquisition and forgetting. Journal of Verbal Learning and Verbal Behavior, 15(4), 347–364. MacLeod, C. (1988). Forgotten but not gone: Savings for pictures and words in long-term memory. Journal of Experimental Psychology: Learning, Memory and Cognition, 14(2), 195–212. MacSwan, J. (1999). A minimalist approach to intrasentential code switching. New York: Garland. MacSwan, J. (2005). Codeswitching and generative grammar: A critique of the MLF model and some remarks on “modified minimalism.” Bilingualism: Language and Cognition, 8(1), 1–22. MacSwan, J. (2016). Code-switching in adulthood. In E. Nicoladis & S. Montanari (Eds.), Bilingualism across the lifespan: Factors moderating proficiency (pp. 183–200). Berlin: Walter de Gruyter & American Psychological Association. MacWhinney, B. (2000). The CHILDES project: Tools for analyzing talk (3rd ed.). Mahwah, NJ: Lawrence Erlbaum. Mady, C. (2015). Examining immigrants’ English and French proficiency in French immersion. Journal of Immersion and Content-Based Language Education, 3(2), 268–284. Majone, G. (1989). Evidence, argument, and persuasion in the policy process. New Haven, CT: Yale University Press. Makoe, P. & McKinney, C. (2014). Linguistic ideologies in multilingual South African suburban schools. Journal of Multilingual & Multicultural Development, 35(7), 658–673. Makoni, S. & Pennycook, A. (2007). Disinventing and reconstituting languages. Clevedon: Multilingual Matters. Malcolm, N. (1995). Wittgenstein: The relation of language to instinctive behaviour. Philosophical Investigations, 5(1), 3–22. Mancilla-Martinez, J. & Lesaux, N. (2011). The gap between Spanish speakers’ word reading and word knowledge: A longitudinal study. Child Development, 82(5), 1544–1560. Marchman, V. A. & Bates, E. (1994). Continuity in lexical and morphological development: A test of the critical mass hypothesis. Journal of Child Language, 21(2), 339–366. Marchman, V. A. & Fernald, A. (2008). Speed of word recognition and vocabulary knowledge in infancy predict cognitive and language outcomes in later childhood. Developmental Science, 11(3), F9–F16. Marchman, V. A., Fernald, A., & Hurtado, N. (2010). How vocabulary size in two languages relates to efficiency in spoken word recognition by young Spanish–English bilinguals. Journal of Child Language, 37(4), 817–840.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

612

References

Marchman, V. A., Martı´nez-Sussmann, C., & Dale, P. S. (2004). The language-specific nature of grammatical development: Evidence from bilingual language learners. Developmental Science, 7(2), 212–224. Marchman, V. A., Martı´nez, L. Z., Hurtado, N., Gru¨ter, T., & Fernald, A. (2017). Caregiver talk to young Spanish–English bilinguals: Comparing direct observation and parent-report measures of duallanguage exposure. Developmental Science, 20(1), e12425. Marcon, R. A. & R. C. Coon (1983). Communication styles of bilingual preschoolers in preferred and non-preferred languages. Journal of Genetic Psychology, 142(2), 189–202. Marcotte, K. & Ansaldo, A. I. (2014). Age-related behavioural and neurofunctional patterns of second language word learning: Different ways of being successful. Brain & Language, 135, 9–19. Marian, V., Blumenfeld, H. K., & Kaushanskaya, M. (2007). The Language Experience and Proficiency Questionnaire (LEAP-Q): Assessing language profiles in bilinguals and multilinguals. Journal of Speech, Language, and Hearing Research, 50(4), 940–967. Marinis, T. & Armon-Lotem, S. (2015). Sentence repetition. In S. ArmonLotem, J. de Jong, & N. Meir (Eds.), Assessing multilingual children: Disentangling bilingualism from language impairment (pp. 95–122). Bristol: Multilingual Matters. Marinis, T., Armon-Lotem, S., & Pontikas, G. (Eds.). (2017). Language impairment in bilingual children [Special Issue]. Linguistic Approaches to Bilingualism, 7(3/4), 265–475. Marinova-Todd, S. H., Colozzo, P., Mirenda, P., Stahl, H., Kay-Raining Bird, E., Parkington, K., Cain, K., Scherba de Valenzuela, J., Segers, E., MacLeod, A. A. N., & Genesee, F. (2016). Professional practices and opinions about services available to bilingual children with developmental disabilities: An international study. Journal of Communication Disorders, 63, 47–62. Marinova-Todd, S. H. & Mirenda, P. (2016). Language and communication abilities in bilingual children with autism spectrum disorders. In J. Patterson & B. Rodriguez (Eds.), Multilingual perspectives on child language disorders (pp. 31–48). Bristol: Multilingual Matters. Markwardt, A. H. (1948). Motives for the study of modern languages. Language Learning, 1(1), 1–11. Marley, D. (2013). The role of online communication in raising awareness of bilingual identity. Multilingua, 32(4), 485–506. Marriot, H. (1995). The acquisition of politeness patterns by exchange students in Japan. In B. Freed (Ed.), Second language acquisition in a study abroad context (pp. 197–224). Amsterdam: John Benjamins. Marschak, J. (1965). The economics of language. Behavioral Science, 10(2), 135–140. Marsden, E., Mackey, A., & Plonsky, L. (2016). The IRIS Repository: Advancing research practice and methodology. In A. Mackey &

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

613

E. Marsden (Eds.), Advancing methodology and practice: The IRIS Repository of instruments for research into second languages (pp. 1–21). London: Routledge. Marszalenko, J. E. (2014). Three stages of interpreting in Japan’s criminal process. Language and Law = Linguagem e Direito, 1(1), 174–87. º Martensson, J., Eriksson, J., Bodammer, N. C., Lindgren, M., Johansson, M., Nyberg, L., & Lo¨vde´n, M. (2012). Growth of language-related brain areas after foreign language learning. NeuroImage, 63(1), 240–244. Martinez, R. & Schmitt, N. (2012). A phrasal expressions list. Applied Linguistics, 32(3), 299–320. Maryns, K. (2006). The asylum speaker: Language in the Belgian asylum procedure. Manchester: St. Jerome Publishing. ¨ zc¸alıskan, S., & Hoff, E. (2016). Parental translation of child Mateo, V., O gesture helps bilingual vocabulary development. In J. Scott & D. Waughtal (Eds.), Proceedings of the 40th Boston University Conference on Language Development (pp. 241–252). Sommerville, MA: Cascadilla Press. Mathis, N. (2015). Plurilingual literacy practices in a creative writing workshop with adult second language learners. In J. Simpson & A. Whiteside (Eds.). Adult language education and migration: Challenging agendas in policy and practice (pp. 138–148). London: Routledge. Matras, Y. (2009). Language contact. Cambridge: Cambridge University Press. May, S. (Ed.). (2014). The multilingual turn: Implications for SLA, TESOL and bilingual education. London: Routledge. Mayberry, R. I. & Lock, E. (2003). Age constraints on first versus second language acquisition: Evidence for linguistic plasticity and epigenesis. Brain and Language, 87(3), 369–384. Mayer, C. & Leigh, G. (2010). The changing context for sign bilingual education programs: Issues in language and the development of literacy. International Journal of Bilingual Education and Bilingualism, 13 (2), 175–186. Mayer, R. E. (2009). Multimedia learning (2nd ed.). Cambridge: Cambridge University Press. Mayr, R. & Montanari, S. (2015). Cross-linguistic interaction in trilingual phonological development: The role of the input in the acquisition of the voicing contrast. Journal of Child Language, 42(5), 1006–1035. Mayr, R., Price, S., & Mennen, I. (2012). First language attrition in the speech of Dutch–English bilinguals: The case of monozygotic twin sisters. Bilingualism: Language and Cognition, 15(4), 687–700. Mazzucelli, C. (1997). France and Germany at Maastricht: Politics and negotiations to create the European Union. London: Routledge. McArthur, T. (1992). The Oxford companion to the English language. Oxford: Oxford University Press. McAuliffe, K. (2012). Language and law in the European Union: The multilingual jurisprudence of the ECJ. In P. M. Tiersma &

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

614

References

L. M. Solan (Eds.), The Oxford handbook of language and law (pp. 201–216). Oxford: Oxford University Press. McCarty, T. & Lee, T. (2014). Critical culturally sustaining/revitalizing pedagogy and Indigenous education sovereignty. Harvard Educational Review, 84(1), 101–124. McClure, E. (1977). Aspects of code-switching in the discourse of bilingual Mexican-American children. In M. Saville-Troike (Ed.), Latino language and communicative behavior (pp. 92–115). Washington, DC: Georgetown University Press. McClure, E. & Wentz, J. (1975). Functions of code switching among Mexican-American children. In R. Grossman, L. J. San, & T. Vance (Eds.), Papers from the parasession on functionalism (pp. 421–432). Chicago: Chicago Linguistic Society. McDonald, J. L. (2006). Beyond the critical period: Processing-based explanations for poor grammaticality judgment performance by late second language learners. Journal of Memory and Language, 55(3), 381–401. McDonald, M. (1994). Women and linguistic innovation in Brittany. In P. Burton, K. K. Dyson, & S. Ardener (Eds.), Bilingual women: Anthropological approaches to second language use (pp. 85–110). Exeter: Short Run Press. McEwan-Fujita, E. (2010). Ideology, affect, and socialization in language shift and revitalization: The experiences of adults learning Gaelic in the Western Isles of Scotland. Language in Society, 39(1), 27–64. McHugh, M. & Challinor, A. E. (2011). Improving immigrants’ employment prospects through work-focused language instruction. Washington, DC: Migration Policy Institute. McLaughlin, B. (1992). Myths and misconceptions about second language learning: What every teacher needs to unlearn. Educational Practice Report 5, University of California. http://cmmr.usc.edu/FullText/ McLaughlinMyths.pdf, last accessed January 15, 2018. McLaughlin, B. & Nayak, N. (1989). Processing a new language: Does knowing other languages make a difference? In H. W. Dechert & M. Raupach (Eds.), Interlingual processes (pp. 5–16). Tu¨bingen: Gunter Narr. McLeod, S. (2014). Resourcing speech-language pathologists to work with multilingual children. International Journal of Speech-Language Pathology, 16(3), 208–218. McLeod, S. & Goldstein, B. (2012). Multilingual aspects of speech sound disorders in children. Bristol: Multilingual Matters. Meakins, F. & O’ Shannessy, C. (Eds.) (2016). Loss and renewal: Australian languages since colonisation. Berlin: De Gruyter Mouton. Meara, P. (2005). LLAMA Language Aptitude Tests. Swansea: Lognostics. Mechelli, A., Crinion, J. T., Noppeney, U., O’Doherty, J., Ashburner, J., Frackowiak, R. S., & Price, C. J. (2004). Neurolinguistics: Structural plasticity in the bilingual brain. Nature, 431(7010), 757.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

615

Mehler, J., Jusczyk, P., Lambertz, G., Halsted, N., Bertoncini, J., & AmielTison, C. (1988). A precursor of language acquisition in young infants. Cognition, 29(2), 143–178. Meisel, J. M. (1989). Early differentiation of languages in bilingual children. In K. Hyltenstam & L. K. Obler (Eds.), Bilingualism across the lifespan: Aspects of acquisition, maturity and loss (pp. 13–40). Cambridge: Cambridge University Press. Meisel, J. M. (2004). The bilingual child. In T. K. Bhatia & W. C. Ritchie (Eds.), The handbook of bilingualism (pp. 91–113). Malden, MA: Blackwell. Melby-Lervaº g, M. & Lervaº g, A. (2014). Reading comprehension and its underlying components in second-language learners: A meta-analysis of studies comparing first- and second-language learners. Psychological Bulletin, 140(2), 409–433. Me´litz, J. (2016). Langues, commerce, bien-eˆtre et francophonie. In C. Carre`re (Ed.), L’impact e´conomique des langues (pp. 50–70). Paris: Economica. Me´litz, J. & Toubal, F. (2014). Native language, spoken language, translation and trade. Journal of International Economics, 93(2), 351–363. Menezes, E. (2014). Ansiedad y disposicio´n a comunicarse en el aprendizaje del ingle´s como segunda lengua: Estudio de las influencias del modelo formative (AICLE y Ensen˜anza Formal). PhD Dissertation. University of the Balearic Islands. Available from: www.tdx.cat/bitstream/handle/10803/291816/ tesm1de1.pdf?sequence=1. Menezes, E. & Juan-Garau, M. (2015). English learners’ willingness to communicate and achievement in CLIL and formal instruction contexts. In M. Juan-Garau & J. Salazar-Noguera (Eds.), Content-based language learning in multilingual educational environments (pp. 221–236). Berlin: Springer. Meng, X. & Meurs, D. (2009). Intermarriage, language, and the economic assimilation process. International Journal of Manpower, 30(1), 127–144. Mercado, L. A. (2015). Integrating classroom learning and autonomous learning. In D. Nunan & J. C. Richards (Eds.), Language learning beyond the classroom (pp. 190–201). London: Routledge. Mettewie, L. & Van Mensel, L. (2006). Entreprises bruxelloises et langues e´trange`res. Pratique et couˆt d’une main d’oeuvre ne maıˆtrisant pas les langues e´trange`res. Brussels: TIBEM. Migration Policy Institute. (no date). International migration statistics (data hub). Retrieved from www.migrationpolicy.org/programs/data-hub/ international-migration-statistics, last accessed February 1, 2018. Mihaljevic´ Djigunovic´, J. (2011). Impact of learning conditions on young FL learners’ motivation. In M. Nikolov (Ed.), Early learning of modern foreign languages: Processes and outcomes (pp. 75–89). Bristol: Multilingual Matters.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

616

References

Miller, J. & Iglesias, A. (2012). Systematic Analysis of Language Transcripts (SALT), Research Version 2012 [Computer Software]. Middleton, WI: SALT Software, LLC. Miller Amberber, A. & Cohen, H. (Eds.). (2012). Clinical and research applications of the Bilingual Aphasia Test [Special Issue]. Journal of Neurolinguistics, 25(6), 515–654. Milroy, L. & Li Wei. (1995). A social network perspective on code-switching and language choice: The example of the Tyneside Chinese community. In L. Milroy & P. Muysken (Eds.), One speaker, two languages: Crossdisciplinary perspectives on code-switching (pp. 136–157). Cambridge: Cambridge University Press. Mishina-Mori, S. (2011). A longitudinal analysis of language choice in bilingual children: The role of parental input and interaction. Journal of Pragmatics, 43(13), 3122–3138. Moeketsi, R. (1999). Discourse in a multilingual and multicultural courtroom: A court interpreter’s guide. Hatfield: J. L. van Schaik. Mohanty, A. K. (2006). Multilingualism of the unequals and predicaments of education in India: Mother tongue or other tongue? In O. Garcı´a, T. Skutnabb-Kangas, & M. Torres-Guzma´n (Eds.), Imagining multilingual schools: Languages in education and glocalization (pp. 262–283). Clevedon: Multilingual Matters. Molnar, M., Gervain, J., & Carreiras, M. (2014). Within-rhythm class native language discrimination abilities of Basque–Spanish monolingual and bilingual infants at 3.5 months of age. Infancy, 19(3), 326–337. Montero Pe´rez, M., Van Den Noortgate, W., & Desmet, P. (2013). Captioned video for L2 listening and vocabulary learning: A meta-analysis. System 41(3), 720–739. Montrul, S. (2002). Incomplete acquisition and attrition of Spanish tense/ aspect distinctions in adult bilinguals. Bilingualism: Language and Cognition, 5(1), 39–68. Montrul, S. (2008). Incomplete acquisition in bilingualism: Re-examining the age factor. Amsterdam: John Benjamins. Montrul, S. (2012). Is the heritage language like a second language? EUROSLA Yearbook, 12, 1–29. Morales, J., Calvo, A., & Bialystok, E. (2013). Working memory development in monolingual and bilingual children. Journal of Experimental Child Psychology, 114(2), 187–202. Morford, J. O., Kroll, J. F., Pin˜ar, P., & Wilkinson. E. (2014). Bilingual word recognition in deaf and hearing signers: Effects of proficiency and language dominance on cross language activation. Second Language Research, 30(2), 251–271. Morgan-Short, K., Sanz, C., Steinhauer, K., & Ullman, M. T. (2010). Second language acquisition of gender agreement in explicit and implicit training conditions: An event-related potential study. Language Learning, 60(1), 154–193.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

617

Morris, R. (2008). Missing stitches: An overview of judicial attitudes to interlingual interpreting in the criminal justice systems of Canada and Israel. Interpreting, 10(1), 34–64. Moyer, A. (1999). Ultimate attainment in L2 phonology: The critical factors of motivation and instruction. Studies in Second Language Acquisition, 21(1), 81–108. Moyer, A. (2013). Foreign accent: The phenomenon of non-native speech. Cambridge: Cambridge University Press. Mukadam, N., Sommerlad, A., & Livingston, G. (2017). The relationship of bilingualism compared to monolingualism to the risk of cognitive decline or dementia: A systematic review and meta-analysis. Journal of Alzheimer’s Disease, 58(1), 45–54. Mu¨ller, N. (2009). Aging with French: Observations from South Louisiana. Journal of Cross Cultural Gerontology, 24(2), 143–155. Mun˜oz, C. (2006). The effects of age on foreign language learning: The BAF project. In C. Mun˜oz (Ed.), Age and rate of foreign language learning (pp. 1–40). Clevedon: Multilingual Matters. Mun˜oz, C. (2012). Intensive exposure experiences in second language learning. Bristol: Multilingual Matters. Mun˜oz, C. & Singleton, D. (2011). A critical review of age-related research on L2 ultimate attainment. Language Teaching, 44(1), 1–35. Mun˜oz, M. L. & Marquardt, T. P. (2008). The performance of neurologically normal bilingual speakers of Spanish and English on the short version of the Bilingual Aphasia Test. Aphasiology, 22(1), 3–19. Mun˜oz, M. L., Marquardt, T. P., & Copeland, G. (1999). A comparison of the codeswitching patterns of aphasic and neurologically normal bilingual speakers of English and Spanish. Brain and Language, 66(2), 249–274. Murakami, A. & Alexopoulou, T. (2016). L1 influence on the acquisition order of English grammatical morphemes: A learner corpus study. Studies in Second Language Acquisition, 38(3), 365–401. Murphy, V. & Evangelou, M. (2016). Introduction. In V. A. Murphy & M. Evangelou (Eds.), Early childhood education in English for speakers of other languages (pp. 4–18). London: British Council. Murphy, V. A., Macaro, E., Alba, S., & Cipolla, C. (2015). The influence of learning a second language in primary school on developing first language literacy skills. Applied Psycholinguistics, 36(5), 1133–1153. Muysken, P. (1981). Halfway between Quechua and Spanish: The case for relexification. In A. Highfield & A. Valdman (Eds.), Historicity and variation in creole studies (pp. 52–78). Ann Arbor, MI: Karoma. Muysken, P. (2000). Bilingual speech: A typology of code-mixing. Cambridge: Cambridge University Press. Muysken, P. (2013). Language contact outcomes as the result of bilingual optimization strategies. Bilingualism: Language and Cognition, 16(4), 709–730.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

