ELF Research Methods and Approaches to Data and Analyses: Theoretical and Methodological Underpinnings 9780367898793, 9781003021650

Table of contents :
Cover
Half Title
Series
Title
Copyright
Contents
List of figures
List of tables
List of contributors
Acknowledgements
1 ELF research methods and different approaches to data and analyses: introduction
Part I ELF research: some theoretical underpinnings and development
2 Research perspectives on ELF: linguistic usage and communicative use
3 Researching ELF communication: focus on high-stakes encounters
4 ELF and translanguaging: covert and overt resources in a transnational workplace
Part II Spoken and written corpora in ELF research
5 Good texts in non-standard English: ELF and academic writing
6 Exploring diachronic changes in research articles from an ELF perspective: a small corpus study
7 Tracing the emergence of situational multilingual practices in a BELF meeting: micro-diachronic analysis and implications of corpus design
8 Asian Corpus of English (ACE): features and applications
9 Interpersonal formulaic sequences in ELF academic lectures: methodological challenges
Part III Conversation analytic approaches to ELF interactions
10 Applying conversation analysis to ELF interaction data
11 Analysing multilingual/lingua franca interactions using conversation analysis: notes on transcription and representability
12 Accounting for asymmetries in ELF interactions: three disparate Asian settings
Part IV Narrative approaches in ELF research: exploring attitudes towards ELF
13 ‘Place-reflexivity’ as an imaginary kaleidoscope to explore methodological issues in ELF research
14 Understanding the ELF phenomenon through narrative inquiry: a diary study on identities of Japanese ELF users
15 Rigour in ELF language attitude research: an example of a conversational interview study
Index

Citation preview

ELF Research Methods and Approaches to Data and Analyses

This book systematically explores and discusses English as a Lingua Franca (ELF) research methods frequently deployed by ELF researchers in analysing their data. It mainly covers three different approaches: corpus-­based, both written and spoken, conversation analytic and narrative approaches. In addition to exploring these different approaches to ELF data, the volume also introduces case studies that utilise them in analysing data in both academic and workplace settings, which facilitates not only the understanding of the ways in which research is conducted but also its findings. Furthermore, the book discusses theoretical underpinnings of ELF research and its recent development in its first part. It is comprehensive both in understanding theory and exploring research methods which can be deployed in conducting ELF research. The book, therefore, will be of great interest and use for both ELF researchers and educators as well as undergraduate and postgraduate students who are about to embark on their ELF and ELF-­related research, and also to those who are new to the field. Kumiko Murata is a retired professor, formerly with the School of Education and Graduate School of Education in the Faculty of Education and Integrated Arts and Sciences, Waseda University, Japan.

Routledge Research in Language Education

The Routledge Research in Language Education series provides a platform for established and emerging scholars to present their latest research and discuss key issues in Language Education. This series welcomes books on all areas of language teaching and learning, including but not limited to language education policy and politics, multilingualism, literacy, L1, L2 or foreign language acquisition, curriculum, classroom practice, pedagogy, teaching materials, and language teacher education and development. Books in the series are not limited to the discussion of the teaching and learning of English only. Books in the series include: Developing Critical Cultural Awareness in Modern Languages A Comparative Study of Higher Education in North America and the United Kingdom Elinor Parks TESOL and the Cult of Speed in the Age of Neoliberal Mobility Osman (Othman) Barnawi Multilingual Approaches for Teaching and Learning From Acknowledging to Capitalising on Multilingualism in European Mainstream Education Edited by Claudine Kirsch and Joana Duarte ELF Research Methods and Approaches to Data and Analyses Theoretical and Methodological Underpinnings Edited by Kumiko Murata Language Teacher Identity in TESOL Teacher Education and Practice as Identity Work Bedrettin Yazan and Kristen Lindahl For more information about the series, please visit www.routledge.com/Routledge-­ Research-­in-­Language-­Education/book-­series/RRLE

ELF Research Methods and Approaches to Data and Analyses Theoretical and Methodological Underpinnings Edited by Kumiko Murata

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, Kumiko Murata; individual chapters, the contributors The right of Kumiko Murata to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-­in-­Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-­in-­Publication Data A catalog record for this book has been requested ISBN: 978-­0-­367-­89879-­3 (hbk) ISBN: 978-­1-­003-­02165-­0 (ebk) Typeset in Times New Roman by Apex CoVantage, LLC

For the three ‘founding mothers’ of the research field of ELF, Jennifer Jenkins, Anna Mauranen and Barbara Seidlhofer, who have pioneered this exciting and important field and greatly contributed to its development, both theoretically and methodologically, exploring research methods, producing rich results and inspiring a great number of students and scholars worldwide, including, of course, the contributors to this book.

Contents

List of figuresx List of tablesxi List of contributorsxii Acknowledgementsxv   1 ELF research methods and different approaches to data and analyses: introduction

1

KUMIKO MURATA

PART I

ELF research: some theoretical underpinnings and development

19

  2 Research perspectives on ELF: linguistic usage and communicative use

21

HENRY WIDDOWSON

  3 Researching ELF communication: focus on high-­stakes encounters

29

BARBARA SEIDLHOFER

  4 ELF and translanguaging: covert and overt resources in a transnational workplace

38

ALESSIA COGO

PART II

Spoken and written corpora in ELF research55   5 Good texts in non-­standard English: ELF and academic writing ANNA MAURANEN

57

viii  Contents   6 Exploring diachronic changes in research articles from an ELF perspective: a small corpus study

81

SAYAKO MASWANA

  7 Tracing the emergence of situational multilingual practices in a BELF meeting: micro-­diachronic analysis and implications of corpus design

97

MARIE-­L UISE PITZL

  8 Asian Corpus of English (ACE): features and applications

126

LIXUN WANG

  9 Interpersonal formulaic sequences in ELF academic lectures: methodological challenges

143

YING WANG

PART III

Conversation analytic approaches to ELF interactions

159

10 Applying conversation analysis to ELF interaction data

161

JAGDISH KAUR

11 Analysing multilingual/lingua franca interactions using conversation analysis: notes on transcription and representability

179

KAISA S. PIETIKÄINEN

12 Accounting for asymmetries in ELF interactions: three disparate Asian settings

197

ALAN THOMPSON

PART IV

Narrative approaches in ELF research: exploring attitudes towards ELF

219

13 ‘Place-­reflexivity’ as an imaginary kaleidoscope to explore methodological issues in ELF research

221

MASUKO MIYAHARA

Contents ix 14 Understanding the ELF phenomenon through narrative inquiry: a diary study on identities of Japanese ELF users

241

YOKO NOGAMI

15 Rigour in ELF language attitude research: an example of a conversational interview study

258

TOMOKAZU ISHIKAWA

Index

276

Figures

  6.1 Title structures in 1995 and 2015   7.1 Schematic representation of Individual Multilingual Repertoires (IMRs) and Multilingual Resource Pool (MRP) of the bilateral TIG in PBmtg3   7.2 Number of utterances per speaker (PBmtg3; n = 4374): holistic view   7.3 Number of utterances per speaker (PBmtg3; n = 4,204): micro-­diachronic view (9-­minute segments)   7.4 Number of L1/LN tags per speaker (PBmtg3; n = 179): holistic view   7.5 Functions of L1/LN elements in PBmtg3 (n = 179): holistic view   7.6 L1 tags used in side sequences in PBmtg3 (n = 131): micro-­diachronic view   7.7 Gaps in transcription due to L1 side sequences (PBmtg3): micro-­diachronic view (9-­minute segments)   8.1 Screenshot of the ACE online homepage   8.2 Screenshot of the Browse Corpus interface   8.3 Screenshot of concordances for “yeah” in the Leisure sub-­section of ACE   9.1 Distribution of the four broad functional categories in the sub-­corpora 13.1 Conceptualization of language learner, to user, and developing a sense of ownership

88 107 107 108 111 117 118 119 131 131 132 148 237

Tables



6.1 6.2 6.3 7.1 7.2 7.3 8.1 8.2

9.1 9.2 9.3 9.4 12.1 12.2 12.3 12.4 12.5 15.1

Overview of the introduction corpora Percentage of verb tenses and modals per move Number of words per sentence ‘Foreign words’ in the professional domains of VOICE ‘Foreign words’ in VOICE business meetings Overview of non-­*English elements (L1/LN tags) in PBmtg3 Top 20 most frequent action/mental verbs in ACE Everybody/everyone/nobody/he/she/it + verb (non-­marking vs. marking of the 3rd person singular in ACE) Data used for the study Distribution of interpersonal FSs across disciplines (normalised frequency per 1,000 words in brackets) Congruency with the IDIOM Search results Types of FSs that are not identified by IDIOM Search The three settings investigated, event types, and nationalities of participants Structure of clauses in S3’s speech in Excerpt 4 Most frequent modality markers in the corpus subset Use of the modality markers kind of/kinda and like arranged by decreasing participation by speaker Use of the modality markers kind of/kinda and like by two individuals, comparing dominant participation with peripheral participation Speech function moves

85 86 87 101 104 110 134 134 145 148 152 152 201 207 211 212 212 264

Contributors

Alessia Cogo is Senior Lecturer in applied linguistics at Goldsmiths, University of London, where she is director of the MA in Multilingualism, Linguistics and Education. She supervises PhD students in Global Englishes, English as a Lingua Franca (ELF) and implications for language teaching. She is Editor-­ in-­Chief of ELT Journal. Her current research concerns ELF multilingual practices in professional, academic and migration contexts. Her latest projects concern critical pedagogy and English-­medium instruction (EMI) policy and practices. Tomokazu Ishikawa is Assistant Professor at Tamagawa University’s Center for English as a Lingua Franca and a postdoctoral member at the University of Southampton’s Centre for Global Englishes. He has published on ELF mainly in relation to English within multilingualism, language attitudes and ideologies and English language teaching. Jagdish Kaur is Senior Lecturer at the Department of English Language, Faculty of Languages and Linguistics, University of Malaya. She conducts research on English as a Lingua Franca, intercultural pragmatics and English-­medium instruction. She has published her work in journals such as World Englishes, Journal of Pragmatics, Intercultural Pragmatics and Text&Talk. Sayako Maswana  is Associate Professor of  English education at Tokyo University of Science, Japan. She is currently conducting research on academic writing and  genre analysis. She has published articles in journals such as the Journal of Asia TEFL, Ampersand and the Asian ESP Journal. Anna Mauranen is Professor of English, University of Helsinki. Research foci include ELF, language processing, corpus linguistics and academic discourses. She is Co-­editor of Applied Linguistics, founding co-­editor of Journal of English as a Lingua Franca (JELF), director of ELFA and WrELFA corpora. Her current project is on processing: “Chunking in language: units of meaning and processing (CLUMP)”. Major publications include Exploring ELF (2012). Masuko Miyahara is a tenured lecturer in the English for Liberal Arts Program, International Christian University, Tokyo. She holds an MA in TESOL and a

Contributors xiii PhD from the Institute of Education, University of London. She served as an honorary associate professor for IOE (Institute of Education, University of London) in 2018. Her research interests include learner/teacher identities, emotions, autonomy and EMI as well as issues on research methodology. She has published on such topics including her book Emerging Self-­identities and Emotion in Foreign Language Learning: A Narrative-­oriented Approach (2015). Kumiko Murata is a retired professor, formerly with the School of Education and Graduate School of Education in the Faculty of Education and Integrated Arts and Sciences, Waseda University, Japan. Her research interests include ELF, conversation and discourse analyses, pragmatics, intercultural communication and language teaching. Her recent edited books include Exploring ELF in Japanese Academic and Business Contexts (Routledge, 2016) and English-­ Medium Instruction from an English as a Lingua Franca Perspective: Exploring the higher education context (Routledge, 2019). Yoko Nogami is Assistant Professor at Kwansei Gakuin University, Japan. She is the author of Identity and Pragmatic Language Use: A Study of Japanese English as a Lingua Franca Users (forthcoming) and a co-­editor of JACET ELF SIG Journal. Kaisa S. Pietikäinen received her PhD from the University of Helsinki and is currently working as Postdoctoral Fellow at NHH Norwegian School of Economics. She is passionate about ELF-­CA inquiry and has recently proposed a Special Issue on the topic to the Journal of Pragmatics. Her research focuses on ELF pragmatics in the most intimate, long-­term context possible: couples and families. Marie-­Luise Pitzl is one of the creators of Vienna-­Oxford International Corpus of English (VOICE) and has published widely on many ELF topics (e.g. Creativity in English as a Lingua Franca: Idiom and Metaphor, 2018). Her current research project “English as a lingua franca in Transient International Groups” (FWF V747) seeks to develop conceptual and methodological tools for describing the micro-­diachronic development of ELF communication in (short-­lived) group interaction. Barbara Seidlhofer is Professor of English and Applied Linguistics at the University of Vienna. Her research and teaching focus on English as an international language, intercultural communication and multilingualism and their implications for language teacher education. She is the founding director of the freely available Vienna-­Oxford International Corpus of English (VOICE) and author of Understanding English as a Lingua Franca. She is past editor of the International Journal of Applied Linguistics and founding editor of the Journal of English as a Lingua Franca. Alan Thompson has a PhD from the University of Toronto and teaches at the Faculty of Foreign Languages, Gifu Shotoku Gakuen University, in central Japan.

xiv  Contributors His main research interests are in language contact and change, translation and the uses of literature in English language learning. Lixun Wang is Associate Professor in the Department of Linguistics and Modern Language Studies at the Education University of Hong Kong. His major research interests include corpus linguistics, computer-­assisted language learning and multilingual education. He is one of the leaders of the Asian Corpus of English (ACE) project. Ying Wang is Senior Lecturer (docent) in English at Karlstad University, Sweden, where she teaches academic writing and applied linguistics courses. Her main research interests lie in areas of corpus linguistics and EAP, with a focus on formulaic language use in spoken and written academic discourse. Henry Widdowson was a lecturer at the University of Indonesia and an English language officer with the British Council in Sri Lanka and Bangladesh before becoming a lecturer in applied linguistics at the University of Edinburgh in 1979. He subsequently held professorial positions at the Universities of London, Essex and Vienna. He was a founding editor of the journal Applied Linguistics and for 30 years acted as applied linguistics adviser to Oxford University Press. He had lectured and written extensively on applied linguistics, discourse analysis and language teaching, and his more recent publications include Defining Issues in English Language Teaching (2003), Text. Context, Pretext (2004) and Discourse Analysis (2007) – a book in the Oxford Introductions to Language Study series, of which he was editor. He was Honorary Professor at the University of Vienna.

Acknowledgements

I would like to thank Aina Tanaka and Koichi Saito for their assistance in checking and adjusting the format of the manuscripts.

1 ELF research methods and different approaches to data and analyses Introduction Kumiko Murata This edited volume investigates English as a lingua franca (ELF) research from a methodological perspective, focusing on different types of data, methods and approaches to the analyses of data in order to contribute to furthering ELF research and deepening our understanding of ELF, its use, communication and research and, furthermore, that of neighbouring fields. This will be pursued by exploring various ways of conducting ELF research, systematically approaching data and analysing them, which is simultaneously supported with firm theoretical foundations and development and thus, in turn, could contribute to further development and expansion of the field. In the past 20  years or so, the field of ELF and its research have witnessed unprecedented development due mainly to the acceleration of globalisation and the drastic increase of communication across borders at the global level. In this context, English has increasingly been used as a lingua franca among people from diverse linguacultural backgrounds as, often, the only means of communication (see Seidlhofer 2011 for the definition of ELF among many others). Meanwhile, it has also been revealed that there is often a huge gap between what people have learnt in their English language classrooms and what they often encounter in their actual ELF communication (see, for example, Murata 2016b, 2019a, 2019b; Murata and Iino 2018; Otsu 2017). To better equip our students and future ELF users with the necessary communicative capability (Seidlhofer 2011; Widdowson 2003, 2012, 2016), it is of vital importance for researchers to investigate what is actually happening in lingua franca communication as well as lingua franca users’ attitudes towards its use. A lingua franca in this situation is often reported to be English in the current globalised world (Graddol 2006; Jenkins 2019; Mauranen 2016, this volume; Seidlhofer 2016, this volume), although depending on local conditions of interaction and accessibility to ELF, other lingua francas could also be used (see Kubota 2016; García and Li 2014, for example). To investigate this situation, ELF scholars started conducting research on ELF communication prompted by Seidlhofer’s (2001) call at an early stage of the development of ELF research for the need for more descriptive analyses based on data. This has resulted in research deploying mainly three different methods and approaches: corpus-­based analysis (see Kirkpatrick 2010, 2016; Mauranen 2012, 2016, this volume; Seidlhofer 2004, 2009, 2011 among others), which includes

2  Kumiko Murata the analyses of both spoken and written corpora; conversation analysis (CA) (see, for example, works by Cogo, Kaur, Konakahara, Matsumoto, O’Neal and Pietikäinen), whose analyses are based on recordings of interaction in various settings, including business, workplace, academic, in particular, often in EMI (English-­ medium instruction) classrooms and informal meetings at more specific levels; and thirdly, the narrative approach, which often bases its enquiry on questionnaire surveys and interviews (Ishikawa 2017, this volume; Jenkins 2007, 2014, 2019; Takino 2016; Konakahara et al. 2019; Murata et al. 2017, 2018, 2019; Nogami 2019, this volume), including also e-­mail-­based diary studies (see Nogami, this volume, and also Miyahara, this volume, for e-­mail-­based communication), thus often leading to the incorporation of new technology in its investigation. It is therefore timely at this stage, when ELF-­related research is flourishing and also becoming influential on the development of research in neighbouring fields of applied linguistics, such as language policy and planning, English language teaching (ELT), second language acquisition (SLA) and testing, to name a few, to overview major ELF research methods so far deployed as well as types of data frequently used in investigating ELF, ELF communication and related fields. Furthermore, the investigation also includes exploring various types of interactions in both business/workplace and academic settings as well as attitudes towards English and ELF use and users’ identity construction in the process of using English and English as a lingua franca. Accordingly, this enquiry will benefit both ELF scholars and those in neighbouring fields. This volume thus explores the aforementioned research methods and investigates the extent to which they are effective as well as what challenges have arisen in their use by also discussing the frameworks and theoretical underpinnings, simultaneously introducing case studies by ELF scholars who have employed these methods in their specific ELF research. For this purpose, the volume is divided into four parts. Part I, entitled ELF research: some theoretical underpinnings and development, explores theoretical backgrounds and recent development of ELF research, including the phenomenon of translanguaging in multilingual settings, which is increasingly attracting researchers’ attention, while Part II, Spoken and written corpora in ELF research, investigates ELF corpus data, both spoken and written, how to compile data to start with and then the deployment of these data to conduct corpus-­based research. Part III, Conversation analytic approaches to ELF interactions, explores the possibility of CA in analysing talk-­in-­interaction in ELF research, and finally, Part IV, Narrative approaches in ELF research: exploring attitudes towards ELF, introduces various narrative-­based approaches, which are frequently used in investigating ELF users’ attitudes towards English, ELF and its communication as well as construction of identity in relation to them. Despite the categorisation presented presented, however, clear demarcation between the deployment of different approaches and methods is often difficult as some researchers base their research and analyses on multidisciplinary or mixed research methods and approaches (see, for example, Cogo, Pitzl and Thompson, this volume, who incorporate both corpus-­based (Pitzl, Thompson)

ELF research methods and approaches  3 and conversation analytic approaches as well as ethnographic observation and, in the case of Thompson (and Y. Wang, this volume, for that matter), the perspective of Hallidayan systemic functional grammar (SFG), in their contributions. Accordingly, it is fair to say that the categorisation is based on the different weight placed on disparate approaches in respective contributions. In what follows, each of the four parts (Parts I–IV) will be discussed in a slightly more detailed manner. Part I, ELF research: some theoretical underpinnings and development, aims to explore the theoretical background and recent development of ELF research. Nearly 20  years after Seidlhofer (2001) voiced her opinion about the need for ELF descriptive research, research is now flourishing not only in the European context, where the first corpus-­based ELF research projects, such as VOICE (Vienna Oxford International Corpus of English) (Seidlhofer 2004, 2011, see also Pitzl, this volume) and ELFA (English as a Lingua Franca in Academic Settings) (Mauranen 2006, 2012) were initiated, but also in the Asian context, where a major project such as ACE (Asian Corpus of English, Kirkpatrick 2010, 2012, 2016, see also L. Wang, this volume) as well as various other projects, albeit on a smaller scale, were conducted or are currently being undertaken. Under these circumstances, it is relevant to further deepen our understanding of the meaning of ELF research as well as its role, possibility of and implications for investigating real-­world issues and language pedagogy. These are the themes to be pursued in Part I of this volume, and it contains three contributions by Henry Widdowson, Barbara Seidlhofer and Alessia Cogo. The first contribution by Henry Widdowson overviews major ELF research so far conducted by utilising his own original and firmly established categorisation of use and usage (see Widdowson 1978) and classifying ELF research into two types of use and usage-­based. Although admitting that both types of research are of great value, Widdowson maintains that the value of ELF research in the recent globalised world resides more on the former, that is, investigation of how people use ELF for the solution of real-­world problems, which, according to Widdowson, is a more applied linguistic concern, rather than linguistic, with which the latter, i.e., usage-­based one, is more concerned. Seidlhofer in the following chapter further explores this use-­based, more applied linguistic-­concerned research by discussing the possibility and meaning of ELF research in the real-­world settings, particularly focusing on ‘high-­stakes’ gate-­keeping encounters, where one party holds more power over the other. On the other hand, Cogo, using the framework of CA, analyses her own business ELF interaction data in Europe and explicates the relationship between multilingualism and ELF, in particular, translanguaging among multilingual ELF users, which is recently attracting great attention (see also Cogo 2012; Jenkins 2015; Hülmbauer and Seidlhofer 2013). The three contributions in Part I thus enhance our understanding of the nature of ELF and its research, while at the same time demonstrating the possibility of its further development on real-­world issues in the globalised world. Upon understanding the nature of ELF, its research and recent development, Parts II–IV investigate different research methods and approaches to data and their analyses in conducting research.

4  Kumiko Murata Part II entitled Spoken and written corpora in ELF research investigates ELF corpora and their role in conducting ELF research. As already acknowledged in various ELF-­related publications, large-­scale ELF-­specific spoken corpora developed at Vienna (VOICE) and Helsinki (ELFA, and most recently WrELFA for written ELF) played a pivotal role in the development of ELF research and its conceptualisation (see Mauranen 2012, 2016, 2018; Seidlhofer 2011, 2016 among many others). Furthermore, following VOICE, an Asian version of ELF corpus, ACE has also been developed (see Kirkpatrick 2012, 2016; L. Wang, this volume). In what follows, the nature of spoken and written corpus research, starting with the latter, will be briefly discussed. Written corpus data are represented by Mauranen and her colleagues’ WrELFA. Written ELF has hitherto been less explored compared to spoken ELF, but the exploration of written corpora is particularly important and pertinent in academic settings, where, with the increasing mobility of students and academics in the globalised world, more and more of them are writing their term papers, dissertations, theses and research papers and publishing their articles in international journals in English used as a lingua franca (see D’Angelo 2017; Jenkins 2014; Kuteeva 2017; Kuteeva and Mauranen 2014; Mauranen 2012, this volume; Murata 2016a; Turner 2017, 2018 among many others). Written ELF, compared with spoken ELF, is less researched partly because written language is still very much based on the norms of standard language (see, for example, Kuteeva 2017; Turner 2017) and thus also strongly influenced by standard language ideology (see Seidlhofer 2018). Another reason is because written communication is less interactive compared to spoken one (but see, for example, Widdowson 1975, 1979, 1984), ELF researchers have traditionally investigated negotiation of meaning more in face-­to-­face communication (see, for example, Cogo 2009, 2012; Ehrenreich 2009, 2011; Jenkins 2015; Kaur 2009, 2011a, 2011b, 2016; Konakahara 2016; Mauranen 2012; Seidlhofer 2011) than the written one. This difference, however, is becoming increasingly blurred due to the recent advancement of technological innovation, some written electronic communication simultaneously showing characteristics of both spoken and written language. The standard language ideology has strongly been maintained until recently, particularly in international academic journals, although very slowly it has started showing certain changes, in particular, in the field of science (see D’Angelo 2017; Kuteeva 2017; Maswana and Mauranen, both this volume). The increase of international scholars contributing to publications has greatly impacted this change, which is partly due to the encouragement by ‘international’ universities of their publishing in international journals to survive global competition. This is often assessed by the global ranking of their academic achievement, the measurement of which includes the number of publications. The second contributor in Part II, Maswana, investigates the effect of these changes observed in international journals in two different fields of economics and medicine at specific periods of time. Similar changes, however, are less reported to be taking place in the practice of academic writing at international universities as reported, for example, by Jenkins (2014, 2019) and also by Sadoshima (2017) as well as Turner (2017, 2018).

ELF research methods and approaches 5 The latter half of Part II on spoken corpora focuses on three spoken corpora: VOICE and ACE, represented by Pitzl and L. Wang’s contributions, respectively, and ELFA, which is a corpus compiled by Mauranen and her colleagues at Finnish universities, and utilised here by Y. Wang for her analysis. Pitzl’s paper demonstrates the ways in which the data from VOICE could be deployed in the analysis of multilingual practices among multinational business people, using business data from VOICE and introducing its full use for the description of ‘what is happening’ in business negotiation both qualitatively and quantitatively to draw its comprehensive picture. Pitzl thus highlights what is possible with the use of the corpus simultaneously introducing the nature and characteristics of the pioneering VOICE project, in which Pitzl played an important part as one of the project’s full-­time researchers. On the other hand, the contribution by L. Wang, who was in charge of the technical side of the compilation of ACE, which is an Asian version of VOICE and compiled in the same manner as VOICE to make the two corpora comparable (see Kirkpatrick 2012, 2016, L. Wang, this volume), concentrates more on the introduction of the technical side of designing, compiling and running the corpus as well as how to access it, which, in turn, is useful for also understanding the design of VOICE, the two being compiled in a similar manner. Furthermore, L. Wang also conducts a small scale study himself to demonstrate the ways in which the corpus data can be deployed on the basis of some earlier findings of the VOICE project, which reported some lexico-­grammatical aspects of ELF interaction, carrying out the comparison of some features of ELF use between the two, illustrating how the both corpora can be comparable. The final contribution regarding spoken ELF-­based research is represented by Y. Wang’s work, who deploys the ELFA corpus in her investigation of the formulaicity of spoken academic ELF communication. Thus, Part II extensively explores the possibility of ELF-­related corpora, both spoken and written, as well as methods of deploying these data, which cover both more typical corpus linguistic-­based approach, that is, what Widdowson (this volume) terms a usage-­based and more socially situated use-­based approach, which leads to implications for such high-­stakes encounters as international publication and multinational business encounters in the increasingly globalised world (see Seidlhofer, this volume). Spoken ELF is also investigated in smaller-­scale particularised corpora of talk-­in-­interaction, for the analysis of which a CA perspective is employed, and will be discussed in Part III, Conversation analytic approaches to ELF interactions. A CA perspective is adopted by a substantial number of ELF scholars for detailed description of ELF face to face interactions (Cogo 2009, 2012, 2018; Cogo and Dewey 2012; Firth 1996, 2009; House 2003, 2016; Klimpfinger 2009; Kaur 2009, 2011a, 2011b, 2016, this volume, Konakahara 2016; Matsumoto 2014, 2018; O’Neal 2019; Pietikäinen 2014, 2018, this volume) and has produced fruitful results in revealing what is going on in ELF interaction in a detailed manner from an emic perspective. Part III contains both methodological issues on the analysis and description of ELF interactions, and actual examples of the analyses, using small-­scale particularised conversational corpora collected by each researcher for their respective research purposes. Kaur, Pietikäinen and Thompson’s corpora all

6  Kumiko Murata fall under this category, Kaur’s, in academic settings, Thompson’s, mainly in business and workplace settings, but also including a corpus from a business-­related EMI setting, while Pietikäinen’s data come from private ELF couples’ talks. CA is a powerful tool to depict the local context of actual face-­to-­face talk-­in-­interaction from the participants’ emic perspective (Sacks et al. 1974; Hutchby and Wooffitt 2008), not only in NES talks in which it originated but also in ELF interactions. It can delineate participants’ moment by moment negotiation of meaning in its local context, which is pertinent to the description of ELF interactions, where people from diverse linguacultural backgrounds negotiate meaning for achieving understanding and their specific communicative goals, while utilising various multilingual and multimodal resources available. The first contributor in Part III, Kaur, discusses the role and efficacy of CA in ELF research in general, reviewing ELF research based on CA to date as well as more recent development of CA-­based ELF research, which often incorporates both multimodal and multilingual perspectives (see Konakahara 2016; Matsumoto 2018, for example), also providing some examples from her own data, while the second contributor, Pietikäinen, explores an important but not necessarily extensively discussed issue of transcribing ELF/multilingual lingua franca data from a CA perspective. On the other hand, Thompson, on the basis of three different workplace and business EMI-­related academic setting data in Asian contexts, demonstrates how conversation analytic approaches combined with the use of analytical categories from Hallidayan systemic functional grammar (SFG) could effectively and in a detailed manner describe what’s happening in communication among ELF speakers. More specifically, Thompson investigates various strategies used by the participants in these different settings in securing understanding and enhancing group solidarity. Another type of approach, which is increasingly used in exploring attitudes towards and beliefs about ELF, and ELF users’ identity, is narrative enquiry. Part IV is entitled Narrative approaches in ELF research: exploring attitudes towards ELF. Narrative approaches are increasingly utilised recently by ELF researchers to obtain ELF users’ own accounts for their attitudes, beliefs in using ELF and also their identity construction in ELF interaction (see, for example, Jenkins 2007, 2009, 2014, 2019; Iino and Murata 2016; Konakahara et al. 2019; Murata et al. 2017, 2018, 2019; Takino 2016). They also owe greatly to earlier studies utilising the notion of ‘community of practice’ (Lave and Wenger 1991; Wenger 1998; Ehrenreich 2018) and relatively recent qualitative SLA research, mainly sociocultural approach (Norton 2000; Norton and Early 2011). For this type of enquiry in Part IV, the data are mainly collected through interviews (Ishikawa and Miyahara), questionnaire (Ishikawa) and diary study via e-­mail exchanges (Nogami, and also partly used by Miyahara). Here, the use of new technology along with narrative enquiry is noteworthy, opening a new dimension to ELF research. Through detailed analyses of the data, these scholars have successfully elicited information on ELF users’ experiences in language learning and using and their attitudes towards and beliefs about ‘English’ and ELF use together with the information on their identity formation in the process.

ELF research methods and approaches  7 As briefly introduced earlier, these mainly three types of methods and approaches to data are typically used in ELF research. However, as already touched on, some scholars utilise not only one of these approaches exclusively but also combine different methods and approaches in their effort to comprehensively describe what is actually happening in ELF interaction in their specific context. To further investigate this, the following section details how it is put into practice by introducing each of these contributions to the volume in a more detailed manner in order of the four sub-­divisions discussed earlier.

Contributions in the volume As stated in the preceding section, this volume is divided into four parts, the first one, entitled ELF Research: some theoretical underpinnings and development includes three papers by Widdowson, Seidlhofer and Cogo. Widdowson, the first contributor, under the title of Research perspectives on ELF: linguistic usage and communicative use sets the tone for the whole volume, which investigates ELF research methods in addition to its theoretical underpinnings and recent development. In so doing Widdowson revisits his original and insightful binary notions of ‘use’ and ‘usage’ (see Widdowson 1978, among others), which have inspired communicatively oriented ELT (English language teaching) practitioners for the past 40  years, and explicates that ELF research can be approached from both ‘usage’ and ‘use’ perspectives. The former is represented, for example, by corpus linguistic approach (but see Pitzl, this volume, who deploys corpus data for the investigation of detailed business interaction), which is more linguistically orientated research, whereas the latter, according to Widdowson, is related more to the investigation of real-­world issues, which are concerns of applied linguistic research. Widdowson maintains that the focus of ELF research should not only be placed on forms but also on how people ‘use’ it in this globalised world to solve real-­world problems, which, Seidlhofer, the second contributor in Part I, directly and in a more detailed manner discusses. Seidlhofer expands the paradigm of ELF research into ‘real-­world’ problems, which, Widdowson in the preceding chapter clearly states, are of applied linguistic concerns. Seidlhofer first points out the fact that ELF research contexts have up to now tended to cover the contexts which are relatively ‘consensual’ by listing the contexts in which her own VOICE and the Asian equivalent of ACE data were collected and calls for the need for collecting and analysing more diverse data in the future research. She then moves on to exploring possible current and future contributions of ELF research to broaden its scope to social issues and how ELF plays (or is playing) a critical role in ‘high-­stakes’ gate-­keeping encounters such as the one involving asylum seekers, where both asylum seekers and immigration officials use ELF extensively as a means of communication (see Guido 2012, for example). This is followed by the discussion about cases of international publishing as well as teaching and testing, which can also play the role of gate-­keeping for academics from diverse linguacultural backgrounds and learners for their academic achievement. Accordingly, both Widdowson and Seidlhofer pave the way

8  Kumiko Murata for new directions of future ELF research as representative of applied linguistic enterprise, which pursues the quest for solving ‘real-­world’ issues. On the other hand, Cogo, the third contributor in Part I, explicates the necessity of reconceptualising ELF as multilingual practices, introducing the concepts covert and overt practices, the former being, as the term indicates, more covertly exploited practices in the form of, for example, translation from one resource language to another, which often retains some features of the original code in the new one. On the other hand, overt practices include translanguaging practice, in which multilinguals freely exploit their multilingual resources when communicating in ELF, and which, according to Cogo, also incorporates more traditionally termed code-­switching, thus quite often the use of multilingual resources being obvious (see also Cogo 2012; Li 2016, 2018; Garcia and Li 2014). Cogo illustrates, listing some examples from a European multilingual business setting, how translanguaging practices are commonly deployed, while at the same time explicating that the participants of her research also intentionally exploit the ideology of separate languages to exert their power, at times monolingually deploying their multilingual resources. Thus, Cogo, in this chapter illustrates the ways in which the analysis of ELF interaction could more explicitly be located in a multilingual paradigm. This is an intriguing observation, and the paradigm has potential for furthering research in multilingual practice in ELF communication. Parts II-­IV discuss mainly three types of ELF research methods, namely corpus-­based, CA and narrative enquiry, together with illustrative examples of case studies deploying respective methods. In the first contribution in Part II, Mauranen, on the basis of her expertise in compiling spoken ELF in academic settings (ELFA), explicates the compilation of the first written ELF corpus in academic settings (WrELFA). In this contribution, Mauranen discusses WrELFA from three perspectives: macro-­social, micro-­social and cognitive. The macro-­ social perspective includes the discussion of historical backgrounds as well as the recent changing nature of publication due to open-­access and digitalisation, which, Mauranen asserts, makes it very difficult for native English speakers (NESs) to control the language. In discussing the second perspective, micro-­ social, Mauranen states, referring also to Hynninen (2016), that ‘spontaneous language regulation’ is also found in written text, and what is interesting is that she further explicates that specific research ‘communities negotiate their own norms’, not necessarily being regulated by NS ones, although they ‘do not stray far from standard English’. By introducing two written examples from her corpus, Mauranen persuasively discusses what is meant by ‘good texts’ in successful writing. In the third perspective, the cognitive one, Mauranen demonstrates some evidence of ‘approximation’ and ‘fixing’ with ample examples from her WrELFA and illustrates how some changes of use (or rather usage according to Widdowson’s categorisation this volume), might already be taking place. This, Mauranen explicates, will be further enhanced by micro-­interaction and macro-­social level aspects. Supported by evidence from these three different perspectives, Mauranen reiterates that effective writing does not mean achieving NS standards. That is, changes take place not led by NES users but more by discourse communities of

ELF research methods and approaches  9 academics promoted also by transdisciplinary needs of scientific investigation and digitalisation. The second contributor in Part II, Maswana, primarily on the basis of ESP (English for Specific Purposes) genre-­specific analysis, or more specifically, taking an approach informed by EAP (English for Academic Purposes), also investigates the issue of writing from an ELF perspective, examining some characteristics of articles published in international journals. In so doing, she focuses on two different disciplinary fields of medicine and economics in 2 specific years of 1995 and 2015, considering the changes in academic and social environments in these 20 intervening years. Maswana investigates, in particular, Introductions and Titles of each article, examining such features as the changes in rhetorical organisation, word count, language structure, the use of tense and modals, and has found that more changes are in general detected in medical journals in the 2 different years. This, Maswana asserts, is due to an increase of the number of scholars from various linguacultural backgrounds, co-­authoring and contributing to the journals during the period. In this context, what Maswana has found is that internationally published articles particularly in the field of medicine seem to have been establishing their own disciplinary standard not necessarily rigidly constrained by native speaker (NS) norms, although this newly appropriated standard, according to Maswana, seems to have NES standards as its basis. Accordingly, the field of academic writing seems to be heading for the establishment of discipline-­specific standard of its own with the increase of ELF writers, irrespective of nativeness of the standards (see also Kuteeva 2017; Mauranen 2018, this volume; Turner 2017). As stated in the preceding section, Part II is divided into two types of ELF corpora, spoken and written. The second part of Part II on spoken corpora includes three contributions by Pitzl, L. Wang and Y. Wang. The first two introduce large-­ scale ELF corpora, VOICE and ACE respectively, while the third contribution by Y. Wang demonstrates the deployment of the ELFA corpus compiled by Mauranen and her colleagues and is widely available on their website. The first of these, Pitzl’s contribution, is firmly based on the data from VOICE and is an excellent example of how a corpus can be deployed in exploring multilingual practices among business people from different linguacultural backgrounds, using English as a lingua franca, while simultaneously utilising their own multilingual resources. The corpus annotation system, for example, enables readers to overview the whole lengthy interaction between two different L1 (German and Korean) groups of business people communicating for business purposes, when there is no shared LF apart from English. It has also made it possible to list what is happening in the phase of respective L1 communication diachronically and also at individual participatory level. Furthermore, Pitzl also deploys a CA perspective, which has enabled her to illustrate an overview of the use of multilingual resources in the interaction in detail. Thus, the use of both qualitative and quantitative analyses is very effective. By introducing what she calls micro-­diachronic analysis, Pitzl succeeds in how multilinguals negotiate the use of their L1 in BELF communication as time passes by, establishing context specific multilingual practices (or what she terms ‘etiquette’) in the use of multilingual resources, securing the other party’s comprehension. The analysis is thus very