618

References

Myers-Scotton, C. (1993). Duelling languages: Grammatical structure in codeswitching. Oxford: Oxford University Press. Myers-Scotton, C. (2002). Contact linguistics: Bilingual encounters and grammatical outcomes. Oxford: Oxford University Press. Nakamura, J. (2015). Nonnative maternal input: Language use and errors in a Thai mother’s interactions in Japanese with her child. Japan Journal of Multilingualism and Multiculturalism, 21(1), 10–26. Nakamura, J. (2016). Hidden bilingualism: Ideological influences on the language practices of multilingual migrant mothers in Japan. International Multilingual Research Journal, 10(4), 308–323. Nakamura, J. & Quay, S. (2012). The impact of caregivers’ interrogative styles in English and Japanese on early bilingual development. International Journal of Bilingual Education and Bilingualism, 15(4), 417–434. Nance, C. (2015). “New” Scottish Gaelic speakers in Glasgow: A phonetic study of language revitalisation. Language in Society, 44(4), 553–579. Nance, C., McLeod, W., O’Rourke, B., & Dunmoreb S. (2016). Identity, accent aim, and motivation in second language users: New Scottish Gaelic speakers’ use of phonetic variation. Journal of Sociolinguistics, 20(2), 164–191. Napier, J. & Spencer, D. (2008). Guilty or not guilty? An investigation of deaf jurors’ access to court proceedings via Sign Language Interpreting. In D. Russell & S. Hale (Eds.), Interpreting in legal settings (pp. 72–122). Washington, DC: Gallaudet University Press. Nardy, A., Chevrot, J.-P., & Barbu, S. (2013). The acquisition of sociolinguistic variation: Looking back and thinking ahead. Linguistics, 51(2), 255–284. Nardy, A., Chevrot, J.-P., & Barbu, S. (2014). Sociolinguistic convergence and social interactions within a group of preschoolers: A longitudinal study. Language Variation and Change, 26(3), 273–301. National Academies of Sciences, Engineering, and Medicine [NASEM]. (2017). Promoting the educational success of children and youth learning English: Promising Futures. Washington, DC: National Academies Press. Nation, I. S. P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University Press. Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? Canadian Modern Language Review, 63(1), 59–82. National Center for Education Statistics. (2017). English Language Learners in public schools. https://nces.ed.gov/programs/coe/indicator_cgf.asp, last accessed January 18, 2018. Nelson, T. (1978). Detecting small amounts of information in memory: Savings for nonrecognized items. Journal of Experimental Psychology: Human Learning and Memory, 4(5), 453–468. Nely, N. T. R. & Ayonghe, S. L. (2015). Audiovisual translation as a tool for teaching English language to French-speaking students in Cameroon. Journal of African Studies and Development, 7(8), 200–206.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

619

Nespoulous, J.-L. & Virbel, J. (2007). From the study of language dysfunction and handicap to a better understanding of linguistic processing in normality. In M. J. Ball & J. S. Damico (Eds.), Clinical aphasiology. Future directions (pp. 107–124). Hove: Psychology Press. Netten, J. & Germain, C. (2004). Introduction: Intensive French. Canadian Modern Language Review, 60(3), 263–273. Ng, K. H. (2009). The common law in two voices: Language, law, and the postcolonial dilemma in Hong Kong. Stanford, CA: Stanford University Press. Nguyen, N. (2012). How English has displaced Russian and other foreign languages in Vietnam since “Doi Moi.” International Journal of Humanities and Social Science, 2(23), 259–266. Nicoladis, E. (2002). Some gestures develop in conjunction with spoken language development and others don’t: Evidence from bilingual preschoolers. Journal of Nonverbal Behavior, 26(4), 241–266. Nicoladis, E. & Gavrila, A. (2015). Cross-linguistic influence in Welsh– English bilingual children’s adjectival constructions. Journal of Child Language, 42(4), 903–916. Nicoladis, E., Mayberry, R. I., & Genesee, F. (1999). Gesture and early bilingual development. Developmental Psychology, 35(2), 514–526. Niedzielski, N. & Preston, D. (2000). Folk linguistics. Berlin: Mouton de Gruyter. Nikolov, M. (2000). The Critical Period Hypothesis reconsidered: Successful adult learners of Hungarian and English. IRAL: International Review of Applied Linguistics in Language Teaching, 38(2), 109–124. Nikolov, M. (2016). Trends, issues, and challenges in assessing young language learners. In M. Nikolov (Ed.) Assessing young learners of English: Global and local perspectives (pp. 1–18). Berlin: Springer. Nikolov, M. & Curtain, H. (2000). Introduction. In M. Nikolov & H. Curtain (Eds.), An early start: Young learners and modern languages in Europe and beyond (pp. 5–12). Strasbourg: Council of Europe. Norris, J. M. & Ortega, L. (2000). Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis. Language Learning, 50(3), 417–528. Nortier, J. & B. Svendsen (Eds.). (2015). Language, youth and identity in the 21st Century. Linguistic practices across urban spaces. Cambridge: Cambridge University Press. Norton, B. (2006). Not an afterthought: Authoring a text on adult ESOL. Linguistics and Education, 17(1), 91–96. Novoa, L., Fein, D., & Obler, L. K. (1988). Talent in foreign languages: A case study. In L. K. Obler & D. Fein (Eds.), The exceptional brain: Neuropsychology of talent and special abilities (pp. 294–302). New York: Guilford Press. Nycz, J. (2013). Changing words or changing rules? Second dialect acquisition and phonological representation. Journal of Pragmatics, 52, 49–62.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

620

References

Ochs, E. & Schieffelin, B. (2011). The theory of language socialization. In A. Duranti, E. Ochs, & B. Schieffelin (Eds.), The handbook of language socialization (pp. 1–21). Malden, MA: Blackwell. Odlin, T. (2014). Rediscovering prediction. In Z. Han & E. Tarone (Eds.), Interlanguage: Forty years later (pp. 27–45). Amsterdam: John Benjamins. Ohara, Y. (2001). Finding one’s voice in Japanese: A study of pitch levels of L2 users. In A. Pavlenko, A. Blackledge, I. Piller, & M. Teutsch-Dwyer (Eds.), Multilingualism, second language learning, and gender (pp. 231–254). Berlin: Mouton de Gruyter. Okita, T. (2002). Invisible work: Bilingualism, language choice and childrearing in intermarried families. Amsterdam: John Benjamins. Opitz, C. (2013). A dynamic perspective on late bilinguals’ linguistic development in an L2 environment. International Journal of Bilingualism, 17(6), 701–715. Organisation for Economic Co-operation and Development [OECD]. (2017). Education at a glance 2017. OECD indicators. Paris: OECD Publishing. www.oecd.org/education/education-at-a-glance-19991487.htm, last accessed January 10, 2018. Oriyama, K. (2016). Community of practice and family language policy: Maintaining heritage Japanese in Sydney – ten years later. International Multilingual Research Journal, 10(4), 289–307. Ortega, L. (2013). SLA for the 21st century: Disciplinary progress, transdisciplinary relevance, and the bi/multilingual turn. Language Learning, 63 (S1), 1–24. Ortega, L. (2014). Ways forward for a bi/multilingual turn in SLA. In S. May (Ed.), The multilingual turn: Implications for SLA, TESOL, and bilingual education (pp. 32–53). London: Routledge. Ortega, L. (2016). Multi-competence in second language acquisition: Inroads into the mainstream? In V. Cook & Li Wei (Eds.), The Cambridge handbook of linguistic multicompetence (pp. 50–76). Cambridge: Cambridge University Press. Ortega, L. (2018). SLA in uncertain times: Disciplinary constraints, transdisciplinary hopes. Working Papers in Educational Linguistics, 33(1), 1–30. https:// repository.upenn.edu/wpel/vol33/iss1/1, last accessed January 20, 2018. Ortega, L. & Iberri-Shea, G. (2005). Longitudinal research in SLA: Recent trends and future directions. Annual Review of Applied Linguistics, 25, 26–45. Orwell, G. (1949). Nineteen eighty-four. London: Secker and Warburg. Oshita, H. (2000). What is happened may not be what appears to be happening: A corpus study of “passive” unaccusatives in L2 English. Second Language Research, 16(4), 293–324. Otheguy, R., Garcı´a, O., & Reid, W. (2015). Clarifying translanguaging and deconstructing named languages: A perspective from linguistics. Applied Linguistics Review, 6(3), 281–307. Otsuji, E. & Pennycook, A. (2010). Metrolingualism: Fixity, fluidity and language in flux. International Journal of Multilingualism, 7(3), 240–254.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

621

Ozolins, U. (1993). The politics of language in Australia. Cambridge: Cambridge University Press. Paap, K. R. & Greenberg, Z. I. (2013). There is no coherent evidence for a bilingual advantage in executive processing. Cognitive Psychology, 66(2), 232–258. Paap, K. R., Johnson, H. A., & Sawi, O. (2014). Are bilingual advantages dependent upon specific tasks or specific bilingual experiences? Journal of Cognitive Psychology, 26(6), 615–639. Paap, K. R., Johnson, H. A., & Sawi, O. (2015). Bilingual advantages in executive functioning either do not exist or are restricted to very specific and undetermined circumstances. Cortex, 69, 265–278. Paap, K. R., Myuz, H. A., Anders, R. T., Bockelman, M. F., Mikulinsky, R., & Sawi, O. M. (2017). No compelling evidence for a bilingual advantage in switching or that frequent language switching reduces switch cost. Journal of Cognitive Psychology, 29(2), 89–112. Paap, K. R. & Sawi, O. (2014). Bilingual advantages in executive functioning: Problems in convergent validity, divergent validity, and the identification of the theoretical constructs. Frontiers in Psychology, 5, Article 962. Padilla, A. M., Fan, L., Xu, X., & Silva, D. (2013). A Mandarin/English two-way immersion program: Language proficiency and academic achievement. Foreign Language Annals, 46(4), 661–679. Page, B. R. & Putnam, M. (2015). Moribund Germanic languages in North America. Theoretical perspectives and empirical findings. Leiden: Brill. Pallier, C., Dehaene, S., Poline, J.-B., LeBihan, D., Argenti, A.-M., Dupoux, E., & Mehler, J. (2003). Brain imaging of language plasticity in adopted adults: Can a second language replace the first? Cerebral Cortex, 13(2), 155–161. Palmer, D. K. (2008). Building and destroying students’ “academic identities”: The power of discourse in a two-way immersion classroom. International Journal of Qualitative Studies in Education, 21(6), 647–667. Palviainen, A˚ . & Boyd, S. (2013). Unity in discourse, diversity in practice: The One Person One language policy in bilingual families. In M. Schwartz & A. Verschik (Eds.), Successful family language policy: Parents, children and educators in interaction (pp. 223–248). Berlin: Springer. Pandey, A. (2016). Monolingualism and linguistic exhibitionism in fiction. Basingstoke: Palgrave Macmillan. Panicacci, A. & Dewaele, J-M. (2018). Do interlocutors or conversation topics affect migrants’ sense of feeling different when switching languages? Journal of Multilingual & Multicultural Development, 39(3), 240–255. Paradis, J. (2010). The interface between bilingual development and specific language impairment. Applied Psycholinguistics, 31(2), 227–252.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

622

References

Paradis, J. (2011). Individual differences in child English second language acquisition: Comparing child-internal and child-external factors. Linguistic Approaches to Bilingualism, 1(3), 213–237. Paradis, J., Emmerzael, K., & Duncan, T. S. (2010). Assessment of English language learners: Using parent report on first language development. Journal of Communication Disorders, 43(6), 474–497. Paradis, J., Genesee, F., & Crago, M. (2011). Dual language development and disorders: A handbook on bilingualism and second language learning (2nd ed.). Baltimore Brookes. Paradis, J. & Navarro, S. (2003). Subject realization and crosslinguistic interference in the bilingual acquisition of Spanish and English: What is the role of the input? Journal of Child Language, 30(2), 371–393. Paradis, J., Nicoladis, E., Crago, M., & Genesee, F. (2011). Bilingual children’s acquisition of the past tense: A usage-based approach. Journal of Child Language, 38(3), 554–578. Paradis, M. (1987). The assessment of bilingual aphasia. Mahwah, NJ: Lawrence Erlbaum. Paradis, M. (1993). Linguistic, psycholinguistic, and neurolinguistic aspects of “interference” in bilingual speakers: The Activation Threshold Hypothesis. International Journal of Psycholinguistics, 9(2), 133–145. Paradis, M. (1998). Aphasia in bilinguals: How atypical is it? In P. Coppens, Y. Lebrun, & A. Basso (Eds.), Aphasia in atypical populations (pp. 35–66). London: Routledge. Paradis, M. (2001). Bilingual and polyglot aphasia. In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (2nd ed., pp. 69–91). Amsterdam: Elsevier. Paradis, M. (2007). L1 attrition features predicted by a neurolinguistic theory of bilingualism. In B. Ko¨pke, M. S. Schmid, M. Keijzer, & S. Dostert (Eds.), Language attrition: Theoretical perspectives (pp. 121– 133). Amsterdam: John Benjamins. Paradis, M. (2009). Declarative and procedural determinants of second languages. Amsterdam: John Benjamins. Paris, D. (2012). Culturally sustaining pedagogy: A needed change in stance, terminology, and practice. Educational Researcher, 41(3), 93–97. Park, D. C. & Reuter-Lorenz, P. (2009). The adaptive brain: Aging and neurocognitive scaffolding. Annual Review of Psychology, 60, 173–196. Park, J. S.-Y. (2009). The local construction of a global language: Ideologies of English in South Korea. Berlin: De Gruyter Mouton. Park, J. S.-Y. (2011). The promise of English: Linguistic capital and the neoliberal worker in the South Korean job market. International Journal of Bilingual Education and Bilingualism, 14(4), 443–455. Park, J. S.-Y. & Wee, L. (2012). Markets of English: Linguistic capital and language policy in a globalizing world. London: Routledge. ¯ ., Green, D. W., Grogan, A., Pliatsikas, C., Filippopolitis, K., Parker Jones, O Ali, N., Lee, H. L, Ramsden, S., Gazarian, K., Prejawa, S., Seghier, M. L.,

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

623

& Price, C. J. (2012). Where, when and why brain activation differs for bilinguals and monolinguals during picture naming and reading aloud. Cerebral Cortex, 22(4), 892–902. Patrinos, H. A. & Hurst M. E. (2007). Indigenous language skills and the labor market in a developing economy: Bolivia. In B. R. Chiswick & P. W. Miller (Eds.), The economics of language. International analyses (pp. 473–489). London: Routledge. Patterson, J. (2016). Child language disorders across languages, cultural context and syndromes. In J. Patterson & B. Rodriguez (Eds.), Multilingual perspectives on child language disorders (pp. 1–30). Bristol: Multilingual Matters. Pavlenko, A. (2005). Emotions and multilingualism. Cambridge: Cambridge University Press. Pavlenko, A. (2008). “I’m very not about the law part”: Nonnative speakers of English and the Miranda warnings. TESOL Quarterly, 42(1), 1–30. Pavlenko, A. (2018). Superdiversity and why it isn’t: Reflections on terminological innovation and academic branding. In S. Breidbach, L. Kuster, & B. Schmenk (Eds.), Sloganizations in language education discourse: Conceptual thinking in the age of academic marketization (pp. 142–168). Bristol: Multilingual Matters. Pearson, B. Z. (2010). We can no longer afford a monolingual norm. Applied Psycholinguistics, 31(2), 339–343. Pearson, B. Z., Ferna´ndez, S. C., Lewedeg, V., & Oller, D. K. (1997). The relation of input factors to lexical learning by bilingual infants. Applied Psycholinguistics, 18(1), 41–58. Pearson, B. Z., Ferna´ndez, S. C., & Oller, D. K. (1993). Lexical development in bilingual infants and toddlers: Comparison to monolingual norms. Language Learning, 43(1), 93–120. Peirce, B. N. (1995). Social identity, investment, and language learning. TESOL Quarterly, 29(1), 9–31. Pekarek Doehler, S. (2013). Social-interactional approaches to SLA: A state of the art and some future perspectives. Language, Interaction and Acquisition/Langage, Interaction et Acquisition, 4(2), 134–160. Pen˜a, E., Gutie´rrez-Clellen, V., Iglesias, A., Goldstein, B., & Bedore, L. (2014). Bilingual English–Spanish assessment. San Diego: AR Clinical Publications. Penfield, W. & Roberts, L. (1959). Speech and brain mechanisms. Princeton, NJ: Princeton University Press. Penn, C. (2012). Towards cultural aphasiology: Contextual models of service delivery in aphasia. In M. R. Gitterman, M. Goral, & L. K. Obler (Eds.), Aspects of multilingual aphasia (pp. 292–306). Bristol: Multilingual Matters. Perani, D., Paulesu, E., Galles, N. S., Dupoux, E., Dehaene, S., Bettinardi, V., Cappa, S. F., Fazio, F., & Mehler, J. (1998). The bilingual brain: Proficiency and age of acquisition of the second language. Brain, 121 (10), 1841–1852.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

624

References

Pe´rez Ba´ez, G. (2013). Family language policy, transnationalism, and the diaspora community of San Lucas Quiavinı´ of Oaxaca, Mexico. Language Policy, 12(1), 27–45. Pe´rez-Vidal, C. & Barquin, E. (2014). Comparing progress in academic writing after formal instruction and study abroad. In C. Pe´rez-Vidal (Ed.), Language acquisition in study abroad and formal instruction contexts (pp. 217–234). Amsterdam: John Benjamins. Peterson, J. M., Marinova-Todd, S. H., & Mirenda, P. (2012). Brief report: An exploratory study of lexical skills in bilingual children with autism spectrum disorder. Journal of Autism and Developmental Disorders, 42(7), 1499–1503. ´ Giollaga´in, C., & O ´ Curna´in, B. (2014). Inu´chadh ar an gCumas Pe´terva´ry, T., O Da´theangach: An sealbhu´ teanga i measc ghlu´n o´g na Gaeltachta/Analysis of bilingual competence: Language acquisition among young people in the Gaeltacht. Dublin: An Chomhairle um Oideachas Gaeltachta agus Gaelscolaı´ochta. Petitto, L. A., Katerelos, M., Levy, B. G., Gauna, K., Tetreault, K., & Ferraro, V. (2001). Bilingual signed and spoken language acquisition from birth: Implications for the mechanisms underlying early bilingual language acquisition. Journal of Child Language, 28(2), 453–496. Petrovic, J. E. (2005). The conservative restoration and neoliberal defenses of bilingual education. Language Policy, 4(4), 395–416. Pettitt, N. & Tarone, E. (2015). Following Roba: What happens when a low-educated multilingual adult learns to read. Writing Systems Research, 7(1), 20–38. Pfenninger, S. E. & Singleton, D. (2017). Beyond age effects in instructional L2 learning: Revisiting the age factor. Bristol: Multilingual Matters. Phillipson, R. (2017). Myths and realities of “global” English. Language Policy, 16(3), 313–331. Pickering, M. & S. Garrod, S. (2013). An integrated theory of language production and comprehension. Behavioral and Brain Sciences, 36 (4), 1–64. Piller, I. (2000). Language choice in bilingual, cross-cultural interpersonal communication. Linguistik Online, 5(1). https://bop.unibe.ch/linguistikonline/article/view/1015/1675, last accessed January 29, 2018. Piller, I. (2002). Bilingual couples talk: The discursive construction of hybridity. Amsterdam: John Benjamins. Piller, I. (2015). Monolingual ways of seeing multilingualism. Journal of Multilingual Discourses, 11(1), 25–33. Pitres, A. (1895). Etude sur l’aphasie chez les polyglottes. Revue de Me´decine, 15, 873–899. Pizer, G., Walters, K., & Meier, R. P. (2013). “We communicated that way for a reason”: Language practices and language ideologies among hearing adults whose parents are deaf. Journal of Deaf Studies and Deaf Education, 18(1), 75–92.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

625

Place, S. & Hoff, E. (2016). Effects and noneffects of input in bilingual environments on dual language skills in 2½-year-olds. Bilingualism: Language and Cognition, 19(5), 1023–1041. Polinsky, M. (2016). Bilingual children and adult heritage speakers: The range of comparison. International Journal of Bilingualism. doi: 10.1177/ 1367006916656048 Polinsky, M. & Kagan, O. (2007). Heritage languages: In the “wild” and in the classroom. Language and Linguistics Compass, 1(5), 368–395. Poplack, S. (1980). Sometimes I’ll start a sentence in Spanish Y TERMINO ˜ OL: Toward a typology of code-switching. Linguistics, 18(7–8), EN ESPAN 581–618. Poplack, S. & Tagliamonte, S. (2001). African American English in the diaspora. Malden, MA: Blackwell. Porte, G. (2003). English from a distance: Code-mixing and blending in the L1 output of long-term resident overseas EFL teachers. In V. Cook (Ed.), Effects of the second language on the first (pp. 103–119). Clevedon: Multilingual Matters. Portocarrero, J. S., Burright, R. G., & Donovick, P. J. (2007). Vocabulary and verbal fluency of bilingual and monolingual college students. Archives of Clinical Neuropsychology, 22(3), 415–422. Poulin-Dubois, D., Bialystok, E., Blaye, A., Polonia, A., & Yott, J. (2013). Lexical access and vocabulary development in very young bilinguals. International Journal of Bilingualism, 17(1), 57–70. Prat, C. S. (2011). The brain basis of individual differences in language comprehension abilities. Language and Linguistics Compass, 5(9), 635–649. Preece, S. (Ed.) (2016). The Routledge handbook of language and identity. London: Routledge. Prevoo, M. J. L., Malda, M., Mesman, J., & van IJzendoorn, M. H. (2016). Withinand cross-language relations between oral language proficiency and school outcomes in bilingual children with an immigrant background: A meta-analytic study. Review of Educational Research, 86(1), 237–276. Pring, L. (2007). Savant talent. Developmental Medicine & Child Neurology, 47(7), 500–503. Prinz, P. & Prinz, E. (1981). Acquisition of ASL and spoken English by a hearing child of a Deaf mother and a hearing father: Phase II, Early combinatorial patterns. Sign Language Studies, 30(1), 78–88. Prior, A. (2012). Too much of a good thing: Stronger bilingual inhibition leads to larger lag-2 task repetition costs. Cognition, 125(1), 1–12. Prior, A. & MacWhinney, B. (2010). A bilingual advantage in task switching. Bilingualism: Language and Cognition, 13(2), 253–262. Proctor, C. P., Daley, S., Louick, R., Leider, C. M., & Gardner, G. L. (2014). Motivation and engagement predict reading comprehension among native English-speaking and English-learning middle school students with disabilities in a remedial reading curriculum. Learning and Individual Differences, 36, 76–83.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

626

References

Proctor, L. M. (2014). English and globalization in India: The fractal nature of discourse. Journal of Linguistic Anthropology, 24(3), 294–314. Pufahl, I. & Rhodes, N. (2011). Foreign language instruction in U.S. schools: Results of a national survey of elementary and secondary schools. Foreign Language Annals, 44(2), 258–288. Pulinx, R., Van Avermaet, P., & Ag˘irdag˘, O. (2017). Silencing linguistic diversity: The extent, the determinants and consequences of the monolingual beliefs of Flemish teachers. International Journal of Bilingual Education and Bilingualism, 20(5), 542–556. Queen, R. (2012). Turkish German bilinguals and their intonation: Triangulating evidence about contact-induced language change. Language, 88(4), 791–816. Quinto-Pozos, D. (Ed.) (2007). Sign languages in contact. Washington, DC: Gallaudet University Press. Raikes, H., Alexander Pan, B., Luze, G., Tamis LeMonda, C. S., Brooks Gunn, J., Constantine, J., & Rodriguez, E. T. (2006). Mother–child bookreading in low income families: Correlates and outcomes during the first three years of life. Child Development, 77(4), 924–953. Ramos, S., Ferna´ndez Garcı´a, Y., Anto´n, E., Casaponsa, A., & Dun˜abeitia, J. A. (2017). Does learning a language in the elderly enhance switching ability? Journal of Neurolinguistics, 43(A), 39–48. Rampton, B. (2014). Crossing: Language and ethnicity among adolescents (2nd ed.). London: Routledge. Rampton, B. (2015). Contemporary urban vernaculars. In J. Nortier & B. Svendsen (Eds.), Language, youth and identity in the 21st century. Linguistic practices across urban spaces (pp. 24–44). Cambridge: Cambridge University Press. Ramscar, M., Hendrix, P., Shaoul, C., Milin, P., & Baayen, H. (2014). The myth of cognitive decline: Non-linear dynamics of lifelong learning. Topics in Cognitive Science, 6(1), 5–42. Ranta, L. & Meckelborg, A. (2013). How much exposure to English do international graduate students really get? Measuring language use in a naturalistic setting. Canadian Modern Language Review, 69(1), 1–33. Rastelli, S. (2014). Discontinuity in second language acquisition. The switch between statistical and grammatical learning. Bristol: Multilingual Matters. Rathmann, C., Mann, W., & Morgan, G. (2007). Narrative structure and narrative development in deaf children. Deafness & Education International, 9(4), 187–196. Ra¯tima, M. T. & Papesch, T. R. (2014). Te Rita Papesch: Case study of an exemplary learner of Ma¯ori as an additional language. International Journal of Bilingual Education and Bilingualism, 17(4), 379–393. Ratiu, I., Hout, M. C., Walenchok, S., Azuma, T., & Goldinger, S. D. (2017). Comparing visual search and eye movements in bilinguals and monolinguals. Attention, Perception, & Psychophysics, 79(6), 1695–1725.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

627

Rebenstorf, H. (2004). Political interest: Its meaning and general development. In H. Rebenstorf (Ed.), Democratic development? East German, Israeli and Palestinian adolescents (pp. 89–93). Berlin: Springer. Regan, V., Howard, M., & Leme´e, I. (2009). The acquisition of sociolinguistic competence in a study abroad context. Bristol: Multilingual Matters. Rehner, K. & Mougeon, R. (1999). Variation in the spoken French of immersion students: To “ne” or not to “ne,” that is the sociolinguistic question. Canadian Modern Language Review, 56(1), 124–154. Restrepo, A. & Gutie´rrez-Clellen, V. (2012). Grammatical impairments in Spanish–English bilingual children. In B. Goldstein (Ed.), Bilingual language development and disorders in Spanish–English speakers (pp. 213–232). Baltimore: Brookes Publishing. Restrepo, A., Morgan, G., & Thompson, M. (2013). The efficacy of a vocabulary intervention for dual-language learners with language impairment. Journal of Speech, Language, and Hearing Research, 56(2), 748–765. Ribot, T. (1881). Les maladies de la me´moire. Paris: Baille`re. Ribot, T. (1887). Diseases of memory: An essay in the positive psychology. New York: D. Appleton and Company. Ricento, T. (2005). Problems with the “language as resource” discourse in the promotion of heritage languages in the USA. Journal of Sociolinguistics, 9(3), 348–368. Ricento, T. (Ed.). (2000). Ideology, politics and language policies: Focus on English. Amsterdam: John Benjamins. Ricento, T., Peled, Y., & Ives, P. (Eds.). (2014). Language policy and political theory. [Thematic issue]. Language Policy, 13(4). Richland, J. B. (2008). Arguing with tradition: The language of law in Hopi Tribal court. Chicago: University of Chicago Press. Rickford, J. R. & King, S. (2016). Language and linguistics on trial: Hearing Rachel Jeantel (and other vernacular speakers) in the courtroom and beyond. Language, 92(4), 948–988. Ringbom, H. (2007). Cross-linguistic similarity in foreign language learning. Clevedon: Multilingual Matters. Rivera, K. M. & Huerta-Macı´as, A. (2008). Adult bilingualism and biliteracy in the United States: Theoretical perspectives. In K. M. Rivera & A. Huerta-Macı´as (Eds.), Adult biliteracy: Sociocultural and programmatic responses (pp. 3–28). Mahwah, NJ: Lawrence Erlbaum. Rivera Gaxiola, M., Silva Pereyra, J., & Kuhl, P. K. (2005). Brain potentials to native and non native speech contrasts in 7 and 11 month old American infants. Developmental Science, 8(2), 162–172. Rivers, D. J. & Ross, A. S. (2013). Idealized English teachers: The implicit influence of race in Japan. Journal of Language, Identity & Education, 12(5), 321–339. Roberts, C. (2010). Language socialization in the workplace. Annual Review of Applied Linguistics, 30, 211–227.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

628

References

Roberts, J. (1997). Acquisition of variable rules: A study of (-t, d) deletion in preschool children. Journal of Child Language, 24(2), 351–372. Roberts, P. M. (2008). Issues in assessment and treatment for bilingual and culturally diverse patients. In R. Chapey (Ed.), Language intervention strategies in aphasia and related neurogenic communication disorders (5th ed., pp. 245–275). Philadelphia: Wolters Kluwer/Lippincott Williams & Wilkins. Robinson, P. (2002). Learning conditions, aptitude complexes and SLA: A framework for research and pedagogy. In P. Robinson (Ed.), Individual differences and instructed language learning (pp. 113–133). Amsterdam: John Benjamins. Rock, F. (2017). Shifting ground: Exploring the backdrop to translating and interpreting. The Translator, 23(2), 217–236. Rogers, V. E., Meara, P., Aspinall, R., Fallon, L., Goss, T., Keey, E., & Thomas, R. (2016). Testing aptitude: Investigating Meara’s (2005) LLAMA tests. In S. A. Liszka, P. Leclercq, M. Tellier, & G. D. Ve´ronique (Eds.), EUROSLA Yearbook, 2016 (pp. 179–210). Amsterdam: John Benjamins. Rogers, V. E., Meara, P., Barnett-Legh, T., Curry, C., & Davie, E. (2017). Examining the LLAMA aptitude tests. Journal of the European Second Language Association, 1(1), 49–60. Romaine, S. (1984). The language of children and adolescents: The acquisition of communicative competence. Oxford: Basil Blackwell. Romaine, S. (1989). Bilingualism. Oxford: Blackwell. Romaine, S. (2013). The bilingual and multilingual community. Basil In T. K. Bhatia & W. C. Ritchie (Eds.), The handbook of bilingualism and multilingualism (2nd ed., pp. 445–465). Malden, MA: Wiley-Blackwell. Rosenhouse, J. (2013) Bilingualism/multilingualism in the Middle East and North Africa: A focus on cross-national and diglossic bilingualism/ multilingualism. In T. K. Bhatia & W. C. Ritchie (Eds.), The handbook of bilingualism and multilingualism (2nd ed., pp. 899–919). Malden, MA: Wiley-Blackwell. Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86(3), 638–641. Ross, M. D. (1996). Contact-induced change in Oceanic languages in NorthWest Melanesia. In A. Y. Aikhenvald & R. M. W. Dixon (Eds.), Areal diffusion and genetic inheritance (pp. 134–166). Oxford: Oxford University Press. Ross, S. (1998). Self-assessment in second language testing: A meta-analysis and analysis of experiential factors. Language Testing, 15(1), 1–20. Rosselli, M., Ardila, A., Araujo, K., Weekes, V. A., Caracciolo, V., Padilla, M., & Ostrosky-Solı´, F. (2000). Verbal fluency and repetition skills in healthy older Spanish–English bilinguals. Applied Neuropsychology, 7(1), 17–24. Rossi, E., Prystauka, Y., & Diaz, M. (2019). Investigating L1 attrition and language change: Neuroimaging perspectives. In M. S. Schmid &

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

629

B. Ko¨pke (Eds.), The Oxford handbook on language attrition. Oxford: Oxford University Press. Rothman, J. & Treffers-Daller, J. (2014). A prolegomenon to the construct of the native speaker: Heritage speaker bilinguals are natives too! Applied Linguistics, 35(1), 93–98. Rowe, M. (2012). A longitudinal investigation of the role of quantity and quality of child-directed speech in vocabulary development. Child Development, 83(5), 1762–1774. Roy, S. (2010). Not truly, not entirely . . . pas comme les francophones. Canadian Journal of Education, 33(3), 541–563. Royal College for Speech & Language Therapists (2006). Communicating quality 3: RCSLT’s guidance on best practice in service organisation and provision. London: Royal College of Speech & Language Therapists. Rubdy, R. (2001). Creative destruction: Singapore’s speak good English movement. World Englishes, 20(3), 341–355. Rubinstein, A. (2000). Economics and language. Cambridge: Cambridge University Press. Rubio-Ferna´ndez, P. (2017). Why are bilinguals better than monolinguals at false-belief tasks? Psychonomic Bulletin & Review, 24(3), 987–998. Ruiz, R. (1984). Orientations in language planning. NABE Journal, 8(2), 15–34. Ruiz, R. (2010). Reorienting language-as-resource. In. J. Petrovic (Ed.), International perspectives on bilingual education: Policy, practice, and controversy (pp. 155–172). Charlotte, NC: Information Age Publishing. Ruiz de Zarobe, Y. (2015). The effects of implementing CLIL in education. In M. Juan-Garau & J. Salazar-Noguera (Eds.), Content-based language learning in multilingual educational environments (pp. 51–68). Berlin: Springer. Ruiz-Felter, R., Cooperson, S. J., Bedore, L. M., & Pen˜a, E. D. (2016). Influence of current input-output and age of first exposure on phonological acquisition in early bilingual Spanish-English-speaking kindergarteners. International Journal of Language and Communication Disorders, 51(4), 368–383. Runnqvist, E., Strijkers, K., Alario, F.-X., & Costa, A. (2012). Cumulative semantic interference is blind to language: Implications for models of bilingual speech production. Journal of Memory and Language, 66(4), 850–869. Samimy, K. K. (2008). Achieving the advanced oral proficiency in Arabic: A case study. Foreign Language Annals, 41(3), 401–414. Sandwall, K. (2010). “I learn more at school”: A critical perspective on workplace-related second language learning in and out of school. TESOL Quarterly, 44(3), 542–574. Sanoudaki, E. & Thierry, G. (2015). Language non-selective syntactic activation in early bilinguals: The role of verbal fluency. International Journal of Bilingual Education and Bilingualism, 18(5), 548–560.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

630

References

Sanz, C. (2013). Multilingualism and metalinguistic awareness. In C. A. Chapelle (Ed.), The encyclopedia of applied linguistics. Blackwell Publishing Ltd. doi: 10.1002/9781405198431.wbeal0802 Schepens, J., van der Slik, F., & van Hout, R. (2016). L1 and L2 distance effects in learning L3 Dutch. Language Learning, 66(1), 224–256. Schiffrin, D. (1994). Approaches to discourse. Oxford: Blackwell. Schleef, E. (2017). Developmental sociolinguistics and the acquisition of T-glottalling by immigrant teenagers in London. In G. De Vogelaer & M. Katerbow (Eds.), Acquiring sociolinguistic variation (pp. 311–347). Amsterdam: John Benjamins. Schleppegrell, M. J. (2013). The role of metalanguage in supporting academic language development. Language Learning, 63(S1), 153–170. Schmeißer, A., Hager, M., Arnaus Gil, L., Jansen, V., Geveler, J., Eichler, N., Patuto, M., & Mu¨ller, N. (2016). Related but different: The two concepts of language dominance and language proficiency. In C. Silva-Corvala´n & J. Treffers-Daller (Eds.), Language dominance in bilinguals: Issues of operationalization and measurement (pp. 36–65). Cambridge: Cambridge University Press. Schmid, M. S. (2002). First language attrition, use, and maintenance: The case of German Jews in Anglophone countries. Amsterdam: John Benjamins. Schmid, M. S. (2007). The role of L1 use for L1 attrition. In B. Ko¨pke, M. S. Schmid, M. Keijzer & S. Dostert (Eds.), Language attrition: Theoretical perspectives (pp. 135–153). Amsterdam: John Benjamins. Schmid, M. S. (2011a). Contact x time: External factors and variability in L1 attrition. In M. S. Schmid & W. Lowie (Eds.), Modeling bilingualism: From structure to chaos (pp. 155–176). Amsterdam: John Benjamins. Schmid, M. S. (2011b). Language attrition. Cambridge: Cambridge University Press. Schmid, M. S. (2014). The debate on maturational constraints in bilingual development: A perspective from first language attrition. Language Acquisition, 21(4), 386–410. Schmid, M. S. & Keijzer, M. C. J. (2009). First language attrition and reversion among older migrants. International Journal of the Sociology of Language, 200, 83–101. Schmid, M. S. & Ko¨pke, B. (2017). The relevance of first language attrition and to theories of bilingual development. Linguistic Approaches to Bilingualism, 7(6), 637–667. Schmidt, L. B. (2009). The effect of dialect familiarity via a study abroad experience on L2 comprehension of Spanish. In J. Collentine, M. Garcia, B. Lafford, & F. Marin (Eds.), Selected proceedings of the 11th Hispanic Linguistics Symposium (pp. 143–154). Somerville, MA: Cascadilla Proceedings Project. Schmidt, R. W. (1983). Interaction, acculturation, and the acquisition of communicative competence. In N. Wolfson & E. Judd (Eds.), Sociolinguistics and language acquisition (pp. 137–174). Rowley, MA: Newbury House.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

631

Schmitt, E. (2010). When boundaries are crossed: Evaluating language attrition data from two perspectives. Bilingualism: Language and Cognition, 13(1), 63–72. Schmitt, N. (1998). Tracking the incremental acquisition of second language vocabulary: A longitudinal study. Language Learning, 48(2), 281–317. Schneider, V. I., Healey, A. F., & Bourne, L. E., Jr. (2002). What is learned under difficult conditions is hard to forget: Contextual interference effects in foreign vocabulary acquisition, retention, and transfer. Journal of Memory and Language, 46(2), 419–440. Schneiderman, E. I. & Desmarais, C. (1988). A neuropsychological substrate for talent in second-language acquisition. In L. K. Obler & D. Fein (Eds.), The exceptional brain: Neuropsychology of talent and special abilities (pp. 103– 126). New York: Guilford Press. Schrauf, R. W. (2009). English use among older bilingual immigrants in linguistically concentrated neighborhoods: Social proficiency and internal speech as intracultural variation. Journal of Cross-Cultural Gerontology, 24(2), 157–179. Schroeder, S. R. & Marian, V. (2014). Bilingual episodic memory: How speaking two languages affects remembering. In R. Heredia, & J. Altarriba (Eds.) Foundations of bilingual memory (pp. 111–132). Berlin: Springer. Schultz, R. (2006). Reevaluating communicative competence as a major goal in postsecondary language requirement courses. Modern Language Journal, 90(2), 252–255. Schumann, J. (1978). Research on the acculturation model for second language acquisition. Journal of Multilingual & Multicultural Development, 7(5), 379–392. Schu¨ppert, A. & Gooskens, C. (2012). The role of extra-linguistic factors for receptive bilingualism: Evidence from Danish and Swedish preschoolers. International Journal of Bilingualism, 16(3), 332–347. Schwab, J. F., Schuler, K. D., Stillman, C. M., Newport, E. L., Howard, J. H., & Howard, D. V. (2016). Aging and the statistical learning of grammatical form classes. Psychology and Aging, 31(5), 481–487. Schwartz, M. (2010). Family language policy: Core issues of an emerging field. Applied Linguistics Review, 1(1), 171–192. Segalowitz, N. & Freed, B. F. (2004). Context, contact, and cognition in oral fluency acquisition. Studies in Second Language Acquisition, 26(2), 173–199. Seliger, H. W. & Vago, R. M. (1991). The study of first language attrition: An overview. In H. W. Seliger & R. M. Vago (Eds.), First language attrition (pp. 3–15). Cambridge: Cambridge University Press. Selinker, L. (1972). Interlanguage. IRAL: International Review of Applied Linguistics in Language Teaching, 10(1–4), 209–232.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