10  Kumiko Murata comprehensive, integrating both corpus-­based quantitative analysis and a CA-­based qualitative approach, exploring the possibility of extending and combining methods to explore multilingual TIG (Transient International Groups, see Pitzl 2018) interactions, deploying the corpus data in full and capturing what is happening in the extended interactions, questioning simultaneously traditional linguistic research, which quite often starts with pre-­existing categorisation and conceptualisation, classifying analysed data accordingly. Pitzl’s contribution is an excellent example of the recent development in the theorisation of ELF from a multilingual paradigm, which has resonance with Cogo’s contribution, both of them simultaneously deploying a CA perspective in analysing their multilingual data. The fourth contribution of Part II by L. Wang, on the other hand, concentrates on the explication of the ACE project. It is a very practical, a step by step introduction of ACE. Accordingly, by reading this paper readers will be thoroughly familiarised with the characteristics of ACE and how it can be accessed and utilised for possible future research. It also means that even readers who are not familiar with VOICE will also be familiarised with it by reading the detailed description of ACE, as the latter is compiled to produce an Asian version of VOICE. Furthermore, L. Wang also briefly demonstrates what kind of corpus linguistic research is possible utilising ACE by conducting his own small-­scale survey on the basis of some findings from the early phase of VOICE-­based research, comparing some lexico-­grammatical features found in the VOICE data with those from ACE. While listing some features common to both corpora, L. Wang has also identified some ACE specific features of lexico-­grammatical usage. Although ELF research has moved on since its earlier phase of research and its interest, recognising more dynamic and fluid nature of ELF communication and thus focusing more on the analysis of actual interaction and the negotiation of meanings, L. Wang’s demonstration, which follows the tradition of a typical corpus linguistic-­based approach, is useful especially for those who are more interested in investigating formal linguistic features of ELF usage. Furthermore, the contribution is also useful in understanding how to compile this kind of corpus, offering information not only on ACE, but also on VOICE. It thus offers invaluable information to those who are planning to compile a similar kind of ELF corpus themselves. Both VOICE and ACE contain one million words in their respective corpora, they can therefore be classified as large-­scale ELF corpora. Another spoken ELF corpus containing one million words is ELFA, which has been compiled by Mauranen and her colleagues, prior to WrELFA introduced earlier in this chapter. The final contribution in Part II by Y. Wang utilises this corpus and examines linguistic features of ELF lectures in European contexts, paying special attention to the use of formulaic sequences (FSs) in three broad disciplinary domains  – social sciences (SS), natural sciences (NS) and medicine – in nine lectures extracted from the ELFA corpus. Given scarce research into FSs in spoken academic discourse in ELF contexts, Y. Wang scrutinises frequency and distribution of four types of FSs with interpersonal functions, namely modality, evaluation, commitment and engagement across the three disciplines, based on Halliday’s (2014) SFG classification. The analysis was conducted by using a computer-­assisted ‘manual’

ELF research methods and approaches  11 approach. This enabled her to explore formal variations in potential FSs, which otherwise would have been overlooked if an automatic identification had been adopted. Although the corpus for the current study is relatively small, interesting difference in patterns in the use of what Y. Wang terms ‘interpersonal’ FSs is found, in particular, between SS and NS, and medicine. The results also resonate with Maswana’s (this volume) study on the comparison of written academic discourse in the two disciplines of economics and medicine, the latter showing distinct difference from economics. Y. Wang’s study thus provides an invaluable insight into the use of FSs in the academic ELF lectures, revealing cross-­disciplinary differences and similarities as well as features specific to spoken academic discourse and those in ELF lectures. Although more investigation is necessary, particularly to further explore ELF academic discourse, it is a promising field for future research. Methodologically as well, Y. Wang’s claim that relying only on the frequency count is not satisfactory is convincing. Among the five contributors in Part II on corpus analytic approach, one of them, Pitzl simultaneously deployed a CA perspective in addition to her corpus-­based approach as briefly stated previously, which will be discussed in a more detailed manner in Part III. Part III, entitled Conversation analytic approaches to ELF interactions, comprises three papers by Kaur, Pietikäinen and Thompson. The first contribution in Part III, Kaur’s, discusses the relevance and appropriateness of CA as methodology for investigating ELF interaction, starting with its characteristics in analysing talk-­in-­interaction, moving on to the discussion and introduction of the recent development of ELF-­related research which has deployed a conversation analytic perspective. It is a comprehensive overview of recent CA-­based ELF research, quoting some concrete examples of the analysis of ELF interaction from existing works to demonstrate what CA approach can offer to deepen our understanding of ELF research, after thoroughly reviewing basic principles of CA and why it is appropriate for analysing and describing ELF interaction data. Furthermore, Kaur simultaneously points out how CA could also benefit from the analysis of ELF interaction, which is being recognised as talk-­in-­interaction among multilinguals, whereas CA has originated in and founded on the analysis of NES-­NES interaction (see Sacks et  al. 1974, among others) and thus is not necessarily equipped with detailed techniques to cope with diverse data. Accordingly, there emerges a challenging issue of how to analyse multilingual franca talk-­in-­interaction, such as ELF interaction, which is the main concern of Pietikäinen in the following chapter. The second contributor in Part III, Pietikäinen, explores the appropriateness and effectiveness of CA’s distinctive method of transcribing recorded data of talk-­in-­interaction, more specifically, multilingual ELF talk-­in-­interaction. She very appropriately points out the limitation of the original method of transcription established by the founding members of CA, such as Sacks et  al. (1974), the informants of whose original analysis having been native English speakers (NESs). In particular, Pietikäinen problematises the inappropriateness and ineffectiveness of the current system of CA transcription based on ‘eye dialect’ in depicting ELF interaction among multilingual participants, which, Pietikäinen asserts, cannot cope with transcribing increasingly diverse communication among

12  Kumiko Murata ELF speakers from different linguacultural backgrounds, and suggests an alternative way of transcribing multilingual ELF data. This is convincingly demonstrated with illustrative examples from the author’s data between multilingual ELF couples, describing the ways in which information on the pronunciation particularities of the interactants from different linguacultural backgrounds leads to deeper and more persuasive analysis. Against this background, the use of some International Phonetic Alphabet (IPA) symbols together with eye dialect is recommended to produce more accurate transcripts of multilingual data, where otherwise varied pronunciation of speakers is not identifiable. The contribution is invaluable and of practical use, considering the increasing use and relevance of CA in analysing ELF talk-­in-­interaction in multilingual settings, which simultaneously satisfies CA practitioners’ needs in general. The final contributor in Part III, Thompson, discusses the results of data analyses from three different workplace and academic settings in Asian contexts, paying attention to asymmetries in interaction in which interactants from various linguacultural backgrounds and social roles communicate with each other. In so doing, Thompson deploys CA combined with Halliday’s SFG categorisation together with a quantitative analysis of corpus-­based investigation of frequency of occurrences of some patterning in the use of modality markers. In addition, he further employs ethnographic observations and note-­taking at the phase of data collection. Thus, Thompson explores the possibility of combining various methods available in comprehensively depicting what is going on in ELF interaction in specific goal-­oriented institutional settings where asymmetries between interactants are negotiated by use of various strategies. This enables him to describe the interactions in a more integrative manner, whereas deploying only CA would have shown some limitations. Thompson thus asserts that more focus should be placed in the diversity of interactants and asymmetries of communication, which the deployment of his mixed methods has made possible. Thus, the three contributions in Part III on CA all emphasise the importance of CA analysts to be aware of and be able to cope with multilingual and diverse realities of ELF interactions. Having explored some theoretical and methodological issues related to CA and highlighted some examples of CA approaches in different contexts, another powerful method of investigation, narrative enquiry, which is also increasingly employed by ELF scholars, particularly by those who work on the issue of attitudes and identity in ELF interaction, will be discussed in Part IV. Part IV, entitled Narrative approaches in ELF research: exploring attitudes towards ELF, includes contributions by Miyahara, Nogami and Ishikawa. The first contributor in Part IV, Miyahara, explores narrative enquiry head-­on supported also by some examples of analysis, paying specific attention to a hitherto relatively disregarded aspect of the influence of researchers’ own reflexivity on the interpretation of participants’ narrative, in particular, focusing on the idea of ‘place-­reflexivity’ in analysing data. The term ‘place’, however, seems a misnomer as it sounds as if it is static entity, whereas what Miyahara does under the term is discussing interaction types which take place in the ‘place’. Despite this, the introduction of the new dimension in the process of interpreting participants’

ELF research methods and approaches  13 narrative acutely reminds us of the nature of narrative as being co-­constructed by both interviewees and interviewers as researchers in situ, the latter directly or indirectly influencing not only the content and direction of narrative itself but also the type and extent of the analysis of the data. Miyahara’s main focus in this chapter is placed more on the investigation of narrative approach itself from this new dimension of reflexivity and its potential as a tool for investigating informants’ attitudes towards English language learning and ELF use. Another contributor who also bases her analysis on users’ voices, albeit from a slightly different perspective and using a different kind of data and method, is Nogami. Nogami, the second contributor of Part IV, utilises a narrative approach in her research on Japanese ELF users’ attitudes and identity in ELF communication, in her case, using ‘longitudinal diary study’. This is carried out mainly through e-­mail exchanges between the author and nine participants, who were mostly students in their 20s and 30s at the time of her research, but some with experiences of using ELF in business contexts. Nogami specifically pays attention to the identity construction of these ELF users in her analysis, simultaneously reviewing the field also from the perspectives of intercultural communication, learner and user identity and the notion of ‘community of practice’ (Lave and Wenger 1991; Wenger 1998). In addition to the review of ELF and identity and its related fields as well as the detailed description of her version of narrative enquiry, Nogami also demonstrates the ways in which the approach could produce intriguing results by presenting one detailed case study from her extended investigation. That is, Nogami introduces one participant’s voice in her identity construction and meaning-­making process in negotiation with a business ELF interactant, which is also insightful from an intercultural perspective. On the other hand, Ishikawa, the third and final contributor in Part IV and in fact the whole volume, investigates Japanese students’ attitudes towards English through a questionnaire survey and unstructured interviews, collecting narrative of ELF learners/users. The methods are successful in eliciting the result which reveals that most of the informants’ attitudes towards Japanese English is negative, reflecting, according to Ishikawa, their mainly North American NES-­norm-­ based English education at secondary level and their not being informed of the use of English as a lingua franca (ELF). The findings from the interviews with the informants, however, simultaneously indicate how some of their attitudes towards English change after the limited space of the interview time in conversation with the researcher, their coming to appreciate the role of ELF and its use. This could be a positive sign which indicates attitudes might change after the provision of appropriate information and education in line with the use of English in the globalised world. Ishikawa’s results thus simultaneously give important implications not only for ELF research methods but also for language pedagogy. Part IV has thus focused on narrative enquiry as a methodology for investigating attitudes towards ELF. The volume accordingly focuses specifically on methodological issues of ELF research, which is timely as ELF is now a firmly established, important area of applied linguistic research, its importance being increasingly recognised also by researchers in neighbouring fields such as SLA and language testing (see, for

14  Kumiko Murata example, McNamara 2012, 2018; Ortega 2018; Shohamy 2019, among others). As the number of researchers in the field increases, it is essential that we consider the further possibility and development of ELF research by reviewing its methods hitherto deployed for different types of ELF research as well as deepening understanding of the theoretical background and its recent development. The volume therefore targets not only scholars and students relatively new to the field, but also established ELF scholars for further advancing and enriching their research and exploring the possibility of the investigation of research goals and targets from new perspectives and approaches in search of a new insight into the ever-­evolving and growing field of ELF and its research. This is even more important with an increasing number of border-­crossing people worldwide often using English as a lingua franca for both their interactional and transactional purposes. To investigate these dynamic and fluid situations and diverse people communicating in ELF, more flexible approaches to data collection and diverse methods for the analyses of data or combination of both are required. It is hoped that this volume contributes to further development of ELF research, theory-­building and pedagogy through the investigation into different methods and approaches to data in ELF research.

References Cogo, A. 2009. Accommodating difference in ELF conversations: A  study of pragmatic strategies. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Cambridge: Cambridge Scholars Publishing, pp. 254–273. Cogo, A. 2012. ELF and super-­diversity: A case study of ELF multilingual practices from a business context. Journal of English as a Lingua Franca 1(2), 287–313. Cogo, A. 2018. ELF and multilingualism. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. Abingdon, Oxon: Routledge, pp. 357–368. Cogo, A. and M. Dewey. 2012. Analysing English as a Lingua Franca: Corpus-­driven investigation. London: Continuum. D’Angelo, J. 2017. Managing an academic journal from an ELF-­informed perspective: The case of Asian Englishes. Waseda Working Papers in ELF 6, 175–186. Ehrenreich, S. 2009. English as a Lingua Franca in multinational corporations. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 126–151. Ehrenreich, S. 2011. The dynamics of English as a business Lingua Franca: A language contact perspective. In A. Archibald, A. Cogo and J. Jenkins (eds), Latest trends in ELF research. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 11–34. Ehrenreich, S. 2018. Communities of practice and English as a Lingua Franca. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. Abingdon, Oxon: Routledge, pp. 37–50. Firth, A. 1996. The discursive accomplishment of E “normality”: On “Lingua Franca” English and conversation analysis. Journal of Pragmatics 26, 237–259. Firth, A. 2009. The Lingua Franca factor. Intercultural Pragmatics 6(2), 147–170. García, O. and Li Wei. 2014. Translanguaging: Language, bilingualism and education. Basingstoke: Palgrave. Graddol, D. 2006. English next. London: British Council.

ELF research methods and approaches  15 Guido, M. 2012. ELF authentication and accommodation strategies in cross-­cultural immigration encounters. Journal of English as a Lingua Franca 1(2), 219–240. Halliday, M.A.K. 2014. Halliday’s introduction to functional grammar (4th ed., revised by Christian M.I.M. Matthiessen). Abingdon, Oxon: Routledge. House, J. 2003. English as a Lingua Franca: A threat to multilingualism? Journal of Sociolinguistics 7(4), 556–578. House, J. 2016. Own language use in academic discourse in English as a Lingua Franca. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts -­conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 59–69. Hülmbauer, C. and B. Seidlhofer. 2013. English as a Lingua Franca in European multilingualism. In A.-­C. Berthoud, F. Grin and G. Lüdi (eds), Exploring the dynamics of multilingualism. Amsterdam: John Benjamins, pp. 387–406. Hutchby, I. and R. Wooffitt. 2008. Conversation analysis (2nd ed.). Cambridge: Polity Press. Hynninen, N. 2016. Language regulation in English as a Lingua Franca. Berlin: de Gruyter Mouton. Iino, M. and K. Murata. 2016. Dynamics of ELF communication in an English-­medium academic context in Japan – from EFL learners to ELF users In K. Murata (ed), Exploring ELF in Japanese academic and business contexts -­conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 111–131. Ishikawa, T. 2017. Japanese university students’ attitudes towards their English: Conversational interview study. Waseda Working Papers in ELF 6, 110–125. Jenkins, J. 2007. English as a Lingua Franca: Attitude and identity. Oxford: Oxford University Press. Jenkins, J. 2009. Exploring attitudes towards English as a Lingua Franca in the East Asian context. In K. Murata and J. Jenkins (eds), Global Englishes in Asian contexts. Abingdon, Oxon: Palgrave Macmillan, pp. 40–58. Jenkins, J. 2014. English as a Lingua Franca in the international university – the politics of academic language policy. Abingdon, Oxon: Routledge. Jenkins, J. 2015. Repositioning English and multilingualism in English as a Lingua Franca. Englishes in Practice 2(3), 49–85. Jenkins, J. 2019. The internationalisation of higher education: But what about its Lingua Franca? In K. Murata (ed), English-­medium instruction from an English as a Lingua Franca perspective -­exploring the higher education context. Abingdon, Oxon: Routledge, pp. 15–31. Kaur, J. 2009. Pre-­empting problems of understanding in English as a Lingua Franca. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Cambridge: Cambridge Scholars Publishing, pp. 107–123. Kaur, J. 2011a. “Doing being a language expert”: The case of the ELF speaker. In A. Archibald, A. Cogo and J. Jenkins (eds), Latest trends in ELF research. Cambridge: Cambridge Scholars Publishing, pp. 53–75. Kaur, J. 2011b. Raising explicitness through self-­repair in English as a Lingua Franca. Journal of Pragmatics 43(11), 2704–2715. Kaur, J. 2016. Using pragmatics strategies for effective ELF communication: Relevance to classroom practice. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts -­conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 240–254. Kirkpatrick, A. 2010. English as a Lingua Franca in ASEAN: A multilingual model. Hong Kong: Hong Kong University Press.

16  Kumiko Murata Kirkpatrick, A. 2012. English in ASEAN: Implications for regional multilingualism. Journal of Multilingual and Multicultural Development 33(4), 331–344. Kirkpatrick, A. 2016. The Asian corpus of English  – introduction to the special issue. Journal of English as a Lingua Franca 5(2), 225–228. Klimpfinger, T. 2009. “She’s mixing the two languages together” – forms and functions of code-­ switching in English as a Lingua Franca. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 348–371. Konakahara, M. 2016. The use of unmitigated disagreement in ELF casual conversation: Ensuring mutual understanding by providing correct information. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts -­conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 70–89. Konakahara, M., K. Murata and M. Iino. 2019. “English”-­medium instruction in a Japanese university: Exploring students and lecturers’ voices from an ELF perspective. In K.  Murata (ed), English-­medium instruction from an English as a Lingua Franca perspective -­ exploring the higher education context. Abingdon, Oxon: Routledge, pp. 157–175. Kubota, R. 2016. “Language is only a tool”: Japanese expatriates working in China and implications for language teaching. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts -­conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 156–179. Kuteeva, M. 2017. Written ELF and EAP instruction: A  North European perspective. Waseda Working Papers in ELF 6, 40–55. Kuteeva, M. and A. Mauranen. 2014. Writing for international publication in multilingual contexts: An introduction to the special issue. Journal of English for Academic Purposes 13, 1–4. Lave, J. and E. Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. Li Wei. 2016. New Chinglish and the post-­multilingualism challenge: Translanguaging ELF in China. Journal of English as a Lingua Franca 5(1), 1–25. Li Wei. 2018. Translanguaging as a practical theory of language. Applied Linguistics 39, 261. Matsumoto, Y. 2014. Collaborative co-­constructions of humorous interaction among ELF speakers. Journal of English as a Lingua Franca 3(1), 81–107. Matsumoto, Y. 2018. Functions of laughter in English-­as-­a-­lingua-­franca classroom interactions: A multimodal ensemble of verbal and nonverbal interactional resources at miscommunication moments. Journal of English as a Lingua Franca 7(2), 229–260. Mauranen, A. 2006. Signaling and preventing misunderstanding in English as Lingua Franca communication. International Journal of the Sociology of Language 2006(177), 123–150. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Mauranen, A. 2016. English as a global Lingua Franca: Changing language in changing global academia. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts: Conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 29–46. Mauranen, A. 2018. Conceptualizing ELF. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 7–24. McNamara, T. 2012. English as a Lingua Franca: The challenge for language testing. Journal of English as a Lingua Franca 1(1), 199–202.

ELF research methods and approaches  17 McNamara, T. 2018. A  challenge for language testing: The assessment of English as a Lingua Franca. Waseda Working Papers in ELF 7, 13–29. Murata, K. (ed). 2016a. Exploring ELF in Japanese academic and business contexts: Conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge. Murata, K. 2016b. ELF research  – its impact on language education in Japan and East Asia. In M.L. Pitzl and R. Osimk-­Teasdale (eds), English as a Lingua Franca: Perspectives and prospects. Contributions in honour of Barbara Seidlhofer. Berlin: de Gruyter Mouton, pp. 77–86. Murata, K. (ed). 2019a. English – medium instruction from an English as a Lingua Franca perspective: Exploring the higher education context. Abingdon, Oxon: Routledge. Murata, K. 2019b. The realities of the use of English in the globalised world and the teaching of English: A discrepancy? JACET Journal 63, 7–26. Murata, K. and M. Iino. 2018. EMI in higher education: An ELF perspective. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of ELF. London: Routledge, pp. 400–412. Murata, K., M. Iino and M. Konakahara. 2017. An investigation into the use of and attitudes towards ELF (English as a Lingua Franca) in English-­Medium Instruction (EMI) classes and its implications for English language teaching. (In Japanese). Waseda Review of Education 31(1), 21–38. Murata, K., M. Iino and M. Konakahara. 2019. Realities of EMI practices among multilingual students in a Japanese university. In J. Jenkins and A. Mauranen (eds), Linguistic diversity in international universities. Abingdon, Oxon: Routledge, pp. 149–171. Murata, K., M. Konakahara, M. Iino and N. Toyoshima. 2018. An investigation into attitudes towards English as a Lingua Franca (ELF) in English-­Medium Instruction (EMI) and business settings and its implications for English language pedagogy. (In Japanese). Waseda Review of Education 32(1), 55–75. Murata, K., M. Konakahara, M. Iino and N. Toyoshima. 2019. An exploration into changes of students’ attitudes towards ELF in EMI and business people’s attitudinal differences towards the use of ELF and its implications for English language teaching. (In Japanese). Waseda Review of Education 33(1), 19–38. Nogami, Y. 2019. Identity and pragmatic language use among East Asian ELF speakers: Its implication for English-­medium education. In K. Murata (ed), English – medium instruction from an English as a Lingua Franca perspective: Exploring the higher education context. Abingdon, Oxon: Routledge, pp. 176–197. Norton, B. 2000. Identity and language learning. Harlow: Pearson Education. Norton, B. and M. Early. 2011. Researcher identity, narrative inquiry, and language teaching research. TESOL Quarterly 45(3), 415–439. O’Neal, G. 2019. Phonetic and phonological accommodation in ELF interactions: Interactional intelligibility and sufficiency from an English as a Lingua Franca perspective and its implications for English pronunciation pedagogy. Unpublished PhD thesis, submitted to Waseda University. Ortega, L. 2018. Ontologies of language, second language acquisition, and world Englishes. World Englishes 37(1), 64–79. Otsu, A. 2017. The gap between educational policies and actual workplace communication in ELF: A Japanese case. European Journal of Language Policy 9(1), 105–118. Pietikäinen, K. 2014. ELF couples and automatic code-­switching. Journal of English as a Lingua Franca 3(1), 1–26. Pietikäinen, K. 2018. Misunderstandings and ensuring understanding in private ELF talk. Applied Linguistics 39(2), 188–212.

18  Kumiko Murata Pitzl, M.-­L. 2018. Transient International Groups (TIGs): Exploring the group and development dimension of ELF. Journal of English as a Lingua Franca 7(1), 25–58. Sacks, H., E. Schegloff and G. Jefferson. 1974. A simplest systematics for the organization of turn-­taking for conversation. Language 50(4), 696–735. Sadoshima, S. 2017. Fostering ELF writers’ independence: Challenges and possibilities of the Waseda university writing center. Waseda Working Papers in ELF 6, 163–174. Seidlhofer, B. 2001. Closing a conceptual gap: The case for a description of English as a Lingua Franca. International Journal of Applied Linguistics 11(2), 133–158. Seidlhofer, B. 2004. Research perspectives on teaching English as a Lingua Franca. Annual Review of Applied Linguistics 24, 209–239. Seidlhofer, B. 2009. Common ground and different realities: World Englishes and English as a Lingua Franca. World Englishes 28(2), 236–245. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Seidlhofer, B. 2016. ELF: English in a global context. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts: Conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 17–28. Seidlhofer, B. 2018. Standard English and the dynamics of ELF variation. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. Abingdon, Oxon: Routledge, pp. 85–100. Shohamy, E. 2019. Critical language testing and English Lingua Franca: How can one help the other? In K. Murata (ed), English – medium instruction from an English as a Lingua Franca perspective: Exploring the higher education context. Abingdon, Oxon: Routledge, pp. 271–285. Takino, M. 2016. Using narrative inquiry in ELF research: Exploring BELF users’ perspective. Waseda Working Papers in ELF 5, 117–135. Turner, J. 2017. ELF and heterogeneity in written English in the contemporary international university. Waseda Working Papers in ELF 6, 56–74. Turner, J. 2018. On writtenness: The cultural politics of academic writing. London: Bloomsbury. Wenger, E. 1998. Communities of practice: Learning, meaning, and identity. Cambridge: Cambridge University Press. Widdowson, H.G. 1975. Stylistics and the teaching of literature. London: Longman. Widdowson, H.G. 1978. Teaching language as communication. Oxford: Oxford University Press. Widdowson, H.G. 1979. Explorations in applied linguistics. Oxford: Oxford University Press. Widdowson, H.G. 1984. Explorations in applied linguistics 2. Oxford: Oxford University Press. Widdowson, H.G. 2003. Defining issues in English language teaching. Oxford: Oxford University Press. Widdowson, H.G. 2012. ELF and inconvenience of established concepts. Journal of English as a Lingua Franca 1(1), 5–26. Widdowson, H.G. 2016. Competence and capability: Rethinking the subject English. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts: Conceptualization, research and pedagogic implications. Abingdon, Oxon: Routledge, pp. 213–223.

Part I

ELF research Some theoretical underpinnings and development

2 Research perspectives on ELF Linguistic usage and communicative use Henry Widdowson

What counts as valid academic research and appropriate ways of conducting it are determined by how the object of enquiry is conceptualised. The appearance of English as a lingua franca (ELF) as a pervasive global phenomenon posed, and still poses, a problem because it does not fit the conceptual categories that conventionally inform sociolinguistic and pedagogic accounts of language. It is generally conceded to have something of the appearance of English, to be a kind of English, but of an ad hoc reduced kind, used expediently by people who have not learned it properly. One reaction is to deny that it is worthy of serious attention as an object of research. Another reaction is to seek to categorise it in terms of established concepts, thereby lending it research respectability. If regularities could be identified in the non-­conformist linguistic features of ELF usage, then one can claim that ELF is a variety, a relatively stable état de langue, a kind of dialectal version of the language. It would then be regarded as a valid area of study. Inconveniently, the data reveal no such regularities: the variety concept does not fit. But then the question arises: if description of the formal linguistic properties of ELF does not reveal a variety, what does it reveal? Where ELF communication manifests abnormalities these can of course be described in reference to conventionalised norms. But what does this tell us about how these linguistic forms are used to realise the communicative purposes that motivate recourse to ELF in the first place? And how does it account for communication between speakers of different first languages where linguistic abnormalities are not in evidence? Is one to assume that in that case they are not using English as a lingua franca at all? ELF is a kind of language behaviour, a parole or performance phenomenon. This is conventionally defined as actualisation of a stabilised langue or competence, assumed to represent the linguistic knowledge of a particular community of native speakers. ELF performance can be described accordingly by reference to the linguistic form it manifests in relation to an established communal norm. But performance can also refer to what communicative function is realised in producing it. This is the crucial distinction between text and discourse, which I have discussed elsewhere, most extensively in Widdowson 2004. Much earlier, in Widdowson 1978, I used the term usage to refer to the linguistic forms that are manifested and use to refer to the communicative function that these forms

22  Henry Widdowson realise. I will employ these terms in the present discussion to distinguish two ways of thinking that have informed ELF research. Although research in ELF has been concerned with both kinds of performance, and the relationship between them, ELF has been seen primarily as a usage phenomenon, marked by the textual manifestation of linguistic forms which do not conform to the standard language or norms of native speaker usage. This is not surprising since it is these non-­conformities that are most readily observable in non-­ native speaking ELF users and to describe this usage is to follow well-­established sociolinguistic practice, which is essentially comparative and normative in orientation, identifying lectal (i.e. dialectal and sociolectal) variants by using some norm or other as benchmark. The value of such descriptions is that they provide convincing evidence that ELF users are capable of communicating effectively without conforming to the rules and conventions which are said to define ‘proper’ English. It is a usage way of thinking that informs the conceptualising of ELF in the lead chapter of the recently published Handbook (Jenkins et al. 2018), where ELF is represented as a similect-­based contact language (Mauranen 2018). Similects, as Mauranen explains, are the usages of English which bear the traces of the speakers’ first languages and which in popular parlance have been wryly referred to in such hybrid terms as Spanglish, Finglish, Japlish, and so on. ELF happens when different similect speakers interact: ELF, then, embodies contact between speakers from different similects . . . ELF, then, means contact between these hybrid contact-­based lects – that is ELF is a higher-­order, or second-­order language contact. (Mauranen 2018: 10) Presumably what this means is that, when similects come into contact, they merge to become hybridised lectal variants of English so that Spanglish and Japlish, for example would presumably merge to result in ELFish. In such a conceptualisation, ELF is defined in reference to established norms of usage, and where there are lectal non-­conformities, these are described as approximations to such norms: Approximations that are sufficiently close to their target may not pose too much difficulty for a hearer to construct the meaning from the elements that are there. (Mauranen 2016: 38) The assumption here, it would seem, is that meaning is inscribed in linguistic form and that what ELF users mean by what they say can only be re-­constructed by reference to the target norm whereby the normal form of words can be recovered from their approximations. As Mauranen puts it: By approximating intended expressions well enough, speakers can achieve communicative success. (Mauranen 2018: 18)

Linguistic usage and communicative use  23 Communication in ELF, then, is conceptualised as a function of usage conformity. This conceptualisation, of course, defines ELF as used only between similect speakers and so runs counter to that proposed for example in Seidlhofer 2011 that includes native speakers as ELF users. Although Mauranen concedes that native speakers may also be involved, they are nevertheless excluded from consideration in the definition of ELF: ENL [English as a native language] speakers of different origins participate in ELF communities. The distinctive feature of ELF is nevertheless its character as a hybrid of similects. (Mauranen 2012: 29–30) This description of ELF as linguistic usage is consistent with the traditional concern in historical linguistics for identifying variation as evidence of incipient change. What is focused on here is the effect that ELF variants might have on the development of ‘the’ English language as a formal construct. This is made explicit in the title given to the 10th annual ELF conference: ELF and Changing English. The same interest in ELF variants as vectors of systemic change is evident too in the work of Schneider (2012), where they are represented as interim stages of development towards variety norms. One limitation of using ELF data as evidence only of departures from ‘normal’ English is, as has been pointed out in the literature for some time (e.g., Hülmbauer and Seidlhofer 2013), that non-­ENL ELF users do not only draw on English as a linguistic resource but on other languages, not necessarily their own L1. This observation has led to the proposal (Jenkins 2015) that it is just these multilingual occurrences in ELF that are its defining features, so that ELF should be re-­conceptualised as English as a multilingua franca. This is proposed as a new theoretical perspective in which it is multilingualism that is the primary area of study, where: multilingualism rather than English, is the ‘superordinate’. ELF is conceived as within a framework of multilingualism (versus multilingualism within a framework of ELF) So to an extent, it is a question of emphasis, evolution, not revolution: reducing the size of ‘English’ in ELF and focusing more on the multilingualism of most ELF users. (Jenkins 2015: 75) It is not entirely clear how this change of emphasis and this shift of focus constitute a re-­theorisation of ELF. One can readily accept that the observed overt and covert presence of other languages in ELF interactions justifies describing them as multilingual and that accounting for this presence calls for some theory of multilingualism. But such a theory would, of course, need to go beyond the occurrence of multilingual features and explain how they are evidence of cognitive and communicative processes that are essentially different from those of monolingual users.

24  Henry Widdowson There has been a good deal of theorising about multilingualism over recent years which, as Jenkins herself makes eloquently clear, has given rise to a bewildering array of different terms to label the phenomenon: plurilingualism, metrolingualism, code-­switching, code-­meshing, translanguaging, multilanguaging and so on. But what should be noted is that, although the significance of the phenomenon has been discussed in various ways, it is, as the terms indicate, the manifestation of linguistic forms that identify and define it. To make multilingualism a ‘superordinate’ defining feature of ELF is similarly to focus on what is linguistically distinctive about it. As Jenkins points out, this conceptualisation of ELF as a multilingual franca is consistent with that of Mauranen, with the linguistic features now identified as lingual rather than lectal, deriving from different languages (linguae) rather than ‘similects’. But in both cases, the focus is on usage, as it was in the beginning phase of ELF study – in what Jenkins refers to as ELF 1. The description of ELF performance in usage terms of course follows the orthodox line in mainstream descriptive linguistics and dialectology. As such it has the obvious advantage of giving it a prima facie disciplinary acceptability. But as I pointed out at the start of this paper, there is an alternative perspective on language performance and so an alternative conceptualisation: one that focuses primarily not on the usage but the use of ELF. Here the focus of attention shifts from linguistic forms that constitute the texts that users produce to the discourse they enact in the communicative process of producing them. This shift can be described as a re-­alignment of the different factors that are involved in this process – the factors that Roman Jakobson proposed as necessary constitutive elements of what he referred to as the speech event. ADDRESSER

CONTEXT MESSAGE CONTACT CODE

ADDRESSEE (Jakobson 1960: 353)

The term ‘message’ is ambiguous. It could denote linguistic form; in which case, a particular expression is an instance of usage with reference only to how it relates to the code factor. But the term can, of course, also refer not to its linguistic form of an expression but its communicative function as use, to the pragmatic meaning that can be assigned to it. In this case, its description involves reference to the other factors. What is meant by a linguistic expression obviously depends on such factors as the context of its occurrence and how it is intended and interpreted by the participants, addresser and addressee. Although the description of message forms in lectal or lingual terms as code manifestations is undeniably a disciplinary enterprise of major significance, it needs to be noted that it is not as usage but as use that people actually experience language. And ELF users, like everybody else, draw on whatever encoding resources are available to them to communicate their meanings and relate to others as best they can to serve their purposes and to meet the contingencies of context. What concerns them is not the occurrence of linguistic forms as

Linguistic usage and communicative use  25 such and what they semantically encode but what pragmatic significance they might have in the communicative process of meaning making. So in research on the multilingual features of ELF from a use perspective (see Cogo 2018 for a brief survey) what is significant is not so much the occurrence of non-­English elements, but what motivates the user in resorting to them: how they function pragmatically in a particular discourse. They may be co-­operatively motivated in that they serve to get the user’s meaning across, accommodate to the other, and so maintain participation in a contextualised interaction. Or they may be used unco-­operatively in that they are intended, and interpreted, as a violation of Grice’s maxim of manner and so to bring about implicatures (Grice 1975) to express an attitude, create the effect of amusement, or to assert socio-­cultural identity. As with any text, multilingual or not, what concerns the language user, as distinct from the linguistic analyst, is what discourse is pragmatically realised from it. The same distinction also, of course, applies to the concept of approximation. As usage, it can be described as a relationship between the factors of message form and code  – the extent to which a particular expression conforms to what is conventionally accepted as the encoded norm. But if we think of what might induce users to approximate, all the other Jakobson factors need to be considered. Users may produce approximate expressions in a particular context because they serve their pragmatic purposes: to get their meaning across more directly, to mediate the relationship with their interlocutors, to create an implicature. In short, they may approximate because it is communicatively appropriate to do so. The general point that needs to be made, then, is that ELF is experienced as communicative use, which cannot be accounted for in the linguistics of ELF as usage. And this experience is far from easy to cope with. The essential difficulty about ELF interactions for its users, whether they are non-­native speakers of English or not, is that they cannot reliably depend on the communal lingua-­cultural knowledge that is mutually presupposed in the communication they are accustomed to. This means that they frequently have to create the requisite conditions for communication by negotiating common ground as they go along. Whether they are business people trying to do a deal, diplomats negotiating conflict resolution, refugees seeking asylum, all have the problem of finding ways of using English as an expedient resource to negotiate an understanding favourable to themselves with interlocutors who have different, often opposing, attitudes, beliefs and values. The less transient the encounter, and the more familiar participants are with the kind of speech event they are engaged in, the greater of course is the possibility of finding common ground. And in some recurrent situations, in business transactions or diplomatic negotiation for example, the problem is further eased if it is in the interest of participants to converge on an agreement. Here ELF users may have room for manoeuvre in positioning themselves to their advantage. In other situations, however, the problem is exacerbated, as in unequal encounters, when ELF users are disadvantaged to begin with, refugees seeking asylum for example, who have positions imposed upon them and where there may be little if any allowable room for manoeuvre.

26  Henry Widdowson The spread of English as an expedient linguistic resource in combination with digitalisation affords the unprecedented possibility of unbounded communication across the confines of lingua-­cultural communal borders. In this respect, the phenomenon of ELF is to be recognised as a positive development in human affairs. But we also need to recognise that the realisation of this possibility in actual use creates problems. And these are just the problems that are the concern of applied linguistics, which, in the often-­quoted definition of Christopher Brumfit, claims to be: The theoretical and empirical investigation of real-­world problems in which language is a central issue. (Brumfit 1995: 27) With regard to ELF, how it can be accounted for as usage is doubtless an interesting problem, even a central issue, for linguistic description, but as such it is of little direct relevance to the real world of ELF users. What is problematic for them is how they can cope with situations in which they have to make do with using whatever linguistic resource they have at their disposal to negotiate common ground and a favourable communicative outcome with others with whom they have little in common in the way of lingua-­cultural background. What ELF users experience as a central issue is what applied linguistics is centrally concerned to investigate, namely: problems in the world in which language is implicated.

(Cook 2003: 5)

It is the reality of problems in the world that ELF users have to confront and deal with – social, economic and political problems that globalisation has set in train, that have to do with exploitation, oppression, human rights and social and individual identity. ELF is relevant only because this is the language resource that is implicated as the communicative means for engaging with these problems. The language that is manifested is an epiphenomenon, a by-­product of this process. From this perspective, the focus of ELF research shifts quite radically to become essentially an enquiry into how people cope with unfamiliar communicative conditions. What then is of central relevance is not how the message form relates to the code factor in Jakobson’s scheme, but how it relates to the other factors in the communicative process – how it connects with context, what the addresser intends by it, how it serves to position participants in an interaction, and so on. Communicative success depends on adjusting the relationship between these factors and this is far from easy, even in favourable conditions when users are of the same lingua-­cultural community. It poses particular problems when they are not. Resolving these problems may on some occasions, with some participants, be relatively easy, and failure to do so may not matter all that much. But there are other occasions where the problems are more intractable and failure to resolve them may well, and all too often does, have serious consequences for the

Linguistic usage and communicative use  27 ELF user, as is the case in the ‘high-­stakes encounters’ discussed by Seidlhofer (this volume). These, to quote from the description of the symposium on that theme that she convened at the 2014 AILA World Congress, raise ‘vital issues of misunderstanding, alienation, inequity, and disenfranchisement’. These issues may indeed be matters of life and death, as, for example, when talks on conflict resolution fail or when misunderstandings arise in encounters between asylum seekers and immigration officials (Guido 2012, 2018). In this conceptualisation, ELF is intrinsically implicated in problems in the contemporary world that in one way or another globalisation has either given rise to or has exacerbated. What now becomes significant about ELF is not its global occurrence in variable textual manifestations, but its involvement in extended networks of communicative interaction, how it provides both the potential means and the actual challenge for people from different lingua-­cultural backgrounds, with different values, beliefs, assumptions to negotiate relationships, reconcile differences, make decisions and so on. It is not what form their language takes that is of primary interest, but how it is used as a resource for achieving their communicative purposes and the kind of problems they encounter in the process. The focus of research now shifts from the form that ELF takes as language contact to how people use it to make contact with each other. Conceived of in this way, it is not multilingualism but global communication that is the ‘superordinate’, and ELF research is then not located in the field of linguistics but falls squarely into that of applied linguistics as an investigation of what people in the real-­world experience in using ELF as an expedient resource for communicative problem solving.