632

References

Serrano, R., Llanes, A`., & Tragant, E. (2011). Analyzing the effect of context of second language learning: Domestic intensive and semi-intensive courses vs. study abroad in Europe. System, 39(2), 133–143. Serratrice, L. (2013). Cross-linguistic influence in bilingual development: Determinants and mechanisms. Linguistic Approaches to Bilingualism, 3(1), 3–25. Serratrice, L. & Herve´, C. (2015). Referential expressions in bilingual acquisition. In L. Serratrice & S. Allen (Eds.), The acquisition of reference (pp. 311–333). Amsterdam: John Benjamins. Shafto, M. A., James, L. E., Abrams, L., Tyler, L. K., & Cam-CAN (2017). Agerelated increases in verbal knowledge are not associated with word finding problems in the Cam-Can cohort: What you know won’t hurt you. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 72(1), 100–106. Shahidullah, S. & Hepper, P. G. (1994). Frequency discrimination by the foetus. Early Human Development, 36(1), 13–26. Sharwood Smith, M. (1983). On explaining language loss. In R. Felix & H. Wode (Eds.), Language development at the crossroads (pp. 49–59). Tu¨bingen: Gunther Narr Verlag. Sherman, T. & Homola´cˇ, J. (2017). “The older I got, it wasn’t a problem for me anymore”: Language brokering as a managed activity and a narrated experience among young Vietnamese immigrants in the Czech Republic. Multilingua, 36(1), 1–29. Shintani, N. & Ellis, R. (2010). The incidental acquisition of English plural – s by Japanese children in comprehension-based lessons: A processproduct study. Studies in Second Language Acquisition, 32(4), 607–637. Shook, A. & Marian, V. (2012). Bimodal bilinguals co-activate both languages during spoken comprehension. Cognition, 124(3), 314–324. Siegal, M. (1996). The role of learner subjectivity in second language sociolinguistic competency: Western women learning Japanese. Applied Linguistics, 17(3), 356–382. Siegel, J. (2010). Second dialect acquisition. Cambridge: Cambridge University Press. Siguan, M. (1980). Changement de langue dans le couple et dans la famille. In P. H. Nelde (Ed.), Sprachkontakt und Sprachkonflikt (pp. 283–285). Wiesbaden: Franz Steiner. Silva-Corvala´n, C. (1994). Language contact and change: Spanish in Los Angeles. Oxford: Clarendon Press. Silva-Corvala´n, C. & Treffers-Daller, J. (Eds.). (2016). Language dominance in bilinguals: Issues of operationalization and measurement. Cambridge: Cambridge University Press. Silve´n, M., Voeten, M., Kouvo, A., & Lunde´n, M. (2014). Speech perception and vocabulary growth: A longitudinal study of Finnish–Russian bilinguals and Finnish monolinguals from infancy to three years. International Journal of Behavioral Development, 38(4), 323–332.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

633

Silverman, M. & Salisbury, R. F. (1977). A house divided? Anthropological studies of factionalism. St. John’s, Newfoundland: Institute of Social and Economic Research, Memorial University of Newfoundland. Simon-Cereijido, G. & Gutie´rrez-Clellen, V. F. (2009). A cross-linguistic and bilingual evaluation of the interdependence between lexical and grammatical domains. Applied Psycholinguistics, 30(2), 315–337. Simons, G. F. & Lewis, M. P. (2013). The world’s languages in crisis: A 20year update. In E. Mihas, B. Perley, G. Rei-Doval, & K. Wheatley (Eds.), Responses to language endangerment: In honor of Mickey Noonan (pp. 3–19). Amsterdam: John Benjamins. Simpson, J. & Whiteside, A. (Eds.). (2015). Adult language education and migration: Challenging agendas in policy and practice. London: Routledge. Singh, L., Liederman, J., Mierzejewski, R., & Barnes, J. (2011). Rapid reacquisition of native phoneme contrasts after disuse: You do not always lose what you do not use. Developmental Science, 14(5), 949–959. Singleton, D. (1987). Mother and other tongue influence on learner French: A case study. Studies in Second Language Acquisition, 9(3), 327–345. Singleton, D. & Mun˜oz, C. (2011). Around and beyond the Critical Period Hypothesis. In E. Hinkel (Ed.), Handbook of research in second language teaching and learning (Vol. II, pp. 407–425). London: Routledge. Singleton, J. L., Morgan, D., DiGello, E., Wiles, J., & Rivers, R. (2004). Vocabulary use by low, moderate, and high ASL-proficient writers compared to hearing ESL and monolingual speakers. Journal of Deaf Studies and Deaf Education, 9(1), 86–103. Sinka, I. (2000). The search for cross-linguistic influences in the language of young Latvian–English bilinguals. In S. Do¨pke (Ed.), Cross-linguistic structures in simultaneous bilingualism (pp. 149–174). Amsterdam: John Benjamins. Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press. Skehan, P. (2016). Foreign language aptitude. In G. Granena, D. O. Jackson, & Y. Yilmaz (Eds.), Cognitive individual differences in L2 processing and acquisition (pp. 381–395). Amsterdam: John Benjamins. Skutnabb-Kangas, T. (1981). Bilingualism or not: The education of minorities. Clevedon: Multilingual Matters. Skutnabb-Kangas, T. (2000) Linguistic genocide in education – or worldwide diversity and human rights? Mahwah, NJ: Lawrence Erlbaum. Skutnabb-Kangas, T. & Phillipson, R. (Eds.). (2016). Language rights. London: Routledge. Slabakova, R. (2016). Second language acquisition. Oxford: Oxford University Press. Slavkov, N. (2017). Family language policy and school language choice: Pathways to bilingualism and multilingualism in a Canadian context. International Journal of Multilingualism, 14(4), 378–400.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

634

References

Smith, A. K., Andrews. J. F., Ausbrooks, M., Gentry, M. A., & Jacobowitz, E. L. (2013). A metalinguistic awareness test for ASL/English bilingual deaf children: The TASLA-R. Journal of Language Teaching and Research, 4(5), 885–899. Smith, J., Durham, M., & Richards, H. (2013). The social and linguistic in the acquisition of sociolinguistic norms: Caregivers, children, and variation. Linguistics, 51(2), 285–324. Smith, L. B. & Thelen, E. (2003). Development as a dynamic system. Trends in Cognitive Sciences, 7(8), 343–348. Smith, L. C. & Baker, W. (2011). L2 dialect acquisition of German vowels: The case of Northern German and Austrian dialects. Poznan Studies in Contemporary Linguistics, 47(10), 120–132. Smith, N., Tsimpli, I., Morgan, G., & Woll, B. (2011). The signs of a savant. Language against the odds. Cambridge: Cambridge University Press. Smithson, L., Paradis, J., & Nicoladis, E. (2014). Bilingualism and receptive vocabulary achievement: Could sociocultural context make a difference? Bilingualism: Language and Cognition, 17(4), 810–821. Snow, C. E. & Hoefnagel-Ho¨hle, M. (1978). The critical period for language acquisition: Evidence from second language learning. Child Development, 49(4), 1114–1128. Solan, L. M. & Tiersma, P. M. (2005). Speaking of crime: The language of criminal justice. Chicago: University of Chicago Press. Søndergaard, H.-P. (2017). Language learning in refugees: An account of a Swedish study. Tema BABYLONIA, 1, 15–18. http://babylonia.ch/ fileadmin/user_upload/documents/2017-1/Sondergaard_x_sito.pdf, last accessed January 30, 2018. Song, L., Tamis-LeMonda, C. S., Yoshikawa, H., Kahana-Kalman, R., & Wu, I. (2012). Language experiences and vocabulary development in Dominican and Mexican infants across the first 2 years. Developmental Psychology, 48(4), 1106. Spiegel, M. & Sunderland, H. (2006). Teaching basic literacy to ESOL learners. London: LLU+. Spivak, G. (1988). Can the subaltern speak? In C. Nelson & L. Grossberg (Eds.), Marxism and the interpretation of culture (pp. 66–111). Basingstoke: Macmillan Education. Spolsky, B. (2012). What is language policy? In B. Spolsky (Ed.), The Cambridge handbook of language policy (pp. 3–16). Cambridge: Cambridge University Press. Spotti, M. (2011). Modernist language ideologies, indexicalities and identities: Looking at the multilingual classroom through a post-Fishmanian lens. Applied Linguistics Review, 2(2), 29–50. Stam, G. A. (2010). Can an L2 speaker’s patterns of thinking for speaking change? In Z. Han & T. Cadierno (Eds.), Linguistic relativity in SLA: Thinking for speaking (pp. 59–83). Bristol: Multilingual Matters.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

635

Stavans, A. & Hoffmann, C. (2015). Multilingualism. Cambridge: Cambridge University Press. Steele, J. L., Slater, R. O., Zamarro, G., Miller, T., Li, J., Burkhauser, S., & Bacon, M. (2017). Effects of dual-language immersion programs on student achievement: Evidence from lottery data. American Educational Research Journal, 54(1S), 282S–306S. Stein, M., Federspiel, A., Koenig, T., Wirth, M., Strik, W., Wiest, R., Brandeis, D., & Dierks, T. (2012). Structural plasticity in the language system related to increased second language proficiency. Cortex, 48(4), 458–465. Steinberg, L., Bornstein, M. H., Vandell, D. L., & Rook, K. (2011). Life-span development: Infancy through adulthood. Boston: Wadsworth Cengage Learning. Stell, G. & Yakpo, K. (Eds.). (2015). Code-switching between structural and sociolinguistic perspectives. Berlin: De Gruyter. Sternberg, R. J. (2002). The theory of successful intelligence and its implications for language aptitude testing. In P. Robinson (Ed.), Individual differences and instructed language learning (pp. 13–43). Amsterdam: John Benjamins. Stevenson, P. (2015). The language question in contemporary Germany: The challenges of multilingualism. German Politics and Society, 33(1/2), 69–83. Street, J. & Da˛browska, E. (2010). More individual differences in language attainment: How much do adult native speakers of English know about passives and quantifiers? Lingua, 120(8), 2080–2094. Stro¨mmer, M. (2016). Affordances and constraints: Second language learning in cleaning work. Multilingua, 35(6), 697–721. Subtirelu, N. C. (2014). A language ideological perspective on willingness to communicate. System, 42(1), 120–132. Subtirelu, N. C. (2013). “English . . . it’s part of our blood”: Ideologies of language and nation in United States congressional discourse. Journal of Sociolinguistics, 17(1), 37–65. Sullivan, R. (2004). The challenges of interpreting multilingual, multijural legislation. Brooklyn Journal of International Law, 29, 986–1066. Sun, H., Steinkrauss, R., Tendeiro, J., & de Bot, K. (2016). Individual differences in very young children’s English acquisition in China: Internal and external factors. Bilingualism: Language and Cognition, 19(3), 550–566. Sundara, M., Polka, L., & Genesee, F. (2006). Language-experience facilitates discrimination of /d-/ in monolingual and bilingual acquisition of English. Cognition, 100(2), 369–388. Sundara, M., Polka, L., & Molnar, M. (2008). Development of coronal stop perception: Bilingual infants keep pace with their monolingual peers. Cognition, 108(1), 232–242.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

636

References

Sundara, M. & Scutellaro, A. (2011). Rhythmic distance between languages affects the development of speech perception in bilingual infants. Journal of Phonetics, 39(4), 505–513. Surrain, S. & Luk, G. (2017). Describing bilinguals: A systematic review of labels and descriptors used in the literature between 2005–2015. Bilingualism: Language and Cognition. https://doi.org/10.1017/ S1366728917000682. Swain, M. (1988). Manipulating and complementing content teaching to maximize second language learning. TESL Canada Journal, 6(1), 68–83. Swanwick, R., Hendar, O., Dammeyer, J., Kristoffersen, A.-E., Salter, J., & Simonsen, E. (2014). Shifting contexts and practices in sign bilingual education in Northern Europe: Implications for professional development and training. In M. Marschark, G. Tang, & H. Knoors (Eds.), Bilingualism and bilingual Deaf education (pp. 292–310). Oxford: Oxford University Press. Swanwick, R., Wright, S., & Salter, J. (2016). Investigating deaf children’s plural and diverse use of sign and spoken languages in a super diverse context. Applied Linguistics Review, 7(2), 117–147. Swigart, L. (2000). The limits of legitimacy: Language ideology and shift in contemporary Senegal. Journal of Linguistic Anthropology, 10(1), 90–130. Sylve´n, L. K. & Sundqvist, P. (Eds.). (2017). Computer-assisted language learning (CALL) in extracurricular/extramural contexts. [Special issue]. CALICO Journal, 34(1). Sylve´n, L. K. & Thompson, A. S. (2015). Language learning motivation and CLIL: Is there a connection? Journal of Immersion and Content-Based Language Education, 3(1), 28–50. Sze, F. & Tang, G. (2016). Metalinguistic awareness in the bimodal-bilingual acquisition of locative sentences in Chinese and Hong Kong Sign Language by deaf/hard of hearing children. Rivista di Psicolinguistica Applicata/Journal of Applied Psycholinguistics, 16(2), 101–132. Taguchi, N. (Ed.). (2014). English-medium education in the global society. [Special issue]. International Review of Applied Linguistics, 52(2). Tang, G. (2017). Sign bilingualism in Deaf education: From Deaf schools to regular school settings. In O. Garcı´a, A. M. Y. Lin, & S. May (Eds.), Bilingual and multilingual education (3rd ed., pp. 191–203). Berlin: Springer. Tang, G., Lam, S., & Yiu, C. K-M. (2014). Language development of Deaf children in a sign bilingual and co-enrollment environment. In M. Marschark, G. Tang, & H. Knoors (Eds.), Bilingualism and bilingual Deaf education (pp. 313–341). Oxford: Oxford University Press. Tang, G., Yiu, C. K.-M., & Lam, S. (2015). Awareness of sign language and visual communication system in a sign bilingual and co-enrollment setting. In H. Knoors & M. Marschark (Eds.), Teaching Deaf learners: Psychological and developmental perspectives (pp. 117–148). Oxford: Oxford University Press.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

637

Tare, M. & Gelman, S. A. (2010). Can you say it another way? Cognitive factors in bilingual children’s pragmatic language skills. Journal of Cognition and Development, 11(2), 137–158. Tarone, E. M., Bigelow, M., & Hansen, K. (2009). Literacy and second language oracy. Oxford: Oxford University Press. Tedick, D. & Wesely, P. (2015). A review of research on content-based foreign/second language education in US K-12 contexts. Language, Culture and Curriculum, 28(1), 25–40. ten Thije, J. D. & Zeevaert, L. (2007). Receptive multilingualism: Linguistic analyses, language policies and didactic concepts. Amsterdam: John Benjamins. Thomas, M. A. M. (2018). Research capacity and dissemination among academics in Tanzania: Examining knowledge production and the perceived binary of “local” and “international” journals. Compare: A Journal of Comparative and International Education, 48(2), 281–298. Thomason, S. (1997). Contact languages: A wider perspective. Amsterdam: John Benjamins. Thomason, S. & Kaufman, T. (1988). Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Thompson, A. S. (2013). The interface of language aptitude and multilingualism: Reconsidering the bilingual/multilingual dichotomy. Modern Language Journal, 97(3), 685–701. Thordardottir, E. (2010). Towards evidence-based practice in language intervention for bilingual children. Journal of Communication Disorders, 43, 523–537. Thordardottir, E. (2011). The relationship between bilingual exposure and vocabulary development. International Journal of Bilingualism, 15(4), 426–445. Thordardottir, E., Cloutier, G., Me´nard, S., Pelland-Blais, E., & Rvachew, S. (2015). Monolingual or bilingual intervention for primary language impairment? A randomized control trial. Journal of Speech, Language, and Hearing Research, 58(2), 287–300. Thordardottir, E., Rothenberg, A., Rivard, M.-E., & Naves, R. (2006). Bilingual assessment: Can overall proficiency be estimated from separate measurement of two languages? Journal of Multilingual Communication Disorders, 4(1), 1–21. Thornton, R. & Little, L. L. (2006). Language comprehension and production in normal aging. In J. E. Birren, K. W. Schaie, R. P. Abeles, M. Gatz, & T. A. Salthouse (Eds.) Handbook of the psychology of aging (pp. 261–287). Amsterdam: Elsevier. Todeva, E., & Cenoz, J. (Eds.) (2009). The multiple realities of multilingualism: Personal narratives and researchers’ perspectives. Berlin: Mouton de Gruyter. Torres Cacoullos, R. & Travis, C. E. (2015). Gauging convergence on the ground: Code-switching in the community: Introduction. International Journal of Bilingualism, 19(4), 365–386.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

638

References

Tragant, E., Mun˜oz, C., & Spada, N. (2016). Maximizing young learners’ input: An intervention program. Canadian Modern Language Review, 72 (2), 234–257. Tran, C. D., Arrendondo, M. M., & Yoshida, H. (2015). Differential effects of bilingualism and culture on early attention: A longitudinal study in the U.S., Argentina, and Vietnam. Frontiers in Psychology, 6, Article 795. Treffers-Daller, J. (2009). Code-switching and transfer: An exploration of similarities and differences. Cambridge: Cambridge University Press. Treffers-Daller, J. (2016). Language dominance: The construct, its measurement and operationalization. In C. Silva-Corvala´n & J. TreffersDaller (Eds.), Language dominance in bilingualism: Issues of measurement and operationalization (pp. 235–265). Cambridge: Cambridge University Press. ¨ zsoy, A. S., & van Hout, R. (2007). (In)complete acquisiTreffers-Daller, J., O tion of Turkish among Turkish–German bilinguals in Germany and Turkey: An analysis of complex embeddings in narratives. International Journal of Bilingual Education and Bilingualism, 10(3), 248–276. Treffert, D. A. (2009). The savant syndrome: An extraordinary condition. A synopsis: past, present, future. Philosophical Transactions of the Royal Society B, 364(1522), 1351–1357. Treffert, D. A. (2011). Daniel Tammet – Brainman: Numbers are my friends. Retrieved from: www.wisconsinmedicalsociety.org/professional/ savant-syndrome/profiles-and-videos/profiles/daniel-tammet-brain man/, last accessed January 30, 2018. Trenchs-Parera, M. & Juan-Garau, M. (2014). A longitudinal study of learners’ motivation and beliefs in at home and study abroad contexts. In C. Pe´rez-Vidal (Ed.), Language acquisition in study abroad and formal instruction contexts (pp. 111–136). Amsterdam: John Benjamins. Trinch, S. L. (2001). The advocate as gatekeeper: The limits of politeness in protective order interviews with Latina survivors of domestic abuse. Journal of Sociolinguistics, 5(4), 475–506. Tsao, F. M., Liu, H. M., & Kuhl, P. K. (2004). Speech perception in infancy predicts language development in the second year of life: A longitudinal study. Child Development, 75(4), 1067–1084. Tschirner, E. (2016). Listening and reading proficiency levels of college students. Foreign Language Annals, 49(2), 201–223. Tseng, V. & Fuligni, A. (2000). Parent-adolescent language use and relationships among immigrant families with East Asian, Filipino, and Latin American backgrounds. Journal of Marriage and the Family, 62(2), 465–476. Tsimpli, I. (2007). First language attrition from a minimalist perspective: Interface vulnerability and processing effects. In B. Ko¨pke, M. S. Schmid, M. Keijzer, & S. Dostert (Eds.), Language attrition. Theoretical perspectives (pp. 83–98). Amsterdam: John Benjamins.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