References Brumfit, C. 1995. Teacher professionalism and research. In G. Cook and B. Seidlhofer (eds), Principle and practice in applied linguistics. Oxford: Oxford University Press, pp. 27–42. Cogo, A. 2018. ELF and multilingualism. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 357–368. Cook, G. 2003. Applied linguistics. Oxford introductions to language study. Oxford: Oxford University Press. Grice, H.P. 1975. Logic and conversation. In P. Cole and J.L. Morgan (eds), Syntax and semantics, volume 3: Speech acts. New York: Academic Press, pp. 41–58. Guido, M.G. 2012. ELF authentication and accommodation strategies in crosscultural immigration encounters. Journal of English as a Lingua Franca 1(2), 219–240. Guido, M.G. 2018. English as a Lingua Franca in migrants’ trauma narratives. London: Palgrave Macmillan. Hülmbauer, C. and B. Seidlhofer. 2013. English as a Lingua Franca in European multilingualism. In A. Berthoud, F. Grin and G. Lüdi (eds), Exploring the dynamics of multilingualism: The DYLAN project. Amsterdam: John Benjamins, pp. 409–428. Jakobson, R. 1960. Closing statement: Linguistics and poetics. In T.A. Sebeok (ed), Style in language. Cambridge, MA: MIT Press. Jenkins, J. 2015. Repositioning English and multilingualism in English as a Lingua Franca. Englishes in Practice 2(3), 49–85.

28  Henry Widdowson Jenkins, J., W. Baker and M. Dewey (eds). 2018. The Routledge handbook of English as a Lingua Franca. London: Routledge. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Mauranen, A. 2016. English as a global Lingua Franca: Changing language in changing global academia. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts. London and New York: Routledge, pp. 29–46. Mauranen, A. 2018. Conceptualising ELF. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London and New York: Routledge, pp. 7–24. Schneider, E. 2012. Exploring the interface between world Englishes and second language acquisition-­ and implication for English as a Lingua Franca. Journal of English as a Lingua Franca 1(1), 57–91. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Widdowson, H. 1978. Teaching language as communication. Oxford: Oxford University Press. Widdowson, H. 2004. Text, context, pretext: Critical issues in discourse analysis. Oxford: Wiley Blackwell.

3 Researching ELF communication Focus on high-­stakes encounters Barbara Seidlhofer1

As I have discussed elsewhere (e.g. Seidlhofer 2011), research on English as a lingua franca (ELF)2 leads to a reconsideration of established concepts in sociolinguistics, such as community and competence. This necessarily also involves methodological issues since how competence or community is conceived will determine what kind of data will be taken as relevant to its description and how they are interpreted as evidence. ELF research also has major conceptual and methodological implications for work in applied linguistics, and these will be the focus of this paper. Brumfit’s definition of applied linguistics is well known: the theoretical and empirical investigation of real-­world problems in which language is a central issue. (Brumfit 1995: 27) What this calls for, according to the editorial in the International Journal of Applied Linguistics (Seidlhofer and Breivik 2004: iii), is developing “an awareness of the way language works, how it affects peoples’ lives” and proposing “what interventions it is desirable and feasible to make in different domains of language use and learning”. First, awareness: how do ideas, observations, results, suggestions emerging from ELF research relate to ‘real-­world’ problems? Then intervention: how could or should these be followed up, or followed through, to practical proposals? The first point to make is that today the most obvious real-­world problems are those in and of a globalised world and the prevalent means of communication of this globalised world is ELF. It is this that is used in the domains of international business, scientific research, and diplomacy, and which directly affects people’s lives at the most basic level in such areas as peace keeping, conflict resolution, and immigration control. In many of these cases, the interactions are unequal encounters. The empirical research on ELF up till now has focused mainly (but by no means exclusively) on encounters with fairly ‘symmetrical’ participant relationships. Power differential is often in evidence, but generally participant roles and relationships are not characterised a priori by gate-­keeping. Thus, the domains in VOICE (Vienna-­Oxford International Corpus of English; VOICE 2013) and in ACE, the Asian Corpus of

30  Barbara Seidlhofer English [as a Lingua Franca] (ACE 2014; see also Pitzl and L. Wang, this volume), are socially defined situations or areas of activity: ED (educational) LE (leisure) P (professional) PB (professional business) PO (professional organisational) PR (professional research and science) The interactions recorded in these corpora are predominantly of a consensual kind: it is in the interest of all parties to cooperate, and they do so willingly in the co-­construction of agreed meanings. Hence the description of these interactions “gives the impression of ELF talk being overtly consensus-­oriented, cooperative and mutually supportive”(Seidlhofer 2004: 218). In the literature, there has even been an over-­generalising discussion about, and sometimes misrepresentation of, ELF interactions having been presented as always being particularly cooperative and this co-­operativeness being a “unique feature” of ELF talk (Jenks 2012: 388). It was recognised and pointed out early on, however, that the cooperative character observed in many ELF interactions may well be a function of the kinds of interaction that were recorded and that the findings available to date result from research on a fairly limited database. It is therefore conceivable that further research might show some of the present findings to be a function of the type and purpose of the interactions investigated. (Seidlhofer 2004: 219) While this caveat as such is ignored in Jenks (2012), the observation and conjecture made in this quotation are actually borne out by his findings: the participants in the voice-­based online chatroom interactions he investigated, in which interactants cannot see each other (as opposed to the face-­to-­face communication hitherto predominantly described in ELF research) do indeed engage in un-­cooperative behaviour and highlight, rather than ignore, communication problems. Walkinshaw and Kirkpatrick (2014), in their account of Asian ELF interactions captured in the ACE corpus, describe both mutual face preservation as well as teasing and direct contradictions and counterclaims. A particularly shocking example of reprehensive verbal behaviour is a highly unequal ELF encounter between a US soldier and several Iraqi boys in Iraq presented in Bysouth et al. (2015), referred to in Jenks (2018): here the US soldier abuses his power and the boys’ limited understanding of English by contemptuously making fun of them with a pseudo-­offer, in order to amuse himself and a fellow soldier. It is, then, clearly high time to engage in further research, to take account of a great deal of diverse ELF data and to further develop our methodological procedures for considering unequal high-­stakes encounters. In such interactions

Focus on high-­stakes encounters  31 willingness to co-­operate is often not a given, instead there is a unilateral (im)position of power, and the use of ELF, or the way it is conceived, can be conflictual in its consequences. Even in the fairly ‘equal encounters’ that have hitherto been the focus of investigation, the study of ELF communication has called for a rethinking of the traditional notions of community and competence, predicated as they usually are on the assumption of the primacy of the native speaker. This rethinking becomes a matter of immediate practical urgency in these high-­stakes encounters, where the participants from different communities and with different competences must also be seen as entirely legitimate users of the language. Such high-­stakes encounters and activities include language (education) policy and language planning, testing, international publishing, language and the law, and even interpreting. Particularly striking examples of gate-­keeping encounters are those involving the interrogation of refugees and asylum seekers. These interrogations are very often carried out in English used as a lingua franca being the only means of communication available to both sides. But such a situation is not conducive to mutually supportive co-­operation, and ELF interaction can frequently result in misunderstanding, alienation, inequity, and disenfranchisement. A particularly clear example is the one of ELF encounters on the southern shores of Europe, where every year, many thousands of displaced people and other refugees seek the protection of various states and apply for political asylum. Asylum seekers are interviewed by immigration officials and their claims examined by national commissions. After the interview a judgement is made, on the basis of a short text summarising their story. The fate of the asylum seekers is entirely dependent on the transcripts of their verbal performance and how they are interpreted. But of course all transcripts of speech are approximate, and furthermore, what is transcribed in these cases is frequently not what the asylum seekers actually say (let alone what they meant) but what the interviewer understands them as saying, and frequently what is taken to be their ‘imperfect’ English is normalised. The transcript is then subjected to interpretation by officials by reference to their own schematic assumptions. There is every likelihood in these circumstances that the stories of the asylum seekers will be distorted. The process, as has been pointed out, reveal[s] how public officials, faced with the intrinsic alterity of the asylum seekers, rely on commonsensical, but at times inappropriate, knowledge of social, cultural, and linguistic values to construct, process, and eventually determine the validity of each claim. As a result, this procedure is fraught with unexamined assumptions about language, national identity, and communicative competence, leading to egregious violations of the asylum seekers’ human rights. (Jacquemet 2009: 525) What tends to happen when these interviews are conducted via ELF (i.e. the part that is not conducted through interpreters) is that there is a conflict between two different concepts of English. To take the well documented example of African asylum seekers being interviewed by Italian immigration officers (Guido 2008,

32  Barbara Seidlhofer 2012, 2018; Guido et al. 2018): many asylum seekers come from what Kachru has called the Outer Circle, especially from Nigeria. For them, English has become domesticated as an everyday second language, but it is naturally a version of English which does not conform to the standard language that the immigration officers recognise because they have been taught it as a foreign language. On the one hand, the asylum seekers will be unable to adjust their natural way of speaking so that it conforms to the standard, and on the other hand, the immigration officers bring to the interaction the knowledge about the standard language they have been schooled in and judge the verbal performance of the asylum seekers by reference to it. There often is, therefore, a lack of accommodation on both sides and consequently misunderstandings arise which, because of the power differential, are resolved to the disadvantage of the asylum seekers. What we have here is an obvious ELF situation, with participants in the interaction coming from different linguacultures, but neither side, for different reasons, is actually using the language as a lingua franca. The research recorded in Guido (2012) explores in detail this linguacultural disparity and its consequences and shows how both sides transfer structural features and conventional assumptions from their own linguacultural experience into the English they use. What these unequal encounters indicate quite clearly is the degree to which effective, and equitable, communication depends on the kind of strategies of accommodation and adaptation that ELF researchers have long emphasized as paramount. These encounters also can be said to show the negative effects of an insistence on the conventional normative teaching of standard English in schools, which makes no provision for the subsequent need to use the language strategically in actual communicative situations. The fact is that the immigration officers often do not have what Widdowson (2003, 2016) calls the communicative capability that would suitably qualify them for the work they do. There is a second procedure related to immigration that also reveals a disturbing lack of understanding about the necessary and natural variation of language in actual use that ELF communication so clearly exemplifies. Whereas the first procedure involves the processing of asylum seekers’ talk as communication, this second procedure processes it as what Le Page and Tabouret-­Keller (1985) call ‘acts of identity’. It is called Language Analysis for the Determination of Origin (LADO), and here a particular way of conceiving of the relationship between language and identity determines how courts and government officials interpret the data of the immigrants’ talk as decisive evidence of ‘where they come from’. LADO is one of several methods many states employ to investigate the reliability of information supplied by asylum seekers about their national or regional origin. In practice, LADO is used to determine possible national origins of asylum seekers in order to justify deportation after the rejection of their application. The assumption that informs this investigation is that what speakers say inevitably bears traces of their original first languages, which, rather like fingerprints, reveal their real national identities, whatever nationality they may claim to have. LADO is an officially sanctioned forensic procedure for identifying national origins so that immigrants can be sent back to where they belong.

Focus on high-­stakes encounters  33 The assumption is entirely fallacious, as is pointed out in the PluS project at the University of Vienna, whose full title is “When Plurilingual Speakers Encounter Unilingual Environments. Migrants from African Countries in Vienna: Language Practices and Institutional Communication” (https://bim.lbg.ac.at/en/ when-­plurilingual-­speakers-­encounter-­unilingual-­environments-­0). What the PluS project reveals is that the evidence that LADO presents is unreliable and therefore unsafe also in legal terms, since it is based on simplistic and anachronistic beliefs about the nature of language (see Dorn et al. 2014; Eades 2005; Fraser 2013). As with the immigration interviews, LADO analysts tend to apply ideas about language based on their own particular linguistic experience as if they were universally applicable to all language use. One such idea, particularly prevalent for instance in Europe and Japan, is that monolingualism is the norm and that there are distinct identifiable languages that correspond with, and define, particular communities or nations. But such an idea has little if any validity in the African communities from where the asylum seekers come. Here multilingualism is the norm, and communities may draw on a continuum of linguistic resources without consciously dividing it into separate languages. And of course, many of what might be called different languages or language varieties have features in common, which makes it difficult to identify those which are uniquely defining features of ‘a particular language or variety’. But perhaps the fallacy of the LADO procedure that ELF research reveals most clearly is the assumption of invariability. Particularly in present global circumstances of enormously increased mobility and extended networks of interaction, individuals will encounter various new contexts and purposes and will expand and adapt their linguistic resources accordingly. Their language will naturally bear traces of this variable linguacultural experience so it will be impossible to interpret some traces as original and others not. LADO procedures are given official sanction as providing expert forensic evidence upon which objectively fair judgements can be reliably made. In fact, these procedures are sociolinguistically and pragmatically uninformed, based as they are on the following related misconceptions: 1

Diverse varieties of English are internally consistent and can be delimited clearly from each other. •

2

On the contrary, as the phenomenon of ELF brings out so clearly, what we find in actual use is continual creative and adaptive variation, as people exploit diverse linguistic resources as communicatively appropriate to different social contexts.

The individual is associated with a primary ‘language’ or ‘variety’ (of English or any other language), and this remains as a lifelong marker of origin and identity. •

On the contrary, individuals have a communicative capability which enables them to develop flexible, variable, changeable repertoires, whereby new linguistic features are adopted and others fall out of use.

34  Barbara Seidlhofer 3

Language, community and territory are in some essential primordial way bound together. •

On the contrary, although such a belief might have been tenable in earlier times, and may still be socio-­politically expedient, it no longer holds true in the current globalised world of massive mobility, extended networks of communication and demographic change. As ELF research has made clear, traditional notions about language and community need to be radically revised.

What comes across very clearly in the ELF research reported on earlier in this paper is how an adherence to traditional ideas about English held by gate-­keepers who have power to impose them, put asylum seekers at a severe disadvantage and are in effect discriminatory. The same point can be made about other kinds of gate-­keeping encounter where fixed concepts about the language predominate. One example is in the area of international publishing. On the face of it, this is as remote from the asylum-­seeking situation as it is possible to be. Whereas asylum seekers have no recognised status and are at times hardly treated as human beings at all, academics who submit articles for publication in international learned journals are people who are, in comparison, accorded very high status. But they too can be disadvantaged if they are required to write in English and not their own/strongest language, and since most international journals are in English that is what they are indeed required to do. That English, furthermore, has to be in conformity with ‘Anglo’ norms – and aspiring authors are routinely advised by editors to get their submission vetted by native speakers of English to ensure that it does conform. And conformity is not only a matter of keeping to the ‘rules’ of Standard English, but also to the idiomatic patterns that have been established as the normal wordings of the rhetorical conventions of the genre as approved by ‘Anglo’ scholarship. As a result, academic work that does not conform to this rhetorical etiquette, no matter how intrinsically high in quality, gets undervalued or rejected completely. This is not only discriminatory, unfair, and demeaning for the individual, but actually is detrimental to the very idea of international scholarship which so-­called international journals claim to be promoting (for further discussion, see Lillis and Curry 2010; Lillis et al. 2010; Curry and Lillis 2019; Mur-­Duenas 2013). But international publishing is in principle an ELF activity: scholars from different primary linguacultural backgrounds and different academic traditions interact with each other by using English as a lingua franca means of communication. So the question again arises, just as in the asylum procedures discussed earlier, as to the appropriateness of imposing traditional ideas about English without regard to the varying contexts and purposes which the language needs to serve as a means of communication (Mauranen 2012, 2016; Seidlhofer 2012). Another example of gate-­keeping where the same kind of issue arises is in the teaching and testing of English as a foreign or other language. Here again the only English that is institutionally approved is that of the so-­called educated native speaker (whatever this phrase may mean). That is the English, for example,

Focus on high-­stakes encounters  35 that is implicitly represented as the norm in the Common European Framework of Reference (CEFR) – now widely established, and not only in Europe, as the authoritative means for the assessment of language achievement. This normative imposition actually inhibits achievement in that it has the effect of putting learners at a disadvantage in two ways. In the first place, it sets standards of conformity that most learners fail to achieve, which has obvious negative effects on motivation. Secondly, these standards are themselves irrelevant for most of the purposes that learners will need English for when they actually use it to communicate in the contexts of the real world. What they need to do is not to conform to the standard language or the norms of native speaker usage, but to develop a communicative capability which will enable them to put the resources of English to adaptive and appropriate use. In other words, the frame of reference for English teaching and learning is not English as a native language (ENL) but English as a lingua franca (ELF). For further discussion of these matters, see Hynninen (2014), Lo Bianco (2014), McNamara (2014), McNamara and Shohamy (2016), Pitzl (2015), Seidlhofer (2011: ch. 8), Widdowson and Seidlhofer (2019), Sifakis and Tsantila (2019) and particularly the exchange between Swan (2012) and Widdowson (2012). Asylum seekers, academics, and English language learners are in most respects very different kinds of people. But all of them, I have argued, are disadvantaged, discriminated against by the imposition by gate-­keepers and stakeholders of traditional normative ideas about what English is or should be – ideas which determine what data are selectively focused on as evidence. The global significance of ELF research is that it opens up alternative ways of thinking, more suited to the changed circumstances of the contemporary world. ELF research has so far revealed much about the nature of communication that is overlooked if we adhere to (‘western’/’northern’) traditional assumptions of monolingual speakers in stable communities with native competence in a particular language. Such adherence is particularly damaging in the case of English precisely because it is so widely and variably used. This is clear from the cases of high-­stakes unequal encounters and discriminatory gate-­keeping that I have discussed here. The applied linguistic significance of ELF study is that it has direct relevance to the understanding and resolution of real-­world problems such as the ones I have focused on in this paper. But much more research into this significance, of course, still needs to be done – research with an extended database and a methodology appropriate to the further exploration of ELF, not primarily as a lingual or multilingual phenomenon, but as a global means of communication.

Notes 1 This chapter is a revised and updated version of the article published in Waseda Working Papers in ELF, Vol.4. 2 Note that I am using “English”, “ELF”, “language”, etc. as shorthand expressions and am so refraining from following the widespread practice of putting an asterisk in front of words denoting ‘linguistic entities’ (which of course do not exist as such) in order to avoid ‘essentialising’ – since if this were done consistently it would lead to ‘infinite regress’.

36  Barbara Seidlhofer

References ACE. 2014. The Asian corpus of English. Director: Andy Kirkpatrick. Project co-­ investigator: Wang Lixun. Available from: http://corpus.ied.edu.hk/ace. Brumfit, C.J. 1995. Teacher professionalism and research. In G. Cook and B. Seidlhofer (eds), Principle and practice in applied linguistics: Studies in honour of H.G. Widdowson. Oxford: Oxford University Press, pp. 27–41. Bysouth, D., K. Ikeda and S. Jeloos-­Haghi. 2015. Collateral damage: An investigation of non-­combatant teasing by American service personnel in occupied Iraq and Afghanistan. Pragmatics and Society 6(3), 338–366. Curry, M.J. and T. Lillis. 2019. Unpacking the lore on multilingual scholars publishing in English: A discussion paper. Publications 7(2), article no. 27. Dorn, N., M. Rienzner, B. Busch and A. Santner-­Wolfartsberger. 2014. “Here I find myself to be judged”: ELF/plurilingual perspectives on language analysis for the determination of origin. Journal of English as a Lingua Franca 3(2), 409–424. Eades, D. 2005. Applied linguistics and language analysis in asylum seeker cases. Applied Linguistics 26(4), 503–526. Fraser, H. 2013. Language analysis for the determination of origin (LADO). In C.A. Chapelle (ed), The Encyclopedia of applied linguistics. Oxford: Wiley Blackwell, pp. 2920–2922. Guido, M.G. 2008. English as a Lingua Franca in cross-­cultural immigration domains. Frankfurt am Main: Peter Lang. Guido, M.G. 2012. ELF authentication and accommodation strategies in cross-­cultural immigration encounters. Journal of English as a Lingua Franca 1(2), 219–240. Guido, M.G. 2018. English as a Lingua Franca in migrants’ trauma narratives. London: Palgrave Macmillan. Guido, M.G., P.L. Iaia and L. Errico. 2018. ELF-­mediated intercultural communication between migrants and tourists in an Italian project of responsible tourism: A multimodal ethnopoetic approach to modern and classical sea-­voyage narratives. In I. Guillén-­Galve and I. Vázquez-­Orta (eds), English as a Lingua Franca and intercultural communication. Implications and Applications in the Field of English Language Teaching. Bern: Peter Lang, pp. 97–123. Hynninen, N. 2014. The common European framework of reference from the perspective of English as a Lingua Franca: What we can learn from a focus on language regulation. Journal of English as a Lingua Franca 3(2), 293–316. Jacquemet, M. 2009. Transcribing refugees: The entextualization of asylum seekers’ hearings in a transidiomatic environment. Text & Talk 29(5), 525–546. Jenks, C. 2012. Doing being reprehensive: Some interactional features of English as a Lingua Franca in a chat room. Applied Linguistics 33(4), 386–405. Jenks, C. 2018. Uncooperative Lingua Franca encounters. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. Abingdon: Routledge, pp. 279–291. Kachru, B.B. (ed.). 1992. The other tongue: English across cultures (2nd ed.). Chicago: University of Illinois Press. Le Page, R. and A. Tabouret-­Keller. 1985. Acts of identity. Cambridge: Cambridge University Press. Lillis, T. and M.J. Curry. 2010. Academic writing in a global context: The politics and practices of publishing in English. London: Routledge. Lillis, T., A. Hewings, D. Vladimiroua and M.J. Curry. 2010. The geolinguistics of English as an academic Lingua Franca: Citation practices across English-­medium national

Focus on high-­stakes encounters  37 and English-­medium international journals. International Journal of Applied Linguistics 20(1), 111–135. Lo Bianco, J. 2014. Dialogue between ELF and the field of language policy and planning. Journal of English as a Lingua Franca 3(1), 197–213. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Mauranen, A. 2016. English as a global Lingua Franca: Changing language in changing global academia. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts. Abingdon: Routledge, pp. 29–46. McNamara, T. 2014. 30 years on – evolution or revolution? Language Assessment Quarterly 11(2), 226–232. McNamara, T. and E. Shohamy. 2016. Language testing and ELF: Making the connection. In M.L. Pitzl and R. Osimk-­Teasdale (eds), English as a Lingua Franca: Perspectives and prospects. Berlin: de Gruyter Mouton, pp. 227–233. Mur-­Duenas, P.  2013. Spanish scholars’ research article publishing process in English-­ medium journals: English used as a Lingua Franca? Journal of English as a Lingua Franca 2(2), 315–340. Pitzl, M.-­L. 2015. Understanding and misunderstanding in the common European framework of reference: What we can learn from research on BELF and intercultural communication. Journal of English as a Lingua Franca 4(1), 91–124. Seidlhofer, B. 2004. Research perspectives on teaching English as a Lingua Franca. Annual Review of Applied Linguistics 24, 209–239. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Seidlhofer, B. 2012. Anglophone-­centric attitudes and the globalization of English. Journal of English as a Lingua Franca 1(2), 393–407. Seidlhofer, B. and L.E. Breivik. 2004. Editorial. International Journal of Applied Linguistics 14(2), ii–iii. Sifakis, N. and N. Tsantila (eds). 2019. English as a Lingua Franca for EFL contexts. Bristol: Multilingual Matters. Swan, M. 2012. ELF and EFL: Are they really different? Journal of English as a Lingua Franca 1(2), 379–389. VOICE. 2013. The Vienna-­Oxford international corpus of English. version 2.0 online. Director: Barbara Seidlhofer; Researchers: Angelika Breiteneder, Theresa Klimpfinger, Stefan Majewski, Ruth Osimk-­Teasdale, Marie-­Luise Pitzl, Michael Radeka. Available from: www.univie.ac.at/voice/. Walkinshaw, I. and A. Kirkpatrick. 2014. Mutual face preservation among Asian speakers of English as a Lingua Franca. Journal of English as a Lingua Franca 3(2), 269–291. Widdowson, H.G. 2003. Defining issues in English language teaching. Oxford: Oxford University Press. Widdowson, H.G. 2012. ELF and EFL: What’s the difference? Comments on Michael Swan. Journal of English as a Lingua Franca 2(1), 187–193. Widdowson, H.G. 2016. Competence and capability: Rethinking the subject English. In H.G. Widdowson (ed), On the subject of English: The linguistics of language use and learning. Berlin and Boston: de Gruyter Mouton, pp. 233–243. Widdowson, H.G. and B. Seidlhofer. 2019. Accounting for communication. In C. Roever and G. Wigglesworth (eds), Social perspectives on language testing. Berlin: Peter Lang, pp. 243–252.

4 ELF and translanguaging Covert and overt resources in a transnational workplace Alessia Cogo

Introduction Traditionally, English language research has been approached separately from other languages research. The two areas of English as a Lingua Franca (ELF) and multilingualism have been seen as two separate disciplines: multilingualism, which concerns languages other than English on one side, and ELF research, which supposedly focuses exclusively on English, on the other. This may in part be due to the tendency to see English as isolated from other languages, especially in the role of “the language of globalization” or “the language of international communication”. However, this separation is paradoxical, to say the least: when English is used as a lingua franca it becomes less foreign, but also “less English” and closer to other languages because of the cross-­linguistic, or trans-­linguistic, influences of the resources in the users’ repertoire or their sociolinguistic contexts. ELF as a phenomenon has always been multilingual  – in the sense that the “lingua franca” aspect of the acronym ELF has always been about a contact language perspective and the key role of multilingual resources. However, especially for people outside this research field, the “English” part of the acronym ELF has become more prominent and the lingua franca aspect has been overlooked or even confused with the “international” perspective, reducing ELF to something like “just English in international contexts”. This chapter is an attempt to balancing out that undue tendency to focus on the first part of the ELF acronym, by re-­situating the issue of multilingualism more centrally in ELF research and conceptualising ELF within a multilingual perspective. So the aims of this chapter are twofold. First, to show that research on multilingualism and research on ELF can and should inform each other, rather than be kept separate. Second, and as a consequence of the first, to demonstrate that ELF is a multilingual phenomenon even when the multilingual aspects are not obvious, i.e. when it seems that it is only English. I start by reviewing the literature on ELF and multilingual research, together with the shift to a multilingual perspective and the view of ELF communication in a translanguaging space. I then illustrate this perspective by analysing data collected in a transnational workplace and exploring the different kind of resources, covert and overt, constructed by ELF speakers. Finally, I point out how

ELF and translanguaging  39 translanguaging does not completely do away with ideologies of language separation, but speakers may still exploit them to make meaning in naturally occurring communication.

The “multilingual turn” and translanguaging Recently, there has been an epistemological shift in applied linguistics towards the “multilingual turn”, that is the need to move away from the monolingual bias that has dominated, and in some cases “contaminated”, the foundational concepts and areas of applied linguistics, among which SLA, TESOL and bilingual education (cf. Conteh and Meier 2014; May 2013). In the old paradigm, bi-­ and multilingualism are viewed through a monolingual linguistic ideology (Piller 2016), which conceives languages as separate and fixed entities, submitted to monolingual norms. In practical terms, this means that individuals’ bi-­and multilingualism tends to be interpreted as the sum of different languages. In the new multilingual turn, instead, languages are not seen as separate, but they are seen as linguistic resources, which are merging, meshing and “languaging” (Jørgensen 2008) in semiotic repertoires. Of course, the fact that languages influence each other through language contact has been an aspect of interest developed by scholars concerned with creoles, vernaculars and similar linguistic phenomena, but their research was not informed by a multilingual perspective, instead underpinned by the monolingual stability of codes in contact. However, research on multilingualism has contributed to bring to light the general monolingual bias in language research (cf. Auer 2007; Auer and Li Wei 2007), which focused on a monolingual approach to language description and analysis and tended to research languages as separate entities. Another contribution of the multilingual turn has been a critical view of the deficit approach to bi-­and multilingualism, a critique of the notion of native speaker and the related concepts of interlanguage and cross-­linguistic interference.1 These are critiques that particularly align the multilingual turn with ELF research. SLA and ELT research and practices have been equally responsible for the monolingual bias and the separation approach. In these areas, there is still a tendency to see monolingualism as unmarked, and a monolingual bias as the yardstick against which all language research, and second language users, should be evaluated. In both fields of research, the foundations of the disciplines lay on comparisons between so-­called native and non-­native speakers, monolinguals and multilinguals, and between participants’ (or subjects’, in traditional SLA terms) competence in different and separate languages. However, lately “the multilingual turn” has started to redress the balance by criticising this monolingual linguistic ideology and proposing an “epistemic reorientation” (Ortega 2013: 48), in the direction of multilingual or translingual practices (Canagarajah 2013) and “translanguaging” (García and Li Wei 2014). The approach also involves a reconceptualisation of multilingualism itself, not as a deficient realisation in any of the languages in the individual’s repertoire, but a holistic approach to the repertoire as a whole. Repertoires are then collections of

40  Alessia Cogo all the varieties, languages, registers etc, that an individual has accumulated in their lifetime. Therefore, these repertoires are also dynamic, since these resources change over time (Blommaert 2010; Blommaert and Backus 2013). Translanguaging, with the emphasis on the prefix trans-­, allows to move away from additive concepts of bilingualism and draw more attention to the transformative perspective, which implies a different view of linguistic resources as in relation to each other, transforming each other and creating new repertoires. In certain contexts, translanguaging is also viewed in a social justice perspective (Conteh 2018) and through transgressive lenses (Prada and Nikula 2018). The concept of “repertoire”, the transformative nature and the related ideological aspects are explored in the next section.

ELF from a translanguaging perspective García (2012) explains that, rather than one main language and other second or third languages, where linguistic hierarchies are reproduced at individual level, “bilinguals have one linguistic repertoire from which they select features strategically to communicate effectively” (ibid: 1, emphasis in original). ELF research so far has focused on the use of and reliance on multilingual resources of various kinds, especially in a language separation/codeswitching perspective, but the emphasis has recently shifted to the repertoire of resources available in the community of speakers (Cogo 2018). While research initially focused particularly on the first languages of the participants, on pointing out where they are from and what language they speak, more recently other linguistic resources in the participants’ repertoire have been explored, particularly in the negotiation of meaning and exploitation of potential meaning-­making aspects (Brunner and Diemer 2018; Kimura and Canagarajah 2018; Pietikäinen 2018). ELF research aligns with the multilingual turn not in terms of separate languages but of linguistic resources and repertoires. Apart from shifting the focus to the repertoire, research has also problematised the relevance of monolingual ideologies in multilingual speakers’ practices. García and Li Wei maintain that “bilingual speakers are often able to recognise themselves only as subjects that speak two separate languages. In so doing, bilingual speakers become complicit in their own domination as they often conform to monolingual monoglossic practices that constrain their own bilingualism to two separate autonomous languages, although at times they may resist by engaging in fluid language practices” (García and Li Wei 2014: 15). The “fluid language practices” which are referred to in this quote are the translanguaging practices that multilingual speakers engage with and that transcend linguistic separations. However, what García and Li Wei also point out is that the same multilingual speakers that engage in fluid, translanguaging practices might also describe their own practices in monolingual terms or constrain their own performance along monolingual lines. ELF practices are constantly brought into being in context of communication, by the participants and their negotiation and co-­construction of the repertoire,

ELF and translanguaging  41 and for that reason they are also not free, simple or egalitarian (see Cogo 2016). Canagarajah reflects on the powerful nature of ideologies, which “have the possibility of being reproduced in social relationships, educational settings, and language interactions” (Canagarajah 2013: 201). Language ideologies also play an important role in ELF communication and language use. In fact, while individual actions may not necessarily be determined by ideological discourses, they are certainly affected by them. The normative character of ideologies has already been demonstrated in empirical studies (cf. Nekvapil and Sherman 2013, among others), but not much research has addressed the extent to which ideologies affect language use, especially how speakers exploit ideologies of monolingualism in multilingual communication. The final part of this chapter is a contribution in that direction. Speakers in the community explored in this chapter also show resistance with their translanguaging practices but, unlike Canagarajah’s participants emphasising the suppressive role of ideologies, they also exploit the monolingual ideological separation of languages and use it to their own advantage  – to co-­ construct meaning and address issues of power relations in communication.

ELF covert and overt multilingual resources Compared to other aspects of investigation, studies on multilingual aspects of ELF are still a minority. And this is quite surprising in light of the fact that ELF is in itself a multilingual phenomenon, originating from language contact and globalisation. As mentioned before, the “Lingua Franca” aspect of ELF is about a contact language perspective and the key role of multilingual resources in negotiating and re-­negotiating the repertoire in communicative interaction. So if ELF is not only about English, it is important to explore in what ways it is multilingual. The multilingual resources that constitute an integral part of the phenomenon are complex, of various kinds, as they depend on the sociocultural context of communication, the constellation of participants and their linguistic repertoires. In this study I am going to explore two main kinds: the covert and the overt multilingual resources. The linguistic resources in an individual repertoire are inevitably constructed, adapted and changed in contact with other linguistic resources and in interaction within a specific context. And how these resources are constructed in an individual’s repertoire can sometimes be difficult to see, not so clear or evident. These covert multilingual influences are of a more cognitive nature, resulting from contact in the repertoire of the participants, but they seem English on the surface. Beneath the surface, however, covert resources offer insights into how speakers “make” their repertoire, what resources they bring with them, and the knowledge and experience that shapes their language. Speakers may draw on similarities in their repertoire to create new words or phrases. For example, Hülmbauer (2011) explores the use of coinages, such as “dictature”; “card”; “overfulled”, which are traditionally described as “false friends”, but she re-­coins as “true friends” (Hülmbauer 2011: 142) because they do not create problems in the conversation and they actually help the participants’

42  Alessia Cogo understanding, relying on the similarities within the participants’ repertoire of resources. In other situations, however, where speakers may not have the same knowledge of linguistic resources or may not share a similar repertoire, then the covert usage may be translated and/or explained or negotiated. Cogo and Dewey (2012) explore some examples of covert resources in their study of discourse strategies used in ELF naturally occurring conversations and show how they are negotiated. Among these is the use of idiomatic expressions translated from the speakers’ own L1, or other languages in their repertoire, into English and explain their meaning. For instance, the expression “step on the stones” is a literal translation from a Japanese idiomatic expression, which is used to warn someone to be attentive, not to lower their guard, because this may cause one to step on stones and fall. Stepping on stones (Cogo and Dewey 2012: 131–132; Cogo and House 2018: 212–213). 01 S1: 02 03 S2: 04 S1: 05 S2: 06 S1: 07 S2: 08 S1: 09 S2: 10 S1: 11

=it’s okay so she’s re-­that’s why I said don’t step on the stones (.) cause so relaxed you-­you might feel (chuckle) you know stepping (.) ano: just do not think anything wrong ahead oh all ri:ght [I see [yeah but is it just a Japanese:: eh:: way of saying? yeah and when we finish something mhm mhm sort of our concentration (.) will not be (.) ah how can I explain (.) not disturbed (.) we can’t concentrate on something

S1 introduces the Japanese expression (line 01) and explains that it can be used if people are too relaxed (line 02) and they might trip (or step) on the stones. Because people do not think anything wrong will happen (line 04), they may lose concentration (line 10), thus falling and getting hurt. S2 does not speak Japanese, and S1 is aware of that – S2 performs a comprehension check explicitly asking S1 whether the expression is a Japanese way of saying (line 07). This might mean that S2 has already guessed that the idiom originates from Japanese and may want to clarify its meaning. S1 confirms this by replying yeah and offering an explanation of the motivation for using the idiom and when we finish something (line 08). In this example the Japanese participant, who translated the idiom, only used English wording and not the Japanese original, but she engaged in a negotiation of the covert multilingual resource, which otherwise would not have been understood by the interlocutors. So, although covert resources may not be obviously multilingual, or multilingual in form, they are representations of the repertoire of the speakers and may thus contribute to negotiating and/or developing understanding. Overall, the studies mentioned previously (and others that have analysed multilingualism and ELF, such as Hülmbauer 2009; Pölzl and Seidlhofer 2006;

ELF and translanguaging  43 Klimpfinger 2009; Cogo 2009, 2010; Seidlhofer 2011; Vettorel 2014) operate mainly from a framework of language separation, rather than translanguaging. However, what is included in covert resources are processes of translation, or rather, transformation, whereby the linguistic resources used are not necessarily the result of direct translation from one language to another (or a switch from language A to language B), but of contact and therefore transformation into something different and new. According to Li Wei, the transformative capacity of the Translanguaging process is “not only for language systems, but also for individuals’ cognition and social structures” (Li 2016: 8, emphasis mine).2 Thus, exploring covert resources can provide a glimpse into the cognitive repertoires of the users and their understanding. The second kind of multilingual resources, overt resources, include expressions of multilingualism that are obviously recognised as mixing of different linguistic resources, such as codeswitching and translanguaging. Even though these two language alternation phenomena are part of different traditions of research and different paradigms of reference, they both have in common the obvious display of heterogeneity. Although, in line with García (2009), I  see codeswitching as part of translanguaging (also cf. later in this chapter), traditionally codeswitching is considered a different phenomenon, one that involves language alternation of traditionally separate languages. Codeswitching has probably been the most researched overt phenomenon of multilingual alternation, and there has been an increasing number of studies on codeswitching in ELF contexts too. Cogo’s (2009) study shows how codeswitching can be used as an accommodation strategy, and speakers’ creative use of their multilingual repertoire is helpful in maintaining social relations and constructing group membership. Sociocultural aspects are also central in Klimpfinger’s 2009 study, which analyses codeswitching examples in the VOICE (Vienna-­Oxford International Corpus of English) corpus from a conversation analytic perspective, and finds that signalling culture is one of the common functions of codeswitching, together with specifying an addressee, appealing for assistance and introducing a new idea. Vettorel’s (2014) study of blogging in ELF online communities found that codeswitching is a common practice among bloggers. All this research has also shown how ELF viewed from a multilingual perspective can be “liberating since for many speakers, its use seems to encourage a relatively unconstrained exploitation of the resources of English and a readiness to draw on resources available through plurilingual channels” (Hülmbauer and Seidlhofer 2013: 400). What all this ELF research points us to is that while there has been an attempt to explore the multilingual aspect, the idea of English resources separated from other languages is still quite widespread even within ELF. Lately the approach focusing on systematic codeswitching has been under renewed scrutiny, also by ELF researchers (see Jenkins 2015), when considered in relation to another type of language alternation, i.e. Translanguaging. The phenomenon of Translanguaging “goes beyond what has been termed code-­ switching” and “includes it, as well as other kinds of bilingual language use and bilingual contact” (García 2009: 45). In translanguaging, linguistic resources are

44  Alessia Cogo not separated or treated as distinct systems, they are instead creatively transformed into new linguistic realities. This translanguaging space, where multilinguals transform language separation into new possibilities, “has its own transformative power” and “it entails tension, conflict, competition, difference and change in a number of spheres, ranging from ideologies, policies and practices to historical and current contexts” (García and Li Wei 2014: 24). Adopting a translanguaging perspective for ELF means approaching the field in a way that the interplay between multilingualism and “English” defies the countability usually associated with individual languages, and possibly emphasises translanguaging in more super-­diverse contexts (cf. Cogo 2012). While the translanguaging approach seems incompatible with the term codeswitching, the idea of language separation may still play a role in ELF communication. Even if codeswitching may seem to be an outdated approach for researchers, multilingual speakers may still be invoking traditional language alternation emphasising separate languages, as I  will explore later. The rest of the paper will hopefully contribute to clarifying the relation between translanguaging and codeswitching in a specific community of practice of ELF speakers and hopefully also respond to Jenkins’ (2015) call for a development of the concept of translanguaging in relation to ELF.