639

Tsimpli, I., Peristeri, E., & Andreou, M. (2016). Narrative production in monolingual and bilingual children with specific language impairment. Applied Psycholinguistics, 37(1), 195–216. Tuller, L. (2015). Clinical use of parental questionnaires in multilingual contexts. In S. Armon-Lotem, J. de Jong, & N. Meir (Eds.), Assessing multilingual children: Disentangling bilingualism from language impairment (pp. 301–330). Bristol: Multilingual Matters. Tupas, R. (2015). Inequalities of multilingualism: Challenges to mother tongue-based multilingual education. Language and Education, 29(2), 112–124. Ullman, M. T. (2005). A cognitive neuroscience perspective on second language acquisition: The declarative/procedural model. In C. Sanz (Ed.), Mind and context in adult second language acquisition: Methods, theory, and practice (pp. 141–178). Washington, DC: Georgetown University Press. United Nations Education, Scientific and Cultural Organization (UNESCO). (2003). Education in a multilingual world. Paris: United Nations Educational, Scientific and Cultural Organization. http://unesdoc. unesco.org/images/0012/001297/129728e.pdf, last accessed January 19, 2018. United Nations. (1948). Universal Declaration of Human Rights (UDHR). www.un .org/en/universal-declaration-human-rights/index.html, last accessed January 19, 2018. United Nations. (2016). Sustainable Development blog. www.un.org/sustainable development/blog/2016/01/244-million-international-migrants-livingabroad-worldwide-new-un-statistics-reveal, last accessed January 19, 2018. United Nations. (2017). International Migration Report 2017 (ST/ESA/SER.A/ 403). New York: United Nations. www.un.org/en/development/desa/ p o p u l a t i o n/ m i g r a t i o n / p u b l i c a t i o n s / m i g r a t i o n r e p o r t / d o c s / MigrationReport2017, last accessed January 19, 2018. Universal Declaration of Linguistic Rights Follow-up Committee. (1998). Universal declaration of linguistic rights. www.culturalrights.net/descar gas/drets_culturals389.pdf, last accessed January 19, 2018. Unsworth, S. (2013). Assessing the role of current and cumulative exposure in simultaneous bilingual acquisition: The case of Dutch gender. Bilingualism: Language and Cognition, 16(1), 86–110. Unsworth, S. (2014). Comparing the role of input in bilingual acquisition across domains. In T. Gru¨ter & J. Paradis (Eds.), Input and experience in bilingual development (pp. 181–202). Amsterdam: John Benjamins. Unsworth, S., Persson, L., Prins, T., & de Bot, K. (2015). An investigation of factors affecting early foreign language learning in the Netherlands. Applied Linguistics, 36(5), 527–548. Urry, J. (2007). Mobilities. Cambridge: Polity Press.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

640

References

Uylings, H. B. M. (2006). Development of the human cortex and the concept of “critical” or “sensitive” periods. Language Learning, 56(S1), 59–90. Vaillancourt, F. (1985). E´conomie et langue. Que´bec: Conseil de la langue franc¸aise. Valde´s, G. (Ed.). (2002). Expanding definitions of giftedness: The case of young interpreters from immigrant communities. Mahwah, NJ: Lawrence Erlbaum. Valian, V. (2015). Bilingualism and cognition: A focus on mechanisms. Bilingualism: Language and Cognition, 18(1), 47–50. Valls-Ferrer, M. & Mora, J. C. (2014). L2 fluency development in formal instruction and study abroad: The role of initial fluency level and language contact. In C. Pe´rez-Vidal (Ed.), Language acquisition in study abroad and formal instruction contexts (pp. 111–136). Amsterdam: John Benjamins. van Coetsem, F. (1988). Loan phonology and the two transfer types in language contact. Dordrecht: Foris. van den Bogaerde, B. (2000). Input and interaction in deaf families. Utrecht: LOT Publications. van Heuven, W. J. B., Dijkstra, T., & Grainger, J. (1998). Orthographic neighborhood effects in bilingual word recognition. Journal of Memory and Language, 39(3), 458–483. van Tuijl, C., Leseman, P., & Rispens, J. (2001). Efficacy of an intensive home-based educational intervention program for 4–6 year old ethnic minority children in the Netherlands. International Journal of Behavioral Development, 25(2), 148–159. Vanek, J., King, K., & Bigelow, M. (2018). Social presence and identity: Facebook in an English language classroom. Journal of Language, Identity & Education. doi:10.1080/15348458.2018.1442223 Vanek, N. & Hendriks, H. (2015). Convergence of temporal reference frames in sequential bilinguals: Event structuring unique to second language users. Bilingualism: Language and Cognition, 18(4), 753–768. Vanhove, J. (2013). The critical period hypothesis in second language acquisition: A statistical critique and a reanalysis. PLoS One, 8(7), e69172. Vanhove, J. & Berthele, R. (2015). The lifespan development of cognate guessing skills in an unknown related language. IRAL: International Review of Applied Linguistics in Language Teaching, 53(1), 1–38. VanPatten, B. (Ed.). (2004). Processing instruction: Theory, research, and commentary. Mahwah, NJ: Lawrence Erlbaum. Velliaris, D. M. & Willis, C. R. (2014). International family profiles and parental school choice in Tokyo. Journal of Research in International Education, 13(3), 235–247.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

641

Velupillai, V. (2015). Pidgins, creoles and mixed languages: An introduction. Amsterdam: John Benjamins. Ventureyra, V., Pallier, C., & Yoo, H.-J. (2004). The loss of first language phonetic perception in adopted Koreans. Journal of Neurolinguistics, 17(1), 79–91. Verhoeven, L. & van Leeuwe, J. (2012). The simple view of second language reading throughout the primary grades. Reading and Writing, 25(8), 1805–1818. Verreyt, N., Woumans, E., Vandelanotte, D., Szmalec, A., & Duyck, W. (2016). The influence of language switching experience on the bilingual executive control advantage. Bilingualism: Language and Cognition, 19(1), 181–190. Verspoor, M. (2017). Complex Dynamic Systems Theory and L2 pedagogy: Lessons to be learned. In L. Ortega & Z. Han (Eds.), Complexity theory and language development: In celebration of Diane Larsen-Freeman (pp. 143–162). Amsterdam: John Benjamins. Vertovec, S. (2006). The emergence of super-diversity in Britain. University of Oxford: Centre on Migration, Policy & Society. Vihman, M. M. (2014). Phonological development: The first two years (2nd ed.). Malden, MA: Wiley-Blackwell. Villareal, I. & Garcı´a Mayo, M. P. (2009). Tense and agreement morphology in the interlanguage of Basque Spanish bilinguals: CLIL versus nonCLIL. In Y. Ruiz de Zarobe & R. M. Jime´nez Catala´n (Eds.), Content and language integrated learning. Evidence from research in Europe (pp. 157–175). Bristol: Multilingual Matters. Vinkhuyzen, A. A. E., van der Sluis, S., Posthuma, D., & Boomsma, D. I. (2009). The heritability of aptitude and exceptional talent across different domains in adolescents and young adults. Behavioral Genetics, 39(4), 380–392. Vinogradov, P. (2009). Balancing top and bottom: Learner-generated texts for teaching phonics. In I. van de Craats & J. Kurvers (Eds.) Loweducated second language and literacy acquisition: Proceedings of the 4th symposium (pp. 133–144). Utrecht: LOT Publications. Voegelin, C. F. & Harris. Z. S. (1951). Methods for determining intelligibility among dialects of natural languages. Proceedings of the American Philosophical Society, 95(3), 322–329. Volterra, V. & Taeschner, T. (1978). The acquisition and development of language by bilingual children. Journal of Child Language, 5(2), 311–326. von Gru¨nigen, R., Kochenderfer-Ladd, B., Perren, B., & Alsaker, F. (2012). Links between local language competence and peer relations among Swiss and immigrant children: The mediating role of social behaviour. Journal of School Psychology, 50(2), 195–213. Wadensjo¨, C. (1998). Interpreting as interaction. London: Longman. Wagner, L., Clopper, C. G., & Pate, J. K. (2013). Children’s perception of dialect variation. Journal of Child Language, 41(5), 1062–1084.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

642

References

Walters, K. (1996). Gender, identity, and the political economy of language: Anglophone wives in Tunisia. Language in Society, 25(4), 515–555. Wang, X.-L. (2016). Maintaining three languages: The teenage years. Bristol: Multilingual Matters. Wartenburger, I., Heekeren, H. R., Abutalebi, J., Cappa, S. F., Villringer, A., & Perani, D. (2003). Early setting of grammatical processing in the bilingual brain. Neuron, 37(1), 159–170. Webb, S. (2015). Extensive viewing: Language learning through watching television. In D. Nunan and J. C. Richards (Eds.), Language learning beyond the classroom (pp. 159–168). London: Routledge. Webb, V. (2002). Language in South Africa. The role of language in national transformation, reconstruction and development. Amsterdam: John Benjamins. Weber, J.-J. & Horner, K. (2012). Introducing multilingualism: A social approach. London: Routledge. Wedin, A. (2010). A restricted curriculum for second language learners – A self-fulfilling teacher strategy? Language and Education, 24(3), 171–183. Wee, L. (2014). Linguistic chutzpah and the Speak Good Singlish movement. World Englishes, 33(1), 85–99. Weinreich, U. (1953). Languages in contact: Findings and problems. The Hague: Mouton. Weiss, C. H. (1983). Ideology, interests, and information. The basis of policy positions. In B. Callahan & B. Jennings (Eds.), Ethics, the social sciences and policy analysis (pp. 213–246). New York: Plenum Press. Weissberger, G. H., Wierenga, C. E., Bondi, M. W., & Gollan, T. H. (2012). Partially overlapping mechanisms of language and task control in young and older bilinguals. Psychology and Aging, 27(4), 959–974. Wen, Z. E. (2016). Working memory and second language learning: Towards an integrated approach. Bristol: Multilingual Matters. Werker, J. F. & Lalonde, C. E. (1988). Cross-language speech perception: Initial capabilities and developmental change. Developmental Psychology, 24(5), 672. Wesche, M. (1993). French immersion graduates at university and beyond: What difference has it made? In J. Alatis (Ed.), Georgetown University Round Table on Languages and Linguistics 1992: Language, communication and social meaning (pp. 208–240). Washington, DC: Georgetown University Press. Westby, C. (1990). Ethnographic interviewing: Ask the right questions to the right people in the right ways. Journal of Childhood Communication Disorders, 13(1), 101–111. Whitford, V. & Titone, D. (2016). Eye movements and the perceptual span during first- and second-language sentence reading in bilingual older adults. Psychology and Aging, 31(1), 58–70. Whiting, E., Chenery, H. J., & Copland, D. A. (2011). Effect of aging on learning new names and descriptions for objects. Aging, Neuropsychology, and Cognition, 18(5), 594–619.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

References

643

Wiig, E., Semel, E., & Secord, W. (2006). Clinical evaluation of language fundamentals – Fourth edition, Spanish. San Antonio, TX: Pearson. Wiig, E., Semel, E., & Secord, W. (2009). Clinical evaluation of language fundamentals preschool-2 Spanish. San Antonio, TX: Pearson. Wildavsky, A. (1992). Finding universalistic solutions to particularistic problems: Bilingualism resolved through a second language requirement for elementary public schools. Journal of Policy Analysis and Management, 11(2), 310–314. Winford D. (2003). An introduction to contact linguistics. Oxford: Blackwell. Winke, P. (2013). An investigation into second language aptitude for advanced Chinese language learning. Modern Language Journal, 97(1), 109–130. Winsler, A., Burchinal, M. R., Tien, H. C., Peisner-Feinberg, E., Espinosa, L., Castro, D. C., & De Feyter, J. (2014). Early development among dual language learners: The roles of language use at home, maternal immigration, country of origin, and socio-demographic variables. Early Childhood Research Quarterly, 29(4), 750–764. Winsler, A., Kim, Y., & Richard, E. (2014). Socio-emotional skills, behavior problems, and Spanish competence predict the acquisition of English among English language learners in poverty. Developmental Psychology, 50(9), 2242–2254. Woolard, K. A. (1998). Introduction: Language ideology as a field of inquiry. In B. B. Schieffelin, K. A. Woolard, & P. V. Kroskrity (Eds.) Language ideologies: Practice and theory (pp. 3–47). Oxford: Oxford University Press. Woolard, K. A. & Schieffelin, B. B. (1994). Language ideology. Annual Review of Anthropology, 23(1), 55–82. Woumans, E., Surmont, J., Struys, E. & Duyck, W. (2016). The longitudinal effect of bilingual immersion schooling on cognitive control and intelligence. Language Learning, 66(52), 76–91. Wu, S. L. & Ortega, L. (2013). Measuring global oral proficiency in SLA research: A new elicited imitation test of L2 Chinese. Foreign Language Annals, 46(4), 680–704. Wu, Y. J. & Thierry, G. (2010). Chinese–English bilinguals reading English hear Chinese. Journal of Neuroscience, 30(22), 7646–7651. Wucherer, B. V. & Reiterer, S. M. (2018). Language is a girlie thing, isn’t it? A psycholinguistic exploration of the L2 gender gap. International Journal of Bilingual Education and Bilingualism, 21(1), 118–134. Wyman, L. T. (2013). Indigenous youth migration and language contact. International Multilingual Research Journal, 7(1), 66–82. Yag˘mur, K. (1997). First language attrition among Turkish speakers in Sydney. Tilburg: Tilburg University Press. Yamamoto, M. (1995) Bilingualism in international families. Journal of Multilingual & Multicultural Development, 16(1&2), 63–85.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

644

References

Yao, Y. & Chang, C. (2016). On the cognitive basis of contact-induced sound change: Vowel merger reversal in Shanghainese. Language, 92(2), 433–467. Yashima, T. (2009). International posture and the ideal L2 self in the Japanese EFL context. In Z. Do¨rnyei & E. Ushioda (Eds.), Motivation, language identity and the L2 Self (pp. 144–163). Bristol: Multilingual Matters. Yates, L. (2011). Interaction, language learning and social inclusion in early settlement. International Journal of Bilingual Education and Bilingualism, 14(4), 457–471. Zanini, S., Angeli, V., & Tavano, A. (2011). Primary progressive aphasia in a bilingual speaker: A single case study. Clinical Linguistics & Phonetics, 25(6–7), 553–554. Zell, E. & Krizan, Z. (2014). Do people have insight into their abilities? A metasynthesis. Perspectives on Psychological Science, 9(2), 111–125. Zentella, A. C. (2014). TWB (Talking while Bilingual): Linguistic profiling of Latina/os, and other linguistic torquemadas. Latino Studies, 12(4), 620–635. Zentz, L. (2014). “Love” the local, “use” the national, “study” the foreign: Shifting Javanese language ecologies in (post-) modernity, postcoloniality, and globalization. Journal of Linguistic Anthropology, 24(3), 339–359. Zhao, X. & Li, P. (2010). Bilingual lexical interactions in an unsupervised neural network model. International Journal of Bilingual Education and Bilingualism, 13(5), 505–524. Zhou, W. (2015). Assessing birth language memory in young adoptees. Nijmegen: Max Planck Institute for Psycholinguistics. Zimmerman, I., Steiner, V., & Pond, E (2012). Preschool Language Scales-5 Spanish. San Antonio, TX: Pearson.

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:05:14, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.029

Language Index

AAVE. See African American Vernacular English (AAVE) Aboriginal languages, 530–531 African American English, 122, 126–127, 513 African American Vernacular English (AAVE) bidialectalism and, 516–517 interpreters, 138–139 language contact and, 532 Afrikaans cognates in, 388 functional linguistic relativity and, 146 Algonquian languages, 529 Alsatian, 511 American English, 518 American Indian languages, 548 American Russian, 534 American Sign Language (ASL), 83, 488–489, 491–492, 493, 498–499 American Swedish, 537 Amerindian languages, 529 Andean Spanish, 532–533 Antwerp dialect, 336 Appalachian English, 126–127 Arabic generally, 3 adult bilingualism and, 78, 88–89 bidialectalism and, 513–514 bilingual advantage hypothesis and, 462–463 convergent choice principle and, 330 Egyptian Arabic, 78 language analysis for determination of origin (LADO) and, 149–150 language contact and, 529–530 “language economics” and, 187 Modern Standard Arabic, 547 Moroccan Arabic, 71–72, 321–322, 547 multilingualism and, 547 Palestinian Arabic, 78 vernacular Tunisian Arabic, 68 young “global hybrids” and, 68, 71–72 Argentinian Spanish, 532 Australian Aboriginal English, 138 Austrian Sign Language, 124 Austronesian languages, 67, 534–535

Basque adult bilingualism and, 86 bilingual advantage hypothesis and, 453–454 developmental bilingual education and, 397–398 educational programs supporting, 49 education and, 398 familiarity and, 467 immersion programs and, 44, 214, 215 multilingualism and, 557 speech discrimination in, 17–18 Berber generally, 3 multilingualism and, 546 Bernese, 64–65 Black English. See African American Vernacular English (AAVE) Brazilian Portuguese, 498–499 Brazilian Sign Language (LIBRAS), 498–499 Breton immersion programs and, 215 young “global hybrids” and, 66–67 British English, 302 Bulgarian, 353 Callahuaya, 529–530 Cantonese functional linguistic relativity and, 146–147 input and, 206 language attrition and, 275 sign language and, 489–490, 496–498, 500, 502–504 Caribbean Creole, 330 Castilian Spanish, 264 Catalan adult migrants and, 264 bilingual advantage hypothesis and, 451 bilingual language control and, 437, 446, 447, 448–449 cognates in, 388 immersion programs and, 44, 214–215 lexical learning in early childhood and, 29 speech discrimination in, 17–18 Cebuano, 138

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:09:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.030

646

Language Index Cherokee immersion programs and, 215 language policy and, 164 Chinese. See also Cantonese; Mandarin; Putonghua acculturation and, 42 adult bilingualism and, 91 Australia Asia Project and, 155–156 bilingual language control and, 443–444 clinical linguistics and, 373 foreign language learning and, 243, 245 functional linguistic relativity and, 146–147 gifted bilinguals and, 322 immersion programs and, 215–216 language attrition and, 275 language contact and, 527, 534 manually coded, 503–506 measurement of bilingual abilities and, 303–304 migration and, 256 multilingualism and, 546, 553 older bilinguals and, 102, 111–112 signed, 486 sign language and, 486, 504–506 Coptic, 546 Corsican ideologies of bilingualism and, 128 immersion programs and, 215 Cree, 529 Czech bilingual language control and, 469–470 language brokering and, 57 Danish multilingualism and, 548 polylanguaging and, 130 Dutch academic achievement gaps and, 53–54 adolescent bilingualism and, 36–37 adult bilingualism and, 78–79, 92–93 bilingual advantage hypothesis and, 454–455 bilingual language control and, 437 CLIL and, 221–222 cognates in, 388 convergent choice principle and, 330–331 crosslinguistic influence (CLI) and, 418 exposure and, 199, 200–203, 204 first language attrition and, 358–359, 361 gifted bilinguals and, 311, 321–322 language attrition and, 106, 107–108, 274–275, 279 language choice in bilingual interaction and, 327, 331–333, 335, 340 language contact and, 527, 533, 534, 536–537, 539–541 lexical learning in early childhood and, 22–23 measurement of bilingual abilities and, 291, 293–294, 296–297, 301 multilingualism and, 551 older bilinguals and, 102, 105, 106 savings paradigm and, 281 sign language and, 494 Ecuadorean Quechua, 529 Efik, 189 Egyptian Arabic, 78

English generally, 3, 6 academic achievement gaps and, 52–53 adolescent bilingualism and, 36–37 adult bilingualism and, 88, 89, 92–93 adult migrants and, 260–261 advantages of choosing, 45–46 African American English, 122, 126–127, 513 age of acquisition and, 468 age-related language changes and, 103 American English, 518 Appalachian English, 126–127 Australian Aboriginal English, 138 bidialectalism and, 517–518 bilingual advantage hypothesis and, 451, 462–463 bilingual language control and, 437, 439–441, 443–445, 447, 448–449 British English, 302 CLIL and, 216, 221–222, 226–227 clinical linguistics and, 371, 373, 377–379, 385–386, 387–388 code-switching and, 109–111, 209–210 cognates in, 388 comprehension based instruction (CBI) and, 241–242 convergent choice principle and, 330–331 crosslinguistic influence (CLI) and, 415–417 dual language programs and, 46–48 early childhood bilingualism and, 35 education and, 397, 398 exposure and, 197–198, 199–200, 202, 203, 204–205 familiarity and, 467, 473–475 first language attrition and, 353, 354, 358–359, 361–363 foreign language learning and, 243–246 functional linguistic relativity and, 145–147 gesture in, 29–31 Global English, 131–133 ideologies of bilingualism and, 128 immersion programs and, 44, 215–216, 217, 219–221, 225–226 input and, 205–206, 207 instructional time and, 239–240 interpreters and, 139–140, 142 language attrition and, 106–107, 108, 270–272 language choice in bilingual interaction and, 334, 335, 337–338, 340, 345–346 language contact and, 525–526, 530–531, 533, 534, 536–539 “language economics” and, 178–179, 180–181, 183–184, 187 language policy and, 168, 169 legal profession and, 137, 147–148 lexical learning in early childhood and, 23–25, 27 as Lingua Franca, 65, 132, 137, 176–177, 197–198, 545 linguistic relativity and, 457–459 measurement of bilingual abilities and, 296–297, 299–301, 303–304 memory retrieval and, 460 migration and, 255, 256 multilingualism and, 545, 546, 551, 553, 555, 556–557 normative monolingualism and, 123 Northern American English, 517–518

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:09:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.030

Language Index

older bilinguals and, 102, 105, 106–107, 113 savings paradigm and, 281–282 second language acquisition (SLA) and, 408, 409, 413–415, 419–420, 425–426, 429–431 Signed Exact English, 486 sign language and, 488, 491–492, 493, 498–499, 503 speech discrimination in, 17 vocabulary, relation to grammar, 32–34 Esperanto, 316, 319 Finnish lexical learning in early childhood and, 25 multilingualism and, 547 young “global hybrids” and, 63–64, 71 Finnish Sign Language (FinSL), 210 Fongbe, 530 French generally, 3, 319 academic achievement gaps and, 54 adult migrants and, 264 age of acquisition and, 468 age-related language changes and, 103 bidialectalism and, 513–514, 518, 519–520 bilingual advantage hypothesis and, 454–455 clinical linguistics and, 373, 377–378, 386 code-switching and, 210 cognates in, 388 comprehension based instruction (CBI) and, 241–242 convergent choice principle and, 331 crosslinguistic influence (CLI) and, 415–416, 417 early childhood bilingualism and, 35 education and, 398 exposure and, 199, 200, 202–203, 204 familiarity and, 467 foreign language learning and, 243 functional linguistic relativity and, 145–146 gesture in, 30–31 ideologies of bilingualism and, 128–129 immersion programs and, 44, 214, 215, 217, 218, 220–221, 224–226, 230 instructional time and, 239–240 language attrition and, 108, 273–274 language choice in bilingual interaction and, 331–333, 345–346 language contact and, 529, 530, 531 “language economics” and, 178–179, 180–181, 183–184, 187 lexical learning in early childhood and, 22–24 measurement of bilingual abilities and, 301, 302 migration and, 257 Moroccan French, 3 multilingualism and, 174–175, 546, 547, 551, 553, 557 older bilinguals and, 102, 105, 112–113 plurilingualism and, 174–175 savings paradigm and, 281–282 second language acquisition (SLA) and, 408, 413–415, 419–420 sign language and, 493–494 young “global hybrids” and, 66–67 Frisian adult bilingualism and, 86 exposure and, 199 Friulian, 470

647

Gaelic adult bilingualism and, 79, 87 educational programs supporting, 49 immersion programs, 44 language choice in bilingual interaction and, 334, 337 German generally, 319 academic achievement gaps and, 54 adult bilingualism and, 80–81, 87–88 bidialectalism and, 518, 520 bilingual language control and, 469–470 crosslinguistic influence (CLI) and, 416, 417 education and, 398 exposure and, 202 familiarity and, 467 first language attrition and, 358, 362–363 functional linguistic relativity and, 146 Heritage German, 536 immersion programs and, 215 language attrition and, 107–108, 279, 283–284 language choice in bilingual interaction and, 332, 340, 341, 345–346 language contact and, 534 “language economics” and, 183–184 legitimacy and, 38 linguistic relativity and, 458–459 measurement of bilingual abilities and, 296–297, 302 multilingualism and, 557 savings paradigm and, 281 second language acquisition (SLA) and, 426 Standard German, 69, 520 Swiss German, 551 Volga German, 535–536 young “global hybrids” and, 66–67, 69 Germanic heritage languages, 535, 541 Global English, 131–133 Greek, 371 Gurindji, 530–531 Haitian Creole, 530 Hawaiian, 215 Hazargi Dari, 149 Hebrew age-related language changes and, 104 bilingual advantage hypothesis and, 451 clinical linguistics and, 371 code-switching and, 209 exposure and, 197–198, 202 language attrition and, 106–107 multilingualism and, 556 older bilinguals and, 102, 106–107 second language acquisition (SLA) and, 408, 413–415 Heritage German, 536 Heritage Swedish, 537 Hindi Australia Asia Project and, 155–156 multilingualism and, 548–549 savings paradigm and, 282 Hindustani, 548–549 Hiri Motu, 529–530 Hmong, 33–34 Hong Kong Sign Language (HKSL), 489–490, 496–498, 500, 502–506 Hopi, 147

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:09:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.030

648

Language Index Hungarian adult bilingualism and, 88 critical period hypothesis (CPH) and, 310 young “global hybrids” and, 66–67 Icelandic, 319 Ilocano, 138 Indo-Iranian languages, 533–534 Indonesian Australia Asia Project and, 155–156 memory retrieval and, 460 Inuktitut exposure and, 202 functional linguistic relativity and, 146 Irish immersion programs and, 214–215, 225–226 language policy and, 168 Italian age of acquisition and, 468 bilingual language control and, 470 familiarity and, 467 first language attrition and, 363 immersion programs and, 215 multilingualism and, 556–557 sign language and, 488, 494–496, 500 Italian Sign Language (LIS), 488, 494–496, 500 Jaffna Tamil, 127 Jamaican Patois, 138 Japanese generally, 3, 6–7 Australia Asia Project and, 155–156 bidialectalism and, 518 code-switching and, 209 crosslinguistic influence (CLI) and, 416–417 exposure and, 202, 204–205 immersion programs and, 215 input and, 207 language choice in bilingual interaction and, 333–334 linguistic relativity and, 457–458 savings paradigm and, 281 second language acquisition (SLA) and, 421 Study Abroad (SA) and, 224 young “global hybrids” and, 67–68, 72 Kashmiri age-related language changes and, 104 older bilinguals and, 102 Kazakh, 398 Kinyankole, 149 Kinyarwanda, 149 Korean adolescent bilingualism and, 36–37 Australia Asia Project and, 155–156 crosslinguistic influence (CLI) and, 416–417 interpreters and, 139–140 language attrition and, 270–271, 273–275 savings paradigm and, 281 Kriol, 530–531 Kurdish language choice in bilingual interaction and, 344 multilingualism and, 546 Langue des Signes Que´be´coise (LSQ), 210, 493–494 Latin, 112, 548

Latvian, 38 Light Warlpiri, 530–531 Lingua dei Segni Italiana (LIS), 488, 494–496, 500 Lı´ngua de Sinais Brasiliera (LIBRAS), 498–499 Lithuanian, 319 LSQ (Langue des Signes Que´be´coise), 210, 493–494 Luganda, 63–64 Luxembourgish, 398 Ma’a, 529–530 Malay, 218–219 Mandarin. See also Chinese; Putonghua advantages of choosing, 45–46 education and, 397 immersion programs and, 44, 219 language attrition and, 275 language contact and, 539–541 measurement of bilingual abilities and, 304 second language acquisition (SLA) and, 421 sign language and, 503 social media and, 55–56 Ma¨nti, 319 Manually coded Chinese, 503–506 Ma¯ori adult bilingualism and, 86–87 developmental bilingual education and, 397–398 educational programs supporting, 49 immersion programs and, 44, 215 language policy and, 168 Mbugu, 529–530 Media Lingua, 529–530, 531 Medniy Aleut, 529–530 Michif, 529–531 Mina, 284 Mixtec, 138 Mock Spanish, 131 Modern Standard Arabic, 547 Moroccan Arabic, 71–72, 321–322, 547 Moroccan French, 3 Nahuatl, 527–528 Native American languages, 49 Navajo, 44 Ndebele, 545 Ndyuka-Trio, 529–530 Nederlandse Gebarentaal (NGT), 210, 494 New Mexican Spanish, 535 New Tiwi, 530–531 Nguni group, 545 Northern American English, 517–518 Norwegian language contact and, 524–525 multilingualism and, 548 Occitan, 215 Ojibwe adult bilingualism and, 87 immersion programs and, 215 revitalization of, 50 Palestinian Arabic, 78 Panjabi, 330 Papuan, 67 Pennsylvania Dutch, 536, 542 Persian, 344 Pidgin Delaware, 529–530

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:09:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.030

Language Index

Portuguese generally, 1 cognates in, 388 education and, 398 “language economics” and, 187 workplace, bilingualism in, 62–63 young “global hybrids” and, 62–63 Putonghua, 189, 503. See also Chinese; Mandarin Quechua, 532–533 Quichua, 124, 129, 529 Romanian cognates in, 388 first language attrition and, 360 Romansch, 557 Runyankole, 149 Russian code-switching and, 209 education and, 398 exposure and, 197–198, 202 immersion programs and, 215 language choice in bilingual interaction and, 341 language contact and, 534 lexical learning in early childhood and, 25 multilingualism and, 546, 556 police interrogations and, 144 second language acquisition (SLA) and, 408, 413–415, 419 Sámi, 547 Sango, 529–530 Sanskrit, 548–549 Scottish Gaelic. See also Gaelic adult bilingualism and, 87 language contact and, 537 Sepedi, 545 Serbian, 54 Sesotho, 545 Setswana, 545 Signed Chinese, 486 Signed Exact English, 486 Sign Language of the Netherlands (NGT), 210, 494 Singlish, 126 Somali, 43 Sotho group, 545 Spanglish, 131 Spanish generally, 6, 319 academic achievement gaps and, 52 adult bilingualism and, 79, 91 age-related language changes and, 104 Andean Spanish, 532–533 Argentinian Spanish, 532 bidialectalism and, 518 bilingual advantage hypothesis and, 451, 453–454 bilingual language control and, 437, 439–441, 443, 445, 446, 447, 448–449 Castilian Spanish, 264 clinical linguistics and, 371, 377–379, 385–386, 387–388 code-switching and, 109–110 cognates in, 388 comprehension based instruction (CBI) and, 242

649

crosslinguistic influence (CLI) and, 415–417 dual language programs and, 47–48 early childhood bilingualism and, 35 education and, 398 exposure and, 198–200, 203, 204 familiarity and, 467, 473–475 first language attrition and, 354, 361–362 gesture in, 30 ideologies of bilingualism and, 126, 128 immersion programs and, 44, 215, 219–220, 225–226 input and, 205 instructional time and, 240 interpreters and, 138, 139–140, 142 language attrition and, 107, 271–272 language brokering and, 57 language choice in bilingual interaction and, 340, 345 language contact and, 525–526, 527–528, 529, 531, 532–533, 534, 537–539 “language economics” and, 181, 187 language policy and, 169 legal profession and, 137, 147–148 lexical learning in early childhood and, 24–25, 27, 29 measurement of bilingual abilities and, 299–300, 302 Mock Spanish, 131 multilingualism and, 556–557 New Mexican Spanish, 535 normative monolingualism and, 124 older bilinguals and, 102, 107, 112–113 police interrogations and, 143 speech discrimination in, 17–18 vocabulary, relation to grammar, 32–33 Standard German, 69, 520 Swahili, 526, 529–530 Swazi, 545 Swedish generally, 1 academic achievement gaps and, 52 adult bilingualism and, 79 immersion programs and, 214, 217 language choice in bilingual interaction and, 337–338, 344 language contact and, 537 multilingualism and, 548, 557 young “global hybrids” and, 65–66, 71 Swiss Alemannic, 520 Swiss German, 551 Tagalog interpreters and, 138 speech discrimination in, 17 Taiap, 343 Tajik, 332 Takia, 534–535 Tamil, 126, 127 Thai generally, 6–7 exposure and, 202 input and, 207 Tok Pisin, 343 Turkic languages, 533–534 Turkish adolescent bilingualism and, 36–37 adult bilingualism and, 80–81 first language attrition and, 360

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:09:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.030

650

Language Index Turkish (cont.) language choice in bilingual interaction and, 345 language contact and, 527 measurement of bilingual abilities and, 297 Ukrainian, 119–120 Urban Wolof, 128–129 Urdu age-related language changes and, 104 multilingualism and, 548–549 older bilinguals and, 102 Uzbek, 330 Venetian, 511 Vernacular Tunisian Arabic, 68 Vietnamese deaf education in, 124 immersion programs and, 215 language brokering and, 57 multilingualism and, 546

Volga German, 535–536 Vulgar Latin, 548 Warlpiri, 530–531 Waskia, 534–535 Welsh developmental bilingual education and, 397–398 immersion programs and, 214 measurement of bilingual abilities and, 302 plurilingualism and, 129 Wolof, 128–129 Xhosa, 545 Yiddish, 408, 409–410, 413–415, 419 Yup’ik, 50 Zapotec, 343 Zulu language attrition and, 282 multilingualism and, 545

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:09:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.030

Index of Place Names

Afghanistan, 149 Africa. See specific country or region Alaska, 50 Algeria, 547 Alsace, 215, 511 Argentina, 64–65, 67, 216, 223 Asia. See specific country or region Australia adult migrants in, 259, 260–261 first language attrition in, 360 foreign language learning in, 243 heritage languages in, 49 immersion programs in, 215–216 interpreters in, 139–140 language attrition in, 279 language contact in, 530–531 language policy in, 154–156, 169 older bilinguals in, 105, 106 second language acquisition (SLA) in, 425–426, 429 sign bilingualism in, 487 universal second language requirement in, 161 Austria, 66–67, 222, 227 Barcelona, 264 Basque Autonomous Region in Spain, 49, 86, 226, 398, 454 Basque Country. See Basque Autonomous Region in Spain Belarus, 408 Belgium, 150 Bergamo, 518 Bern, 69 Bolivia, 181, 529–530 Brazil, 67, 215–216 Brittany, 67 Brussels, 185, 332–333, 338 Bulgaria, 353 California, 27, 199–200, 204, 219 Cameroon, 243, 409 Canada adolescent bilingualism in, 42, 55–56 adult bilingualism in, 93

adult migrants in, 259, 264 Canadian Charter of Rights and Freedoms, 135 comprehension based instruction (CBI) in, 241–242 early childhood bilingualism in, 23–24, 35 exposure in, 198 first language attrition in, 360, 363 heritage languages in, 49 immersion programs in, 44, 214–215, 217, 220–221, 225, 230 input in, 206 instructional time in, 239–240 language contact in, 529 “language economics” in, 178–179 law in, 145–146 multilingualism in, 549 older bilinguals in, 103, 105, 107–108, 111–113 second language acquisition (SLA) in, 426–427, 429 Catalonia, 26, 264 Central African Republic, 529–530 Central Mexico, 527–528 Chicago, 546–547 Chile, 64–65 China (Mainland) education in, 399 immersion programs in, 215–216 language attrition in, 275 language contact in, 539–540 sign language in, 486 Colombia, 67 Corsica, 215 Cuba, 395 Czech Republic, 57, 432, 469 Dakar, 129 Denmark, 130, 420 East Africa, 43 East Timor, 398 Ecuador, 529 Ecuadorean Andes, 124 England, 254. See also United Kingdom Europe. See specific country or region

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:06:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.031

652

Index of Place Names European Union foreign language learning in, 234–235, 236 law in, 145, 146 multilingualism in, 546 Finland, 21, 25, 63–64, 214, 217, 547 Flanders, 53–54, 293–294, 327, 331–333, 551 Florida, 24–25 France adult bilingualism in, 90–91 adult migrants in, 264 Civil Law, 145–146 immersion programs in, 215 migration and, 257 young adult bilingualism in, 64–65 Friesland, 86, 199 Gapun, Papua New Guinea, 343 Germany adult bilingualism in, 80–81 adult migrants in, 262 language attrition in, 283–284 measurement of bilingual abilities in, 297 multilingualism in, 549, 550 Willkommenskultur, 345 Gothenburg, 261–262 Great Britain, 353. See also United Kingdom Guantanamo, 330 The Hague, 141–142 Haute-Savoie Region, 516 Hiroshima, 333–334 Hong Kong, 146–147, 215–216, 218–219, 256, 399 Hungary, 88, 431 Illinois, 515 India education in, 398 Global English and, 132 multilingualism in, 545–546, 548–549 Indonesia, 123–124 Interlaken, Switzerland, 69 Ireland, 79, 89–90, 168, 214, 225–226 Israel, 107, 110, 143, 197–198, 451 Italy, 149–150, 363, 470, 511, 518 Japan education in, 399 immersion programs in, 215–216, 217, 225 language choice in, 142–143 native speaker ideologies in, 132 second language acquisition (SLA) in, 421 young adult bilingualism in, 72 Jordan, 549 Kansas, 535–536 Kazakhstan, 398 Kenya, 526 Korea, 273–274, 281 Latin America. See specific country or region Lebanon, 264, 549 London, 254, 520 Los Angeles, 537–539, 546–547 Louisiana, 108 Luxembourg, 398

Mainland China. See China (Mainland) Malaysia, 218–219 Mexico, 89, 343, 527–528 Miami, 52 Middle East. See specific country or region Minnesota, 219 Misio´n La Paz, Argentina, 67 Montreal, 112–113, 200, 225 Morocco, 3, 71–72, 78–79, 264, 547 Moselle, 215 Netherlands adolescent bilingualism in, 36–37 adult bilingualism in, 78–79, 92–93 CLIL in, 227 foreign language learning in, 245 gifted bilinguals in, 311 language attrition in, 274–275, 281 language contact in, 527, 534, 539–540 older bilinguals in, 107–108 Newcastle, 527 New York City, 408, 419, 526, 547 New York State, 139 New Zealand, 49, 86–87, 168, 215, 361 North America. See specific country or region Northern Africa, 546, 547 Northern American English, 517–518 Northern Argentina, 67 Northern Europe, 251 Northern Italy, 470, 511, 518 North West Amazon Basin, 67 Nunavut, 146 Oaxaca, 343 Oberwart, Austria, 66–67 Ontario, 42 Oregon, 138 Ottawa, 225 Pakistan, 149, 548–549 Papua New Guinea, 343, 529–530, 534–535 People’s Republic of China. See China (Mainland) Peru, 243–244 Philippines, 138, 398 Poland, 426 Portugal, 1 Prague, 469–470 Provence, 3 Puerto Rico, 107 Qatar, 399 Quebec, 145–146, 178–179, 180–181, 184 Russia, 332, 408, 534, 536 Rwanda, 149 Saudi Arabia, 251 Scandinavia, 487, 548 Scotland, 49, 87, 334. See also United Kingdom Senegal, 128–129 Shanghai, 55–56, 421 Siberia, 529–530 Sichuan, 421 Singapore, 52, 126, 251 Slovakia, 431 Solomon Islands, 67, 129 South Africa, 129, 139, 145–146, 168, 399, 545 South America. See specific country or region

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:06:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.031

Index of Place Names

South Asia, 251 Soviet Union, 408 Spain, 26, 86, 111, 214, 240, 242, 251, 272, 398, 431, 450–451 Spanish Basque Country. See Basque Autonomous Region in Spain Sri Lanka, 126, 127, 431 Sub-Saharan Africa, 49 Surinam, 529–530 Sweden, 65–66, 142–143, 261–262, 549 Switzerland bidialectalism in, 520 early childhood bilingualism in, 54 earnings in, 183–184 language contact in, 524–525 multilingualism in, 174–175, 551 young adult bilingualism in, 69 Syria, 251, 549 Tanzania, 431, 529–530 Tennessee, 515 Texas, 21 Togo, 284 Tokyo, 72 Toronto, 62, 111, 546 Tunisia, 68, 547 Turkey, 78–79, 80–81, 297, 426, 549 Tyneside, 527 Uganda, 63–64, 149 Ukraine, 119–120, 122 United Kingdom adult migrants in, 259, 262 Common Law, 145–147 deaf communities in, 83 interpreters in, 142 migration and, 256 multilingualism in, 550, 558–559 second language acquisition (SLA) in, 426, 429 sign bilingualism in, 487 universal second language requirement in, 161 young adult bilingualism in, 71–72