Methodological observations The data collected and analysed in this chapter comes from a research project into ELF and multilingualism in a professional community of practice. The empirical data collection was conducted in a small Information Technology company (under the pseudonym of “IT Services”, also cf. Cogo 2012) based in London, UK. The company provides IT services for small businesses and individuals locally and in Europe. With a total number of eight members employees (including the two directors), the company’s internal communication takes place in their UK office, but communication with clients is mainly virtual and takes place from wherever location the IT Services staff and clients happen to be. The fieldwork conducted in the company included observational data, document collection, interviews and focus groups with the employees and directors, audio recordings of naturally occurring conversations and retrospective interviews with the participants to the recorded conversations. The overall aim of the project was to investigate the sociolinguistic practices of a transnational workplace from a linguistic ethnographic perspective (Copland and Creese 2015), and the extent to which multilingual resources were part of the sociolinguistic practices of this ELF professional community. The conversations analysed in this chapter took place over Skype, a voice-­ over protocol system that allows for both video and audio calls. They involve three main participants in a long-­term project concerning the set-­up of an online invoicing system for a German-­based client. The participants involved in the conversations are (here identified with pseudonyms) Pedro (P), the IT consultant; Maria (M), the director of the German company; and Helmut (H), the accountant

ELF and translanguaging  45 working for the German company. They all developed quite rich and complex repertoires of resources. P and M are from Spanish-­speaking backgrounds and H from German background, but they have been living in different places and experienced different sociolinguistic realities, which added to their knowledge of other linguistic resources. According to the interviews carried out with the three participants, and for the purposes of this chapter and the conversations analysed later in this chapter, it is important to point out that P speaks a bit of German, M speaks German rather well and H speaks Spanish to a decent level. They all speak English for their work practices. The complexity of the participants’ repertoires creates challenges from a transcription perspective. When representing multilingual resources in their translanguaging practices it seems contradictory to mark clear boundaries between participants’ resources. Translanguaging is about crossing boundaries and merging them, to the extent that linguistic resources sometimes do not clearly pertain to one language or the other. In the data analysis and transcription in this chapter, however, I  have still used boundary marks for multilingual resources because the purposes of the chapter is about showing that these resources are used and how they are used, though I hope the analysis will show that it is important to go beyond these forced compartmentalisations.

Data analysis Covert resources: the ELF translanguaging space This section provides two examples of exploiting covert multilingual resources in ELF communication. As covert multilingual resources, the following excerpts may sound like “English” in form, but they are what traditionally would be seen as translations or transformations from another language. In other words, they may be seen or heard like “English”, but they are covert representations of the participants’ multilingual repertoire, or transformations resulting from their multilingual pool of resources.

Excerpt 1 – P (Spanish-­speaking background) and M (Spanish-­ speaking background) are discussing an invoice with H (German background). M refers to her customers’ tendency to complain about payment. 1 2 3 4

P: M: →

she want for each type eh method of payments a different thing at the end of the invoice here yes no but this here it say please send me the money and they reclamate always

The verb “reclamate” in line 4 is particularly important here because the participant is using it to make an argument. She is asking P to mention the methods of

46  Alessia Cogo payment on the invoice on the grounds that if they are not included customers would complain (in her words, reclamate).3 Moreover, if we take into account the repertoire of the ELF speakers involved in this exchange it will become obvious how the creative use of “reclamate” can be understood by the participants involved. The verb “reclamate” could be related to Spanish, where the verb “reclamar” means to complain/to express dissatisfaction. This is used in conjunction with – ate, which is a suffix for verb formation normally added to words of Latin origin (as in “locate”, “alleviate”, “activate” and similar). Spanish is the first language of P and M, but for the German accountant (H), who is also present in the conversation, this term could also have been understood because of his knowledge of Spanish, which he developed by working with M. The term “reclamate” can therefore be assumed to be shared by the speakers in the conversation because of their multilingual repertoires, either as coming from their first language or because they are part of a community of practice, which has seen this term being used in their common repertoire. There is no evidence in the analysis of the conversation that shows that “reclamate” might create problems, rather, because of its covert multilingual nature, it can even be assumed that this term would facilitate the process of understanding. In the second example, I am going to analyse another covert resource, which is something of a prepositional nature.

Excerpt 2 – H (German background) and P (Spanish background) are describing the list of items included in the invoice. H wants to draw P’s attention to an item towards the end of the list. 1 H: down under it says first two 2 P: [yes 3 H: [position one position two position one is the 4 article no? In line 1, “down under” could come from the German “unten drunter” (or “weiter unten”), or “underneath”, and possibly the German “unterhalb”, which means “below, downward”. H explained that he meant “further down”, not sequentially but jumping some of the items in the list. The use of “down under” may not seem that important in the light of the whole conversation, but when we consider that the participants are in a virtual meeting, sitting at their desks, located in different offices, and looking at a document (the invoice) on a computer screen, the use of this expression becomes much more relevant. The speakers cannot point to an item in the invoice because they do not see each other, but they can try to use more visual or spatial expressions to describe what they are looking at. So, while the expression in line 1 has potentially been translated or transformed from German, which might be considered as problematic for people who encounter it for the first time, its use in the conversation does not seem to hinder the participants’ comprehension and in fact it helps P understand what they are talking

ELF and translanguaging  47 about (confirmed by “yes” in line 2) and thus continue to address H’s concern with the invoice. The expression “down under” is a creative use of the participant’s multilingual repertoire to make the explanation more effective. In other words, just using the adverb “down” would not have been as clear as “down under” because the item they are referring to is further down in the list and not the last one. Using the multilingual repertoire in translation helped H express himself more effectively. While previous examples pertain mainly to the lexico-­grammatical level, instances of covert resources within pragmatics and phonetics/phonology can also be found. The pragmatic covert resources, however, are more difficult to identify since they are not necessarily translation or transformations of multilingual resources, but interpretations of intercultural strategies or discourse conventions. The phonological covert resources, instead, can be identified more easily throughout the data. One example is when an English word is pronounced in a German way, or a Spanish way, by the participants.

Excerpt 3 – H (German background) and P (Spanish background) are talking about the term “invoice maturity”. 1 2 3

H: P: H:

when you go down you know yes what is invoice maturity(/məˈturəti/) P

In line 3, H pronounces the term “maturity” as /məˈturəti/, with a clear German influence in the pronunciation of the dental plosive as /t/, rather than /tʃ/, and in the pronunciation of the back vowel as /u/, rather than /ʊ/. Pronunciation is indeed a very important level of analysis because its covert multilingual influence actually becomes overt in the very moment when pronunciation takes place and the word, the turn or utterance are produced. In that sense, pronunciation examples cut across the divide between covert and overt resources, clearly showing how it may be rather difficult for ELF multilingual resources to fall clearly within one category or the other. Finally, the implications for this kind of covert creativity, go beyond facilitating understanding and are much more profound. First of all, participants are multilingual, and it is difficult to clearly demarcate where one language begins and another ends. The translation is not the result of a clearly identifiable language A translated into language B, but the result of language contact and transformation, where influences of German and Spanish resources, all developed in different ways and to different extent in an individual’s repertoire, get transformed into some other linguistic resource, something that is called “English”. Secondly, we could argue that the notion of “community of practice” here is clearly grounded on a notion of repertoire which is not unitary, or pertaining to one language “English”, but on a shared translanguaging space, where linguistic resources coexist, interact and make meaning with non-­linguistic, multimodal resources. Finally,

48  Alessia Cogo participants in these exchanges are not trying to find the “correct English word” but are using whatever resources available in their multilingual repertoire and transforming them to make them accessible and understandable in context. The translanguaging processes at play in these covert resources examples are based on transformations and possibly negotiations, rather than on adherence to unitary concepts of language. Overt resources: translanguaging and boundaries While the previous three examples explored the use of covert multilingual resources, the following are examples of explicit, therefore more visible and obvious, multilingual resources. These are excerpts concerning ELF translanguaging, whereby “[L]anguage users employ whatever linguistic features are at their disposal to achieve their communicative aims as best as they can, regardless of how well they know the involved languages” (Jørgensen 2008: 163). In excerpt 4, it is interesting to observe how multilingual resources are used. Here the Spanish and German parts are not used only by the Spanish and German “native” speakers respectively. M and P (both of Spanish speaking background) use German in their negotiation of what is included in the invoice, and H, who speaks very little Spanish, still uses Spanish resources in the conversation with them (though not visible in this excerpt, we can see how H uses Spanish resources in Excerpt 5, line 8).

Excerpt 4 – M (Spanish background), H (German background) and P (Spanish background) are discussing the list of items that needs to appear on their invoice. 1 2 3 4 5 6 7 8 9 10 11 12 13

P

H P M H P H P

ah: ok ok yeah (aspiration) ah: (2.5) and then the netto (3) pfourteen (2) one two so the position is important one two (2) and then we have the euros (1) and then we have the the porcentage? (.) but has another name says usl u u eh u s t ist unxxx steuer was ist dis? betrag (.) betrag? como prefieres? no betrag is whole amount Pedro ah ok betrag is amount amount ok

From lines 1 to 5, P provides a summary of the different items that should be mentioned in the invoice and their positions on the document and tries to double check with M and H that this is fine. P’s use of German specifically refers to how the

ELF and translanguaging  49 items appear on the invoice. P’s long initial turn is interspersed with long pauses, which are taken on by the participants’ observing the invoice document. In lines 5, P is unsure about the acronym UST and the accountant provides an explanation in the following turn in German. After that, P returns to using German in line 7 when he asks, in a German-­Spanishy way, what the item is. German linguistic resources are part of P’s repertoire and, though in his interview he underestimates his competence in German, he seems to play with it (as in line 7) and uses it actively in the discussion of the invoice. In line 8, M picks up the turn in German and then Spanish to help clarify the matter with H, who adds the translation of one of the items (line 10) and, after repeated confirmation from P (lines 12 and 13), the conversation moves on. This ‘German-­Spanishy way of speaking’ is how P describes his translanguaging practices. In the professional community where he works, and for the specific project with H and M, it is quite common to play with linguistic resources while carrying out their working tasks. The German and Spanish words have been learnt during the frequent conversations among the participants, who have created their “common repertoire of resources”. Methodologically, of course, these observations are possible not because of the corpus data (including the transcript and analysis provided earlier), but because the research has been conducted through fieldwork in the company, observations and retrospective interviews with the participants. As a matter of fact, it is in the retrospective interviews carried out with the participants that the idea of using any linguistic resources available in the participants’ repertoire is repeatedly mentioned. Participants in this community seem to be able to effectively use their multilingual repertoire and enjoy doing that. They show flexible practices of translanguaging – calibrating their repertoire according to how they see their interlocutors’ linguistic repertoire. For instance, H uses only minimum resources of German, knowing that P does not speak much German, though M is fluent. These explanations of the resources participants are familiar with come from the fieldwork and interviews with them and their own comments. However, participants’ conceptualisations of their linguistic resources are not done in terms of repertoire. Instead, their understanding of translanguaging is often explored in terms of language separations, justifying what language they speak to whom and why, and therefore also constructing a representation of “languages” in terms of unitary and separated resources. Thus, though data shows how participants engage in a translanguaging space, they may still perceive languages as separated by clear boundaries and, not only explain those boundaries in their interviews, but they also use them in their practices to certain effects. In Excerpt 5, for instance, participants carry out a clear division of labour between Spanish and English, whereby P’s turns are in English and M’s turns are in Spanish. P knows that in order to include H in the conversation he would need to speak mainly in English, while M knows that in order to exclude H from the conversation and only address P she would have to use only Spanish. The following example of language choices in the discussion demonstrate the participants’ selection of multilingual resources for communicative

50  Alessia Cogo effectiveness. Switching into certain languages to select/include/exclude certain participants is a way of exercising power in the conversational exchange, i.e. decide who to include or exclude.

Excerpt 5 – H (German background), P (Spanish background) and M (Spanish background) are talking about tax on goods. M wants P to have a look at another example of how to deal with this aspect, but P wants to discuss the issue with H, i.e. the accountant. 1 2 3 4 5 6 7 8 9 10 11

P: M: P: M: H: P:

I try to adapt the the(.)the programme to that (.) okay? ya para te estoy mandando ahorita una bu-­ but why if it’s the same just don’t don’t bring it es si otro esta mas ordenado que el nuestro para que veas el dice he says seven percent tax on (.) the the line ninety you see that what my-­ -­ yeah yeah yeah yeah I have it in front of mine I I print it so

Excerpt 5 is an example of the tension between P and M concerning their role and relation to H, the accountant. P wants to include H in their conversation about the tax issue, because he is the accountant and would know how to approach the issue, but M continues to intervene and prevents H from taking the floor. This struggle is clearly emphasised by their language choices: P talks in English while M replies in Spanish, the language that H is less comfortable with. By using Spanish, M is preventing H from actively joining the conversation, though he has a passive understanding of it, as indicated by his turn in line 8, where he manages to start with a bit of Spanish before continuing in English. The fight for including or excluding a participant is here clearly signified by choosing to use or not the linguistic resources that make it possible for that participant to contribute to the talk. While translanguaging offers individual resistance to the monolingual ideological positioning of societal requirements, adherence to the societal monolingual norm is still an option for speakers and something they can use to exercise resistance to the status quo. The individual speakers’ repertoire is here not only interacting with the other speakers’ repertoires, but also with the institutional requirements and the societal discourses and ideologies about language as a monolingual system.

Conclusion It would, of course, be legitimate to think that the community and communication analysed in this paper mainly come from a European context and participants share linguistic resources from European languages, which also makes the

ELF and translanguaging  51 possibility of metalinguistic knowledge and awareness more likely. Furthermore, most studies included in the literature discussed in this chapter also explore multilingualism in a European perspective. However, the Asian continent is the largest one on earth and has the largest population and, consequently, a wealth of languages exist within the Asian territory, with variation and heterogeneity of languages within the continent and in the single states. With this wealth of linguistic resources, it would just seem likely that overt and covert multilingual phenomena, like the ones described in this chapter, would be possible, if not quite common. And while research into multilingual education and the role of ELF in Asia is well under-­way (Canagarajah and Ashraf 2013; Kirkpatrick 2007, 2012; Lim and Low 2009), we need more research exploring the sociolinguistic multilingual practices of ELF in naturally occurring communication involving Asian languages. The aim of this chapter has been to demonstrate that ELF is a translanguaging phenomenon even when the multilingual resources are mainly covert, such as when the participants do not share their linguistic repertoires, and do not have other languages in common other than English. Even then, ELF is multilingual in their use of covert resources. With such a perspective, speakers’ resources are not viewed as narrow discrete systems, but as larger communicative repertoires that transform, combine and overlap in rich and complex ways. Even when on the surface we may be dealing with English only, ELF always relies on multilingual resources and multilingual transformations, which also require a multilingual understanding. This approach also includes understanding language from the participants’ perspectives and the ideologies that they exploit to make meaning. Data shown in this paper demonstrates the flexible and transformative nature of ELF speakers’ repertoires, but the excerpts have also shown that speakers may resist by exploiting the monolingual ideological separation of languages and using it to their own advantage  – to make meaning and invoke power asymmetries. Finally, what I hope this chapter has contributed to point out is that research on ELF and research on multilingualism should not be seen as part of separate disciplines but have much in common and more to contribute to each other.

Appendix Transcription conventions

= latching (i.e. speech following the previous turn without a pause) (.) short pause (unmeasured) (0.5) measured pause [abc] overlapping speech [. . .] omitted transcript @ laughing utterances in the speaker’s first language (L1) are put between tags indicating the speaker’s L1, such as de = German, sp = Spanish. utterances in languages which are neither English nor the speaker’s first language are marked LN with the language indicated.

Notes 1 For lack of space, I will not be able to address these critiques here, but suffice to say that they have been extensively explored in many publications, among which Auer 2007; Auer and Li Wei 2007; Canagarajah 2013; Conteh and Meier 2014; Ortega 2013. 2 This cognitive nature of covert resources also aligns with some conceptualisations of Receptive multilingualism, when users make use of their L1s or some acquired knowledge of each other’s language for meaning making and achieving understanding in communication (see, for instance, ten Thije and Zeevaert 2007). In Receptive multilingualism, as in ELF covert resources, the receptive knowledge of the language and/or variety of the interlocutor is activated. 3 The participants involved in the meeting must have understood the point made by M, and therefore the use of “reclamate”, because nobody is raising the issue. In other words, from a conversational perspective, participants did not start a negotiation of meaning: they did not ask for clarifications or similar. It is not a let-­it-­pass strategy either because the point made is of particular importance to the discussion of the invoice.

References Auer, P. 2007. Monolingual bias in bilingualism research – or: Why bilingual talk is (still) a challenge for linguistics. In M. Heller (ed), Bilingualism: A social approach. Houndmills: Palgrave, pp. 319–339. Auer, P. and Li Wei. 2007. Introduction: Multilingualism as a problem? Monolingualism as a problem? In P. Auer and Li Wei (eds), Handbook of multilingualism and multilingual communication. Berlin: de Gruyter Mouton, pp. 1–12.

ELF and translanguaging  53 Blommaert, J. 2010. The sociolinguistics of globalization. Cambridge: Cambridge University Press. Blommaert, J. and A. Backus. 2013. Superdiverse repertoires and the individual. In I. Saint-­Georges and J.J. Weber (eds), Multilingualism and multimodality. Rotterdam, The Netherlands: Sense Publishers, pp. 11–32. Brunner, M.L. and S. Diemer. 2018. “You are struggling forwards, and you don’t know, and then you . . . you do code-­switching . . .” – code-­switching in ELF Skype conversations. Journal of English as a Lingua Franca 7(1), 59–88. Canagarajah, S. 2013. Translingual practice: Global Englishes and cosmopolitan relations. London: Routledge. Canagarajah, S. and H. Ashraf. 2013. Multilingualism and education in South Asia: Resolving policy/practice dilemmas. Annual Review of Applied Linguistics 33, 258–285. Cogo, A. 2009. Accommodating difference in ELF conversations: A  study of pragmatic strategies. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 254–273. Cogo, A. 2010. Strategic use and perceptions of English as a Lingua Franca. Poznań Studies in Contemporary Linguistics 46(3), 295–312. Cogo, A. 2012. ELF and super-­diversity: A case study of ELF multilingual practices from a business context. Journal of English as a Lingua Franca 2(2), 287–313. Cogo, A. 2016. Visibility and absence: Ideologies of “diversity” in BELF. In M.-­L. Pitzl and R. Osimk-­Teasdale (eds), English as a Lingua Franca: Perspectives and prospects. Contributions in honour of Barbara Seidlhofer. Berlin: de Gruyter Mouton, pp. 39–48. Cogo, A. and M. Dewey. 2012. Analysing English as a Lingua Franca: A corpus-­driven investigation. London: Continuum. Cogo, A. and J. House. 2018. ELF and multilingualism. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 357–368. Conteh, J. 2018. Key concepts: Translanguaging. ELT Journal 72(4), 445–447. Conteh, J. and G. Meier (eds). 2014. The multilingual turn in languages education: Opportunities and challenges. Clevedon: Multilingual Matters. Copland, F. and A. Creese. 2015. Linguistic ethnography. London: Sage Publications. García, O. 2009. Bilingual education in the 21st century. Oxford: Wiley Blackwell. García, O. 2012. Theorizing translanguaging for educators. In C. Celic and K. Seltzer (eds), Translanguaging: A CUNY-­NYSIEB guide for educators. Available from: www.nysieb. ws.gc.cuny.edu/files/2012/06/FINAL-­Translanguaging-­Guide-­With-­Cover-­1.pdf. García, O. and Li Wei. 2014. Translanguaging: Language, bilingualism and education. Basingstoke: Palgrave. Hülmbauer, C. 2009. “We don’t take the right way: We just take the way we think you will understand” – the shifting relationship of correctness and effectiveness in ELF communication. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars, pp. 323–347. Hülmbauer, C. 2011. Old friends? Cognates in ELF communication. In A. Archibald, A. Cogo and J. Jenkins (eds), Latest trends in ELF research. Newcastle upon Tyne: Cambridge Scholars, pp. 139–161. Hülmbauer, C. and B. Seidlhofer. 2013. English as a Lingua Franca in European multilingualism. In A.-­C. Berthoud, F. Grin and G. Lüdi (eds), Exploring the dynamics of multilingualism. Amsterdam: John Benjamins, pp. 387–406. Jenkins, J. 2015. Repositioning English and multilingualism in English as a Lingua Franca. Englishes in Practice 2(3), 49–85.

54  Alessia Cogo Jørgensen, J.N. 2008. Polylingual languaging around and among children and adolescents. International Journal of Multilingualism 5(3), 161–176. Kimura, D. and S. Canagarajah. 2018. Translingual practice and ELF. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 295–308. Kirkpatrick, A. 2007. The communicative strategies of ASEAN speakers of English as a Lingua Franca. In D. Prescott (ed), English in Southeast Asia: Literacies, literatures and varieties. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 121–139. Kirkpatrick, A. 2012. English in ASEAN: Implications for regional multilingualism. Journal of Multilingualism and Multicultural Development 33(4), 331–344. Klimpfinger, T. 2009. “She’s mixing the two languages together” – forms and functions of code-­switching in English as a Lingua Franca. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Press, pp. 348–371. Li, W. 2016. New Chinglish and the post-­multilingualism challenge: Translanguaging ELF in China. Journal of English as a Lingua Franca 5(1), 1–25. Lim, L. and E.-­L. Low. 2009. Multilingual, globalizing Asia. AILA Review 22. Amsterdam: John Benjamins. May, S. (ed). 2013. The multilingual turn: Implications for SLA, TESOL, and bilingual education. London: Routledge. Nekvapil, J. and T. Sherman. 2013. Language ideologies and linguistic practices: The case of multinational companies in Central Europe. In E. Barat, P. Studer and J. Nekvapil (eds), Ideological conceptualizations of language. Bern: Peter Lang, pp. 85–118. Ortega, L. 2013. Ways forward for a bi/multilingual turn in SLA. In S. May (ed), The multilingual turn: Implications for SLA, TESOL, and bilingual education. London: Routledge, pp. 32–53. Pietikäinen, K. 2018. ELF in social contexts. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 321–332. Piller, I. 2016. Monolingual ways of seeing multilingualism. Journal of Multicultural Discourses 11(1), 25–33. Pölzl, U. and B. Seidlhofer. 2006. In and on their own terms: The “habitat factor” in English as a Lingua Franca interaction. International Journal of the Sociology of Language 177, 151–176. Prada, J. and T. Nikula. 2018. Introduction to the special issue: On the transgressive nature of translanguaging pedagogies. EuroAmerican Journal of Applied Linguistics and Languages 5(2), 1–7. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. ten Thije, J.D. and L. Zeevaert (eds). 2007. Receptive multilingualism: Linguistic analyses, language policies and didactic concepts. Amsterdam: John Benjamins. Vettorel, P. 2014. English as a Lingua Franca in wider networking. blogging practices. Berlin: de Gruyter Mouton.

Part II

Spoken and written corpora in ELF research

5 Good texts in non-­standard English ELF and academic writing Anna Mauranen

Introduction Writing up research is increasingly international, and international publishing in effect means publishing in English. Universities have oriented to internationalisation all over the world, and as Franzmann et al. (2015) point out from a sociology of science perspective, research science has probably never been as international as it is today. In brief, academic writing in higher education and in research has undergone a major transformation since the end of the 20th century from a largely national regional practice which it had turned into over the last two centuries or so (see also Crawford 1985), to a thoroughly international one, which was closer to its origins. This has also meant that writers, whether university students or authors of research publications, are widely using English as a lingua franca, which for most of them is not their local language. This situation poses a challenge to the study of English: how can we best make sense of English of the kinds that are in contemporary use around the world? What frameworks should we apply to investigating English that has become de-­nationalised, highly variable, and seemingly very unstable in its contemporary global context? The study of English as a lingua franca (ELF) has reversed the traditional order of language study, where research on the written mode has preceded that of the spoken. This has been for good reason: speech changes faster than writing, which tends to be more conservative and norm-­preserving. Educational institutions stress correctness and the standard language, and published texts go thorough editing cycles, often by several people, before they come out. The spoken mode, by contrast, tends to show variation and incipient change much faster than writing, and therefore characteristics of ELF are more readily observable in speech. We have learned to appreciate the wide situational variability in ELF speech (Jenkins 2015), and at the same time to notice new regularities that also emerge in speech, amongst all the variation (e.g., Mauranen 2006, 2012). Recent research, typically using large databases, has helped us get a grasp of how ELF compares to spoken varieties of World Englishes (Paulasto 2015), while research into written ELF is still lagging behind research into speaking. The situation is nevertheless changing. Writing, especially research writing, always of interest to scholars and practitioners, has begun to attract interest in its more traditional

58  Anna Mauranen forms (for instance Carey 2013; Ingvarsdóttir and Arnbjörnsdóttir 2013; Laitinen 2016, 2018), as well as in digital formats (Hynninen 2018; Laitinen and Lundgren forthcoming; Mauranen 2013; McGrath 2016; Vetchinnikova and Hiltunen forthcoming). The first electronic corpus of written ELF was completed in 2015 (WrELFA 2015), as an international collaboration. In this chapter I discuss some issues of academic writing in ELF, and how they may be approached with different methodological orientations even if the same databases are drawn on. I  make use of the WrELFA corpus to explore written academic ELF, and my approach involves three perspectives (cf. Mauranen 2012, 2018): macro-­social, micro-­social, and cognitive.

The framework, data, and methods A research object like English as a lingua franca can be approached from many different perspectives, without a need to establish one ‘best’ approach. Like any phenomenon involving human activity, it will not stay still, because arguably all ongoing activity adds to the aggregate of instances of activity ad infinitum. In this sense there is no way ELF could be captured in its entirety – not because it is large, a worldwide phenomenon, but because it is language activity. We could never hope to capture all language activity even in the smallest language community in existence as long as it is alive. Neither do language communities maintain steady, unchanging or unvarying practices or forms for their communicative purposes, or, normally, abstain from communicating with people outside their communities, which leads them to exchange influences with other language communities. In this way, language is not only infinite in quantity, but varies and changes continually in use, so that any description can only be partial and limited. At the same time, the unlimited nature of language phenomena offers in principle limitless possibilities for research topics and perspectives. Obviously, any particular study needs to focus on something to keep manageable, but it is free to select which perspective it adopts. What I  am looking into in this chapter is a comparatively narrow choice of genre  – written texts in academic environments – but seen from three different perspectives. Each perspective entails different choices of data and different theoretical considerations, but together they yield a more rounded picture than any of them would on their own and contribute to a holistic understanding of what is at stake in writing texts in the broad context of academia. Data This chapter is concerned with different dimensions of one phenomenon, writing in academia, using English as a lingua franca. In order to relate the different vantage points to language in use, it draws on corpus data. This section describes the corpus used, and the next discusses how such corpus data can be used to tease out target phenomena.

Good texts in non-­standard English  59 The database used in this chapter is Written English as a Lingua Franca in Academic Settings, or the WrELFA corpus, which is the very first corpus of written ELF. The compilation began in late 2011 and was completed in early 2015. The texts cover high-­stakes academic genres in different research fields, and over 500 unique authors with at least 37 first languages (approximate figures are due to the indeterminacy of certain types of texts, see later in this chapter). The aim was to cover academic writing practices on a broad scale, ranging from texts circulated within academia to texts that reach the wider public. Among the target text types are evaluative reports, such as examiners’ and peer reviewers’ reports, and digital media such as research blogs with their comment threads. The target author is the academic user of English as a lingua franca (ELF), and texts have not undergone professional proofreading or checking. It is thus a corpus of second-­language use (SLU) in written scientific communication. The corpus consists of 1.5 million words drawn from three academic text types: unedited research papers, PhD examiner reports, and research blogs. The three parts of WrELFA, or its subcorpora, thus are 1

2 3

The SciELF corpus, a stand-­alone subcorpus of research papers that have not undergone professional proofreading services or checking for example by a native speaker of English (759k words, 50% of total). The subcorpus was compiled as an international collaboration PhD examiner reports for submitted doctoral theses from six faculties in the University of Helsinki over a 2-­year period (402k words, 26%). These are longish texts (3–10 pages) in the Finnish tradition and naturally unedited Academic research blogging, a sample of posts and discussions from 40 different research blogs, all of which are maintained by L2 users of English (372K words, 24%). It is important to note that these are not science journalists’ blogs, but researchers’ own blogs

The primary criterion for inclusion was that the author is neither a native speaker of English nor based professionally in an L1-­English country. We also excluded science journalism, which is mostly written by professional journalists. Thus, all 40 bloggers are independent academics who blog about their own fields of interest and are based outside of an ‘Inner Circle’ country. Posts were only considered that were related to the blogger’s research field, including conference reviews, discussions on professional life, and ‘metablogging’ about research blogging itself, to the exclusion of blog posts on non-­professional interests or personal hobbies. Longer, more developed posts were preferred, together with those including comments and discussion. PhD examiner reports is a virtually unexplored genre of academic writing, and therefore some examiner reports were included that were written by native speakers of English, as well as L2 users based in L1 English countries. The SciELF texts were collected based on two main requirements: the author(s) should not have English as an L1, and the

60  Anna Mauranen text should not have undergone professional proofreading services or language checking by an English native speaker. The large majority of these texts were obtained as drafts. In a few cases, we accepted published articles under the condition that the authors could verify that language revisions had not taken place as a condition for publication. The corpus was designed as a written complement to the spoken  ELFA corpus (www.helsinki.fi/elfa) with similar markup and metadata. WrELFA employs a broad binary categorisation of texts into the sciences (category “Sci”) and disciplines in social sciences and humanities (category “SSH”). Natural sciences are the largest group among the Sci texts, (63% of words) followed by medicine (22%) and agriculture & forestry (11%). The SSH texts include social sciences (44%), humanities (36%), and behavioural sciences (18%). At least 37 unique L1s are represented, along with an undetermined number of blog commenters whose identities cannot be verified. As with other ELF corpora, English native speakers are included among the authors (8%), in this case among the blog commenters and PhD examiners. Finnish is the largest L1, but with only 14% of total words. For a more detailed description of the corpus, visit the corpus website (www.helsinki. fi/elfa/wrelfa). Methods This chapter seeks to show that corpus data can offer very useful material for linguistic research on different scales. For charting the big picture of linguistic tendencies and possible ELF-­specific uses and features, large databases and corpus linguistic work provide the basis for spotting macro level tendencies, many of which have been established in sociolinguistic research on language variation and change. At the level of micro level tendencies, that is, social interaction, corpus data is particularly useful for analyses of spoken discourse. However, certain features of written text, even academic, are best seen as interactional, relating to the reader– writer relationship (e.g., Ädel 2006; Hyland 2005; Smart 2016). Since interactional aspects of written monologues comprise only a limited set of linguistic features, such analyses are obviously more restricted than analyses of spoken dialogue. Monologues in speech and writing are very close to each other in terms of interactional expressions, while spoken and written dialogues, in turn, show notable similarities (Mauranen forthcoming). The main methodological use of corpus data for interactional research is that if a corpus is compiled of discussion events, these can be selected for qualitative scrutiny as independent episodes that are available as transcripts, soundtracks, and accompanying field notes if the metadata has been carefully compiled. Written dialogue data, such as blog comment threads (Mauranen 2013, forthcoming) or message boards (Smart 2016) can be used in the same way as speech events, in other words as discourses analysed by qualitative means. For cognitive analyses, a qualitative approach is mostly the most appropriate, because the interpretation of a speech extract cannot realistically be based on a

Good texts in non-­standard English  61 very extended stretch of language at a time, with the restrictions of the speakers’ working memory. Obviously this constraint does not apply to written text, but in effect written, digital dialogues do not use very long turns compared to spoken dialogue (Mauranen forthcoming). Cognitive analyses, however, also benefit from the fact that corpora have large amounts of instances from a large number of authors or speakers. To avoid arbitrary or excessively speculative interpretations of individual occurrences of particular items or forms, for example, the availability of other similar instances lends credibility to explanatory reasoning, while their absence renders individual instances less interesting; they could, after all be just random slips of tongue. Interestingly, instances that lend themselves to qualitative analysis at micro-­ social or cognitive levels, also reflect the macro level trends in the language if they occur repeatedly. In this way, corpus data can be brought to bear on distinct but interrelated phenomena and illuminate their interconnections.

A macro-­social perspective: the wider context for research writing today The major large-­scale phenomena that typify our era are undoubtedly globalisation and mobility. English is the language of both. It is the first really global lingua franca in history, and currently the default language in a wide range of cross-­border encounters. Many scholars have been alarmed by this development (Phillipson 1992; Ammon 2007; papers in Holmes and Dervin 2016), while some predict that it will be soon replaced by another language, such as Chinese, and others like Graddoll (2006) that it will give way to multilingualism. Perhaps two things are worth keeping in mind with regard to such views: first, objections to English do not alter the fact that it is exceptionally widespread, and thus of great interest to linguistic research, and secondly ELF, which is the largest use of English today, in itself already incorporates multilingualism in many different ways (Cogo 2012; Jenkins 2015). Awareness of ubiquitous mobility has already changed social sciences (Sheller and Urry 2006; Urry 2007), and the same awareness has rapidly been making its way to language sciences (e.g., Milroy 2002). Both research fields have traditionally been based on an assumption of stable communities, with consequences to selection of research foci and theory development. Despite changes over the last decade or so, it is still true that most language research is based on comparing, contrasting, and otherwise exploring relatively stable, distinct, systems of language and community. Adopting a macro-­level perspective on ELF in academia, work from social sciences and higher education research becomes relevant, along with historical language studies. ELF in academia Academia is a prime example of contemporary mobility: university ‘internationalisation’ and student mobility have exploded after the turn of the millennium

62  Anna Mauranen in Europe, which is reflected in exponential growth in the number of English-­ taught programmes in countries with various national and official languages (Wächter and Maiworm 2014). Similar trends can be seen in other continents. Competition for students rises as age cohorts and resources for higher education diminish in the Western world, and success indicators in some university rankings include numbers of international students and staff. The effect has meant changes in patterns of language use on campuses (see, Jenkins and Mauranen 2019). Research has undergone a transformation along the same lines: it is permeated with mobility, collaboration across country boundaries, and innumerable international research projects. Scientists are moving between laboratories and research centres in different countries and continents throughout their careers. Such individuals, as Franzmann et al. (2015) point out, are disconnected from any national community. These developments have also changed the way scientists communicate. In addition to international meetings and publications, English is the default language for communicating in immediate research environments, within their multilingual groups. Research communication and the exchange of ideas have become both global and immediate. The rapid internationalisation of the academic world has also led to protective reactions in many countries, where active groups seek to protect the local university systems and local languages against the perceived threat of competition from international researchers and English. Internationalism is nevertheless not new to academia. Universities were first founded in mediaeval Europe in the 11th and 12th centuries when all over Europe Latin was the language of learning which smoothed the path of international scholars. However, even though there was this shared language for actual scholarly activities, the actual linguistic landscape must have been much more varied. Given the variety of languages spoken by the participants, local and other shared languages must have been used beyond the core academic practices in various combinations (as for example in Finland, see Kajanto 2000). Thus, while it is not unreasonable to say that universities were built on a single lingua franca, it is important to bear in mind that international institutions necessarily harbour a multitude of languages, and that their everyday activities involve switching, crossing, and mixing language resources (see also Jenkins and Mauranen 2019). The linguistic consequences of the most recent wave of internationalisation and the consequent formation of international communities for example in business and in research have been the growing detachment of the common language, English, from any of its national roots. ELF, then, has become the new normal in English. Linguistically, ELF is a complex, second-­order contact between ‘similects’ (Mauranen 2012, 2018) or, roughly, a contact language between parallel contact varieties that are specific to a given L1-­L2 combination. Socially, it is the default language of academic activities amidst a multitude of other languages used beside it, crossing it, and alternating with it.

Good texts in non-­standard English  63 Research publishing Research publication has got its share of the strive towards internationalisation in universities: individual researchers are under pressure to publish in English, because universities compete with the number and quality of their international publications. Different research fields appear to take the comparatively new language situation on board somewhat differently (Gnuzmann and Rabe 2014; Kuteeva and Airey 2014), with the sciences largely accepting the situation without much questioning. As a physicist colleague put it to me recently: “in sciences, bad English is our normal everyday language. It’s what we use”. In the humanities and social sciences different attitudes tend to prevail; unlike sciences, most of their publications are single-­authored, and the thrust of the argument is expressed verbally, less by formal, numerical, or visual means. A strong orientation to the verbal gives rise to a commitment to ‘good language’, thus a preference for what authors feel are their strongest languages, and to foreign languages being equated with native, near-­native, or native-­like use. As a high-­stakes social activity, academic publishing is heavily regulated, with its primary goal to ensure research quality. This is what motivates quality control methods like peer review. Even though language control is also practised in many research journals, and many still ask non-­natives to have their texts checked by a native speaker, this is getting less common. Top-­ranking journals simply wish to put the emphasis on research quality and play down the role of ‘correct’ native-­ like language (Drubin and Kellogg 2012). This is not limited to natural sciences, so for instance Applied Linguistics or JELF do not mention a native speaker check. Linguists have been pointing out for a long time that language revisers or language brokers’ comments can be trivial (Ventola and Mauranen 1991) or downright counterproductive (Owen 2011). With regard to the level of knowledge claims authors make, language revisers can simply assume that the authors know what they are doing (Mauranen 1997) and leave the responsibility to the author, or if they make interventions, these can lead to shifts in the knowledge claims authors wish to make (Lillis and Curry 2010). Writing has also changed as a consequence of digitalisation. Boundaries of the public and the private have shifted, and given rise to new types of writing (e.g., Miller and Shepherd 2009; Myers 2010). Linguistic consequences include more speech-­like qualities in written text, (Grieve et al. 2011), together with a widespread use of ELF and language-­crossing (e.g., Mauranen 2013). Digital changes have also impacted academic writing (e.g., Kuteeva 2016; Luzón 2013; Rowley-­Jolivet and Campagna 2011). Digital publishing has not only enabled the inclusion of data and extra materials in publications, but led to increases in both supply and demand for open access. New practices of collaborative writing have begun to take shape via digital means (see, e.g., McGrath 2016). The most drastic changes have perhaps been seen in the birth and proliferation of video journals, which challenge the self-­evident dominance of the written word in research publishing. This trend will undoubtedly not end there but augur major changes in research publication in the foreseeable future. In brief, our print-­based notions of research writing have begun to erode, and there is no turning back.