653

United States adolescent bilingualism in, 36–37, 38, 50–51, 52–53 adult bilingualism in, 78 adult migrants in, 263 bidialectalism in, 515, 517–518 early childhood bilingualism in, 23, 33–34, 35 East Coast, 529–530 foreign language learning in, 235–236 hegemony in, 119 heritage languages in, 48–49 immersion programs in, 44, 45, 215–216, 219–221, 225–226 internalization in, 119 interpreters in, 139, 142 language choice in, 142–143 language contact in, 524–525, 529–530 language policy in, 169 minoritized youth in, 46–48 multilingualism in, 546–547, 553, 555 normative monolingualism in, 123 older bilinguals in, 108, 112, 113 police interrogations in, 142–144 second language acquisition (SLA) in, 420, 426, 429–430 sign bilingualism in, 487 universal second language requirement in, 161 young adult bilingualism in, 62–63 Veneto, 511 Vienna, 469 Vietnam, 57, 546 Wales, 133, 214. See also United Kingdom Western Isles, Scotland, 334 Yugoslavia, 141–142 Zakarapattia Region, 119–120, 122

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:06:55, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.031

Subject Index

Academic achievement overview, 51 in CLIL programs, 228–229 in EMI programs, 218–219 gaps, 50–54 of immersion students, 217, 219 of immigrant children, 52–53 Accommodation theory, 69, 329–330 A Clockwork Orange, 316 Activation threshold hypothesis, 270–271, 279, 352 Adaptive control hypothesis (ACH), 455–456 Additive immersive programs overview, 213–214, 231 CLIL (See Content and Language Integrated Learning (CLIL)) future research, 228–231 immersion programs (See Immersion programs) methodological challenges, 228–231 Study Abroad (SA) (See Study Abroad (SA)) Adolescence as lifespan stage, 4–5, 36–37, 58 Adolescent bilingualism overview, 36–37, 58 academic achievement gaps, 50–54 becoming literate in new language, 41–43 bidialectalism and language attitudes, 514–515 production, 516–517 contextual and personal perspective, 37–40 language as resource heuristic, 40 predictors of successful outcomes, 39 gifted bilinguals, 320–321 immersion programs, 43–46 language brokering, 56–58 language majority youth, opportunities for bilingualism among, 43–46 lives beyond school, 55–56 minoritized youth, language rights and additive language learning for, 46–50 dual language programs, 46–48 heritage languages, 48–50 indigenous languages, 48–50 social media and, 55–56

students with limited or interrupted formal education (SLIFE), 42–43 Adult bilingualism overview, 76–77, 99–100 bidialectalism and language attitudes, 517–518 production, 518–521 bimodal bilingualism, 82–84 childhood language acquisition versus, 92–95 critical period hypothesis and, 84–86 in deaf communities, 82–84 early start, effect of, 77 foreign languages, 87–89 heritage language speakers, 79–82 identification and, 78–79 indigenous languages, 86–87 language attrition, 280–282 literacy and schooling, role of, 79–82 methodological challenges, 95–99 migrants, 78–79, 89–92 (See also Migration) regional languages, 86–87 social factors, 78–79 stigmatisation and, 78–79 Adulthood as lifespan stage. See also Older bilinguals overview, 4–5, 76–77, 99–100 childhood compared, 92–95, 411–412 later, 4–5 midlife or middle, 4–5, 280–282, 411–412 young, 59–61, 75 Affirmative ethnicity, 159 Age-appropriate attention to language, 238–239 levels in two languages, 3, 10 Age of acquisition exposure and, 194–195 neurolinguistics and, 468–469, 473–476, 477 second language acquisition (SLA) and, 422–423, 428 Alberta Language and Development Questionnaire (ALDeQ), 196 Alberta Language Environment Questionnaire (ALEQ), 196, 197 Alzheimer’s disease, 278

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

Subject Index

American Council on the Teaching of Foreign Languages (ACTFL), 88–89, 236 American Speech-Language-Hearing Association, 376 Aphasia intervention practices, 386–387 language attrition and, 278–279 as language disorder, 374–375, 382–383 neurolinguistics and, 472–473 Aphasia TalkBank, 384 Arab Spring, 546–547 Arizona Department of Education, 133 Asperger’s Syndrome, 317, 319 The Atlantic, 345 Attrition. See Language attrition Australia Asia Project, 155–156 Autism Spectrum Disorder (ASD), 317, 318, 373 +Babbel Magazine, 316 Basic Interpersonal Communicative Skills (BICS), 294 Basic Language Cognition (BLC), 305–306 Basque Center for Cognition, Brain, and Language (BCBL), 450–451, 453–455 Bayes Factor Analysis, 450–451 Benedict XVI (Pope), 315 Berlin Wall, 514 BFLA. See Bilingual first language acquisition (BFLA) BIA+ Model, 441–443 Bidialectalism overview, 510–511, 521–523 in adolescence language attitudes, 514–515 production, 516–517 in adulthood language attitudes, 517–518 production, 518–521 bilingual first language acquisition (BFLA) and, 512–513 contexts for acquisition of dialects, 512–514 definition of dialect, 511–512 in early childhood language attitudes, 514–515 production, 516–517 early second language acquisition (ESLA) and, 512–513 gender and, 519 heritage languages and, 513–514 language versus dialect, 511–512 migration and, 514 second language acquisition (SLA) and, 512–513 settings for acquisition of dialects, 512–514 in young adulthood language attitudes, 517–518 production, 518–521 Bilingual advantage hypothesis overview, 449–450 adaptive control hypothesis (ACH), 455–456 controlled dose hypothesis, 453–455 executive functioning and, 463–464 infrequency of bilingual advantages, 450–451, 456–457 necessary and sufficient experience, search for, 452–456 other cognitive areas, advantages in, 462–464 problems with, 451–452

655

Theory of Mind (ToM) and, 463–464 Bilingual Aphasia Test (BAT), 382–383, 384 Bilingual English-Spanish Assessment, 378–379 Bilingual first language acquisition (BFLA) overview, 15–17 bidialectalism and, 512–513 lexical learning and, 22–23, 28–29 speech discrimination and, 17–19 vocabulary, relation to grammar, 31–33, 34 Bilingual Input Output Survey (BIOS), 377–378 Bilingual Interactive Activation (BIA) Model, 439–441, 445, 446 Bilingual interlocutors, 325–326, 328 Bilingualism definition, 10 multilingualism versus, 3–4, 544 Bilingualism: Language and Cognition, 289–290 Bilingual language control overview, 435–436 BIA+ Model, 441–443 Bilingual Interactive Activation (BIA) Model, 439–441, 445, 446 bimodal language control, inhibition and, 449 code-switching and, 446–447 cognates and, 436–438 Cumulative Semantic Interference (CSI) and, 445–446 general task switching versus code-switching, 447–449 inhibitory control model, 445–446 interlingual homograph interference and, 436–438 lexical decision tasks (LDT), 436–437 neurolinguistics and, 469–470 Revised Hierarchical Model (RHM), 438–439 translation priming and coactivation, 443–445 Bimodal bilingualism overview, 507–509 adult bilinguals and, 82–84 future research, 507–509 measurement of bilingual abilities and, 293 metalinguistic awareness and, 502–506 sign bilingualism versus, 486–488 Bimodal language control, 449 Black English. See African American Vernacular English (AAVE) Borges, Jorge Luis, 307 Burgess, Anthony, 316 Bush, Jeb, 345 CALICO Journal, 420 Calques, 536 Categorical Perception Index (CPI), 458 CDI. See MacArthur-Bates Communicative Development Inventory (CDI) C¸evikkollu, Fatih, 345 Chagall, Marc, 408, 409–410, 413–415, 417, 419 Chancellor Merkel, 345 Childhood as lifespan stage overview, 4–5 early, 15–17, 34–35 (See also Bilingual first language acquisition (BFLA); Early second language acquisition (ESLA)) middle, 36–37, 58 Child Language Data Exchange System (CHILDES), 379

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

656

Subject Index Children of deaf adults (CODAs) overview, 83 code-switching by, 488–490, 492 deaf children versus, 483–484, 485–486 experimentally induced crossmodal coactivation, 491–492 Choice of language. See Language choice in bilingual interaction CLIL. See Content and Language Integrated Learning (CLIL) Clinical Evaluation of Language Fundamentals (CELF), 378–379 Clinical linguistics overview, 369–370, 388–389 adults, language disorders in aphasia, 374–375, 382–383, 386–387 assessment of, 380–384 intervention practices, 386–388 primary progressive aphasia (PPA), 375 assessment of language disorders overview, 376 in adults, 380–384 in children, 376–380 direct assessment, 378–379 discourse assessment, 383–384 dynamic assessment, 380 observations, 378 questionnaires, 377–378 samples, 379–380 semi-structured interviews, 377 children, language disorders in overview, 370 assessment of, 376–380 Autism Spectrum Disorder (ASD), 373 Down Syndrome (DS), 372–373 general developmental disorders, 372–374 intervention practices, 385–386 Specific Language Impairment (SLI), 370–372 intervention practices overview, 384–385 in adults, 386–388 in children, 385–386 cognates and, 388 crosslinguistic transfer (CLT), 386, 387–388 CODAs. See Children of deaf adults (CODAs) Code-switching bilingual language control and, 446–447 by children of deaf adults (CODAs), 488–490, 492 by deaf children, 492–502 education and, 395–396 exposure and, 208–210 general task switching versus, 447–449 inhibition and, 446–447 language attrition and, 553 language choice in bilingual interaction and, 326–328 language contact and, 525–528 Language Synthesis Model and, 500–501, 507 measurement of bilingual abilities and, 295–296 multilingualism and, 547–548 by older bilinguals, 108–111 Cognates, 388, 431, 436–438. See also Interlingual homograph interference Cognitive Ability for Novelty in Acquisition of Language (Foreign) aptitude test, 312

Cognitive Academic Language Proficiency (CALP), 294 Cognitive science overview, 435, 464–465 bilingual advantage hypothesis (See Bilingual advantage hypothesis) bilingual language control (See Bilingual language control) emotions and, 460–462 linguistic relativity and, 457–459 memory retrieval and, 459–460 moral judgments and, 460–462 Common European Framework of Reference (CEFR), 234–235 Communication Accommodation Theory (CAT), 329–330, 346–347 Communicative Development Inventory (CDI). See MacArthur-Bates Communicative Development Inventory Communicative language teaching (CLT), 236–237 Comprehension based instruction (CBI), 241–242 Computer mediated communication, 71–72 Contact-induced language change overview, 531 conditions for, 532–533 debate regarding, 531–532 restrictions on, 533–535 Contact linguistics. See Language contact Content and Language Integrated Learning (CLIL) affective factors, 226–227 education generally, 397 learning outcomes, 221–223 optimal integration, 228–230 program options, 216 Controlled dose hypothesis, 453–455 Convergence, 539 Convergent choice principle, 328–331, 340 Corrective feedback (CF), 237–238 Council of Europe, 258–259, 263, 392 Court Interpreters Act of 1978 (US), 135 Courtroom interactions, 139–142 Crebs, Emil, 317 Critical pedagogy, 265 Critical period hypothesis (CPH), 84–86, 309–311 Crosslinguistic influence (CLI), 296–297, 349, 415–418, 556 Crosslinguistic Lexical Tasks (CLTs), 302 Crosslinguistic transfer (CLT), 386, 387–388 Cultural migrants, 90 Cumulative Semantic Interference (CSI), 445–446 Cummins, Jim, 401 Curie, Marie, 432 Deaf communities overview, 483–484, 507–509 adult bilingualism in, 82–84 assistive technology for, 82–83 bimodal bilingualism (See Bimodal bilingualism) children of deaf adults (CODAs) (See Children of deaf adults (CODAs)) code-switching and, 210 deaf children CODAs versus, 483–484, 485–486

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

Subject Index

code-switching by, 492–502 Language Synthesis Model and, 500–501, 507 education and, 391–392, 394 future research, 507–509 ideologies of bilingualism and, 124 input and, 207 language acquisition and, 485–486 language choice in bilingual interaction and, 335–336 metalinguistic awareness and, 502–506 multilingualism and, 558–559 sign language (See Sign language) simultaneous communication (SIMCOM), 486–487, 488–490 Declarative/procedural (DP) model, 471 Defense Language Institute, 322 Deficit views of bilingualism, 8–9, 290, 291–292, 294–295, 299, 450 Delors Committee, 345 Delors, Jacques, 345–346 Deutsch am Arbeitsplatz, 262 Developing countries, 187 Developmental bilingual education, 397–398, 399 Dialects. See Bidialectalism Dickinson, Emily, 307 Digital media. See also Social media overview, 151, 241, 324–325 adult immigrants’ online communication and, 254 out-of-school experiences and, 242–244 use by adolescents, 55–56 use of technology and, 242–244 Diglossia, 127, 547, 549 Discourses, 120, 121–122, 124–125, 126, 129 Down Syndrome (DS), 372–373 Dual language programs, 46–48. See also Two-way immersion Dynamic bilingualism, 393–394 Early childhood bilingualism overview, 15–17, 34–35 bidialectalism and language attitudes, 514–515 production, 516–517 gesture, relation to language development, 29–31 language attrition, 272–277 lexical learning language differentiation, 28–29 processing speed, role of, 26–28 relative exposure, role of, 26–28 vocabulary, effect of size of, 22–26 neuroscience and, 19–22 phonemic learning and, 18–19 speech discrimination, 17–22 vocabulary, relation to grammar, 31–34 Early second language acquisition (ESLA) overview, 15–17 bidialectalism and, 512–513 vocabulary, relation to grammar, 33–34 Earnings, 182–184 Economics overview, 173–174 access to employment, 184–186 definitions, 177 earnings, 182–184

657

emergence of “language economics,” 177–179 GDP in developing countries, 187 international trade, 186–187 job retention, 186 language policy, application to, 187–189 multilingualism versus, 174–175, 545–547 objective versus subjective diversity, 177 plurilingualism versus, 174–175 translanguaging and, 176–177 value, 179–182 variables in, 176 Education overview, 390–391, 406–407 CLIL, 397 code-switching and, 395–396 communities, performance in, 395–396 deaf communities and, 391–392, 394 developmental bilingual education, 397–398, 399 dynamic bilingualism, 393–394 elite bilingualism, 391 EMI programs, 399 folk bilingualism, 391 foreign language learning and, 397–398 heritage languages and, 397–398 immersion programs, 397 indigenous languages and, 397–398 language attrition and, 399 language policy and, 399–402 minoritized youth and overview, 402 foreign language learning and, 397–398 translanguaging, 402–406 mother tongue-based multilingual education (MTB-MLE), 398, 399 multilingual education (MLE), 398 multilingualism and, 545–547 plurilingualism, 392 recursive bilingualism, 393 schools, learners in, 392–394 transitional bilingual education, 399 translanguaging and assessment, 404–405 minoritized youth, 402–406 pedagogy, 402–404 teachers and materials, 405–406 types of bilingualism, 391–392 world language education, 397 Elderly bilinguals. See Adulthood as lifespan stage; Older bilinguals Electroencephalography (EEG), 19–21, 362 Elicited imitation. See Sentence repetition tasks Elite bilingualism, 391 EMI programs, 399 Emmaus (network), 62 Emotions overview, 460–462 emotional warmth in classroom, 54 L1 maintenance and, 62 language choice and, 335–336, 337, 344 language varieties and, 517–518 migration and, 73–74, 90–91 Employment access to, 184–186 adult migrants and, 256–257 gendered roles and, 428

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

658

Subject Index Employment (cont.) job retention, 186 in language-intensive occupations, 427 migration and, 256–257, 261–262 wages, 186 workplace, 62–63 Erasure, 122 ESLA. See Early second language acquisition (ESLA) Essentialist viewpoints, 122 European Commission, 245, 246 European Convention on Human Rights, 135 European Court of Justice, 146 European Survey of Language Competencies (ESLC), 234–235 Event related potentials (ERPs), 362–364 Evidence, language of bilinguals as, 147–150 Executive functioning. See also Bilingual advantage hypothesis overview, 435, 463–464, 465 central executive functions, 312–313, 456, 475 Exposure, early bilingualism and overview, 193–194, 211–212 absolute exposure, 195–196, 202–205 age of acquisition, 194–195 code-switching and, 208–210 cumulative length of exposure, 198–202 length of exposure, 195, 196–198 relative exposure, 195, 196–202 second language acquisition (SLA) and, 418–422 terminology, 194–196 Extreme Male Brain Theory of Autism, 317, 318 Factionalism, 159–161 Familiarity, 467–468, 473–476. See also Language dominance Family language policy (FLP), 70–71, 74 Fanfiction.com, 55–56 Feminism, 68–69 First language attrition overview, 349–351, 364–365 age of onset, 353–354 investigation of overview, 357–358 brain signals, 362–364 EEG, 362 errors, 360 event related potentials (ERPs), 362–364 fluency, 360 fMRI, 362, 363 populations compared, 358–359 processing, 361–362 in older bilinguals, 106–108 research questions overview, 351 causes of, 351–354 linguistic domains affected by, 354–356 mechanisms of, 356–357 Flanker tasks, 465 Flaubert, Gustave, 307 Folk bilingualism, 391 Foreign language learning overview, 233–234, 248–249 communicative language teaching (CLT), 236–237 contexts for, 234–236 corrective feedback (CF), 237–238

early start, effect of overview, 244 advantages of, 247–248 primary school, 246–247 very young learners, 244–246 education and, 397–398 form, focus on, 236–239 innovative strategies overview, 241 comprehension based instruction (CBI), 241–242 out-of-school experiences, 242–244 production based instruction (PBI), 241–242 Study Abroad (SA), 244 technology and media, use of, 242–244 instructional time, 239–241 meaning, focus on, 236–239 older bilinguals immigrants, 111–112 non-immigrants, 112–113 Foreign Service Institute (FSI), 235–236 Forgetting of languages. See Language attrition; Unlearning of languages Functional linguistic relativity, 145–147 Functional Magnetic Resonance Imaging (fMRI), 94, 362, 363, 474 Gender bidialectalism and, 519 polyglots and, 316–317 second language acquisition (SLA) and, 428 Gesture, relation to language development, 29–31 Gifted bilinguals overview, 307–308, 322–323 ability and, 308 aptitude and, 308 children, 320–321 critical period hypothesis (CPH) and, 309–311 future research, 321–322 giftedness and, 308 grammaticality judgment (GJ) tests, 310 hyperglots, 314–318 interpreters, 320–321 multilingualism and, 311–313, 314 polyglots, 314–318 savants, 318–320 second language acquisition (SLA) and, 308–309 talent and, 308, 313 terminology, 308 working memory (WM) and, 312–313 Global English, 131–133 “Global hybrids” overview, 60–66 as couples, 66–70 family language policy (FLP) and, 70–71, 74 future research, 72–75 migration and, 59 one-person one-language (OPOL) strategy, 71–72 as parents, 70–72 in workplace, 62–63 Global Migration Data Analysis Centre, 59 Grammar, relation to vocabulary, 31–34 Grammaticality judgment (GJ) tests, 310 Great Britain, 353. See also United Kingdom Guantanamo, 330