64  Anna Mauranen ELF is deeply intertwined with the digital world, which knows no borders. While other languages have been increasing their proportions on the Internet as Graddoll (2006) predicted, English is hardly likely to loosen its grip on science and scholarship. The effects of digitalisation are likely to turn the tide towards ELF and away from Standard Englishes, since there are fewer opportunities and less demand for linguistic gatekeeping in a world of open publishing. Texts (and videos) will be increasingly regulated by participants in the often self-­selected and largely open communities that gather around given sites, blogs, or researchers, where moderators, not language experts are gatekeepers. ELF is one of many kinds of English use in the world, and it is large-­scale, usually corpus-­based studies that are best suited for putting these many different varieties, functions, and uses in relation to one another. A multivariate typological analysis of grammaticity and analyticity-­syntheticity was carried out by Laitinen (2018) with regard to ELF, indigenised varieties of English, learner language, some dialects, and ‘core’ Standard Englishes. Importantly, his analyses show that in structural terms, ELF is not particularly similar to learner language, but rather to many Outer Circle Englishes, and perhaps most interestingly, especially written ELF is close to Standard English varieties. Such findings would seem to lend support to the claim that ELF, and Second Language Use (SLU) more generally, are different from SLA, as argued in Mauranen (2012). It also shows how large, comparative corpus research throws light on macro-­level language phenomena that cannot be observed by the naked eye.

A micro-­social perspective: text norms and effectiveness The micro-­social perspective mediates between the macro and the individual and consists in social interaction. On the face of it, the primary domain of social interaction would obviously seem to be confined to face-­to-­face spoken interaction. It need not, however, mean only that, as many written text researchers have argued for a long time. Researchers have used notions like the ‘anticipated reader’ or the ‘hypothetical reader’, for the imagined reader that the writer is presumed to have in mind when writing, “anticipating his/her reactions and adjusting his/ her writing accordingly, to facilitate communication” (Bhatia 1993: 9). Whether writers really operate like that, on the whole it is the case, that, as Biber et al succinctly put it: “All language is dialogic, whether the addressee is present or not” (1999: 213). Written text therefore internalises features of speech, as Sinclair (1982/2004) observes. Social interaction in its prototypical form of interactive speaking plays another role in relation to writing: it surrounds, precedes, and follows written texts, by prompting, evaluating and commenting. Spoken interaction regulates text practices and socialises us into its conventions and norms. Top-­down interactional work takes place by diffusing norms, standards, ideological and conventional positions; it serves to regulate standards, received notions about correctness, and manifests itself very distinctly in educational systems. Socially established

Good texts in non-­standard English  65 authorities play a central role; prestigious grammars and dictionaries are appealed to as the ultimate authorities on correctness, and teachers draw on them in their day-­to-­day work. In brief, top-­down interaction serves to diffuse and establish standards, or what might be called imposed norms. By contrast, bottom-­up interaction works primarily in the co-­construction of what can be seen as spontaneous norms. Even though these primarily apply to the group that jointly constructs them but not necessarily beyond it, spontaneous norms can nevertheless also spread more widely when group members disperse and diffuse them where they move. As Agha (2003) points out, instead of being steered by institutions only, the legitimation processes of language also crucially depend on the activities and discursive interactions of the people who are connected to each other via institutions (see also Hynninen 2016). There is a constant tension between imposed and spontaneous norms, with imposed norms standing for stability and conservation, and spontaneous norms initiating changes. In (1), we can see language norms diffused in interaction in a dialogue from the ELFA corpus. (1)   1  . . . part of it to develop a criteria to separate those two forests and   2 then about the current situation er this is kind of all what i found er   3 at least from F-­A-­O forestry er NF NFP update was that the revision of forest policy statement is underway   4 what is FNC and   5 sorry what   6 yeah   7 what is FNC   8 can you tell FNC repeat the acronym   9 was it forest national corporation or [(xx)] 10 [it’s er] not proper english language but it’s the official name forests national corporation 11 it’s kind of the body that [er] 12 [that] is a parastatal forest [er] [yeah] autho-­ er forestry er 13 organisation forests national corporation that it happens to be its official name 14 and then there’s er three key changes that has er occurred. . . . The example comes from a graduate seminar where one of the speakers (S7) is doing a presentation. Another participant (S14) interrupts to ask the meaning of an acronym that the presenter has used slightly earlier (what is FNC, line 4). The seminar leader (S13) immediately attends to the question (sorry what, line 5), and then invites the presenter to explain it (line 8). The presenter gives the full form but also expresses some uncertainty, upon which the seminar leader confirms the answer (line 10) with commentary on the term. This is the key point of the exchange: the speaker indicates his negative evaluation of the

66  Anna Mauranen appropriateness of the name in terms of linguistic correctness (it’s not proper English) but at the same time reinforces it and passes it on as the term that is to be used on account of its status. In this way, the speaker who is neither a linguist nor a native speaker of English, assumes linguistic authority, apparently on the strength of his authority based on his expertise in his academic field. Moreover, he makes a distinction between a standard (proper English), or imposed norm, and actual use, which is clearly a spontaneous norm in this case, a bottom-­up development in an organisation that has coined this name for itself. The use resembles endonormativity in indigenised Englishes in that it is not derived from Standard English but arises from usage, but unlike them, it is not a local norm in a geographic sense. The organisation is international; thus the norm is translocal. Standard languages are primarily constructed on the basis of the written text and for the written text. It is of course easier to fix one’s attention on something relatively permanent like writing than on the fleeting flow of speech. Traditions of standardising and code fixation (Nevalainen 2003) have thus focused on the written text, and we might also expect discourses of standards to focus on the written form. In the earlier example, the presentation was accompanied by slides, and like in many other presentation episodes from the same database (the ELFA corpus), acceptability discussions were typically anchored in something that appeared in written form (like various acronyms in (1). Niina Hynninen (2016) found similar tendencies in her more extensive research where she looked into spontaneous language regulation in academic ELF: speaking and writing were indeed treated differently in spontaneous ELF language regulation. University teachers as well as students made more comments on written than spoken text; in addition, comments on writing were more often concerned with correctness. What these practices thus indicate is that acceptability is wider for speaking than for writing in academia. Norms of writing are not only imposed top down by authorities, but as illustrated in (1), expert communities negotiate their own norms. Academic research papers taken from the SciELF part of WrELFA show that norms and conventions of writing have taken shape in academic communities. Even though the texts use English in broadly similar ways and do not stray far from Standard English, they also seem to fall into patterns of writing that characterise the research communities that they are writing for. I take a couple of examples to illustrate the writing. I sampled Introduction sections from science fields in SciELF. Introduction sections were selected on account of uniformity: all papers had one. I picked one paper at random from each L1 group, altogether nine papers1. General observations from the analysis are first of all that the dominant text organisation pattern followed very closely the CARS (Create a Research Space) pattern as outlined by Swales (1990). Many papers appeared rhetorically implicit, that is, the patterning was not very explicitly signalled with discourse markers, metadiscourse or

Good texts in non-­standard English  67 other similar indicators. Instead, the patterning rested on conventional text structures, most typically generalisations followed by specifications reporting previous research. The degree to which the papers conformed to Standard English conventions varied more than the text patterning. In the next two examples, I show text patterning in one paragraph from each of two introductions. Example (2) consists of a 4-­sentence paragraph, and I have numbered the sentences with superscripts for ease of reference. (2) 1Apart from the very low cost transfers in [10], there are generally two types of transfers, the direct transfer and the weak stability boundary (WSB) transfer. 2The direct transfer usually takes 3–15 days of flight time and needs about 3600–4100  m/s of manoeuvre velocity according to the libration point, the size of Halo orbits, and the flight time. 3The weak stability boundary transfer which has been applied successfully in saving the Japanese lunar satellite Hiten was first discovered by Belbruno and Miller[11], and then explained by Koon et al [12] using the dynamical systems theory. 4The weak stability boundary transfer to lunar Halo orbits needs about 3200 m/s of manoeuvre velocity; however the flight time is generally about 80–150 days. In this extract, which is the second paragraph of the article introduction, we can see content-­driven organisation. There is little overt marking of the organisation of the text, but nevertheless it follows a basically clear pattern. The opening sentence sets up a generalising distinction (there are generally two types of transfers, the direct transfer and the weak stability boundary transfer), which sets up an expectation in the reader that the distinction will be elaborated somehow, for example by saying something further about the two different kinds of transfer. And indeed, the next sentence picks up the first type (direct transfer) and goes on to specify its apparently two relevant properties, the flight time and the manoeuvre velocity. The third sentence, again as predicted, takes up the second type (weak stability boundary transfer) but now gives a brief description of facts about its history and discovery. Sentence 4 is more directly parallel to Sentence 2, detailing the velocity and the flight time. Here, in the final clause, we have the first discourse marker (however), which highlights the crucial contrast: the latter type of transfer takes considerably longer than the former, thereby posing a problem. As readers we now expect the next section to discuss this uncomfortable issue, and hopefully suggest a solution to the problem. The text shows some non-­standard surface features, but on account of its clear organisation it is easy to follow, and the minimal discourse marking suffices to highlight the central difference between the two alternatives outlined at the outset. The textual organisation is content-­based, but successful. If we now look at a different extract (3), we can see that like the previous one, this is also very content-­based. Similarly, it also constitutes one four-­sentence paragraph, this time the final paragraph of the section. The only text-­organising

68  Anna Mauranen features are the beginnings of sentence 1 (this work presents . . .) and 4 (the objective of this study was . . .). The first sentence introduces the general idea of the study, the determination of heat of absorption for a given absorption solution. Sentence 2 does not specify this but rather provides an addition: what else was subject to absorption heat determination apart from the solution described first. The third sentence adds another factor that sometimes occurs in the circumstances being investigated. The final sentence brings all these factors together, so that the reader can retrospectively appreciate why they were taken up one by one. Placed at the end of the Introduction, the aim of the study serves as a kind of reminder of what the idea behind it was. There is no indication of what was found, nor of why this might be interesting or what might be an interesting problem that is not known anything about yet. (3) 1This work present experimentally determined heat of absorption, using calorimetry, for an absorption solution consisting of an amine, 2-­amino-­2-­methyl-­1-­propanol (AMP), and an organic solvent, N-­methyl-­2-­ pyrrolidone (NMP) or triethylene glycol dimethyl ether (TEGDME). 2The heat of absorption of CO2 in the pure solvents was also determined. 3In the studied system precipitation of the AMP carbamate occurs under certain conditions for some of the solutions studied. 4The objective of this study was to determine the heat of absorption for the described absorption systems and how it was affected by the CO2 loading of the solution, the temperature and precipitation of the AMP carbamate. Altogether, despite providing more clues to its textual organisation, (3) appears less well organised and less clear than (2) to a reader with no expertise in either field. Neither text shows major non-­standard features, but both have some minor ones. Clearly, both extracts display well-­internalised notions of the conventions of their academic genre. These text norms are translocal, and apply to an international community, likely to consist of a majority of researchers to whom English is not an L1 (cf. Gnuzmann and Rabe 2014). What happened at the next stage to these texts is not known to us, but since they were written for publication, they probably went through a peer review. They may also have been through a language revision later or at least copyediting. Looking at these two texts, it would seem that one of them was written in a more effective way, but the dissimilarity was not due to any differences in the standard-­likeness of the English, or even in the textlinguistic indicators of text structure. The effectiveness was related to textual organisation. A  professional language expert would be able to help writers in their textual effectiveness, if they had a good understanding of what constitutes effectiveness in research papers, but the professional’s L1 status would be entirely irrelevant. Interaction around texts of these kinds for a language professional would thus mean an in-­depth understanding of science texts. Conventions that relate to

Good texts in non-­standard English  69 expertise in the discipline that the texts are written for as we saw in example (1) earlier, can be left to the field experts themselves.

Cognitive perspective: processing and entrenchment In the cognitive view, we can assume ELF writing generally to reflect processes of second-­language use. Thus it is reasonable to expect that an additional language is more weakly entrenched than a first language, especially for people who have grown up and been educated in mainly monolingual first-­language environments. This can simply be expected on the basis of exposure to more and less dominant languages in the environment. Language processing is fuzzy in any case, but it is a reasonable guess that less well entrenched features in an individual’s linguistic repertoire tend to be processed with less accuracy, i.e., more fuzziness, than better entrenched features. Therefore, we can hypothesise that processes of approximation and fixing are at work in a writer’s cognitive processes just as they are in a speaker’s processes (e.g., Mauranen 2012, 2017). With regard to ‘fixing’ specifically, it is useful to bear in mind that the term was coined by Vetchinnikova (2014) for observed preferences for certain expressions over others that L2 writers developed. By approximation, then, I mean rough equivalents of target items in the language, which are recognizable on the basis of their form, function, or meaning – or all of these together. Fixing, again, is in some sense the reverse process, settling on a particular expression for a particular meaning. We notice these best when the fixation selects an approximate form instead of a received standard form. The latter of course takes place all the time, and this is what Vetchinnikova first observed, but it happens with approximate forms as well, which is perhaps more noteworthy from our point of view here, since it results in change. The resulting change can be either what can be seen as a simplification of language, or it can be a complexification. The interplay of the processes can thus have apparently opposite consequences. The processes of simplification and complexification can be observed at both cognitive level and at macro-­social level. Simplification, specifically structural simplification, has usually been the sociolinguistically predicted process in SLU, and for post-­childhood second language acquisition (e.g., Trudgill 2011). It has also been predicted for individuals with a large variety of social contacts, or what Granovetter (1973) calls ‘weak ties’ as opposed to speakers in close-­knit and isolated communities (Milroy and Milroy 1985). Moreover, it has been observed in languages used as lingua francas more than in closed-­community languages (Kusters 2003). In morphology, a clear indicator of simplification can with good reason be seen in regularisation. Instances of this are seen in (4). (4) (a) There are evidences that WBV training improves strength and power

70  Anna Mauranen (b) It striked me that the few summaries I checked were very unstructured and short, unlike most summaries I see. In (a), a mass noun is used in the manner of countable nouns. Clearly, countability is a more regular phenomenon than uncountability as it can be determined by a regular principle rather than requiring memorisation of special cases. A consistent marking of the plural is cognitively lighter and more transparent than several possibilities for indicating the same meaning, and it facilitates processing for both writer and reader. The same is true of marking pastness, which is illustrated in (b). Regularisation can also take place in prepositional phrases. A common expression in speech, discuss about, was rare in research papers, with only one case in our sample of introduction sections (5). It occurs also in ‘chattier’, more speech-­ like written genres, and the blog subcorpus has some more instances (6). This is arguably a case of regularisation, since Standard English uses the same meaning with different verbs and the same preposition (speak about, talk about), and also the noun takes the preposition (discussion about). That the verb in the standard language happens to be used without the preposition is an anomaly. (5) . . . the industry was sought to create a forum where discussing about the challenges and opportunities of sustainability, (6) (a) . . . students in the Biotec faculty, who had created a mailing list to discuss about Linux and free software (b) I also used the Perelman’s theorem containing Poincaré’s conjecture to discuss about some properties of quantum gravity (c) And the best thing that can happen is to share a day with those passionate scientists plus journalists and people willing to have a word about science in the mass media, and discuss about these topics with them. We can find an interesting kind of simplification in syntax, which I would like to call ‘syntactic shortcuts’ because they seem to ignore some conventional constraints and in this way cut corners (7). Cases of this kind are harder to detect in spoken data on account of utterance structures being different from sentence structures. (7) (a) After reported first by Wainright et al. in 1995 [4], the application of PA-­PBI has been extensively studied (b) . . . however the flight time was generally much longer that of direct transfer. (c) This complication provokes that the spillage factor be generally scaled from accurate calculations only made for some particular cases. Here, (a) shows an example of using a finite verb in a prepositional clause which would take a non-­finite verb in standard usage. The next example (b) could in principle be just an omission by mistake, but since similar cases (comparatives

Good texts in non-­standard English  71 dropping the comparative indicator) recur in the data, the phenomenon appears worth taking note of. In (c), we see verb complementation take a non-­standard path. Often that-­complementation is used more extensively than in conventional usage. This may be a case of gradual simplification, but at the same time to-­ complementation would also seem to be on the increase, even if not in our present sample. If what is happening, then, is a reshuffling of the domains of these complementation patterns, it is no longer a question of simplification but in effect, of ongoing change, or even complexification, as two alternative patterns seem to share the same ground. complexification is the countertendency to simplification. Among indications of ongoing complexification, several kinds can readily be noticed. To begin with morpohology again, a counter-­tendency to regularisation is what looks like ‘overproductivity’. That is, morphemes are added to stems more freely than received conventions would specify. Especially negative prefixes in unconventional uses are common (8 a, b), but other morpheme combinations also occur, and can result in new compounds (8c). (8) (a) . . . this approach would be rather unappropriate for heliostats located at the south of the tower. . . . (b) This unvisualised “sample” is taken care by the passengers thoroughly. . . . (c) . . . the number of dated landslides throughout the world increased severalfold and introduction of modern dating methods. . . . Morphology also shows traces of approximation: in (9a) we may surmise that quantificate could result from a backformation derived from quantification, as we can see happening with predisposited and predisposition in (9b) where the co-­ text gives away the relationship and the likely process. The final example, (9c), shows an approximation apparently based on the phonological form, not very well entrenched and perhaps encountered in speech more often than writing. (9) (a) . . . and a p-­value for each result, to quantificate how much the alignment between the query sequence and the target is good. (b) . . . bed sediments of headwater channels of Appalachian Mountains are strongly predisposited by an ultra-­fine-­scale dynamic equilibrium with bedrock erodibility . . . variance in resistance of individual geological units was recognised as one of the downstream coarsening predisposition. (c) . . . a collection of top-­nodge articles previously published or accepted for publication Moving towards larger units, flipping word classes is an interesting way of recategorisation as a form of approximation (10). Analogical processes are usual in English (although not unique to it), and the examples show very workable-­looking

72  Anna Mauranen solutions for convenient and effective communication with common adverbs being used as adjectives modifying nouns. (10) (a) . . . differ substantially from lowland rivers due to often occurrence of vertically oscillating bedforms (b) The main proposal of this work is that, with the nowadays computation power. . . . When relationships between words need to be shown, for instance with prepositions, or when references are to be specified with determiners, a considerable amount of approximate uses are in evidence. In (11) the prepositions approximate the conventional form, which in the case of (a) would go without a preposition altogether, and in (b) be a different one. A redundant preposition may not be hazardous for communicative clarity, and probably easy to overlook. Prepositions in roughly the right meaning category, that is, one that makes sense in the context, are found in conversations without causing any overt problems (e.g., Mauranen 2012), as in (11b). (11) (a) . . . unfortunate impression that the thesis now sometimes conveys of has the four thinkers (b) . . . these new cultivars may contribute for the effective reduction of crop losses Articles appear in their typical non-­conventional variants below (12): either they are not in place (a), or they do not take the contextually conventional form (b), or they are superfluous (c). They may be a rather different case from prepositions in that they might mostly just pass without much notice. The reason for assuming that this might be the case is that article systems vary notably between languages, and some languages do not even have article systems. It is possible therefore that these systems are neither easily learnable nor intuitively obvious, so that in cross-­ linguistic communication they are not assigned much weight. (12) (a) relevant for researcher working on nanomagnetism (b) question of an genuine “Austrian philosophy” (c) the idea of a intellectual virtue Moving on to larger wholes, multi-­word units or phraseological units often manifest approximation in speech. Written texts as in (13) show quite similar tendencies. (13) (a) In this way the manuscript has an smooth reading. (b) . . . which could put some constrain to their discussion (c) . . . the blend membranes were studied in details, which were considered to be suitable for PEMs materials In all of these, some details do not fit in with the conventionalised unit in standard usage. That this should be so is intriguing if we think of the cognitive implications,

Good texts in non-­standard English  73 precisely because the units are conventional, and of the kind that are most likely acquired spontaneously, not through schooling. Despite their minor non-­ conformities, they rather add to than diminish a feel of fluency and even a certain English idiomaticity. They are used with skill  – they fit their contexts, express a recognisable meaning, and manifest the same holistic shape as their apparent targets. Phrases of this kind serve as candidates for change, if their approximate forms are picked up by interlocutors (micro-­social level) and gradually diffuse to wider use (macro-­social level). Syntactic phenomena were discussed earlier in terms of the simplification observed in syntax, but that is not the whole picture. Certain phenomena that in FL instruction are normally penalised for as non-­standard, such as lack of plural agreement (14) or 3rd person singular agreement (15) are relatively common in ELF. This is evident in speech, also among ENL speakers (Ranta 2013), but also appears in writing, as our examples show. (14) (a) . . . new particle are introduced (b) . . . the simulations suggests tens of keV O+ energies The examples show that a singular noun may be followed by a verb in the plural, or a plural noun by a verb in the singular. Similarly, the following examples illustrate the case of the third person singular, much discussed in ELF research (e.g., Seidlhofer 2011). What spoken language research tends to show – that most of the time the third person singular is marked in the conventional way, despite notable variability – is repeated in academic writing. The variation is illustrated in (15): (a) shows zero marking, (b) has non-­conforming marking with a plural noun, (c) attests a case of non-­compliant marking following an auxiliary. From a cognitive viewpoint (d) is the most interesting type, showing how individuals’ use may fluctuate even within a single utterance, and with coordinated verbs (addresses . . . and show). Basically this indicates shallow entrenchment, and a degree of fuzziness in the processing. (15) (a) This work present experimentally determined heat of absorption (b) Several studies emphasises the sensitivity of dragonflies (c) This means that government insurance covers potential loss from defaults of such loans and may affects its approval process. (d) Section 5 addresses the economic significance of the different forecasts and show how it varies in different market conditions. Why variability of this kind can be seen as increasing complexity in the cognitive view is that compared to fully regular use, it reduces predictability for interlocutors and proliferates forms that an individual may be faced with. It alters expectations and favours fuzzy processing.

74  Anna Mauranen A final example of complexification through approximation and growing variability is variation in expressing one meaning category: contrast. Let me first start with some examples that simply illustrate the majority usage, which is quite conventional (16). (16) (a) In contrast to the comparative emissions figures contained in corporate annual reports, the CDP data does not have an explicit longitudinal dimension. (b) By contrast, fluorescent imaging with reporter mice identifies spots of cartilage maturation without need to kill animals for dissection diagrams. (c) And, of course, whether or not momentum space is anything more than a mathematical construct, as opposed to being an object with a genuine physical existence. (d) On the contrary it will guarantee the equidistance of reporters from any partisan interest and will help deliver what is currently missing: quality information relevant to the citizen. In addition to the conventional uses, where roughly synonymous options are available to any user, our WrELFA corpus sample also showed the following (17): (17) (a) On the opposite, according to previous studies, AC can act as reaction site, (b) Opposite of some other studies, (c) In contrary to previous investigations on the topic, (d) . . . higher channel gradients, coarser sediments and contrary, pool-­ riffles had greater width/depth ratios What these instances show is that the already existing variation in expressions of contrast is supplemented by a fresh source for candidates. So far, we have seen approximation at work and the consequences of adding to variation. For language change to occur, it is also necessary that there is uptake of the new forms, and that they diffuse. Any or all of our examples could in principle have been individual events that only appeared once. It is now time to turn to the possibility of fixing, which implies a new preference, as different people adopt an expression. The first case (18) is from (13c) earlier, in details. (18)

The cases here are not many, but they come from different texts: not only SciELF papers but also doctoral examination reports.

Good texts in non-­standard English  75 As a second case (19), let us look at what (17) suggested earlier. The first pattern is on the opposite, as in (17a). (19)

Again, these instances come from different authors or author teams. This expression can already be taken to suggest an incipient change in use. The final case (20) is not listed in any of these examples, because in itself it is neither a standard expression for contrast nor in any way unconventional or ungrammatical. What is new about opposite to here is that the use has been extended to cover senses beyond a physical meaning of objects facing each other on opposite sides, or an abstract, say, linguistic use with a specification (‘X’ is opposite in meaning to ‘not X’). (20)

In speech, approximations might be put down to the need for fast processing (see, e.g., Christiansen and Chater 2015), together with hesitations, false starts, repeats, and other phenomena that are known as disfluencies. Seeing very similar processes, often exactly the same approximative phrases, in writing, where far more time and discretion can be used, is intriguing from a processing perspective. Many if not all approximations are likely to be one-­off occurrences, as most spontaneous innovations in language are. But for any of the individual users, these may be their preferred use, and as we have seen, some have spread more widely. The individual’s cognitive processes are connected through social

76  Anna Mauranen interaction to macro-­level changes in the whole language and related social practices.

Conclusion We have looked into academic writing in ELF from different perspectives with reference to corpus data, with most examples drawn from the same corpus. This approach has sought to throw light on the interconnectedness of different levels of language use, which connect language as social practice and language as a cognitive and micro-­level communicative practice, and in doing so, shown the versatility of corpus data. One main thread at the macro-­social level was the ubiquity and dominance of English in academic writing, which has both imposed an apparent trend of unilingual use on the globalised academic world, but at the same time ushered in new kinds of complex multilinguality to it, and undermined the status of Standard Englishes. Top-­down regulation of written English is becoming less effective with so much spontaneous co-­construction of local norms in research communities, university communities, blog communities and other discourse communities, such as conference communities, that arise in cross-­border academic encounters. The negotiability of norms can be captured in observing interaction in educational environments, as was done previously. Expert communities are increasingly translocal, and research fields seem to be determining norms of writing, but not necessarily with reference to Standard English as we saw. Future developments will also include the breaking down and renegotiating traditional disciplinary boundaries in research publishing. Large, multidisciplinary research projects are increasingly called upon to grapple with major challenges to humankind, but in practice they also work on less enormous questions, so that all levels of research practices operate on transnational as well as multilingual bases. On top of these developments, digitalised practices are rapidly altering the conditions of research publishing. This chapter explored current research writing from three perspectives, with the last one extending to the individual cognition, specifically focusing on the processes of approximation and fixing. In speech, approximations might be put down to the need for fast processing (see, e.g., Christiansen and Chater 2015), together with hesitations, false starts, repeats, and other disfluencies. Similar processes, often exactly the same approximative phrases, were found in writing, where lack of time cannot be implicated in the same way. Even if most approximations are one-­off occurrences, like most spontaneous innovations in language, they may be fixed in an individual’s repertoire (cf. Vetchinnikova 2014) and be regularly used by those individuals. Alternatively, they may live on as small-­group usage, as an instrument of social cohesion within, say, a work team, research group, or other community of practice, or a family circle. But there are also approximations that are shared across individuals and regions, as

Good texts in non-­standard English  77 we have seen in this chapter. It is here that the cognitive meets the micro-­ and the macro-­social. ‘Good writing’ is a topic of interest for anyone who wants to publish, and anyone who wants to help others publish. Effectiveness in language use is not dependent on correctness  – there is no evidence in support of this widespread assumption, as Pitzl (2015) points out. We need more research and more professional expertise in coming to grips with effectiveness in texts – in their evolving genres, new contexts, and in view of the changing expectations of the readership. Language professionals have their own kind of expertise, which needs developing and updating just like the expertise of researchers in any field. Ongoing changes are particularly rapid in research writing. Even for those who are committed to ‘good language’, it is unrealistic to seek – or offer – native-­like writing as the goal. We cannot make false promises in teaching where we offer native-­like skills as an alternative: our courses tend to be of short duration, and we could spend the time better than to focus on correctness and some version of ‘native-­likeness’ to the students, who will hardly make significant progress towards such goals during the courses. Language professionals must strive for a deeper understanding of what it is that ‘good texts’ consist of. The language professional’s task will shift from teaching and controlling correctness to teaching effective, competitive rhetoric, that helps authors get their texts accepted in traditional research publishing and noticed in the open but overcrowded world of digital communication.

Note 1 The L1s were: Chinese, Czech, Finnish, French, Italian, Portuguese, Russian, Spanish, and Swedish.

References Ädel, A. 2006. Metadiscourse in L1 and L2 English. Amsterdam: John Benjamins. Agha, A. 2003. The social life of a cultural value. Language & Communication 23, 231–273. Ammon, U. 2007. Global scientific communication: Open questions and policy suggestions. AILA Review 20, 123–133. Bhatia, Vijay K. 1993. Analysing genre. London: Longman. Biber, D., S. Johansson, G. Leech, S. Conrad and E. Finegan. 1999. The Longman grammar of spoken and written English. London: Pearson Education. Carey, R. 2013. On the other side: Formulaic organizing chunks in spoken and written academic ELF. Journal of English as a Lingua Franca 2(2), 207–228. Christiansen, M.H. and N. Chater. 2015. The now-­or-­never bottleneck: A fundamental constraint on language. Behavioral and Brain Sciences. Available on CJO2015. doi:10.1017/ S0140525X1500031X. Cogo, A. 2012. ELF and super-­diversity: A case study of ELF multilingual practices from a business context. Journal of English as a Lingua Franca 1(2), 287–314.

78  Anna Mauranen Crawford, E. 1985. The beginning of the Nobel institution: The science prizes, 1901–1915. Cambridge: Cambridge University Press. Drubin, D.A. and D.R. Kellogg. 2012. English as the universal language of science: opportunities and challenges. Editorial. Mol Biol Cell. 2012 Apr 15; 23(8): 1399. Franzmann, A., A. Jansen and P. Münte. 2015. Legitimizing science: Introductory essay. In A. Franzmann, A. Jansen and P. Münte (eds), Legitimizing science: National and global publics (1800–2010). Chicago: University of Chicago Press, pp. 11–34. Gnuzmann, C. and F. Rabe. 2014. “Theoretical subtleties” or “text modules”? German researchers’ language demands and attitudes across disciplinary cultures. Journal of English for Academic Purposes 13, 31–40. Graddoll, D. 2006. English next? Why global English may mean the end of “English as a foreign language”. London: The British Council. Granovetter, M. 1973. The strength of weak ties. American Journal of Sociology 78(6), 1360–1380. Grieve, J., D. Biber, E. Friginal and T. Nekrasova. 2011. Variation among blog text types: A multi-­dimensional analysis. In A. Mehler, S. Sharoff and M. Santini (eds), Genres on the web: Computational models and empirical studies. New York: Springer-­Verlag, pp. 303–322. Holmes, P. and F. Dervin (eds). 2016. The cultural and intercultural dimensions of English as a Lingua Franca. Clevedon: Channel View Publications. Hyland, K. 2005. Metadiscourse: Exploring interaction in writing. London: Continuum. Hynninen, N. 2016. Language regulation in English as a Lingua Franca. Berlin: de Gruyter Mouton. Hynninen, N. 2018. Impact of digital media tools on the research writing process: A case study of collaborative writing in computer science. In M. Kuteeva and A. Mauranen (eds), Special Issue of Discourse, context and media (vol. 24), pp. 16–23. Ingvarsdóttir, H. and B. Arnbjörnsdóttir. 2013. ELF and academic writing: A perspective from the expanding circle. Journal of English as a Lingua Franca 1(2), 123–146. Jenkins, J. 2015. Repositioning English and multilingualism in English as a Lingua Franca. Englishes in Practice 2(3), 49–85. doi:10.1515/eip-­2015-­0003. Jenkins, J. and A. Mauranen (eds). 2019. Linguistic diversity on the international campus; Insider accounts of the use of English and other languages in ten universities within Asia, Australasia and Europe. London: Routledge. Kajanto, I. 2000. Latina, kreikka ja klassinen humanismi Suomessa keskiajalta vuoteen 1828. Helsinki: SKS. Kusters, W. 2003. Linguistic complexity: The influence of social change on verbal inflection. Utrecht: LOT. Kuteeva, M. 2016. Research blogs, wikis and tweets. In K. Hyland and P. Shaw (eds), Routledge handbook of English for academic purposes. London: Routledge, pp. 431–444. Kuteeva, M. and J. Airey. 2014. Disciplinary differences in the use of English in higher education: Reflections on recent language policy developments. Higher Education 67(5), 533–549. Laitinen, M. 2016. Ongoing changes in English modals: On the developments in ELF. In O. Timofeeva, A.-­C. Gardner, A. Honkapohja and S. Chevalier (eds), New approaches to English linguistics: Building bridges. Amsterdam: John Benjamins, pp. 175–196. Laitinen, M. 2018. Placing ELF among the varieties of English. In S.C. Deshors (ed), Modeling world Englishes: Assessing the interplay of emancipation and globalization of ESL varieties. Amsterdam: Benjamins, pp. 109–131. Laitinen, M. and J. Lundberg. forthcoming. ELF, language change and social networks: Evidence from a third-­generation ELF corpus In A. Mauranen and S. Vetchinnikova

Good texts in non-­standard English  79 (eds), Language change: The impact of English as a Lingua Franca. Cambridge: Cambridge University Press. Lillis, T. and M.J. Curry. 2010. Academic writing in a global context. London: Routledge. Luzón, M.J. 2013. Public communication of science in blogs: Recontextualizing scientific discourse for a diversified audience. Written Communication 30(4), 428–457. Mauranen, A. 1997. Hedging and modality in revisers’ hands. In R. Markkanen and H. Schröder (eds), Hedging and discourse: Approaches to the analysis of a pragmatic phenomenon. Berlin: de Gruyter Mouton, pp. 115–133. Mauranen, A. 2006. Signaling and preventing misunderstanding in English as Lingua Franca communication. International Journal of the Sociology of Language 2006(177), 123–150. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Mauranen, A. 2013. Hybridism, edutainment, and doubt: Science blogging finding its feet. Nordic Journal of English Studies 12(1), 7–36. Mauranen, A. 2017. A  glimpse of ELF. In M. Filppula, J. Klemola, A. Mauranen and S. Vetchinnikova (eds.), Changing English: Global and local perspectives. Berlin: de Gruyter Mouton, pp. 223–254. Mauranen, A. 2018. Second language acquisition, world Englishes, and English as a Lingua Franca (ELF). World Englishes 37(1), 106–119. Mauranen, A. forthcoming. Reflexively speaking – uses of metadiscourse in ELF. Berlin: de Gruyter Mouton. McGrath, L. 2016. Open-­access writing: An investigation into the online drafting and revision of a research article in pure mathematics. English for Specific Purposes 43, 25–36. Miller, C.R. and D. Shepherd. 2009. Questions for genre theory for the blogosphere. In J. Giltrow and D. Stein (eds), Genres in the internet: Issues in the theory of genre. Amsterdam: John Benjamins, pp. 263–290. Milroy, J. and L. Milroy. 1985. Linguistic change, social network and speaker innovation. Journal of Linguistics 21, 339–384. Milroy, L. 2002. Social networks. In J.K. Chambers, P. Trudgill and N. Schilling-­Estes (eds), The handbook of language variation and change. Oxford: Wiley Blackwell, pp. 549–572. Myers, G. 2010. The discourse of blogs and wikis. London: Continuum. Nevalainen, T. 2003. English. In A. Deumert and W. Vandenbussche (eds), Germanic standardizations: Past to present. Impact: Studies in Language and Society 18. Amsterdam and Philadelphia: Benjamins, pp. 127–156. Owen, D. 2011. “Native here and to the manner born”? Academic publishing and “proper” English. English Text Construction 4(2), 279–302. Paulasto, H. 2015. Focus fronting as a potential angloversal. Presentation at the workshop Changing English across time-­scales and Planes of Language Representation. University of Helsinki, 8 June. Phillipson, R. 1992. Linguistic imperialism. Oxford: Oxford University Press. Pitzl, M.-­L. 2015. Understanding and misunderstanding in the common European framework of reference: What can we learn from research on BELF and intercultural communication. Journal of English as a Lingua Franca 1(4), 91–124. Ranta, E. 2013. Universals in a universal language? Exploring verb-­syntactic features in English as a Lingua Franca. Doctoral dissertation, University of Tampere. Rowley-­Jolivet, E. and S. Campagna. 2011. From print to web 2.0: The changing face of discourse for special purposes. LSP Journal 2(2). Available from: http://lsp.cbs.dk

80  Anna Mauranen Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Sheller, M. and J. Urry. 2006. The new mobilities paradigm. Environment and Planning A 38, 207–226. Sinclair, J.M. 1982. Planes of discourse. In Rizvi (ed), The twofold voice, pp.  70–91. Reprinted in Trust the text. 2004. London: Routledge, pp. 51–66. Smart, C. 2016. Discourse reflexivity in linear unit grammar. Amsterdam: John Benjamins. Swales, J. 1990. Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press. Trudgill, P. 2011. Sociolinguistic typology: Social determinants of linguistic complexity. Oxford and New York: Oxford University Press. Urry, J. 2007. Mobilities. Cambridge: Polity. Ventola, E. and A. Mauranen. 1991. Non-­native writing and native revising of scientific articles. In E. Ventola (ed), Functional and systemic linguistics: Approaches and Uses. Berlin: de Gruyter Mouton, pp. 457–492. Vetchinnikova, S. 2014. Second language lexis and the idiom principle. Helsinki: Unigrafia. Wächter, B. and F. Maiworm (eds). 2014. English-­taught programmes in European higher education. ACA Papers on International Cooperation in Education. Bonn: Lemmens.

Corpus: WrELFA. 2015. The corpus of written English as a Lingua Franca in academic settings. Director: Anna Mauranen. Compilation manager: Ray Carey. Available from: www.helsinki.fi/ elfa/wrelfa.html.

6 Exploring diachronic changes in research articles from an ELF perspective A small corpus study1 Sayako Maswana Introduction An academic lingua franca is a language used to communicate scientific knowledge and information among people who do not share the same language (Mauranen et al. 2010). In the contemporary moment, English has played this role in academic contexts. Top peer-­reviewed international journals are, without exception, written in English. English is largely agreed upon as an academic lingua franca (ELFA) by the international research and academic community, and in theory, it does not have any native speakers as such (e.g., Kuteeva 2017; Mauranen 2006). In reality, however, native speaker norms have influenced standard academic English rhetorically, logically, and lexicogrammatically, particularly in written texts; academic English teaching has generally followed this tradition (Heng Hartse and Kubota 2014). Nevertheless, recent corpus-­based English as a lingua franca (ELF) studies (e.g., Mauranen et  al. 2010; WrELFA 2015) have revealed the increasing and strong nature of the lingua franca and its features in spoken and written academic texts. This calls for a reexamination of the traditional notion that academic English must conform to native English speaker norms in practice. Academic English covers a wide range of academic genres, from the conference presentation to the grant proposal to the research paper. The research article is undoubtedly the most studied genre within the area of English for Academic Purposes (EAP), a subcategory of English for Specific Purposes (ESP), broadly defined as an approach to teaching to meet specific needs of the learners (Dudley-­ Evans and St. John 1998), since it is officially the means by which new knowledge is communicated and disseminated among scholars and sometimes directed toward broader public audiences. ESP’s understanding of the genre concept is helpful to investigate academic English which could mean differently depending on the context. While it may be the case that the genre of the research article is more dependent on the norms of native speakers than other academic genres, it is useful to study this signature EAP genre, as Tribble (2017) argues, in this genre, the dichotomy of apprentice versus expert, but not native versus non-­native speaker, offers useful teaching implications. To examine emerging ELF features in the research article, diachronic research would give insights. Several diachronic ESP studies on research articles have been carried out, mostly through using a corpus analytic approach. These ESP

82  Sayako Maswana studies include Biber and Gray’s (2010) analysis, which focused on research articles from the years 1965, 1985, and 2005, and Li and Ge’s (2009) analysis on medical articles, which compared articles published between 1985 and 1989 to those published between 2000 and 2004. More recently, Hyland and Jiang (2016, 2017) have examined engagement and informal features2 during the past 50 years, with interval years of 1965, 1985, and 2015 in four disciplines of applied linguistics, biology, engineering, and sociology. However, to help better understand how and why the academic article genre has been evolving recently, further cross-­ disciplinary comparison using articles published after 2010 and discussion of the results from an ELF perspective are important, considering the recent global use of ELFA. The period between the eve of the advent of online scientific publication expansion and the present time is significant because this period is witnessing accelerated globalization. We can observe that many academic journals have changed their physical and online appearances between 19953 and now. What we have been taught as correct academic English in writing classes might not be always observed in current research papers. Today, academic communities are frequently occupied by non-­native speakers; for this reason, the corpus study of the linguistic evolution of journals over the past two decades is an urgent need. Many general English corpus studies have centered on comparing learners’ (i.e., non-­native speakers) and native speakers’ corpora to establish learning deficiencies (Mauranen 2003). By contrast, in ESP research, language use based on genre is more fundamental than the prescribed standard use of English, and thus ESP corpus studies have compared corpora of novice users and expert users of the target genre (Flowerdew 2003). Flowerdew (2003) thus used an existing general corpus, BNC, to confirm that the particular features of novice-­user corpora are specific to the novice. A corpus is usually used for quantitative analysis and is compiled by a large number of subcategories and their texts. It is supposed that the larger a corpus, the more language patterns are present and types of analyses are possible. On the other hand, a small corpus, as in this study, has the merits of investigating specific materials and pinpointing analyses through manual coding. Small corpus analysis, therefore, is feasible and relevant for an ESP/EAP practitioner who needs to understand a specific text’s characteristics and context on their terms. In an earlier study, I considered features of titles and introductions to medical journal articles published in the years of 1995 and 2015 and found structural and grammatical changes (Maswana 2016). The study was pedagogically motivated to provide teaching guides for those who wished to write articles in the discipline. However, the results I found there suggest the relevance of adopting the perspective of ELF to further study another discipline, to examine changes in the discipline, and to discern details in surrounding contexts. It was not clear whether the changes found in the study could be considered universal in academic contexts or whether they should be attributed either to specific situations within disciplines, to changing circumstances (writers/readers), to personal preference for writing style, or to yet other factors. Discussion of changes in my earlier study was therefore limited to the advancement of online publications as well as to the changing research projects that now require consideration of the wider specialization of

Exploring diachronic changes in articles  83 researchers. Following from this earlier study, this chapter attempts to identify changes in research articles in the fields of medicine and economics, or the lack therein, by comparing articles published in the years 1995 and 2015. The disciplines of medicine and economics were chosen as one of the major applied sciences disciplines and the social sciences disciplines with comparable previous research in genre analysis. The current study therefore should be considered a focused corpus study of limited disciplines and periods and will explore identified changes and also consider suggestions for future research.

Methods This study focuses on introduction sections and titles of research articles in the disciplines of medicine and economics as a preliminary step to understanding changes during this period from an ELF perspective. The study assumes that the 2015 research-­article corpus is more ELF oriented than the 1995 one. There said to be only small quantitative differences between standard English and ELF in academic English (Mauranen et al. 2016), and for the genre of the research article, in particular, we would expect this to be true because articles go through multiple editorial processes. However, given that English is used by members of the ever-­growing international community in respective disciplines, the differences between the two corpora may be explained by ELF features or at least its direct or indirect influence. The study did not consider an author’s L1 because “considering the increasingly global character of academia, determining with any certainty an author’s L1 based on name and home institution is problematic” (McGrath 2016: 87). In addition, in many cases of multiple authors, it is difficult to determine the specific writer who authored the text or to what extent each author has contributed to authoring the text and what his or her writing process consisted of. Furthermore, editing processes influence the text (Fletcher and Fletcher 1997), as Hyland and Jiang (2017: 49) rightly point out: “research articles are carefully refined and polished and so a potential drawback of examining a published corpus is the influence exercised by editors, copy editors and reviewers on the style of the eventual finished text” (see also Mauranen, this volume). Despite these concerns, this small-­scale and exploratory study with a pedagogical orientation aims to provide insights into the teaching of English as an academic lingua franca, the kind of genre analysis that pedagogical researchers could conduct by themselves based on the needs of their own students. There are many features to be considered, but this study uses the method for introduction and title analysis, and the corpus of medicine articles previously described and compiled in Maswana (2016), which was itself largely based on the research method of Li and Ge’s (2009). This study extends the research to the discipline of economics to examine whether changes occur due to the discipline or whether these are common across disciplines. The corpus of each discipline consists of 30 articles from leading international journals for 1995 and 30 for 2015, with a total of 120 article introductions and titles. These journals were selected based on citation indexes, references to previous comparable studies,

84  Sayako Maswana and the accessibility of these articles to the research site. The journals included in the medicine corpus are The BMJ,4 the Journal of Clinical Investigation, and The Lancet. The journals in the economics corpus are The Quarterly Journal of Economics, the Journal of Financial Economics, and The Review of Economic Studies. Introductions and titles were chosen because they are commonly included in all articles and disciplines and because an established method exists for analyzing their structures (see, for example, Soler 2007; Swales 1990 described later). With regard to introductions, the study adopted the Create a Research Space (CARS) model provided by Swales (1990) in his move analysis. A  move in a text refers to a rhetorical unit with a specific communicative purpose; the move unit and its purpose support achievement of the text’s goal. The model consists of three essential moves: move 1 – establishing a territory, move 2 – establishing a niche, and move 3 – occupying the niche. Each move was manually coded based on close reading of the texts primarily by the researcher with verification of randomly selected introductions by an academic-­writing teacher. Verb tense and modality are said to be related to rhetorical functions (e.g., Salager-­Meyer 1992) and, thus, were coded in each move. Voice and sentence length were also examined as they provide the basic information of the sentences and relate to readability of texts. For title analysis, Soler’s (2007) framework was followed to identify the title structures. This framework has three main structures: 1 – nominal group construction (e.g., “Hedge funds and discretionary liquidity restrictions”); 2 – compound construction (e.g., “Liquidity hoarding and interbank market rates: The role of counterparty risk”); and 3 – full sentence construction (e.g., “Lysyl hydroxylase 2 induces a collagen cross-­link switch in tumor stroma”). The study counted an interrogative sentence which was part of a compound title as a compound construction and an independent interrogative sentence as a full-­sentence construction. The study compared differences in these aspects/features between the 1995 and the 2015 articles and attempted to explain the differences, focusing on the disciplinary research environment and the increasing number of non-­native experts in the discourse community, although it acknowledges that there might be changes that would be explained by other perspectives.

Results Introductory section Table 6.1 overviews the introduction corpus word count. The medical article introductions tend to be short, while economics article introductions are long, a marked difference. In both disciplines, the numbers indicate a large increase in article length, length of the introductory section, and total number of sentences in the introduction. This tendency is in accordance with a diachronic study Hyland and Jiang (2016) conducted at three periods over the past 50 years. This change in word count might be taken for granted by scholars in practice, but in analysis, it is significant that the genre of the academic article, one noted for its resistance to change, has in fact changed in a visible way without explicit instructions to do so based on the guidelines of journals.5

Exploring diachronic changes in articles  85 Table 6.1  Overview of the introduction corpora

Medicine

Economics

1995

2015

9,703 words in total 323.4 words/intro (SD = 171.93) 13.0 sentences/intro (SD = 6.60) 30,749 words in total 1,025.0 words/intro (SD = 411.39) 43.2 sentences/intro (SD = 17.14)

13,629 words in total 454.3 words/intro (SD = 186.21)** 17.1 sentences/intro (SD = 7.62)* 52,750 words in total 1,758.3 words/intro (SD = 459.50)** 68.5 sentences/intro (SD = 21.49)**

Note: Significantly different compared to 1995 (**p < .01, *p < .05).

It is easy to understand that the increase in word count in the article introductions naturally correlates with increases in the number of sentences in introductions. The results of each analysis are presented and discussed later. Move structure Here, we look at the structure of moves in the introduction. Overall, three major moves were used in most of the paper introductions in both disciplines. In medicine, moves 1 (Presenting background information), 2 (Reviewing related research), and 3 (Presenting new research conducted by the author(s)) were used in 100% of the samples, except move 1 in 1995 (90%) and move 2 in 2015 (97%). The conventional pattern (moves 1–2–3) was included in 63% of the articles in each year. Other patterns included repetitions of certain moves, e.g., moves 1–2. In economics, move 1 was used in 90% of the articles in 1995 and 100% in 2015; similarly, the usage rate of move 2 was 83% in 1995 and 97% in 2015; that of move 3 was 97% in 1995 and 100% in 2015. The conventional pattern only was found in 20% of 1995 articles and 17% of 2015 articles. Other major patterns included the conventional pattern plus repetitions of moves 2 and 3 (see the example here); this accounted for 17% of 1995 articles and 40% of 2015 articles. Move 3: In our data, we find that skilled workers are overrepresented in occupations that are pervasive at nonroutine abstract tasks, whereas unskilled workers are often in occupations that involve routine tasks.  .  .  .  Move 2: To date, research has largely focused on the consequences of investments in computers and R&D. . . . Move 3: Our findings are consistent with the widespread view that ICT is complementary with human capital. . . . (Ref.1,6 p. 1784) Finally, apart from the fact that introductions were longer in 2015 with the subsequent increase of repetitions of moves 2 and 3 in economics articles, it can be argued that, overall, rhetorical structures remained largely unchanged. Nevertheless, detailed descriptions of the research and background information have become more necessary for current article introductions, as the need to establish respective research territories has grown with the increased volume of publishing.

86  Sayako Maswana Verb tense in each move This subsection reports on the tense use in the introduction. In medicine, as Table  6.2 shows, the tendency of particular tense uses in each move has not changed much over two decades. A dominant use of the present tense in moves 1 and 2 is common, as is the use of past tense in move 3. A slight increase in the degree of the use of present tense in moves 1 and 2 is also apparent. Similarly, the tendency has not significantly changed for economics articles. Unlike in medicine, however, the dominant use of the present tense is observed throughout the introductions in economics, including move 3. A difference occurred in move 3 in 2015, in which more past tense and fewer modals were used compared to 1995; nevertheless, the overall tense distribution remained relatively similar. Sentence structure and length Next, let us look at the structure and length of sentences in the introductory section, first with the number of instances of passive voice in the main clauses. For both disciplines, the use of passive voice significantly decreased (p < .01), although its use used to be recommended in academic writing (Hyland 2002). In 1995, 25% and 15% of the main clauses were in passive voice in medical and economics articles, respectively; in 2015, the rate decreased to 18% and 8%, respectively. It is now considered the use of active voice makes journal articles more readable (e.g., Millar et al. 2013). By contrast, as for the word count in a sentence, Table 6.3 shows it increased significantly in both disciplines, making the texts more difficult to read (e.g., McLaughlin 1969). It is natural to assume that the longer introduction lengths would correspond to a higher number of sentences as revealed in the previous section; while this is true, sentences themselves actually Table 6.2  Percentage of verb tenses and modals per move 1995 introductions

2015 introductions

Move 1 Move 2 Move 3 Move 1 Move 2 Move 3 Past tense Present tense Medicine Present perfect tense Modals Past tense Present tense Economics Present perfect tense Modals

11% 69% 15% 4% 22% 56% 11% 10%

22% 41% 23% 14% 4% 77% 9% 9%

58% 33% 6% 2% 1% 86% 1% 11%

3%** 73% 15% 9% 17% 60% 14% 9%

14% 54%* 22% 9% 2% 79% 10% 9%

64% 27% 5% 5% 5%** 86% 0% 9%*

Note: Significantly different compared to 1995 (**p < .01, *p < .05). The following instances are not listed in the table: 1 past perfect in Move 1, 2015 medicine; 1 past progressive and 1 present progressive in Move 1, 1995 economics; 1 present progressive in Move 2, 1995 economics; 7 past perfect in Move 3, 1995 economics; 1 present perfect progressive in Move 1, 2015 economics; and 12 past perfect and 1 present progressive in Move 3, 2015 economics.

Exploring diachronic changes in articles  87 Table 6.3  Number of words per sentence

Medicine Economics

1995

2015

M = 24.82 (SD = 9.70) M = 23.71 (SD = 11.31)

M = 26.52 (SD = 11.23)* M = 25.66 (SD = 12.34)**

Note: Significantly different compared to 1995 (**p < .01, *p < .05).

grew in length as well (see Table 6.3). As generally recommended for academic writing (e.g., Wallwork 2016) as well as specifically mentioned in one journal in the corpus, the writing would need to be concise; however, this does not necessarily mean the text has to be short, as revealed in this study. While the two disciplines showed an increase in the number of words per sentence, the type of long sentences used seemed different between the disciplines. In medicine, longer sentences tend to employ a “compressed” discourse style, as suggested by Biber and Gray’s (2010) analysis, particularly in the use of noun phrases. The example shows this tendency. We aimed to assess the effect of the prostaglandin analogue latanoprost on the visual field preservation of patients with open-­angle glaucoma in a comparatively short trial. (Ref.2, p. 1296) Biber and Gray (2010) showed that a pattern of compressed style, with nonclausal modifiers embedded in noun phrases, is a characteristic factor that contributes to the structural complexity and comprehension difficulty of academic writing. Snow (2010) explained that these nominalizations are employed to refer to complex processes and pose a challenge for students learning science. In economics articles, longer sentences tend to contain complex structures of a different sort, such as with sentences having an independent clause and a dependent clause or multiple relative pronouns in one sentence. The following example illustrates this. However, because the total progress that needs to be carried out is fixed, the agents benefit from the ability to complete the project quicker, which increases the present discounted value of their reward, and consequently strengthens their incentives. (Ref.3, p. 188) This complex discourse style is also acknowledged to be a reason academic writing is generally judged to be difficult to comprehend (e.g., Gillett n.d.). In terms of sentence structure, both disciplines demonstrated similar tendencies over the two decades in their decreased use of voice on one hand, and the increased sentence length, on the other. However, a close look at sentences gives a suggestion that there may be differences between the two disciplines in terms of complexity

88  Sayako Maswana and compression of sentence structure. Related disciplinary characteristics will be explored in the discussion section. Titles Titles have also shown differences between the two disciplines in their patterns of structures. As Figure  6.1 shows, nominal group construction was dominant in 1995 in both disciplines. In 2015, compound construction was predominantly used in medicine, followed by full-­sentence construction, whereas economics articles used nominal group construction most of all, narrowly followed by compound construction. Because most of the compound construction titles comprised two nominal parts connected with a colon, it may be assumed that many of the nominal group constructions in the medical titles of 1995 were replaced by compound constructions in 2015. Title examples of the dominant construction from the 2015 corpora are shown here: Antibiotic treatment for 6 weeks versus 12 weeks in patients with pyogenic vertebral osteomyelitis: An open-­label, non-­inferiority, randomised, controlled trial (Medicine) Credit markets, limited commitment, and government debt

(Economics)

30 25 20 15 10 5 0 Medicine 1995

Medicine 2015

nominal group construction full-sentence construction

Figure 6.1  Title structures in 1995 and 2015

Economics 1995

Economics 2015

compound construction

Exploring diachronic changes in articles  89 The average numbers of words composing the title were 13.33 (SD  =  4.94) in 1995 and 16.50 (SD = 7.01) in 2015 in medical articles which showed a significant increase (p < .05), while the numbers were 8.8 (SD = 4.39) and 8.3 (SD = 3.06) for respective years with no significant change (p = 0.61) in economics articles. Changes were most likely due to changes in the type of construction of the title, but only medical articles saw a significant increase in the number of title words. The increase in compound construction with no increase in the word count for economics titles does not indicate that compound constructions provide detailed information, as medical titles appear to do (an example of compound construction in the 2015 economics titles reads “Financial asset holdings and political attitudes: Evidence from revolutionary England”). Overall, the findings suggest both broad and discipline-­specific changes in organizing introductions and constructing titles. There were few clear journal guidelines instructing those changes except one medical journal recommending active style. In addition, in close reading of the introductory sections for the rhetorical analysis, the researcher did not find any particular usage of English in the 2015 articles that differs from standard English. The changes reported herein may have been caused by interrelated factors and will be discussed next.

Discussion Differences in the English of the research article genre between the disciplines over the period can be attributed to varying discipline cultures and practices, including the constitution of disciplinary knowledge and the nature of communication between community members. Hyland and Jiang (2016: 32) suggest the relative decline in the use of engagement features in research articles may be “perhaps due to the influence of growing numbers of second language writers schooled in the virtues of objective writing styles”. At the same time, such changes usually depend on particular disciplines and selected features for analysis (Hyland and Jiang 2017). These changes are related to changes occurring in research projects and publications more broadly. Compared to 20 years ago, there is an expansion of research fields that has occurred because of the volume and specialization of research. Although there is disciplinary variance, there has been a significant increase in the number of authors per article, particularly in science disciplines, over the past 20 years. In the field of medicine in the 1995 corpus, for example, the average number of authors was 5 (SD = 2.70); in 2015, the number of authors more than doubled to an average of 12 (SD = 6.35); the list of authors often includes international researchers. During the same period, in economics, an average of 2 authors (SD = 0.57) in 1995 became just 3 (SD = 1.04) in 2015. Ware and Mabe (2015) reported disciplinary differences in the patterns of publishing: the proportion of joint authorship is higher among medical articles than among those in the social sciences, and article output is higher in the hard sciences than in social sciences. Disciplinary differences also exist in reading and article-­ seeking behaviors: the number of articles read by faculty members in medicine is much higher than that of articles read by faculty members in the social sciences,

90  Sayako Maswana although what is meant by reading might vary among respondents (ibid.). This faster paced and multi-­authored publication in medicine may have affected the patterns of verb tense and voice as discussed in the preceding section. Research articles being established as digital sources also seem to have great influence on titles within medicine – titles have become longer and more highly specialized to facilitate searches of online databases such as the medical literature database PubMed (Jacques and Sebire 2010). This reflects the nature of reading in the field of medicine, where the readers often include medical and healthcare practitioners (Demner-­Fushman et  al. 2005). In addition, external restrictions could also be influencing the way researchers write papers, such as changes in citation practices between disciplines (e.g., Wilhite and Fong 2012). The discipline of economics has seen subtler changes in research projects and publications, possibly because the discipline heavily relies on language to construct knowledge and meaning. There is no guideline in economics equivalent to the International Committee of Medical Journal Editors Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals (the ICMJE recommendations), which has helped medical journals standardize the preparation of manuscripts. Because language is more important as a means of knowledge construction in social sciences, as compared to the hard sciences, native English speakers seem to possess advantages over non-­native speakers in the field of economics. For example, Kuteeva and Airey (2014) show that non-­native humanists and social scientists report feeling less confident about using the English language. Kuteeva and Negretti (2016) have indicated that based on their studies of graduate students’ genre knowledge and actual disciplinary writing, “developing genre knowledge in L2 can be more challenging for L2 writers in the humanities and social sciences” (p. 47). Based on such studies, we can assume that English is not as easily adaptable to the authors in economics articles as in the hard sciences and that in economics, language is something the author must master sufficiently to express ideas in depth. It is possible to say the language norm seems still to be dependent on native-­speaker language use in the economics journals, as evidenced in the mention of “language editing” in two economics journals,7 not reflecting the current spread of the use of ELF. Based on these results, I argue that there may be a core and a periphery of academic genre writing. The core is defined by the aspects that do not change, even with accelerated globalization and increasing engagement by non-­native speakers in the discourse community. In the case of research articles, the core consists of the features that characterize the genre, such as its purpose and content regardless of disciplines. Essential move elements and logical argument could vary based on discourse community, and periphery features are subject to changes influenced by English as lingua franca users and its research environment, such as verb tense, voice, and sentence length. This could include changes related to strategies that create a relationship between the reader and the author (Hyland and Jiang 2017). These peripheral changes could be related to standardizing manuscript writing style in research articles for more efficient communication of much information. This standardization may simplify language use, such as the increased use of the present tense even to refer to past studies. It could also lead to complex language use, such as the “compressed”

Exploring diachronic changes in articles  91 English of medical articles, which appears to differ slightly from English use in other disciplines and tends to create more complexity in texts, yet this English use also seems to offer greater flexibility to express detailed and specialized content.

Conclusion English in academic settings is now undoubtedly characterized as the academic lingua franca (Mauranen et al. 2016). In ESP research whose subcategories include EAP, the concept of ELF is not necessarily explicit because English use in ESP tradition is mainly defined by a genre and its discourse community members. However, as it is undeniable that current changes in language use in a particular genre are partly driven by ELF, analysis of language use explicitly from the ELF perspective is called for to establish more reasonable and efficient goals for ESP learners and users because joining professional communities is critical for professional success. Writing is considered more unadaptable to change than speaking (Mauranen 2012), and the research article has been considered among the most change-­resistant of written genres. However, we can observe some marked differences over just the past two decades. Although the linguistic background of the article authors was not examined in this study, these texts were doubtlessly crafted by and to address the rapidly expanding international community of members. This study implies that the researchers who are non-­native English speakers have been and will likely continue to be a driving force to make changes to the genre of the research article. Writing a research paper is context dependent in terms of discipline and journal, and this small-­scale survey, by revealing the changing nature of research article writing, has provided relevant information and an analytical method for teachers to help meet the specific needs of learners. It has shown how contemporary introduction sentences and titles are structured in comparison to those of 20 years ago. It has also shown how teachers or even learners can create their own corpus to obtain information relevant to the journals of interest, instead of relying on general guidebooks for research paper writing or on personal experiences of writing research papers. The modestly focused corpora considered here can be easily updated with the adoption of new data sets corresponding with specific journal articles selected from the target years of publication. This genre-­based approach to academic writing would also allow students to better understand the genre features of a discipline more broadly. Educating students to develop logical arguments based on core features of the research article genre such as the three essential moves found also in the current study irrespective of disciplines and writers’ background should be encouraged – successful writing depends more on professional expertise than just native-­like use of English. During the close reading of introductory sections, the author did not find any English usage that varied significantly from standard English. It may be the case because the target journals were leading journals and their systematic and established screening processes work as a robust gatekeeper to maintain the standard academic English based on native speaker norms. If we examine articles in English journals based in countries where English is not the native language or emerging English-­language journals with less rigorous screening and editing processes and

92  Sayako Maswana less international recognition, the English in these journals might be different from that of this study. However, in medicine, and to lesser extent in economics, there seems to be standardization of the writing style that corresponds with increased participation and agreement of experts in disciplines, regardless of their native status. With clear compartmentalization of content increasingly observed in medical journals,8 English in medical articles has become a neutral linguistic tool that does not favor any particular cultural or linguistic background, a standard that is originally being created based on the native speaker standard but now has become standardized by the discourse community and subject to change with the participation of researchers from various linguistic backgrounds. In comparison with medicine, this tendency is seen less in economics papers, where language is expected to carry meaning and context and in which research projects are carried out by fewer researchers who are more isolated from other researchers than those in medicine. I have used the corpora for specific texts with specific analytical purposes in mind to demonstrate that any practitioner could conduct similar corpus studies with some language features to examine particular genres and publication periods. It is true that with limited texts in a corpus, it is difficult to extract representative patterns of the genre and generalize the results (Flowerdew 2004), but we could say that even if only one article published in reputable journals uses particular language features, it would be a permissible use because the author succeeds in attaining his/her intended communicative purpose, that is, publishing his/her research results in an internationally reputable journal and communicating the results to a wide audience. Having said that, we should also consider obvious shortcomings of a small corpus that can be solved by using existing ELF or general English corpora, such as WrELFA and BNC, by examining whether extracted features are genre-­specific or commonalities. Future research should look at how a small-­scale corpus study could be utilized with existing ELF corpora to provide methodological insights. Furthermore, future research should consider the following. First, while this study used the ESP genre analysis method and selective features in introductions and titles in an attempt to explore an ELF perspective in research articles, there may be other features to consider. Although the study shows a feasible method for ESP writing practitioners to understand the changes in and nature of the research article genre, using small samples and target journals in specific disciplines, more samples focusing on changes in the core and periphery of the research article genre would be necessary. Identified changes in this paper are discussed based on the change in the research environment, including increasing numbers of non-­native experts, although this discussion remains admittedly speculative, other interpretations being possible. Finally, English is the current academic lingua franca, and the English used in leading medical and economics journals still principally follows the norms of native English speakers. However, the study showed the changes over the past two decades, during which there has been a significant increase in researchers whose native language is not English. Given the current research trend, the increased use of English in research articles by non-­native experts as well as changes in its use are expected to continue. Closer investigation is needed to better explain how ELF is used and how it manifests across a wider range of disciplines and contexts in research articles.

Appendix List of references quoted in the examples

1 2

3

Akerman, A., I. Gaarder and M. Mogstad. 2015. The skill complementarity of broadband internet. The Quarterly Journal of Economics 130(4), 1781–1824. Garway-­Heath, D.F., D.P. Crabb, C. Bunce, G. Lascaratos, F. Amalfitano, N. Anand, . . . and T.G. Zeyen. 2015. Latanoprost for open-­angle glaucoma (UKGTS): A randomised, multicentre, placebo-­controlled trial. The Lancet 385(9975), 1295–1304. Georgiadis, G. 2015. Projects and team dynamics. The Review of Economic Studies 82(1), 187–218.

Notes 1 This chapter is a revised version of the article “Changes in research articles from 1995 to 2015: An analysis from an ELF perspective” published in Waseda Working Papers in ELF 6. 2 Engagement is defined as “the ways writers rhetorically acknowledge the presence of their readers in a text” (Hyland and Jiang 2016: 30), whose features are measured by 1) Reader mentions, 2) Questions, 3) Appeals to shared knowledge, 4) Directives, and 5) Personal asides. Relatedly, informality in their study refers to the use of expressions of a personal tenor to build a closer relationship with readers, which are measured by: 1) First person pronouns to refer to the author(s), 2) Unattended anaphoric pronouns that can refer to antecedents of varying length, 3) Split infinitives – an infinitive that has an adverb between to and the verb stem, 4) Sentence initial conjunctions or conjunctive adverbs, 5) Sentence final preposition, 6) Listing expressions, 7) Second person pronouns/determiners to refer to the reader, 8) Contractions, 9) Direct questions, and 10) Exclamations (Hyland and Jiang 2017: 44). 3 For example, three medical journals included in this study started online services in 1995 (The BMJ) and 1996 (The Journal of Clinical Investigation and The Lancet). 4 Formerly known as The British Medical Journal. 5 For example, all medical journals mention the required or recommended maximum number of words in the manuscript in the current guidelines for authors. For economics journals, no specific mention about word count was found in their guidelines. The inclusion of requisite detailed description of research may explain, to some extent, the increase in overall introduction word count. 6 See the Appendix for the references of all the articles whose texts were quoted in the examples. 7 Two economics journals specifically mention language editing, e.g., “Language editing, if your first language is not English, to ensure that the academic content of your

94  Sayako Maswana paper is fully understood by journal editors and reviewers is optional” (The Quarterly Journal of Economics), which seems to be a generic instruction in many journals. Similarly, The Journal of Financial Economics asks authors if “manuscript has been ‘spell checked’ and ‘grammar checked’ ” in the checklist provided for prospective authors. In the medical journals, following the ICMJE recommendations is critical, but the checking of a manuscript by a native English speaker is not specifically mentioned. In addition, all medical journals give input about manuscript writing mostly concerning sections and information to be included. The BMJ also instructs authors to “write in a clear, direct, and active style. The BMJ is an international journal, and many readers do not have English as their first language”. 8 Conforming to the ICMJE recommendations, the information included in a specific section (Introduction, Methods, Results, and Discussion) is standardized, and journals occasionally give recommendations on how to write each section. For example, in a video provided by The BMJ, the journal offers the following tip for introductions: “brief background for this audience, 3–4 paragraphs only. . . .”

References Biber, D. and B. Gray. 2010. Challenging stereotypes about academic writing: Complexity, elaboration, explicitness. Journal of English for Academic Purposes 9(1), 2–20. The BMJ. n.d. Resources for authors. Available from: www.bmj.com/about-­bmj/resources-­ authors [accessed 10 March 2017]. Demner-­Fushman, D., S. Hauser and G. Thoma. 2005. The role of title, metadata and abstract in identifying clinically relevant journal articles. AMIA Annual Symposium Proceedings 2005, 191–195. Dudley-­Evans, T. and M.J. St. John. 1998. Developments in English for specific purposes: A multi-­disciplinary approach. Cambridge: Cambridge University Press. Fletcher, R.H. and S.W. Fletcher 1997. Evidence for the effectiveness of peer review. Science and Engineering Ethics 3(1), 35–50. Flowerdew, L. 2003. A combined corpus and systemic-­functional analysis of the problem-­ solution pattern in a student and professional corpus of technical writing. TESOL Quarterly 37(3), 489–511. Flowerdew, L. 2004. The argument for using English specialized corpora to understand academic and professional language. In U. Connor and T. Upton (eds), Discourse in the professions: Perspectives from corpus linguistics. Amsterdam: John Benjamins, pp. 11–33. Gillett, A. n.d. Using English for academic purposes for students in higher education. Available from: www.uefap.com/reading/underst/difficult/difficult.htm [accessed 21 September 2019]. Heng Hartse, J. and R. Kubota. 2014. Pluralizing English? Variation in high-­stakes academic texts and challenges of copyediting. Journal of Second Language Writing 24, 71–82. Hyland, K. 2002. Options of identity in academic writing. ELT Journal 56(4), 351–358. Hyland, K. and F. Jiang. 2016. “We must conclude that . . .”: A diachronic study of academic engagement. Journal of English for Academic Purposes 24, 29–42. Hyland, K. and F. Jiang. 2017. Is academic writing becoming more informal? English for Specific Purposes 45, 40–51. Jacques, T.S. and N.J. Sebire. 2010. The impact of article titles on citation hits: An analysis of general and specialist medical journals. JRSM Short Reports 1(2), 1–5. The Journal of Financial Economics. n.d. Guide for authors. Available from: www.elsevier. com/journals/journal-­of-­financial-­economics/0304-­405x/guide-­for-­authors [accessed 10 March 2017].

Exploring diachronic changes in articles  95 Kuteeva, M. 2017. Written ELF and EAP: A North European perspective. Waseda Working Papers in ELF 6, 40–55. Kuteeva, M. and J. Airey. 2014. Disciplinary differences in the use of English in higher education: Reflections on recent language policy developments. Higher Education 67, 533–549. Kuteeva, M. and R. Negretti. 2016. Graduate students’ genre knowledge and perceived disciplinary practices: Creating a research space across disciplines. English for Specific Purposes 41, 36–49. Li, L.-­J. and G.-­C. Ge. 2009. Genre analysis: Structural and linguistic evolution of the English-­medium medical research article (1985–2004). English for Specific Purposes 28(2), 93–104. Maswana, S. 2016. A genre analysis of research articles from a chronological perspective. Journal of the Ochanomizu University English Society 6, 27–36. Mauranen, A. 2003. The corpus of English as Lingua Franca in academic settings. TESOL Quarterly 37(3), 513–527. Mauranen, A. 2006. A  rich domain of ELF  – the ELFA corpus of academic discourse. Nordic Journal of English Studies 5(2), 145–159. Available from: http://hdl.handle. net/2077/3152 [accessed 22 September 2019]. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Mauranen, A., N. Hynninen and E. Ranta. 2010. English as an academic Lingua Franca: The ELFA project. English for Specific Purposes 29(3), 183–190. Mauranen, A., N. Hynninen and E. Ranta. 2016. English as the academic Lingua Franca. In K. Hyland and P. Shaw (eds), The Routledge handbook of English for academic purposes. Abingdon, Oxon: Routledge, pp. 44–55. McGrath, L. 2016. Self-­mentions in anthropology and history research articles: Variation between and within disciplines. Journal of English for Academic Purposes 21, 86–98. McLaughlin, G.H. 1969. SMOG grading-­a new readability formula. Journal of Reading 12(8), 639–646. Millar, N., B. Budgell and K. Fuller. 2013. “Use the active voice whenever possible”: The impact of style guidelines in medical journals. Applied Linguistics 34(4), 393–414. The Quarterly Journal of Economics. n.d. Instructions to authors. Available from: https:// academic.oup.com/qje/pages/General_Instructions [accessed 10 March 2017]. Salager-­Meyer, F. 1992. A text-­type and move analysis study of verb tense and modality distribution in medical English abstracts. English for Specific Purposes 11(2), 93–113. Snow, C.E. 2010. Academic language and the challenge of reading for learning about science. Science 328(5977), 450–452. Soler, V. 2007. Writing titles in science: An exploratory study. English for Specific Purposes 26(1), 90–102. Swales, J.M. 1990. Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press. Tribble, C. 2017. ELFA vs. genre: A new paradigm war in EAP writing instruction? Journal of English for Academic Purposes 25, 30–44. Wallwork, A. 2016. English for writing research papers (2nd ed.). Cham: Springer. Ware, M. and M. Mabe. 2015. The STM report: An overview of scientific and scholarly journal publishing (4th ed.). International Association of Scientific, Technical and Medical Publishers. Available from: www.stm-­assoc.org/2015_02_20_STM_Report_2015. pdf [accessed 13 September 2019].

96  Sayako Maswana Wilhite, A.W. and E.A. Fong. 2012. Coercive citation in academic publishing. Science 335(6068), 542–543. WrELFA. 2015. The corpus of written English as a Lingua Franca in academic settings. Director: Anna Mauranen. Compilation manager: Ray Carey. Available from: www. helsinki.fi/elfa/wrelfa.html [accessed 25 March 2017].

7 Tracing the emergence of situational multilingual practices in a BELF meeting Micro-­diachronic analysis and implications of corpus design1 Marie-­Luise Pitzl Introduction Research questions and the methodology chosen to investigate them are supposed to go hand in hand; this is common knowledge and good practice for any academic discipline, including linguistics. Usually the research interest comes first, and the methodology is devised accordingly. Which questions can be investigated empirically, however, is also always influenced by the methods and technologies available at a certain point in time. The availability of electronic corpora has certainly been one of the biggest game changers for linguistics with far-­reaching implications for descriptive work. The study of English as a lingua franca (ELF) interactions is a good example of this. We have come a long way since the early 2000s, when “the need for an ELF corpus” (Mauranen 2003: 514) was something that needed to be fervently argued (and defended against sceptics). The decision to “embark on the compilation of a corpus of English as a Lingua Franca” (Seidlhofer 2001: 145) was, at the time, revolutionary but constituted a crucial step that secured ELF research a place on the map of linguistics. In 2020, corpora like VOICE (Vienna-­Oxford International Corpus of English) and ELFA (English as a lingua franca in academic settings) have been publicly available for over a decade and have been complemented by other (predominantly spoken) ELF corpora like ACE (Asian Corpus of English as a lingua franca) (see L. Wang, this volume). For young scholars and students today, it is normal to have ELF corpora available, although “this was far from obvious before the first ELF corpora came on the scene” (Mauranen 2016: 19). This chapter addresses two themes relating to spoken (ELF) corpora and to the methodology of describing spoken (ELF) interactions: implications of corpus design and principles of micro-­diachronic analysis. The first theme, also discussed in Pitzl (2016a),2 has to do with the importance of corpus-­design principles and the implications of decisions taken in the process of corpus compilation (especially concerning transcription). Posing the question how and to what extent multilingual practices in ELF interactions can be investigated on the basis of the 3rd person evidence as represented in VOICE, the chapter examines this phenomenon in the professional business domain. After a brief contextualization with current (B)ELF research on multilingual practices, it provides an overview of non-­*English speech in BELF meetings in VOICE

98  Marie-­Luise Pitzl and illustrates the development of situational multilingual practices throughout one long BELF meeting. In discussing the long-­term relevance of corpus design, the chapter complements publicly available documentation with project-­internal information. It highlights decisions taken during corpus compilation and the consequences these have for the investigation of multilingual practices on the basis of VOICE, pointing out opportunities and limitations for corpus users. The second theme, not addressed in Pitzl (2016a), concerns the research potential of adopting a micro-­diachronic approach for the analysis of spoken (ELF) interactions. To address this theme, the chapter supplements established methods of data presentation (such as excerpts of transcripts) with visual tools that offer additional and more comprehensive portraits of the data. The proposed micro-­ diachronic approach argues for a combination of conversational, holistic and micro-­diachronic data visualization and suggests that this method of data analysis allows researchers to provide concrete evidence of how patterns of communication and language use are developed by speakers in spoken interaction in real time.