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

Subject Index

Guidelines for communicating rights to non-native speakers of English in Australia, England and Wales, and the USA, 143–144 Hale, Kenneth, 316 Harmonious bilingualism, 38 Hegemony, 119 Heritage language speakers. See also specific language adolescent bilingualism and, 48–50 adult bilingualism and, 79–82 bidialectalism and, 513–514 education and, 397–398 language attrition and, 275–277 language contact and, 535–541 minoritized youth and, 48–50 multilingualism and, 551–552 Heteroglossic ideologies, 393–394, 395–396 Hierarchical multilingualism, 129 Higher Language Cognition (HLC), 305–306 Highly proficient bilinguals. See Gifted bilinguals Hi-Lab, 422–423 Human rights, 250 Hyperglots, 314–318 Iconicity, 122 Identification, cultural and ethnic, 78–79 Ideologies of bilingualism overview, 119–120, 134 deaf communities and, 124 Discourses on, 120, 121–122, 124–125, 126, 129 essentialist viewpoints and, 122 future research, 133–134 Global English, 131–133 hegemony and, 119 hierarchical multilingualism, 129 internalization and, 119 key concepts, 120–123 linguistic nationalism, 129 linguistic pragmatism, 129 market language ideology, 119–120 methodological challenges, 133–134 metrolingualism, 130–131 monoglossic ideologies, 123–125 mother tongue ideology, 123 native speaker ideologies, 131–133 normative monolingualism, 123–124 one-nation-one-language ideology, 119–120, 123, 129 pluralist ideologies, 127–129 plurilingualism, supporting, 129–131 polylanguaging, 130 reciprocal multilingualism, 129 sociolinguistics, 121–122 standard language ideologies, 126–127 terminology, 120–121 translanguaging, 130 (See also Translanguaging) Immersion programs adolescent bilingualism and, 43–46 affective factors, 224–226 education generally, 397 learning outcomes, 217–221 program options, 214–216 Immigration bump, 110 Inclusive approach, 2–4

659

Indigenous community organisations, 155 Indigenous languages. See Heritage languages Inhibition bimodal language control and, 449 code-switching and, 446–447 Inhibitory control model, 445–446 Input, early bilingualism and, 205–208 Instructional time, 239–241 Interdependence hypothesis, 294 Interdisciplinary connections, 8–10 Intergenerational language transmission, maintenance, vitality or loss, 71, 73, 87, 123, 161, 166, 276, 293, 342–343, 350 Interlingual homograph interference, 436–438. See also Cognates Internalization, 119 International adoptees, 16–17, 273–275, 282, 284–285 International Association of Forensic Linguists (IAFL), 136 International Criminal Tribunal for the former Yugoslavia, 141–142 International English Language Testing System (IELTS), 291 International Journal of Speech Language and the Law, 136 International students, 429–431 International trade, 186–187 Interpreters in child language brokering, 56–58, 320–321 gifted bilinguals, 320–321 in legal profession, 138–139 Intersectionality, 253–254 Intervention practices overview, 384–385 in adults, 386–388 in children, 385–386 cognates and, 388 crosslinguistic transfer (CLT), 386, 387–388 Job retention, 186 John Paul II (Pope), 315 Journal of Policy Analysis and Management, 152, 153–154, 161 Jo´venes con Derechos, 48–49 Joyce, James, 316 Kafka, Franz, 432 Kay, John, 307 Kenrich, Donald, 316 Kissinger, Henry, 432 Krebs, Emil, 316, 317 Kristallnacht, 426 Kudrow, Lisa, 307 Laakso, Pertti, 316 Language analysis for determination of origin (LADO), 148–149 Language and Law, 136 Language attitudes in adolescence, 514–515 in adulthood, 517–518 in early childhood, 3, 514–515 parental home language choice and, 6–7 by teachers, 53–54 in young adulthood, 517–518

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

660

Subject Index Language attrition. See also Unlearning of languages age of onset and, 353–354 aphasia and, 278–279 causes of, 351–354 code-switching and, 352–353 codeswitching and, 553 collective unlearning, 166, 270–272 early childhood bilingualism and, 272–277 education and, 399 errors and, 360 in first language, 349–351, 364–365 (See also First language attrition) fluency and, 360 future research, 282–284 in heritage language speakers, 275–277 in international adoptees, 273–275 multilingualism and, 552–554 in multilinguals, 552–554 in older bilinguals, 106–108, 277–280 Universal Grammar (UG) and, 268–269 Language brokering, 56–58, 320–321 Language choice in bilingual interaction overview, 324–325, 346–348 accommodation theory, 69, 329–330 bilingual interlocutors and, 325–326, 328 code-switching and, 326–328 constraints on, 331–334 convergent choice principle and, 328–331, 340 couples and, 66–70 foundational theories of social interaction, 325–328 language mode and, 338–341 language orientation and, 338–341 language socialization and, 328–331, 337 mixed utterances and, 326–328 politeness principle and, 331, 332–333, 344, 346–347 politics and, 344–346 psycholinguistics of, 338–341 social meaning and, 334–338 unilingual utterances and, 326–328 in young families, 341–344 Language contact overview, 524–525, 541–543 calques and, 536 code-switching and, 525–528 contact-induced language change overview, 531 conditions for, 532–533 debate regarding, 531–532 restrictions on, 533–535 contact languages, 528–531 convergence and, 539 heritage languages and, 535–541 lexical borrowing and, 525–528 matter replication and, 542 metatypy and, 534–535 pattern replication and, 542 Language disorders. See Clinical linguistics Language dominance, 30–31, 109–110, 294–295, 300–301, 365, 489, 494, 552–553. See also Familiarity “Language economics.” See Economics Language Environment Analysis System, 20 Language Experience and Proficiency Questionnaire (LEAP-Q), 381–382

Language Exposure Assessment Tool (LEAT), 377 Language Input Diary (LID), 202–203 Language Learning, 410 Language Learning & Technology, 420 Language mixing. See Code-switching Language Mixing Scale, 208–209 Language mode, 338–341 Language orientation, 338–341 Language policy overview, 152–153, 170–172, 189–190 affirmative ethnicity, 159 authority for policy, 167 diplomatic, trade and security elites, 155 economics, application of, 187–189 education and, 399–402 factionalism, 159–161 goals of policy, 168–169 immigrant community organisations, 155 indigenous community organisations, 155 knowledge and power, 162 language professionals, 155, 160–161 model for planning and policy, 165–170 participation in policy, 167 persuasion, 162–164 policy as argument, 164 protagonist voices, 154–155 recommendations for planning and policy, 165–170 universal second language requirement, 158–159, 161 value and risk of becoming policy problem, 165 Language Policy, 154 Language Problems and Language Planning, 154 Language professionals, 155, 160–161 Language socialization, 328–331, 337 Language Synthesis Model, 500–501, 507 Law overview, 135–136, 151 courtroom interactions, 139–142 evidence, language of bilinguals as, 147–150 functional linguistic relativity, 145–147 future research, 150–151 interpreters, 138–139 language analysis for determination of origin (LADO), 148–149 methodological challenges, 150–151 monolingual jurisdictions, options for bilinguals in, 136–138 police interrogations, 142–144 Legitimacy, 38 Lexical borrowing, 525–528 Lexical decision tasks (LDT), 436–437 Lexical learning language differentiation, 28–29 processing speed, role of, 26–28 relative exposure, role of, 26–28 vocabulary, effect of size of, 22–26 Lifespan perspective overview, 4–5 definition of five stages, 4–5 Lingua Franca, English as, 65, 132, 137, 176–177, 197–198, 545 Linguaspheres, 186–187 Linguistic Integration of Adult Migrants (LIAM), 258–259, 263 Linguistic nationalism, 129 Linguistic pragmatism, 129

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

Subject Index

Linguistic relativity, 457–459 Literacy adult bilingualism, role in, 79–82 adult migrants and, 262–263 LLAMA, 422–423 Lomb, Kato´, 316 Lord of the Rings, 316 Lubach, Arjen, 536 Lundgren, Dolph, 307 MacArthur-Bates Communicative Development Inventory (CDI), 22, 24, 32, 299–302 Market language ideology, 119–120 Martinet, Andre´, 316 Matrix Language Frame (MLF) Model, 489, 526 Matter replication, 542 Mean length of utterance (MLU), 30–31, 303–304 Measurement of bilingual abilities overview, 289–292, 304–306 in assessment of language disorders overview, 376 in adults, 380–384 in children, 376–380 direct assessment, 378–379 discourse assessment, 383–384 dynamic assessment, 380 observations, 378 questionnaires, 377–378 samples, 379–380 semi-structured interviews, 377 Basic Language Cognition (BLC), 305–306 bimodal bilingualism and, 293 code-switching and, 295–296 Communicative Development Inventory (CDI). See MacArthur-Bates Communicative Development Inventory (CDI) construct, defining, 292–294 crosslinguistic influence (CLI), 296–297 Crosslinguistic Lexical Tasks (CLTs), 302 dominance versus balance, 294–295 Higher Language Cognition (HLC), 305–306 interdependence hypothesis and, 294 mean length of utterance (MLU), 303–304 multifaceted phenomena, 297–299 Peabody Picture Vocabulary Test (PPVT), 300–302 self-assessments, 298–299 sentence repetition tasks, 304 Specific Language Impairment (SLI), 302 Total Conceptual Vocabulary (TCV), 299–300 uniquely bilingual abilities, 295–297 variability and, 297–299 vocabulary, 299–303 Memory retrieval, 459–460 Metalinguistic awareness, 502–506 Metatypy, 534–535 Metrolingualism, 130–131 Mezzofanti, Joseph, 315 Midlife bilingualism. See Adult bilingualism Migration adult migrants overview, 78–79, 89–92, 250–252, 265–266 characteristics of, 254–256 critical pedagogy, 265 employment and, 256–257, 261–262 first generation adult migrants, 251 intersectionality and, 253–254

661

literacy and, 262–263 multilingual pedagogy, 263–265 participatory pedagogy, 265 pedagogical support, 260–265 policy support, 257–260 super-diversity and, 252–253 translanguaging and, 263–264 bidialectalism and, 514 cultural migrants, 90 immigrant community organisations, 155 immigration bump, 110 multilingualism and, 546–547 older bilinguals, language learning among, 111–112 second language acquisition (SLA) and, 425–429 young “global hybrids,” 59 Minecraft, 55 Minoritized youth education and, 397–398 language rights and additive language learning for, 46–50 dual language programs, 46–48 heritage languages, 48–50 indigenous languages, 48–50 translanguaging and, 402–406 assessment, 404–405 pedagogy, 402–404 teachers and materials, 405–406 Miranda rights, 144 Mirandola, Pico della, 315 Mismatch Response (MMR), 19–21 Mixed utterances, 326–328 Modeling hypothesis, 208 Modern Language Aptitude Test, 308, 310, 422–423 Monoglossic ideologies, 123–125, 393, 394, 395–396 Monolingual bias, 10–11, 413 Monolingualism, 10–11 Moral judgments, 460–462 Mora-timed languages, 17 Mother tongue-based multilingual education (MTB-MLE), 398, 399 Mother tongue ideology, 123 Multilingual Children’s Speech, 379 Multilingual education (MLE), 398 Multilingualism overview, 544, 560 bilingualism versus, 174–175, 545–547, 554–557 code-switching and, 547–548 communities, language use in, 547–549 crosslinguistic influence (CLI) and, 556 deaf communities and, 558–559 economic factors, 545–547 economics and, 174–175 educational factors, 545–547 future research, 557–559 gifted multilinguals, 311–313, 314 heritage languages and, 551–552 historical factors, 545–547 individuals, use by overview, 549–550 interactional strategies in multilingual families, 550–552 language attrition, 552–554 migration and, 546–547

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

662

Subject Index Multilingualism (cont.) political factors, 545–547 receptive multilingualism, 293, 548–549 societal multilingualism, 545–547 Multilingual pedagogy, 263–265 Nabokov, Vladimir, 307 National Association of Judicial Interpreters and Translators (NAJIT), 139 National Journal, 345 Native speaker ideologies, 131–133 Nativespeakerism, 132, 412, 433–434 Neurolinguistics overview, 466–467, 478–479 age of acquisition and, 468–469, 473–476, 477 aphasia and, 472–473 bilingual language control and, 469–470 declarative/procedural (DP) model, 471 familiarity and, 467–468, 473–476 fMRI and, 474 in processing approaches, 470–472 rethinking of current models, 476–478 Neuroscience early childhood bilingualism and, 19–22 electroencephalography (EEG), 19–21 Mismatch Response (MMR), 19–21 whole head magnetocephalography, 19 New Zealand Maori Language Commission, 87 Nineteen Eighty-Four (Orwell), 168–169 Normative monolingualism, 123–124 Oh, Sandra, 307 Older bilinguals overview, 101–102, 113–115 age-related language changes overview, 102–104 selective versus comparable changes, 105–106 code-switching by, 108–111 cognitive skills and, 104 first language attrition in, 106–108, 277–280 foreign language learning immigrants, 111–112 non-immigrants, 112–113 lexical retrieval and, 103–104, 105 variability, challenge of, 114–115 vocabulary and, 103 One-nation-one-language ideology, 119–120, 123, 129 One-person one-language (OPOL) strategy, 71–72, 209 Opportunities for bilingualism, 6–8 Orwell, George, 168–169 Otakutalk, 55 Oxford University Press, 136 Participatory pedagogy, 265 Pattern replication, 542 Peabody Picture Vocabulary Test (PPVT), 300–302 Phonemic learning, 18–19 Pitres, Jean-Albert, 467–470, 473, 477 Pitres’s rule, 467–468 Pluralist ideologies, 127–129 Plurilingualism, 129–131, 174–175, 392 Police interrogations, 142–144 Policy. See Language policy Politeness principle, 331, 332–333, 344, 346–347

Polyglots, 314–318 Polylanguaging, 130 Postmodernism, 68–69 Po¨hl, Karl Otto, 346 Post-traumatic stress disorder (PTSD), 65–66 Po¨tzl, Otto, 469–470 Preschool Language Scales, 378–379 Primary progressive aphasia (PPA), 375 Production based instruction (PBI), 242 Programme for International Student Assessment (PISA), 51–52 Protagonist voices, 154–155 Psycholinguistics, 338–341 Putin, Vladimir, 332 Questionnaire for Parents of Bilingual Children (PaBiQ), 196 Rask, Rasmus Christian, 316 ReCALL Journal, 420 Receptive and Expressive One Word Picture Vocabulary Tests, 378–379 Receptive bi/multilingualism, 77, 293, 336, 548–549 Reciprocal multilingualism, 129 Recursive bilingualism, 393 Recursivity, 122 Regional languages, adult bilingualism and, 86–87 Regression hypothesis, 277–278 Relearning paradigm, 280–282 Reminiscence bump, 110 Research connections, 8–10 Retrieval Induced Forgetting (RIF), 271–272, 279 Revised Hierarchical Model (RHM), 438–439 Rhythmic classes of languages, 17–18 Rights communicating rights to non-native speakers, 143–144 equal language rights, 145 human rights, 250 individual rights, 125 to interpretation, 135–136 language policy and, 152–153 language rights, 133–134, 145, 250 of minority populations, 161, 165 Miranda rights, 144 rights-based bilingualism and, 153 speaking rights, 139 of test takers, 259 Universal Declaration of Human Rights, 250 Universal Declaration of Linguistic Rights, 250 Royal College of Speech and Language Therapists, 290, 376 SA. See Study Abroad (SA) Sapir-Whorf hypothesis, 457 Savants, 318–320 Savings paradigm, 280–282 Scaffolding Theory of Aging and Cognition, 95 Schumann’s Acculturation Model, 91–92 Schwartz, Alexander, 315 Seals of Biliteracy, 45 Second language acquisition (SLA) overview, 408–410, 432–434 in adults, 423–424 age of acquisition and, 422–423, 428 bidialectalism and, 512–513

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

Subject Index

in children, 422–423 contesting notions of success, 422–424 in context of bilingualism, 411–413 crosslinguistic influence (CLI) and, 415–418 as discipline, 410 exposure and, 418–422 future research, 433 gender and, 428 gifted bilinguals and, 308–309 in international students, 429–431 methodological challenges, 433–434 migration and, 425–429 scope of, 413–415 Second Language Research, 410 Semantic Feature Analysis (SFA), 387 Sentence repetition tasks, 304 Sign bilingualism, 486–488 Sign language. See also specific language overview, 483–484, 507–509 children of deaf adults (CODAs) code-switching by, 488–490, 492 deaf children versus, 483–484, 485–486 experimentally induced crossmodal coactivation, 491–492 deaf children CODAs versus, 483–484, 485–486 code-switching by, 492–502 Language Synthesis Model and, 500–501, 507 future research, 507–509 language acquisition and, 485–486 metalinguistic awareness and, 502–506 simultaneous communication (SIMCOM), 486–487, 488–490 Simon effect, 465 Simultaneous communication (SIMCOM), 486–487, 488–490 SLA. See Second language acquisition (SLA) Snapchat, 55 Social media, 43, 55–56, 71–72, 511. See also Digital media Sociolinguistics, 121–122 Specific Language Impairment (SLI), 302, 370–372 Speech discrimination, 17–22 Standard language ideologies, 126–127 Stress-timed languages, 17 Stroop tasks, 450–451, 465 Students with limited or interrupted formal education (SLIFE), 42–43 Studies in Second Language Acquisition, 410 Study Abroad (SA) affective factors, 228 foreign language learning and, 244 learning outcomes, 223–224 program options, 216–217 understanding benefits of, 230–231 Success in bilingualism contesting notions of, 422–424 predictors of, 38, 39 successful learners, 69, 87–89, 94–95, 425–429 Super-diversity, 177, 252–253 Superior Temporal Gyrus (STG), 474–475 Svenska for invandrare, 261–262 Switch Back Through Translation, 387

663

Switch costs, 109–110 Syllable-timed languages, 17 Syrian refugees, 251, 549 Systematic Analysis of Language Transcripts (SALT), 379, 384 Tammet, Daniel, 317 Task schemas, 445, 446 Taylor, Cristopher, 317 Telemundo, 345 Theory of Mind (ToM), 463–464 Theron, Charlize, 307, 432 Thomson and Reuters Journal Citation Reports, 153–154 Tolkien, J. R. R., 316 Total Conceptual Vocabulary (TCV), 299–300, 378–379 Transitional bilingual education, 399 Translanguaging adult migrants and, 254, 263–264 economics and, 176–177 education and, 393–394, 395–396 minoritized youth and, 402–406 assessment, 404–405 pedagogy, 402–404 teachers and materials, 405–406 plurilingualism and, 130 Translation equivalents (TEs), 28–29 Translation priming and coactivation, 443–445 Two-way immersion, 47–48, 215, 217, 219, 220. See also Dual language programs U¨ber das Geda¨chtnis (Ebbinghaus), 280–281 Ulysses, 316 UNESCO, 258 Unilingualism, 157, 158 Unilingual utterances, 326–328 Universal Declaration of Human Rights, 250 Universal Declaration of Linguistic Rights, 250 Universal Grammar (UG), 268–269, 411–412, 488, 500, 526–527 University of Chicago Press, 136 Unlearning of languages. See also Language attrition collective, 166, 270–272 definition, 268–270 individual, 166 relearning paradigm, 280–282 savings paradigm, 280–282 Value in economics, 179–182 Vandewalle, Johan, 315 Vocabulary academic, 429 lexical learning, effect of size of vocabulary on, 22–26 measurement of, 299–303 older bilinguals and, 103 relation to grammar, 31–34 Total Conceptual Vocabulary (TCV), 299–300 total vocabulary size of L1-L2 educated adults, 431 word families benchmark, 430–431

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032

664

Subject Index Wages, 186 Wattpad, 55 Whole head magnetocephalography, 19 Whose Integration? Project, 265 Williams, Harold, 315 Woods, James, 307 Wordbank, 379 Working memory (WM), 312–313 Work on the Speak Good English Movement (Singapore), 126 Workplace, 62–63 World Health Organization (WHO), 381 World language education, 397 World of Warcraft, 55 Wurm, Stephen, 316 Youlden, Matthew, 316 Young adulthood bilingualism overview, 59–61, 75

bidialectalism and language attitudes, 517–518 production, 518–521 as couples, 66–70 family language policy (FLP) and, 70–71, 74 future research, 72–75 migration and, 59 one-person one-language (OPOL) strategy, 71–72 as parents, 70–72 in workplace, 62–63 Youth. See also Adolescence language majority youth and opportunities for bilingualism, 43–46 minoritized youth, 46–50, 397–398, 402 and heritage language speakers, 48–50 and translanguaging, 402–406 Zamenhof, Ludwik Łazarz, 316 Zimmerman, George, 138–139

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 13 Nov 2018 at 16:11:39, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781316831922.032