From code-­switching to multilingual practices in (B)ELF research While in the early years of ELF research only relatively few studies were concerned with the use of other *languages3 and code-­switching in ELF settings (e.g. Pölzl 2003; Pölzl and Seidlhofer 2006), aspects of multilingualism gradually gained more attention. Sociopolitically, the spread of *English was perceived as a threat to linguistic diversity, causing scholars like House (2003) to argue that ELF is not in competition, but complementary to multilingualism. As ELF research developed, scholars increasingly linked ELF communication to multilingualism, seeing ELF as “adopting the role of a ‘partner language’ ” (Hülmbauer et al. 2008: 29) that represents “only one of several components of the multilingual repertoire of speakers” (Hülmbauer et al. 2008: 29). While early ELF studies made use of terms like code-­switching (e.g. Klimpfinger 2007, 2009) or explored the effectiveness of cognates as “true friends” in ELF interactions (Hülmbauer 2009: 341), many recent ELF publications tend to build on and extend translanguaging research (e.g. García and Li 2014) and look at multilingual practices in relation to and as part of ELF (see e.g. Cogo 2012, 2016a, 2016b, 2018; Hülmbauer 2016; Hülmbauer and Seidlhofer 2013; Jenkins 2015). So although the multilingual dimension had always been part of ELF research, it has indeed become more central in recent years, leading Jenkins (2015) to propose that the theorizing of multilingualism with ELF – and of ELF as an integral part of multilingualism research – might well constitute a central third phase in ELF research, which warrants “the need for further retheorisation” (Jenkins 2015: 57) of ELF in light of multilingualism research (see Jenkins 2015 for a detailed discussion). A domain in which multilingual practices have been researched prominently is the area of business interactions and workplace discourse. BELF (English as a lingua franca in business contexts) is “inherently intercultural and . . . necessarily multilingual” (Kankaanranta and Louhiala-­Salminen 2010: 205), since “various languages

Situational multilingual practices in BELF  99 and cultures are always present – at least implicitly, but occasionally also explicitly” (Kankaanranta and Louhiala-­Salminen 2010: 205). As Ehrenreich remarks, [a]lthough English is clearly the dominant language in international business, other languages do not disappear from the business scene but interact with English in many ways. This interaction is played out on the individual, the social, as well as the organizational level. . . . [T]he way it is evaluated across the disciplines varies considerably. (Ehrenreich 2010: 411) Thus, *English and other *languages may have different degrees of public (corporate) visibility which may – or may not – correspond to their internal presence and value in companies and business contexts in general. To illustrate these different “ideologies of diversity” (Cogo 2016b: 39) in business contexts, Cogo (2016b: 42) uses Goffman’s distinction between front-­stage and back-­stage communication to distinguish “front-­stage realizations” of multilingualism (for example, in official documents and on company websites) from “back-­stage practices” (for example, in spoken or written informal business communication). The focus of this chapter is on business meetings, which can be either front-­stage or back-­stage, depending on contextual factors of the meeting in question. Some meetings take place among internal staff employed by the same company. These meetings are often relatively informal, especially if staff members are well acquainted with each other and share office space. Such meetings would be part of BELF back-­stage practices (see internal meetings, Table 7.2). Other business meetings are more formal and involve negotiating with representatives of other companies, such as clients, distributors or suppliers (see external meetings, Table  7.2). Depending on factors such as level of acquaintedness or hierarchy scale of meeting participants, such meetings might be part of back-­stage practices or belong to the front-­stage realm of corporate communication. Before I look at contextual factors of business meetings in VOICE in more detail in the next section, there are two sets of questions that need to be addressed in order to investigate the multilingual nature of BELF communication in meetings. The first has to do with terminology, the second with the type of data collected in VOICE. To begin with terminology, questions that present themselves for this case study are: When would we cease to speak of code-­switching and start using the term multilingual practices? Is the difference between the two a matter of extent, i.e. how much languages are actually mixed? And how would multilingual practices then be conceptually different from terms like code-­mixing, (trans)languaging or crossing? Cogo (2012: 289) summarizes this issue in the following way: The terms used in sociolinguistic research of super-­diverse context have included ‘languaging’ (Creese and Blackledge 2010a), ‘translanguaging’ (Creese and Blackledge 2010b; García 2009) and ‘crossing’ (Quist and Jørgensen 2007; Rampton 1995). These are practices performed by multilinguals accessing different linguistic resources and thus enhancing their

100  Marie-­Luise Pitzl communicative potential. They are phenomena that include codeswitching, but go beyond it. (Cogo 2012: 289, my emphasis) In a footnote to this passage, Cogo remarks that, although scholars have given different definitions, all these terms “refer to phenomena affected by and resulting from contacts in super-­diverse contexts” (Cogo 2012: 310), which is why she uses them interchangeably. For the moment, I will follow Cogo’s lead and use the term ‘multilingual practices’ as a cover term that includes code-­switching, but goes beyond it, as Cogo (2012: 289) says. The second set of questions concerns the data and methodological tools used for investigating multilingual practices. As Cogo (2016a: 366) points out, studies on multilingual practices often focus on the participants’ perspective by analyzing 2nd person elicited data gathered in interviews, questionnaires or surveys.4 Some publications (e.g. Cogo 2016a) rely primarily on such 2nd person evidence, while others (e.g. Cogo 2012) combine elicited with 3rd person observed data and rely on observations, emails, websites, recordings and transcriptions of naturally occurring BELF communication alongside interviews and other elicited data. But what if multilingual practices are to be examined with spoken corpus data as the primary source of evidence? Is a corpus a suitable source of evidence to investigate multilingual practices? VOICE is a somewhat atypical corpus, not only because it is a spoken corpus, but primarily since it was, from the beginning, designed to be suitable for qualitative research. It contains a high degree of structured metadata and contextual information for each speech event (see e.g. Breiteneder et  al. 2006, 2009) and follows fine-­grained transcription conventions (VOICE Project 2007a, 2007b; see also Pitzl 2018a: 86–89). In consequence, VOICE corpus texts are detailed conversation analytic transcripts of a quality rarely found in other corpora. But of course, VOICE is, after all, a corpus, which means that it offers access to 3rd person observed evidence (i.e. ELF interactions), but not to 2nd person elicited data. The first central methodological question of this chapter is therefore: To what extent is it possible to investigate multilingual practices in BELF meetings only on the basis of the particular kind of 3rd person-­recorded evidence captured in VOICE (i.e. corpus texts with accompanying metadata, mark-­up and annotation)? The second central methodological question concerns data analysis: How can the situational emergence and joint development of multilingual practices in spoken interaction best be empirically described and visualized?

Non-­*English speech in BELF meetings in VOICE Transcribing, annotating and retrieving non-­*English speech: the relevance of corpus design Although VOICE predates the recent focus on multilingual practices and translanguaging, it was already obvious during its compilation that “code-­switching is an intrinsic element of ELF talk” (Klimpfinger 2007: 58). Words and utterances

Situational multilingual practices in BELF  101 from *languages other than *English were uttered in almost all speech events recorded for the corpus, which prompted the question how these were to be represented in transcription. Early on, the decision was taken to include “specific tags to mark non-­English speech which also allow for a distinction between the speaker’s first or other languages” (Breiteneder et al. 2006: 182) and to provide “translations into English . . . whenever . . . possible in order to make the non-­ English speech accessible for . . . future users of the corpus” (Breiteneder et al. 2006: 182). As a result of this, all transcripts for VOICE include mark-­up that signals when speakers use words, phrases or utterances that are not *English (see VOICE Project 2007a: 4–5). In addition to specifying whether the switch is into a speaker’s first *language (L1) or another ‘foreign’ *language (LN), the tags also give information about the *language switched into (see VOICE Project 2007a: 4–5). When part-­of-­speech (POS) tagging was added to the corpus between 2010 and 2013, all words that had been marked as non-­*English in the transcripts (as released in 2009) were annotated with a FW (foreign word)5 tag to allow researchers to retrieve individual non-­*English items via the POS interface of VOICE (see Osimk-­Teasdale 2015: 169; VOICE Project 2014). So what does this mean for investigating multilingual practices in VOICE? Initially, after the first release of the corpus (VOICE 2009), it had already been possible to look for passages marked as non-­*English speech by means of reading through individual speech events in the corpus. With the release of the XML version (VOICE 2011), it became possible to extract certain types of switches with more precision (e.g. switches into L1, into LN, into a particular language) using corpus tools like WordSmith or Xaira. With the release of the POS-­tagged version of VOICE (2013b), non-­*English speech in the recorded speech events was also annotated at the word level. Table 7.1 provides an overview of the number of FW items in the professional business (PB) and the professional organizational (PO) domain of VOICE, with a focus on meetings as a speech event type.6 As can be seen, the percentage of FW items ranges from 0.11% to 1.22%. Before offering any further interpretation of these numbers, there are three crucial points to be made from a corpus compiler’s perspective concerning the design principles of VOICE. The first point is that 1,888 FW elements in meetings in the professional business (PB) domain are not to be equated with 1,888 instances of code-­switching. Table 7.1  ‘Foreign words’ in the professional domains of VOICE Corpus section

FW items (VOICE POS)

No. of words (No. of events)

Percentage of FW items

VOICE PB (professional business) domain PO (professional organization) domain Meetings Meetings in PB Meetings in PO

7,903 1,946 827 2,015 1,888 127

1,023,127 203,413 (23 events) 354,581 (41 events) 273,438 (20 events) 154,619 (8 events) 118,819 (12 events)

0.77% 0.96% 0.23% 0.74% 1.22% 0.11%

102  Marie-­Luise Pitzl As is well documented in research, code-­switches in conversations (and thus also in VOICE) may involve individual lexical elements, but also collocations, shorter or longer phrases or entire utterances. In all but the first scenario, one instance of code-­switching (i.e. one 0.05 SS vs. Med: G2 = 18.92, p < 0.0001 NS vs. Med: G2 = 10.08, p < 0.01

feature characteristic of the Med lecturers. The frequencies of interpersonal FSs are similar between the sub-­corpora of Social Sciences (SS) and Natural Sciences (NS); indeed the log-­likelihood test yielded no significant difference between the two (p > 0.05), while the difference between each of these two and Med is statistically significant. Figure 9.1 shows the distribution of the four broad functional categories in the sub-­corpora. The three disciplinary groups have much in common in terms of the categories of Modality and Commitment, accounting for about 30% and 20%, respectively, of all interpersonal FSs in each sub-­corpus. Apart from the similarities, SS stands out in its relatively high proportion of Engagement FSs, while both Med and NS feature Evaluation FSs in particular. 100% 90% 80% 70% 60%

Engagement

50%

Commitment

40%

Modality

30%

Evaluation

20% 10% 0% Social Sciences

Natural Sciences

Medicine

Figure 9.1  Distribution of the four broad functional categories in the sub-­corpora

Interpersonal formulaic sequences  149 Engagement FSs occurring in the SS sub-­corpus include politeness markers (e.g., i’m afraid that, thank you very much, i’m sorry); instructions (e.g., please go ahead, turn the last page); questions or invitation for comments (e.g., do you wanna say something, what do you think); and other ways of engaging/involving the audience (e.g., as you know, you’ve heard about). In terms of politeness markers, while some of them, such as thank you very much, were found in almost all lectures, the three lecturers in SS employed more types and seemed to rely on them to a greater extent than those in the other two groups. Such a tendency can be seen in Excerpts 1 to 3, which were taken from the three SS lectures. Another point to be made is that the use of those politeness formulae, such as thank you very much, i’m afraid that, you are welcome to, if you don’t mind, rarely shows any deviation from native usage (in terms of both meaning and function) or any variation in form, suggesting chunk processing.

Excerpt 1 . . . thank you very much i will continue the er the next lecture what i left erm if you have any questions i will be very happy to er to to (xx) go ahead please, any questions please, yes yeah. (ULECD030)

Excerpt 2 well i’m afraid that this will be mostly er lectures but, er, hopefully er at least at towards the end of er of of the lecture there there is a discussion too er and er you are welcome to interrupt er at any point and er ask for clarification or put some remarks so please be active and aggressive as americans say. (ULEC01A)

Excerpt 3 and if you don’t mind i go this list through. . . .

(ULEC020)

Excerpts 1 and 2 also illustrate the use of some FSs as a common means of encouraging engagement (if you have any questions, any questions please, ask for clarification, put some remarks). The sequence put some remarks is a rare example that does not quite match the target language use. In this case, the choice of delexical verb (put instead of make) reflects a common phenomenon in learner language probably as a result of the operation of the open-­choice principle in cases which can be analysed as normal syntactic constructions but at the same time may have subtle semantic and/or syntactic restrictions in target language usage (see Wang 2016: 6 for a more detailed discussion). Formal variations were more likely to occur with questions or invitations for comments as in does anyone wanna say something, say something please, would you like

150  Ying Wang to add something, you wanna say something, do you wanna say something, are there any questions. Evaluation FSs make up a more considerable proportion in the sub-­corpora of Med and NS than in SS. Such FSs were often used to express the importance or difficulty of the issue in question (e.g., it’s very important to, it’s not very difficult to, it’s hard to, it’s not very hard, a big problem). Again, they were by no means exclusive to any disciplinary group but were more frequently used by the lecturers in Med and NS. As can be seen in Excerpts 4 and 5, most of these FSs were used to comment on particular procedures, materials, and instruments, which are typical entities that hard science fields often deal with (Kashila and Heng 2014).

Excerpt 4 and i think there are three important erm models i think it’s very important to take into account that there are different models because. . . . (ULEC150)

Excerpt 5 . . ., but with optical methods how do you look at plasma membrane it’s not always an easy task because. . . . (ULEC23A) A need to direct the audience’s attention to these entities has led to a large number of textual FSs in the NS lectures in particular, as observed in Wang (2018b). The same need may have prompted the high frequency of Evaluation FSs in the Med and NS sub-­corpora in the present study, which were used to show the speaker’s attitude or assessment towards these entities. Excerpt 4 illustrates another feature of the use of Evaluation FSs, this time regardless of discipline, namely the tendency of the lecturer employing an intensifier or booster, such as very, extremely, absolutely. This may be seen as a characteristic typical of spoken academic discourse in general, in contrast to the written one, where hedging (i.e., linguistic devices used to reduce the force of statements) is crucial when the author expresses his/her own stance on a particular subject (Hyland 1996). On the other hand, one distinction between hard and soft academic domains may be relevant in explaining the different proportions of such FSs in the dataset. As Hyland (1998) explains, researchers in hard sciences generally see their knowledge as building on what has come before in the field, and there are normally clear-­cut criteria of what constitutes a new contribution. In contrast, research in soft domains is often influenced by contextual vagaries, and there are fewer unequivocal bases for accepting claims. As a result, research in soft fields is not normally reported with the same confidence of shared assumptions as seen in hard sciences. While the distinction between the two may be blurred to some extent in spoken discourse, it can still be detected in the present study. However, it remains to be known how much of what was found in this respect is ELF-­specific.

Interpersonal formulaic sequences  151 Excerpt 6 illustrates the use of another common set of Evaluation FSs, namely those involving the word interesting. Once again, such FSs were found in all sub-­ corpora. However, in comparison, the lecturers in SS seemed to have a particular preference for FSs to do with emotion or cognition (e.g., i’m very much satisfied, a big joke, surprisingly small, believe me or not, what was surprising, take X seriously, it’s funny to). More examples are given here:

Excerpt 6 . . . we also have a study in depressive patients where we thought saw that the the temporal structure of beta oscillations has impaired more er than the amplitude so it’s i think it’s interesting if it’s so that looking at the temporal structure can be often more sensitive a more sensitive measure to. . . . (ULEC180)

Excerpt 7 . . . but we were lucky to get i will, let er let this copy go around so you see the the the product this is the best seller of any UNESCO publication. . . . (ULEC01A)

Excerpt 8 . . . i have studied much of those forecasts which has been made in finland in the 20th century and and usually they are badly wrong but the happy side of the story is that things have usually gone better than it has than then has been predicted. (ULEC020) The wide variety of such FSs in the SS sub-­corpus may be taken as adding further support to the view that soft academic domains are by nature more diverse in terms of discourse conventions than hard fields. Excerpt 8 saw one of the lecturers ‘creatively’ using a conventional expression in standard English, the other side of the story, by blending in an ‘emotional’ word happy. Such creations are, nevertheless, rare with interpersonal FSs in general in the dataset. Congruency with frequency-­based identification As introduced earlier, IDIOM Search was used as a pointer to potential FSs on account of frequency of occurrence. The identified FSs were coded YES if they are congruent with IDIOM Search results (e.g., I’m afraid that, more or less, we can say that). Those that were considered formulaic but failed to be identified by IDIOM Search were given the code NO (e.g., with your er permission, as you see from, as usually). In some cases, such as i would more talk about, it also means that, only part of the sequences (talk about, means that) were identified by the

152  Ying Wang frequency-­based program and were coded PARTLY. Table 9.3 presents the proportions of the three sets of interpersonal FSs in the data. Table 9.3 shows that about one third of the FSs that were manually identified were missed completely by the frequency-­based program. Table 9.4 gives some examples of those FSs, which were put in four broad categories. Table 9.3  Congruency with the IDIOM Search results Type

Yes

No

Partly

Total

No. of occurrence (Proportion)

968 (35%)

874 (31%)

935 (34%)

2777 (100%)

Table 9.4  Types of FSs that are not identified by IDIOM Search Type

Examples

Variations

as we often say, now we say, we said, you could say, you say, so i say, i should say that, you can really say, i will say, i would rather say you do see, we do see, as you can see, you see (that), you can really in the raw signals see that, we see, we will see that you all know, as all of you know, we know (that) with your er permission, it was er said in this way that, in the very long run, it’s nice i think to it was typical for finland that, as much hydrogen as possible i’m not alone, in and out, it’s a good questions, in practice as ususally, are more capable to, as the matter of fact

Intervening elements Underlying frames Continuous FSs Derivations

Apart from a small number of continuous FSs that were not identified by the computer, probably because of their low frequency of occurrence in general, and an even smaller set of FSs that contain derivations from the target language use, all the other unrecognised FSs involve intervening elements of some kind. The majority of such FSs contain formal variations. For instance, one very common FS indicating the speaker’s commitment to his/her assessment was found to be i would say in the dataset, as illustrated by Excerpt 9. While i would say and we can say that were identified by IDIOM Search, variations of the form in terms of the personal pronoun, modal verb, and tense, as shown in Excerpts 10–12 (i say, i will say, we would say), were all overlooked by the program.

Excerpt 9 . . . and it’s highlighted by this resolution which very openly and i would say strongly in view of today’s er perspective speaks of er the need of media to contribute to peaceful relations between countries. (ULEC01A)

Interpersonal formulaic sequences  153

Excerpt 10 . . . it w-­it has been believed that it’s all DST is produced by the ring-­current but it’s not true so i i say this works like ring-­current. . . . (ULEC160)

Excerpt 11 . . . what it stands for erm when the er today the world in short form i will say the intensification of economic political and social and cultural relations across borders that’s what’s about globalisation today. . . . (ULECD030)

Excerpt 12 . . . so in the kraft black liquor we would say this goes from about six to eight as a result of condensation. . . . (ULEC080) All these sequences, despite their formal variability, have the same meaning and function as that of i would say – that is, to indicate who is responsible for the upcoming proposition. There is, therefore, no reason to dismiss them as idiosyncratic usage. Using the same methodological approach, Wang (2018a) examines interpersonal FSs in novice and expert academic writing. The results indicate that formulaicity is a feature of successful professional writing in the sense that even those that seem to be semantically transparent and syntactically flexible expressions (e.g., as X argue) have more or less fixed usage patterns that are preferred by expert writers. Novice writers, by contrast, demonstrate a greater degree of formal variability in accordance with the open-­choice principle. This seems to be the case with regard to academic speech in ELF settings – the variations in those seemingly ‘regular’ expressions, which have not attracted much attention in formulaic language research due to their idiosyncratic as well as unmarked nature, may well be a feature of ELF phraseology that is worth closer investigation in future research. Excerpts 13 and 14 illustrate a less common type of FSs failed to be captured by the computer. They are broken due to hesitation, which is a common phenomenon in spoken discourse.

Excerpt 13 . . . and i can then er assu-­assure to you that if something is left sort of, mhm unclear to you in today’s er morning presentation you will you will be er er erm, furnished with er adequate er material in due course. . . . (ULEC01A)

154  Ying Wang

Excerpt 14 . . . what is er still most still more important is that the change which has taken place in finland in the past 200 years. . . . (ULEC020) Excerpt 15 illustrates a case where one FS (it is known that) may be interrupted by another formulaic unit (of course), again making it difficult for automatic identification. In addition, there are underlying formulaic frames that are discontinuous but could have been easily identified with a pre-­set list, given their limited number of types.

Excerpt 15 . . . it is of course known that er very important decrease in beta cell numbers is responsible for the clinical symptoms to appear. . . . (ULEC150) Very seldom do the unrecognised FSs represent deviation from the target norm, such as be capable to in Excerpt 16. Such deviations are also discussed in Mauranen (2012: 144), who argues that they are not likely to be caused by constructing expressions from the smallest components in accordance with the open-­choice principle, but rather represent approximate versions of whole entities that seem to be stored and retrieved as single choices.

Excerpt 16 . . . they are more capable to make innovations and people are capable to calculate the benefits. . . . (ULEC020) A final point can be made in comparison with Wang (2018b), which looks at FSs signalling discourse organisation in ELF academic lectures, using the same data and methodological approach as in the present study. A much higher percentage of textual FSs were missed completely by the computer than that of interpersonal FSs (50% vs. 31%), suggesting that the latter may involve fewer formal variations or/and discontinuous sequences as a whole. In other words, FSs with different functions may display different degrees of formulaicity, a feature that needs to be taken into account in understanding the role of formulaic language in differentiating different social practices and types of language users.

Conclusion In an attempt to overcome some of the inherent limitations of a frequency-­ based approach, especially for the analysis of ELF data, the present study used a

Interpersonal formulaic sequences  155 computer-­assisted manual approach in the identification and annotation of FSs. The results provided additional insights into the nature of formulaicity in ELF academic lectures. In particular, it was found that the lecturers’ preferences for different interpersonal FSs were likely to be associated with different disciplines in terms of how knowledge is constructed and presented (e.g., Evaluation FSs in hard sciences vs. Engagement FSs in soft sciences). While cross-­disciplinary differences have attracted a large amount of attention in previous formulaic language research on written discourse, the results of the present study indicated that more needs to be known about the impact of this factor on spoken academic discourse and on different speaker varieties. The study also revealed some features that were common to the whole group of speakers, regardless of discipline. A large number of the interpersonal FSs were used by the lecturers in the same way as in native usage, suggesting the operation of the idiom principle. At the same time, a great deal of formal variability was associated with what seem to be semantically and syntactically regular FSs, which may well be ELF-­specific. However, given the small quantity of data involved, such a claim requires further validation. With regard to methodology in formulaic language research, the results gave further support to Wang’s previous studies (2018a, 2018b) in that the frequency-­ based approach may be particularly problematic in dealing with spoken and/or ELF data, which inevitably involve irregularities of different kinds. The manual approach clearly has potential to complement our understanding of formulaic language use in different communities of practice derived from the frequency-­based approach. However, manual identification, as any other manual approach, has the limitation of being subject to the researcher’s subjective judgments in difficult cases, and can only be carried out to a certain extent while more data are needed in order to draw more informative conclusions. Therefore, future research needs to keep addressing these methodological challenges in order to effectively capture all FSs, continuous or discontinuous, in large amounts of data.

Notes 1 This study was supported by the Swedish Research Council (437–2014–6696). 2 www.oxfordlearnersdictionaries.com

References Ädel, A. 2008. What uh the folks who did this survey found: Expert attribution in spoken academic lectures. Nordic Journal of English Studies 7(3), 83–102. Ädel, A. and B. Erman. 2012. Recurrent word combinations in academic writing by native and non-­native speakers of English: A  lexical bundles approach. English for Specific Purposes 31(2), 81–92. Biber, D. 2009. A corpus-­driven approach to formulaic language in English: Multi-­word patterns in speech and writing. International Journal of Corpus Linguistics 14(3), 275–311. Biber, D., S. Conrad and V. Cortes. 2004. If you look at . . . : Lexical bundles in university teaching and textbooks. Applied Linguistics 25(3), 371–405.

156  Ying Wang Biber, D., S. Johansson, G. Leech, S. Conrad and E. Finegan. 1999. Longman grammar of spoken and written English. Harlow: Pearson. Buerki, A. 2016. Formulaic sequences: A drop in the ocean of constructions or something more significant? European Journal of English Studies 20(1), 15–34. Cogo, A. and J. House. 2017. Intercultural pragmatics. In A. Barron, Y. Gu and G. Steen (eds), The Routledge handbook of pragmatics. London and New York: Routledge, pp. 168–183. Colson, J. 2016. IDIOM search. Available from: http://idiomsearch.lsti.ucl.ac.be/index.html. Cortes, V. 2004. Lexical bundles in published and student disciplinary writing: Examples from history and biology. English for Specific Purposes 23, 397–423. Durrant, P. 2017. Lexical bundles and disciplinary variation in university students’ writing: Mapping the territories. Applied Linguistics 38(2), 165–193. Granger, S. 1998. Prefabricated patterns in advanced EFL writing: Collocations and formulae. In A.P. Cowie (ed), Phraseology: Theory, analysis, and application. Oxford: Oxford University Press, pp. 145–160. Halliday, M.A.K. 2014. Halliday’s introduction to functional grammar (4th ed., revised by Christian M.I.M. Matthiessen). Abingdon, Oxon: Routledge. House, J. 2014. English as a global Lingua Franca: A threat to multilingual communication and translation? Language Teaching 47(3), 363–376. Hyland, K. 1996. Writing without conviction? Hedging in science research articles. Applied Linguistics 17(4), 433–454. Hyland, K. 1998. Boosting, hedging and the negotiation of academic knowledge. Text 18(3), 349–382. Hyland, K. 2008a. As can be seen: Lexical bundles and disciplinary variation. English for Specific Purposes 27(1), 4–21. Hyland, K. 2008b. Academic clusters: Text patterning in published and postgraduate writing. International Journal of Applied Linguistics 18(1), 41–62. Hyland, K. 2012. Disciplinary identities: Individuality and community in academic discourse. Cambridge: Cambridge University Press. Kashila, H. and C.S. Heng. 2014. Discourse functions of formulaic sequences in academic speech across two disciplines. Journal of Language Studies 14(2), 15–27. Kjellmer, G. 1991. A mint of phrases. In K. Aijmer and B. Altenberg (eds), English corpus linguistics: Studies in honour of Jan Svartvik. London: Longman, pp. 111–127. Martinez, R. and N. Schmitt. 2012. A phrasal expressions list. Applied Linguistics 33(3), 299–320. Mauranen, A. 2008. The transcriptions of the ELFA corpus, downloadable version [text corpus]. Kielipankki  – The Language Bank of Finland. Available from: http://urn.fi/ urn:nbn:fi:lb-­2014052721. Mauranen, A. 2009. Chunking in ELF: Expressions for managing interaction. Intercultural Pragmatics 6(2), 217–233. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. O’Donnell, M. 2013. UAM corpus tool. Version 3.0. Pawley, A. and F.H. Syder. 1983. Two puzzles for linguistic theory: Native-­like selection and native-­like fluency. In J.C. Richards and R.W. Schmidt (eds), Language and communication. London: Longman, pp. 191–225. Schleef, E. 2008. The “Lecturer’s OK” revisited: Changing discourse conventions and the influence of academic division. American Speech 83(1), 62–84.

Interpersonal formulaic sequences  157 Schneider, N., S. Onuffer, N. Kazour, E. Danchik, M.T. Mordowanec, H. Conrad and N.A. Smith. 2014. Comprehensive annotation of multiword expressions in a social web corpus. Proceedings of the 9th Linguistic Resources and Evaluation Conference, Reykjavík, Iceland. Seidlhofer, B. 2009. Accommodation and the idiom principle in English as a Lingua Franca. Intercultural Pragmatics 6(2), 195–215. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Simpson-­Vlach, R. and N.C. Ellis. 2010. An academic formulas list: New methods in phraseology research. Applied Linguistics 31(4), 487–512. Sinclair, J. 1991. Corpus, concordance, collocation. Oxford: Oxford University Press. Wang, Y. 2016. The idiom principle and L1 influence: A contrastive learner-­corpus study of delexical verb + noun collocations. Studies in Corpus Linguistics (vol. 77). Amsterdam and Philadelphia: John Benjamins. Wang, Y. 2017. Lexical bundles in spoken academic ELF: Genre and disciplinary variation. International Journal of Corpus Linguistics 22(2), 187–211. Wang, Y. 2018a. As Hill seems to suggest: Variability in formulaic sequences with an interpersonal function in L1 novice and expert academic writing. Journal of English for Academic Purposes 33, 12–23. Wang, Y. 2018b. Formulaic sequences signalling discourse organisation in ELF university lectures: A disciplinary perspective. Journal of English as a Lingua Franca 7(2), 355–376. Wray, A. 2002. Formulaic language and the lexicon. Cambridge: Cambridge University Press. Wray, A. 2008. Formulaic language: Pushing the boundaries. Oxford: Oxford University Press.

Part III

Conversation analytic approaches to ELF interactions

10 Applying conversation analysis to ELF interaction data Jagdish Kaur

Introduction In what is probably the first empirical study of English as a lingua franca (henceforth ELF), Firth (1990, 1996) adopted conversation analytic methods to examine international trading negotiations taking place over the telephone in what he termed “lingua franca English”. The research was a departure from the deficit-­ oriented approaches to the study of the English used by its non-­native speakers prevalent among applied linguists of that time. While mindful of the challenges the conversation analyst would face when dealing with non-­native speaker interaction data (discussed in further detail later in this chapter), Firth succeeded in demonstrating how conversation analysis (CA) is able to shed light on how speakers in ELF contexts use English “interactively to accomplish social goals” (1990: 269) without the need for recourse to native speaker norms. Firth’s early work is significant in that it paved the way for the application of discourse analytic methods that underlie an overwhelming proportion of empirical research of ELF interaction that has since followed. In the two decades following Firth’s CA study of lingua franca English, ELF as a field of inquiry has undergone exponential growth. ELF itself, which Seidlhofer defines as “any use of English among speakers of different first languages for whom English is the communicative medium of choice, and often the only option” (2011: 7), is fast becoming a global sociolinguistic reality that brings in its wake pertinent questions of how the language is used by speakers who are diverse for purposes that are wide-­ranging in any number of settings and domains (Cogo 2012a). Not unlike Seidlhofer who describes ELF as “a variable way of using it [English]” (2011: 77), Firth asserts that variability is a key feature of ELF (2009). It is precisely this variability in form and use, both within and across speakers, which contributes in part to “the complexity of ELF” (Baird et  al. 2014: 172) that calls for a research methodology that is able to provide the researcher with access to the intricate workings of ELF as it is used in interaction in real-­world settings. In order to explicate “how ELF speakers use the language in and on their own terms” (Seidlhofer 2011: 23), CA continues to be a favoured methodology and analytical tool amongst a growing number of researchers keen to uncover the methods and resources speakers adopt when they interact in ELF.

162  Jagdish Kaur The present chapter explores the use of conversation analysis to study English as a lingua franca talk-­in-­interaction. It begins with a consideration of some of the key principles of CA that make it well suited to shed light on the kinds of interactional work ELF speakers do to accomplish their communicative goals. Selected studies that have adopted the CA methodology in their study of ELF interaction and their findings are then reviewed. The review examines in particular research that has addressed the topics of intersubjective understanding and the interactional management of turn-­taking in ELF. The chapter concludes with a discussion of some of the implications of adopting CA in the study of ELF interaction, both to ELF and CA as fields of inquiry.

CA and ELF interaction data – the issue of compatibility Conversation analysis, which draws from ethnomethodology in sociology, is based on the core assumption that talk is “a deeply ordered, structurally organized phenomenon” (Hutchby and Drew 1995: 183). The concern then is to explicate how this order is achieved in interaction through micro-­analyses of naturally occurring talk. Or as Liddicoat puts it, the task of the conversation analyst is to “discover and describe  .  .  . the sets of procedures which participants in conversation deploy in order to achieve orderly and ordered social interaction” (2011: 6). To do so, analysts adopt an emic or participant-­based perspective where observations are informed by the orientations of the participants themselves as revealed in the details of the talk. In line with the notion of talk as social action, analyses focus on what participants do when they speak and the practices they adopt to accomplish various actions in and through talk. In this regard, the sequential organization of talk is fundamental given that talk proceeds on a turn-­by-­turn basis; as participants take turns to speak, the “sequential placement” (Schegloff and Sacks 1973: 299) of an utterance within a sequence of talk determines the action the utterance performs. The sequential organization of talk also constitutes a key resource that participants rely on to produce and interpret talk. Not only do speakers design their current turn – appropriately and coherently – to fit the prior one but their interlocutors likewise rely on its placement to interpret the communicative action of the turn in question (Hutchby and Wooffitt 1998). Additionally, it is the sequential organization of talk that aids the analyst in his or her interpretations of what participants do in and through talk. Sacks et al. explain this as follows: Since it is the parties’ understandings of prior turns’ talk that is relevant to their construction of next turns, it is their understandings that are wanted for analysis. The display of those understandings in the talk of subsequent turns affords both a resource for the analysis of prior turns and a proof procedure for professional analyses of prior turns. (1974: 729) Analyses that are grounded in what speakers themselves do, on a turn-­by-­turn basis, and the analyst’s reliance on next turn proof procedure to interpret actions

Applying conversation analysis to data  163 have contributed to CA emerging as “one of the most powerful and influential methodologies hitherto developed to analyse talk” (Firth 1996: 237). However, since conversation analysts initially focused on the conversations of first language (L1) speakers of English, a measure of controversy surrounds the use of CA to analyse other types of interaction data including those which involve non-­ native speakers of English as well as speakers of languages besides English (see Pietikäinen, this volume). The main argument against the use of CA to examine non-­native speaker interaction data (e.g., ELF, although ELF interactions may include NES) is the lack of commonality between speakers, particularly in terms of language competence, as well as the absence of membership knowledge – both linguistic and cultural  – on the part of the analyst which is said will adversely impact the reliability of the analysis (Firth 1996, 2012). CA purists seem to regard the “prototypical conversationalist” as “a monolingual speaker in a stable first language setting, preferably the analyst’s own” (Wagner 1996: 232). Firth (2012) also suggests that the initial lack of interest in lingua franca data may in part be due to the lack of accessibility to and familiarity with such data. Given the availability of L1 interactions in English in the settings in which CA was initially developed (i.e., the United States), conversation analysts naturally focused their attention on this type of data. Schegloff (in Wong and Olsher 2000), however, sets the record straight when he asserts that analysts did not deliberately avoid non-­native speaker interaction data. Rather, they seemed not to consider “non-­native” a distinct categorical membership that necessitated investigation; a category is only of interest to the analyst if the speakers themselves make it relevant in their interactions. Schegloff, who regards non-­native speaker interactions as “a sub-­area in the study of talk-­in-­interaction” (Wong and Olsher 2000: 119), maintains that analyses of such data must be conducted no differently from other CA work. As such, the issue of language competence is only evoked if the participants themselves orientate to it in their talk. Applying CA to lingua franca interaction data is, however, not without its share of challenges. Given the diversity in participants’ linguacultural backgrounds and language competence, the analyst may face some difficulty in establishing the interactional significance of various linguistic and interactional phenomena such as silence, laughter, restarts, etc.; specifically, are such phenomena to be treated as they are in English L1 interactions or might they be linked to the participants’ language competence? (Bushnell 2015). The guiding principle in such instances, as Firth asserts, is to focus “analytic attention on the participants’ demonstrable orientations and relevancies as they are publicly displayed and continuously updated” (1996: 244). Seedhouse (1998) likewise emphasizes the need to base all interpretations on the evidence available in the details of the talk. As both Firth (1996) and Seedhouse (1998) illustrate, CA’s methodological principles are amenable to the analysis of non-­L1 English interactions. The methodology has since been adopted to study different types of data including the spoken interactions of speakers of various languages around the world. Sidnell’s (2009) edited volume for example adopts a comparative perspective to examine the similarities and differences in interactional practices across a range of languages including

164  Jagdish Kaur Mandarin, Russian, Japanese, and Finnish. Similarly, the methodology has been applied to conversations involving second language speakers of a language as in the case of the contributions to Gardner and Wagner (2004). Since Firth’s (1990, 1996) CA study of ELF, others have adopted CA, either solely or in combination with other methods, to study ELF talk-­in-­interaction (e.g., Haegeman 1996; Meierkord 2000; Gramkow Andersen 2001; Cogo 2009; Kaur 2009, 2010, 2011, 2012; Jenks 2012; Matsumoto 2011, 2018; Konakahara 2015, 2016, 2017; Kappa 2016; Pietikäinen 2014, 2018a, 2018b; O’Neal 2015; O’Neal and Matsumoto 2018). These studies evidence the suitability of CA both as a methodology and an analytical tool to investigate how speakers in ELF encounters interact with one another and make meaning using a language that for most is just one resource within their linguistic repertoire. Using CA, with its meticulous attention to the details of the talk, researchers are able to meet Seidlhofer’s call for “rich, empirically well founded descriptions” (2011: 23) of what speakers do when they use ELF. Kaur (2016) highlights some of the reasons why CA is especially compatible with the ELF paradigm. Firstly, empirical ELF research continues to focus on spoken usage given the particular interest in how speakers achieve communicative success and effectiveness using ELF. As Mauranen explains, “it is in real interaction that the effectiveness of communication comes out” (2012: 73). CA, with its obvious focus on naturally occurring spoken interaction, provides the analyst with the resources and tools to uncover the practices and procedures speakers adopt to communicate effectively in ELF. Secondly, present-­day ELF research is centred on the use of ELF rather than on the language itself. As Cogo states, “research in this area is not about identifying the core features that make ELF a variety (which it is not), . . . it is about describing the practices involved in lingua franca communication” (2012b: 98). This ties in well with CA, which as Hutchby and Wooffitt explain, “is only marginally interested in language” (1998: 14); rather the interest is in how talk is interactionally accomplished. Finally, a major concern of ELF research is to uncover how speakers of diverse linguacultural backgrounds whose use of English is variable and flexible are able to understand one another in ELF. Similarly, intersubjectivity constitutes a core concern of conversation analysts; Hutchby and Wooffitt confirm this when they say that “one of the most basic issues in CA: [is] the question of how understanding is accomplished and displayed in talk” (1998: 41). Both CA and ELF adopt a similar approach to the study of intersubjectivity in talk in that it is viewed as locally negotiated and accomplished through participants’ use of relevant procedures and strategies. In addition to the aforementioned reasons, Mauranen’s observation that “there is regularity in ELF despite large numbers of non-­standard features” (2012: 43) reinforces the argument that CA is well equipped to handle ELF interaction data. In the following section, CA-­informed ELF studies that investigate intersubjective understanding in ELF interaction are reviewed to not only highlight some of the key features of talk in ELF but also illustrate the use of conversation analytic methods to examine this type of data. The transcription conventions used in all the extracts are listed in the Appendix.

Applying conversation analysis to data  165

Intersubjective understanding in ELF Intersubjectivity, as Sidnell explains, is “joint or shared understanding between persons” (2010: 12) which is accomplished when interacting speakers converge in their knowledge and comprehension of matters. As discussed earlier, the sequential organization of talk provides for the display of speakers’ understanding of each other. Specifically, the second speaker displays in the next turn his or her understanding of a prior turn by producing a relevant next action. As the speaker’s understanding is made public, the first speaker is able to check the sense that was made of his or her prior utterance; if the understanding is found to be wanting, the first speaker may then initiate and/or perform repair to re-­establish intersubjective understanding. The analyst is privy to the same sense-­making procedures that the speakers adopt and in this way is able to avoid the “problems of indeterminacy and interpretation that afflict other approaches to language and discourse” (Sidnell 2010: 14). The interest of researchers in the subject of intersubjective understanding in ELF interaction is neither surprising nor unexpected. Given that ELF contexts are inherently diverse, “hybridity, fluidity, and variability are, therefore, the main characteristics of ELF communication” (Cogo 2012a: 290). How speakers successfully and effectively negotiate and co-­create understanding in the face of variable forms and use of English has understandably become a productive area of inquiry within ELF research. Matsumoto (2011) for example adopted sequential analysis to examine the phonological negotiations that international graduate students at a university in the United States participate in during their dinner table talk. As speakers in ELF contexts display a variety of accents resulting from the influence of their L1, phonological intelligibility becomes all the more important for the accomplishment of intersubjective understanding (see also Jenkins 2000). In her analysis, Matsumoto highlights how speakers deploy various pronunciation negotiation strategies in repair sequences to specifically increase phonological intelligibility when mutual understanding is at stake. In the extract reproduced here, Matsumoto illustrates how a speaker modifies her pronunciation to accommodate her interlocutor when her pronunciation is oriented to as problematic.

Extract 1 (from Matsumoto 2011: 102) 20. 21. 22. 23. 24. 25. 26. 27.

P:→You (.) went to library at weekend? Y:→((nods)) (4.0) P:→So what did you do >{[ǝt wi:kdeı]}not in a< (.) one goal £>y’knowsometimes< you-­(.) [(0.5)= CN2: [yeah TH3: =[take a] brea:k or (.) [c h a : t or] TH4: [yeah] CN2: [brea:k] [when you] when you go to your office everyday. TH3: hhmm::: normall:y (1.3) [eleve: : : : : : : :n¿] TH4: [nine ten eleven¿] TH3: °(not that) (*)° [huh huh] TH4: [hhhuh huh] huh=

This extract, extracted from casual conversations between international university students in Britain, involves three participants – CN2, TH3, and TH4 – who are course-­mates on a doctoral programme. According to Konakahara, in the sequence preceding this extract, CN2 had posed a question regarding the number of hours of study her co-­participants put in in a day. Responding to CN2’s indication of surprise in line 1, TH3, in line 4, clarifies that 6 hours of study is not undertaken in a stretch. She then begins to expand on what she means starting with “sometimes you” followed by a cut-­off and a brief 0.5-­second pause (line 5) which suggests a possible word search. CN2 proffers the word “break” (line 9) which overlaps with TH3’s own continuation of her utterance with “take a break or” (line 7); this is followed by a pause of very brief duration. At this point, an overlap occurs as a result of TH3 applying rule 1c of Sacks et al.’s (1974: 704) turn construction rules where she produces an additional turn-­constructional unit (TCU) while CN2 applies rule 1b where she projects an upcoming TRP and self-­selects for next speakership. Konakahara explains that “this simultaneous application of the turn taking rules results in overlapping” (2015: 45). Thus, while TH3 continues by clarifying what she means (in relation to the message originating in line 4), CN2 begins to pose a question in connection to the developing topic. CN2 repeats the overlapped segment before proceeding with her question while TH3, through relevant embodied actions, “shifts into recipiency” (2015: 45) and hands over the floor to CN2. According to Konakahara, based on gaze orientation, CN2 selects TH3 to respond, which she does in line 11. The overlapping question which claims turn space and

Applying conversation analysis to data  171 results in speaker change before a TRP is not oriented to as disruptive by the participants given the absence of signals that generally accompany competitive or disruptive overlaps such as complaints and the like (see Konakahara 2015). As Konakahara puts it, the speakers “cooperatively develop the ongoing talk without clinging to the overlap” (2015: 46). Like Meierkord (2000), this study highlights how overlaps – particularly those that are perceived as competitive based on their onset place  – may instead function to demonstrate engagement and the active involvement of participants in the interaction. As previously stated, in Sacks et al.’s (1974) organization of turn-­taking, transition from one turn to the next is said to occur with little or no gap. Silence that occurs at a TRP is therefore considered to have interactional relevance, and is generally oriented to as trouble or a violation of order that makes repair relevant. Meierkord (1998), in her analysis of casual conversations in ELF taking place at a student hall of residence, attributed the frequent occurrence of pausing in the speech of speakers with lower levels of competence to production problems. Conversely, speakers with higher levels of competence in the language used pausing strategically; in order to elicit a reaction from co-­participants, speakers paused so that their interlocutors could display understanding or seek clarification. Pietikäinen (2018b), keen to uncover some of the interactional functions of silence at TRP, examined conflict talk in established intercultural couples who use English as their chosen medium of communication. Specifically, Pietikäinen analysed gaps at TRP where the second speaker withholds a response which is then followed by the first speaker self-­selecting. The next extract, reproduced from Pietikäinen (2018b), illustrates one specific function of silence in her data which is to avoid giving a self-­incriminating response.

Extract 5 (Pietikäinen 2018b: 80) 80 S: 81 82 G: 83 → 84 →G: 85 86: S: 87

you wanna hang around >people li-­with ha-­people like< that (1.4) aren’t-­it, aren’t those normal people (2.9) WHA-­hh (2.1) ° it de°PENDS what you say w-­what (.) what you conside:r norm’l

According to Pietikäinen, in the sequence preceding this extract, S had been criticizing the behavior of G’s friends. S’s question in line 80, particularly the segment “people like that”, makes clear her less-­than-­favourable opinion of G’s friends. Instead of responding to S, G, after a pause of 1.4 seconds, counters her question with another which reframes his friends as “normal people”. G’s question in line 82 is a first pair part (FPP) of a question-­answer sequence that makes relevant a second pair part (SPP) in the form of an answer in the next turn. S, however,

172  Jagdish Kaur withholds her response (line 83), and after a gap of 2.9 seconds, G takes up the next turn to press for an answer. After a further delay of 2.1 seconds, S finally takes up the next turn to provide a rather ambivalent response (lines 86–87) that suggests they have different views on what is considered “normal”. Pietikäinen explains that in withholding her response following G’s initial question, S avoids a “self-­incriminating” (2018b: 80) second pair part, which also allows her some time to come up with a more conciliatory response. According to Pietikäinen, in cases such as this extract, the sequence follows the following format: “FPP (question) + withholding a response + urging + SPP (answer)” (2018b: 81). Thus, in conflict talk, the first speaker is more likely to follow the silence at TRP with a formulation that presses for an answer rather than one that seeks to be agreeable. Analyses from the studies reviewed earlier reflect how speakers in ELF interaction use various interactional devices like the turn-­taking mechanism and silence to create orderly talk as they work to achieve their communicative goals. What is clear from the analyses is that speakers of ELF are equally adept and skillful in the manner in which they operate various interactional practices to collaboratively construct conversation. CA, which demands that meticulous attention be given to participants’ actions within and across turns of talk and that analyses be grounded in what participants themselves treat as significant in their interactions, is able to shed light on how speakers in ELF contexts co-­create talk that is meaningful, normal and ordinary (Firth 1996).

Implications of applying CA to ELF interaction data Findings from CA-­informed studies of ELF have added significantly to our knowledge and understanding of what speakers do when they communicate in ELF. Accounts of how speakers using ELF cooperate and support one another as they conjointly negotiate meaning and achieve understanding using various interactional devices can be traced to some of the early CA work on ELF (e.g., Firth 1990, 1996; Haegeman 1996; Meierkord 2000; Gramkow Andersen 2001). Subsequent research based on corpus linguistics methods and discourse analyses of ELF corpora such as the Vienna-­Oxford International Corpus of English (VOICE), the Corpus of English as a Lingua Franca in Academic Settings (ELFA) and the Asian Corpus of English (ACE), for example, (see Seidlhofer 2011; Mauranen 2012; see also Pitzl; L. Wang, this volume) has affirmed the findings of the aforementioned studies, and been consistent with the findings of more recent CA-­research of ELF (e.g., Cogo 2009; Kaur 2009, 2010, 2011, 2012; Matsumoto 2011, 2018; Konakahara 2015, 2016, 2017; Pietikäinen 2014, 2018a, 2018b; O’Neal 2015; O’Neal and Matsumoto 2018). While findings from many CA-­informed ELF studies point to “a high degree of collaboration and cooperation” (Firth 2012: 1045–1046) amongst speakers in ELF interaction, this should not be taken to mean that ELF talk is intrinsically cooperative and mutually supportive (Jenks 2012; Kappa 2016; Pietikäinen 2018b). Kappa, who also adopts CA methods in her study, quite rightly points out that on the topic of solidarity and consensus in ELF, a distinction needs to be made

Applying conversation analysis to data  173 between “avoidance of misunderstandings and non-­understandings from doing relational work” (2016: 17). In the case of the former, speakers in ELF contexts, driven by the need to make themselves understandable as well as to understand their interlocutors as they strive to accomplish communicative goals, display collaborative and accommodative behaviors that contribute to meaning-­making. For example, speakers work to enhance the clarity of expression (e.g., Kaur 2011, 2012) which is “generally taken to be a collaborative strategy  .  .  . a form of accommodation, as it is a way of accommodating to the hearer’s perceived interpretive competence” (Mauranen 2012: 51). Accommodation ties in with a key concept in CA which is “recipient design”, defined as “a multitude of respects in which the talk by a party in a conversation is constructed or designed in ways which display an orientation and sensitivity to the particular other(s) who are the co-­participants” (Sacks et al. 1974: 727). As ELF contexts are characterized by diversity and variability in language use, speakers are compelled to produce recipient-­designed talk that accommodates their interlocutors so as to effectively and successfully construct and re-­construct talk that is orderly, meaningful and comprehensible. This demands that speakers collaborate and support one another as they pursue shared understanding and the accomplishment of goals. In the case of relational work, however, depending on the context of interaction, speakers’ turns may be less than affiliative. Jenks (2012), who examined the use of ELF in online voice-­based chatrooms using CA methods, observed how the participants in his study engaged in reprehensive talk with interactional behavior that was neither supportive nor consensus-­oriented. On the contrary, the participants drew attention to the disfluencies and ungrammaticalities in the speech of their co-­participants through the use of laughter, ridicule and humour (cf. Firth 1996). Jenks explains that supportive behavior “is determined by the norms, expectations and interactional and institutional goals of the communicative context” (2012: 402). This is said to explain the affiliative and mutually supportive behaviors observed in academic and business settings (in which many ELF studies are conducted) where participants are constrained to cooperate in order to achieve institutional goals; however, exceptions have been noted in the case of the latter (e.g., in a professional ELF meeting on the matter of a construction project), where the “wider sociocultural context” may contribute to communication that is more adversarial in nature (see Tsuchiya and Handford 2014: 126). On the basis of a turn-­by-­turn analysis of ELF interactions based on emic perspectives, both Jenks (2012) and Kappa (2016) reveal the presence of dis-­affiliative behaviors among speakers using ELF to communicate in their data. These CA-­informed ELF studies underscore the need to not only take into greater account the settings and situations in which the interactions take place but to also extend ELF research to interactions that involve conflict, discord and disagreement in order to explore the “full scope of possible interactional trajectories” (Kappa 2016: 30; see also Seidlhofer, this volume) taking place in ELF. While research that adopts CA methodological and analytical principles in the study of ELF interaction has contributed significantly to the rapid growth of the ELF paradigm, such research also has much to offer the field of CA. Given CA’s

174  Jagdish Kaur expressed aim to explicate the shared methods and procedures participants use to produce and interpret talk-­in-­interaction, applying CA to ELF interaction data will serve to extend the scope of inquiry of the field. As the numbers of non-­native speakers of English now surpass those of native speakers worldwide, English is increasingly becoming the communication medium of choice in international encounters (Jenkins 2015). A logical, concomitant development would be for conversation analysts to more fully engage with this type of interactional data. The increasing number of ELF researchers that adopt CA methods in their analyses, as cited earlier, bears testimony to the applicability and suitability of the methodology to this type of data, in spite of the variable forms and usage present. As participants in ELF interaction are found to “orient to a patterned way of conducting their encounters” (Firth 2009: 162), there is regularity in the variable forms and usage produced which allows for credible and well-­founded accounts of recurrent phenomena. The earlier concerns about the lack of linguistic and cultural membership knowledge amongst participants, and between participants and the analyst, are largely inconsequential in the context of analyzing ELF interaction data as “norms [are] negotiated ad hoc depending on specific participants’ repertoires and purposes” (Seidlhofer 2011: 80). This, together with CA’s insistence on grounding the analysis in the details of the talk, liberates both participants and the analyst from having to attribute meaning to individual cultural frames, unless oriented to by the participants themselves. Applying CA to ELF interaction data certainly “raises new and interesting questions” (Seedhouse 1998: 101) which can only contribute to developing CA theory and practice. Meanwhile, CA-­informed ELF studies have but scratched the surface of what it means to use ELF to communicate with others of different linguacultural backgrounds. It is perhaps timely for a concerted cross-­disciplinary enterprise to delve deeper into the present sociolinguistic reality that is ELF.

Appendix Transcription conventions

[ onset of overlapping talk ] end of overlapping talk [[ simultaneous start = latched utterance wor-­ cut off (.) micro pause (0.5) timed pause in seconds : stretched or elongated sound ? rising intonation (indicating a question) . falling intonation (indicating end of utterance) , continuing intonation ! animated intonation ¿ a slightly rising intonation. ↑ ↓ rising/declining intonation (extract 3 only) word emphasis or stress h.h.h. laughter .hhh audible inhalation hhh audible exhalation CAPS section spoken louder °soft° section spoken softer >word< section spoken faster section spoken slower £ or $ section spoken in a smiley voice (word) transcriber’s uncertainty of the actual words produced () unrecoverable speech (**) unrecoverable speech (extract 4 only) (()) transcriber’s comment/contextual information/nonverbal behavior Italic code-­switch and translation (Nor) code-­switch into mentioned language (e.g., Nor = Norwegian)

176  Jagdish Kaur

References Baird, R., W. Baker and M. Kitazawa. 2014. The complexity of ELF. Journal of English as a Lingua Franca 3(1), 171–196. Bushnell, C. 2015. Lost in translation? On using conversation analysis to examine cross-­ linguistic data. Area Studies Tsukuba 36, 107–126. Cogo, A. 2009. Accommodating difference in ELF conversations: A  study of pragmatic strategies. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 254–273. Cogo, A. 2012a. ELF and superdiversity: A case study of ELF multilingual practices from a business context. Journal of English as a Lingua Franca 1(2), 287–313. Cogo, A. 2012b. English as a Lingua Franca: Concepts, use, and implications. ELT Journal 66(1), 97–105. Firth, A. 1990. “Lingua Franca” negotiations: Towards an interactional approach. World Englishes 9(3), 269–280. Firth, A. 1996. The discursive accomplishment of normality: On “Lingua Franca” English and conversation analysis. Journal of Pragmatics 26(2), 237–259. Firth, A. 2009. The Lingua Franca factor. Intercultural Pragmatics 6, 147–170. Firth, A. 2012. Conversation analysis and Lingua Franca. In C.A. Chapelle (ed), The encyclopedia of applied linguistics. Oxford: Wiley Blackwell, pp. 1043–1047. Gardner, R. and J. Wagner (eds). 2004. Second language conversations. London: Continuum. Gramkow Andersen, K. 2001. The joint production of conversation. Aalborg: Centre for Languages and Intercultural Studies, Aalborg University. Haegeman, P.  1996. Business English in flanders: A  study of Lingua Franca telephone interaction. PhD dissertation, University of Ghent, Ghent. Hutchby, I. and P. Drew. 1995. Conversation analysis. In J. Verschueren, J.A. Ostman and J. Blommaert (eds), Handbook of pragmatics manual. Amsterdam: John Benjamins, pp. 182–189. Hutchby, I. and R. Wooffitt. 1998. Conversation analysis: Principles, practices and applications. Oxford: Polity Press. Jenkins, J. 2000. The phonology of English as an international language. Oxford: Oxford University Press. Jenkins, J. 2015. Global Englishes: A resource book for students. London: Routledge. Jenks, C.J. 2012. Doing being reprehensive: Some interactional features of English as a Lingua Franca in a chat room. Applied Linguistics 33(4), 386–405. Kappa, K. 2016. Exploring solidarity and consensus in English as Lingua Franca interactions. Journal of Pragmatics 95, 16–33. Kaur, J. 2009. Pre-­empting problems of understanding in English as a Lingua Franca. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 107–125. Kaur, J. 2010. Achieving mutual understanding in world Englishes. World Englishes 29(2), 192–208. Kaur, J. 2011. Raising explicitness through self-­repair in English as a Lingua Franca. Journal of Pragmatics 42(11), 2704–2715. Kaur, J. 2012. Saying it again: Enhancing clarity in English as a Lingua Franca (ELF) talk through self-­repetition. Text & Talk 32(5), 593–613. Kaur, J. 2016. Conversation analysis and ELF. In M.-­L. Pitzl and R. Osimk-­Teasdale (eds), English as a Lingua Franca: Perspectives and prospects. Berlin: de Gruyter Mouton, pp. 161–168.

Applying conversation analysis to data  177 Konakahara, M. 2015. An analysis of overlapping questions in casual ELF conversation: Cooperative or competitive contribution. Journal of Pragmatics 84, 37–53. Konakahara, M. 2016. A conversation analytic approach to ELF communication: Incorporating embodied action in the analysis of interactional achievement. In K. Murata (ed), Waseda working paper in ELF (vol. 5). Tokyo: Waseda ELF Research Group, pp. 78–96. Konakahara, M. 2017. Interactional management of face-­threatening acts in casual ELF conversation: An analysis of third-­party complaint sequences. Journal of English as a Lingua Franca 6(2), 313–343. Liddicoat, A.J. 2011. An introduction to conversation analysis. London and New York: Continuum. Matsumoto, Y. 2011. Successful ELF communications and implications for ELT: Sequential analysis of ELF pronunciation negotiation strategies. The Modern Language Journal 95(1), 97–114. Matsumoto, Y. 2018. Functions of laughter in English-­as-­a-­lingua-­franca classrooms: A multimodal ensemble of verbal and non-­verbal interactional resources at miscommunication moments. Journal of English as a Lingua Franca 7(2), 229–260. Mauranen, A. 2006. Signaling and preventing misunderstanding in English as Lingua Franca communication. International Journal of the Sociology of Language 177, 123–150. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Meierkord, C. 1998. Lingua Franca English: Characteristics of successful non-­native/non-­ native speaker discourse. Available from: http://webdoc.sub.gwdg.de/edoc/ia/eese/eese. html [accessed 21 October 2004]. Meierkord, C. 2000. Interpreting successful Lingua Franca interaction: An analysis of non-­native/non-­native small talk conversations in English. Linguistik Online 5, 1. Available from: http://linguistik-­online.com [accessed 21 October 2004]. O’Neal, G. 2015. Segmental repair and interactional intelligibility: The relationship between consonant deletion, consonant insertion, and pronunciation intelligibility in English as a Lingua Franca in Japan. Journal of Pragmatics 85, 122–134. O’Neal, G. and Y. Matsumoto. 2018. Beyond intelligibility: “Transintelligibility” phenomena in English as a Lingua Franca interactions. International Journal of Applied Linguistics 29, 44–60. Pietikäinen, K. 2014. ELF couples and automatic code-­switching. Journal of English as a Lingua Franca 3(1), 1–26. Pietikäinen, K. 2018a. Misunderstanding and ensuring understanding in private ELF talk. Applied Linguistics 39(2), 188–212. Pietikäinen, K. 2018b. Silence that speaks: The local inferences of withholding a response in intercultural couples’ conflicts. Journal of Pragmatics 129, 76–89. Sacks, H., E.A. Schegloff and G. Jefferson. 1974. A simplest systematics for the organization of turn-­taking for conversation. Language 50, 696–735. Schegloff, E.A., G. Jefferson and H. Sacks. 1977. The preference for self-­correction in the organization of repair in conversation. Language 53, 361–382. Schegloff, E.A. and H. Sacks. 1973. Opening up closings. Semiotica 7, 289–327. Seedhouse, P. 1998. CA and the analysis of foreign language interaction: A reply to Wagner. Journal of Pragmatics 30, 85–102. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Sidnell, J. (ed). 2009. Conversation analysis: Comparative perspectives. Cambridge: Cambridge University Press.

178  Jagdish Kaur Sidnell, J. 2010. Conversation analysis: An introduction. West Sussex: Wiley Blackwell. Tsuchiya, K. and M. Handford. 2014. A corpus-­driven analysis of repair in a professional ELF meeting: Not “letting-­it-­pass”. Journal of Pragmatics 64, 117–131. Wagner, J. 1996. Foreign language acquisition through interaction: A  critical review of research on conversational adjustments. Journal of Pragmatics 26(2), 215–235. Wong, J. and D. Olsher. 2000. Reflections on conversation analysis and nonnative speaker talk: An interview with Emanuel A. Schegloff. Issues in Applied Linguistics 11, 111–128.

11 Analysing multilingual/lingua franca interactions using conversation analysis Notes on transcription and representability1 Kaisa S. Pietikäinen Language can be partially defined as a system of conventional arbitrary vocal symbols. Writing, the use of pothooks to represent that language, is thus merely the symbol of a symbol, a conventionalized convention, arbitrariness squared. Walpole (1974: 191)

Introduction A growing number of researchers working on English as a lingua franca (ELF) have found the methodological framework and rigorous analytic tradition of conversation analysis (CA) suitable for their research, particularly because of the emic approach of the methodology which relies on local orientations of participants rather than researchers’ interpretations of salient phenomena (see Firth 2012; Kaur 2016 and this volume for more detailed argumentations for the use of CA in ELF research). Intrinsic to the methodology is its meticulous transcription system which largely grounds on the conventions developed by Gail Jefferson (Jefferson 1983, 2004). It attempts to record every minute detail of interaction including timed pauses, audible inhalations and exhalations, hesitations, cut-­offs, prosody, and (some) pronunciation particularities. The Jeffersonian system was initially developed based on data collected in the US mainly from native English speakers, which is why it is often found to be inadequate or too stiff for transcribing data as variable and multilingual as ELF interaction.2 Especially the requisite to use what Jefferson calls “ ‘comic book’ orthography” (1996: 160) – commonly known as eye dialect – sometimes challenges researchers dealing with lingua franca/multilingual data. And even more often, this “transcription that attempts to capture pronunciation details with a conventional alphabet” (ibid) confuses readers who attempt to make sense of the transcripts. For example, Jefferson (1996: 160) proposes that a turn uttered by a Danish speaker in ELF data prepared for a conference in Odense would be more accurately transcribed as: (1a) AB: here in Tenmug we have a weddy stormy weather

180  Kaisa S. Pietikäinen than: (1b) AB: here in Denmark we have a very stormy weather or that the word ‘of’ should rather be transcribed as ‘of’, ‘uff’, ‘ohv’, ‘awv’, ‘off’, ‘awf’, or ‘aff’, depending on its pronunciation (ibid: 161). Without a gloss or very obvious contextual cues, eye dialect transcriptions are clearly very difficult to comprehend. This is, however, not the only problem with eye dialect transcripts, as will be discussed in this chapter. The current chapter focuses on whether eye dialect can be used in transcribing lingua franca talk so that the resulting transcripts are representable of the data but at the same time they remain decipherable by the reader. The paper adopts a hands-­on approach by exploring practicalities of transcribing and analysing multilingual interaction and by suggesting general guidelines for transcribing ELF interaction for CA-­informed studies. In the first section, I seek to find a proper definition for eye dialect from previous literature before I outline some pitfalls with its aims to represent “talk as heard”. The second section discusses some alternatives for transcribing multilingual/lingua franca interaction and proposes the use of a simplified version of the International Phonetic Alphabet (IPA) for disambiguating unclear words and trouble sources. The third section proposes a four-­step protocol for transcribing data where language alternation may pose challenges to the analyst, while the last section concludes the paper with general recommendations.

What is eye dialect? Although conversation analysts have traditionally favoured eye dialect in transcription, it is close to impossible to find a proper definition for this practice of orthographic alternation in CA papers. Even Jefferson shows reluctance toward discussing her method of transcription: Although I’d probably rather transcribe than do any other part of the work (analyzing, theorizing, lecturing, teaching, etc.), the one thing I’d rather not do is talk about transcribing. It’s not a topic. You might as well talk about typewriting. Transcribing is just something one does to prepare materials for analysis, theorizing, etc. Do the best you can, but what is there to talk about? Jefferson (2004: 13, original emphasis) What is there to talk about in conversation analytic transcription? A lot, as it turns out (see Bucholtz 2000; Haberland and Mortensen 2016; Hepburn and Bolden 2017; Jefferson 1983, 2004; Jenks 2011; Ochs 1979; Preston 2000, among others). Concerning CA’s orthographic presentation, Preston (2000) criticises eye dialect transcripts for unnecessarily obscuring the data from the reader,3 based on the crucial observation that for eye dialect, “there is no key to what such respellings

Notes on CA transcription  181 indicate, and each reader will have to recreate this authenticity for him or herself” (pp.  616–617). Indeed, few research papers explain explicitly how eye dialect is used in transcripts. For example, which sounds or phonemes4 correlate with which combinations of letters, are they used consistently, and which pronunciation standard (if any) does standard orthography represent? Often authors of CA papers seem to assume that the reader is familiar with eye dialect transcripts, and furthermore, that s/he will be able to decode the transcripts similarly to the transcriber (see Haberland and Mortensen 2016 for transcription as entextualisation and retextualisation). Kowal and O’Connell (2014: 71) argue that in principle, the eye dialect is an even greater deviation from standard orthography than a literary transcription, as it attempts to represent actual pronunciation “in a pseudo-­ phonetic way”. Due to its arbitrariness – that is, in practice, the transcriber decides the spelling of every phone/phoneme based on little more than a gut feeling – it has been widely criticised for inconsistency, phonetic inaccuracy, and poor readability (see, e.g., Preston 1985, 2000; Bucholtz 2000). Consider, for example, the following extract from Jefferson (2004: 30), where Jim’s turn is practically indecipherable: (2a) Jim:

Don’t sound so (h)amp(h)itious fer Ch(h)rise’sake (h)ih suh ·hh sou’ l(h)i’ yuh k(h)uh g(h)o tuh sleep ‘n the pho(h)ne.  . . . Maggie: I j’st ran up th’ stai(hh)rs that’s wh(h)y I’m huffing en puffing. In the search for a proper definition for the term eye dialect, one is compelled to turn to literature and translation studies, where several definitions are found on how the term is used in literary works. According to McArthur (2003), eye dialect was first coined in The English Language in America (Krapp 1925) to describe colloquial or dialectal language use, i.e., the speech of “uneducated, youthful, rustic [speakers], or [those who are] otherwise unlike the readership” (2003). Using eye dialect in literary works was meant to indicate “a sympathetic sense of superiority” over or humorosity of “the humble speaker of dialect” (Krapp, cited in McArthur 2003). Already here a degree of loadedness can be detected toward the speaker whose vernacular is illustrated with eye dialect, as opposed to the more traditional or educated tone of the narrator whose voice is commonly portrayed in standard spelling. Similar stylistic differences appear in virtually every academic CA paper: the author (the researcher) uses academic language in explaining the research; only spoken data (the talk of the researched) is described in eye dialect. One cannot avoid the thought that perhaps some of the “pejorative stereotypes” (Gumperz and Berenz 1993: 96f) that the eye dialect forms of its objects in literature are also present in academic literature. Even if the best intention of the researcher is not to condescend to the speech styles of the researched and to solely present the data as phonetically accurately as possible without the use of the IPA,

182  Kaisa S. Pietikäinen the historical burden of the eye dialect prevents the readers from knowing whether this really is the case. There is also evidence that points to eye dialect causing readers to underestimate the speaker’s status (Preston 1985), which makes the problem an ethical one. I will try to illustrate this problem with an example from the introductory section: Even if Jefferson unlikely meant to mock the Danish speaker of English in transcribing his turn as “here in Tenmug we have a weddy stormy weather”, the reader of the transcript can easily interpret the transcription as indicating that the speaker was grossly mispronouncing the English name of his own country! This problem is particularly relevant for CA researchers who in principle should avoid comparing ELF/multilingual speakers’ pronunciation against an external standard unless it is emically made relevant in the interaction. Of course, there are contradicting views, too. Brett (2009: 49), for instance, claims that the term eye dialect has developed since Krapp’s time and is now used to indicate the transcription of any variation of pronunciation or accent. He argues that some authors would rather use the term “semi-­phonetic spelling” to describe the same practice, presumably to avoid the stigma around the term eye dialect. This does not, however, save the researcher from tackling with a further issue of authenticity, which is evident in trying to capture spoken language in a written form. Even before Krapp coined the term eye dialect, authors have attempted to come up with more accurate ways of spelling spoken language than standard spelling. Walpole (1974) reports George Bernard Shaw’s futile struggle to create a sound transcription system for spoken English. After exhausting the reader with what Walpole refers to as “eye dialect with a vengeance” (p. 195) in Eliza Doolittle’s opening speech in Pygmalion, he continues the text with standard orthography and adds this bracketed comment for the reader: “Here, with apologies, the desperate attempt to represent her dialect without a phonetic alphabet must be abandoned as unintelligible outside London” (Shaw, cited in Walpole 1974: 195). If such a famous playwright abandoned the idea of non-­standard spelling in the face of unintelligibility, why do some conversation analysts insist on using inconsistently produced eye dialect in their transcripts? Whether one chooses to call CA’s “mock phonetic” spelling eye dialect, semi-­phonetic spelling, or non-­standard spelling, the reality is that transcriptions are impressionistic representations of how the researcher perceives the recording and thus inescapably subjective (see e.g., Ochs 1979; Bucholtz 2000). Presenting the transcription with other than standardised sound-­symbol combinations risks further misinterpretations of the data by the reader. Haberland and Mortensen (2016: 586–587) describe this dilemma with eye dialect transcriptions as follows: the researcher-­transcriber has essentially created a private language which probably only (s)he can use in order to reproduce the original sound (s)he remembers from listening to the first-­order entextualization caught on tape.5 The eye dialect is thus little more than a mental note to the transcriber of the way in which the talk was produced by the speaker (or rather, heard by the analyst).

Notes on CA transcription  183 This chapter suggests that CA practitioners should adopt a more reader-­oriented approach to published transcriptions and, regardless of what their private “working transcriptions” may look like, attempt to make their published transcriptions as unambiguous as possible. This can be done by attempting to minimise orthographic alternations that can rather cause misreadings than help readers understand the transcript. The next section will deal with some specific problematics of the eye dialect and suggest ways to deal with these issues. Pronunciation particularities and eye dialect It is generally accepted that CA transcriptions should capture at least pauses (in brackets; measured if longer than 0.2 seconds), overlaps (in square brackets), latching (=), elongations (:), pitch alternations (upward and downward arrows and punctuation), word stress (underlining), loudness (capital letters for loud, degree signs for quiet/soft talk), audible inhalations (.hh) and exhalations (hh), cut-­offs (-­), faster or slower talk (surrounded by > < or < >), and laughter/smiley voice (£, $, or @).6 A key of these symbols is usually given as an appendix. In contrast, few papers explain their conventions of transcribing pronunciation particularities. In their introduction to Second Language Conversations, Wagner and Gardner (2004) discuss the pros and cons of modified orthography. They argue that eye dialect shows details which may later turn out to be significant for the analysis and claim that such particularities may be especially important when transcribing second language (L2) talk (2004: 6). However, why especially L2 interaction would require transcriptions in eye dialect if, presumably then, L1 does not, is not discussed. Nevertheless, they too acknowledge the problem of stereotyping the speaker, as well as issues regarding readability and electronic searchability, concluding that the discussion of transcription standards remains unresolved. According to Jefferson (1983), pronunciation particularities which invoke other-­repair, self-­repair, cause problems of understanding or otherwise “might well account for subsequent activities” (p.  2) should be transcribed to detail. Yet, even Jefferson admits omitting some information while transcribing other sections more meticulously: “Of course there’s a lot of stuff ‘there,’ i.e., in the tapes, and it doesn’t all show up in my transcripts; so it’s because it’s there, plus I think it’s interesting” (Jefferson 2004: 15). This shows how, even at the earliest stages of transcription, the researcher will make decisions which may have essential consequences to the analysis. It is therefore important to bear in mind that analysts should never work with the transcription alone and should always refer back to the original recording when working with the analysis.7 The “working transcripts” that analysts use while assembling the analysis can range from more or less standardised text files for the sake of enabling computer searches, yet tagged with pronunciation particularities, prosody markings, etc. that might otherwise restrict searchability, to extremely detailed transcripts where every minute detail is recorded either in eye dialect or with phonetic symbols. Many conversation analysts who work with ELF data utilise the first type of working transcripts for mapping the phenomena of interest and then produce modified transcripts for

184  Kaisa S. Pietikäinen publication purposes, where the level of detail is consistent and justifiable against the purpose of the analysis. While the transcriptions are adjusted, the analysis is detailed, working in the spiralling fashion recommended by, e.g., ten Have (2007). If, for example, the topic of interest is the use of metaphors in ELF, a less detailed transcript may suffice, whereas a paper on the relationship between sentence prosody and comprehension grants a detailed record of pitch and stress variation.8 Next, I will return to extract (2a) mentioned earlier from Jefferson (2004: 30) in order to discuss what level of detail suffices for transcripts meant for publication purposes, for papers that do not specifically focus on pronunciation or prosody. Jefferson uses the extract to exemplify what she names as “plosiveness”; breathy or airy quality of voice indicating “laughter, crying, breathlessness, etc.”. What we see in (2a), however, is that the plosiveness markers occur not only after plosive consonants (e.g., “(h)amp(h)itious”; “k(h)uh”), but approximants (“l(h)i’ ”; “wh(h)y”) and vowels (“pho(h)ne”; “stai(hh)rs”) as well, indicating that they are the researcher’s observations of the actual moments in time where airy voice quality is heard. This is of course generally compatible with the principle of transcribing talk as heard, but perhaps using the name “plosiveness” is then an unfortunate choice. Furthermore, using ‘(h)’ to indicate breathy, porous voice quality is a widely used CA transcription practice, but here it is rather confusing and obstructing the reader from comprehending Jim’s turn. In addition, as for example, smiley voice is not indicated with the proper symbols (see earlier in this chapter); the porousness of Jim’s voice gives no indication of “the colour” of his voice  – is he copying Maggie’s breathlessness, crying, or is he himself out of breath? If his whole turn is uttered in a breathy manner, it could have been made easier to read with a simple adjustment, by removing the plosive markers and instead adding a comment on Jim’s voice quality: (2b) Jim (breathily): Don’t sound so ampitious fer Chrise’sake ih suh ·hh sou’ li’ yuh kuh go tuh sleep ‘n the phone.  . . . Maggie: I j’st ran up th’ stai(hh)rs that’s wh(h)y I’m huffing en puffing. Now, let us focus on the orthography of the same extract. Even after the removal of the plosive markers from Jim’s turn, it is evident that the transcript is next to impossible to decode. The choice between standard and non-­standard orthography seems arbitrary. In “ampitious”,9 the standard ending -­itious is spelled letter by letter, even if an eye dialect version could just as well be written as “ampishes” if the ending is uttered as [ˈɪʃəs],10 (providing that ‘e’ represents a mid-­central vowel sound /ə/ like it apparently does in the words “fer”, weak form of ‘for’ and “en”, weak ‘and’). A question therefore arises whether the ending is, instead, pronounced as something like [ˈɪtiəs]? Bucholtz (2000) argues that varying between

Notes on CA transcription  185 standard and non-­standard spellings in this way risks the misinterpretation that a form is non-­standard in the context when it is not, and vice versa, as is also the case with our two example words “fer” and “en” in the extract. Jefferson (2004: 28–29) explains that when she uses a dot below a word that has been written in standard spelling, for example, “fọr”, this means “that it is there fully pronounced as ‘for’ ” (p.  29), possibly indicating that the word is pronounced in its strong form, [fɔːr]. However, there are no dots below “ampitious”, so its pronunciation remains a mystery. The dots below and above vowels also function as hardeners and softeners in Jefferson’s transcripts, which is likely why many analysts have decided against following this transcription practice minutely and rather use a modification of it. Also, with multilingual data, the dots immediately become problematic when used with languages whose spelling includes umlauts or diacritical marks. It is therefore recommended that, in transcripts of ELF where the pronunciation variety may not be of ultimate importance and can, if necessary, be indicated with commentary in the analysis (see e.g., Brouwer 2004), both strong and weak forms of words should be transcribed following standard orthography if their pronunciation does not differ drastically from the standard the speaker’s pronunciation most closely follows.11 Since stress is routinely indicated with underlining in CA transcriptions, it should be fairly straightforward for the reader to decode underlined words using strong forms and non-­underlined words as weak forms, as this comes naturally for fluent English speakers. However, using standard forms only is not always an ideal solution either; for example, such informal forms as “cos” or “cause”, “gonna”, and using an apostrophe to mark non-­standard contractions would unnecessarily be standardised, and therefore skew the reader’s understanding of the unfolding of the interaction. It is therefore recommended that common colloquialisms should be transcribed “as heard”, and if these are rendered ambiguous, they can be followed or replaced with a phonetic transcription (see e.g., Matsumoto 2011; O’Neal 2015a, 2015b). The same recommendation concerns mild pronunciation variations that can unambiguously be expressed by varying the standard orthography slightly. For example, quite a common variation in both L1 and L2 English is the second syllable in the word “pronunciation”, [nʌn] uttered as [naʊn]. This can unequivocally be transcribed as “pronounciation”. The suggestion of using mostly standard spelling is not meant to imply that grammatical variations such as the omittance of the third person present tense -­s identified in some ELF literature (e.g., Cogo and Dewey 2006) should be “corrected” by the transcriber – quite the contrary – but that markings of accented talk should not override comprehensibility. As an example, “he speak” uttered in Northern English accent should not be transcribed as “hey speyk” but should not be corrected to “he speaks” either. In Jefferson’s example transcript, cut-­offs and contractions have also resulted in problematic transcription decisions. These decisions include the adding of a vowel in “fer Chrise’sake” which apparently intends to prevent the misreading of a contracted form of “for Christ’s sake”. Also, in the second line, what is assumed to mean “sounds like” has been transcribed as “sou’ li’ ”, which can easily lead to

186  Kaisa S. Pietikäinen misreading, as the syllables that remain can be read as [soʊ] or [səʊ] and [lɪ] rather than [saʊ] and [laɪ]. Furthermore, “ih suh” in the second line remains a complete mystery, as does the pronunciation of these syllables. They have not been marked with hyphens as cut-­offs, so they cannot be straightforwardly interpreted as hesitation markers; neither do they fit the sentence structure, following immediately after the interjection “fer Chrise’sake”. Perhaps they represent a breathy, cut-­off start of the next turn “it so-­” (‘it so- .hh sou(nds) li(ke)’)? For disambiguation, the transcript should have offered a clarification in double brackets or on a separate line of either the pronunciation of the syllables or the words of which they are presumed to be parts. The issues discussed here have related specifically to the problematicity of impressionistic eye dialect, especially present in Jefferson’s extract that (presumably) dealt with native English speaker data. Analysts dealing with ELF data will, however, often analyse data recordings where several languages are alternated. In the next section, I focus on issues of transcribing multilingual data and accents and variation.

Transcribing multilingual talk for the English-­speaking reader Conversation analytic transcription conventions vary for different languages, and different traditions remain between different schools. It is the analyst’s burden to choose a convention that suits the purpose best. Analysts may have to decide upon questions such as: 1) will non-­Latinate languages be provided with a Latinate script or just a translation, 2) will translations be offered with grammatical information or just as a pragmatic translation, and 3) will languages that use right to left scripts such as Arabic, Hebrew and Urdu be translated line-­for-­line. Hepburn and Bolden (2017: 141) remark that the transcriber might not need to identify the different languages when they are familiar to the readers. Concerning ELF research, it would be hard to imagine a language besides English that all international readers would be able to understand, and if it is not specifically marked, readers may not correctly understand transcriptions where language alternation occurs, as will be shown in the next example. I  will not be giving recommendations on how non-­Latinate languages should be transcribed, but I will offer a suggestion concerning the marking of different languages that use Latin script in transcriptions of multilingual ELF interactions. Consider this example: (3a) 01 Thomas: 02 Laura: 03 Thomas: 04 Laura:

is (.) wild viileä is cool a:h ja: villi (.) is wild

This extract does not differentiate distinct languages or provide a translation to English. The reader may be able to gather that viileä (lines 01 and 02) and

Notes on CA transcription  187 villi (line 04) are not English words, and even perhaps that ja: (line 03) may not exactly be English. However, the topic of the discussion – the proper translation of the Finnish word viileä into Dutch  – is completely missed in this transcription. In order to help the reader understand the multilingual interaction better, in (3b), I have italicised those words that are some other language than English, provided an approximation of the pronunciation of wild in double brackets, and in a separate line, provided a translation for these words and a gloss of the language in which they are uttered. I have not, however, assumed which language ja: may be,12 nor have I provided a translation for viileä in line 02, as its translation has already occurred in the previous line. (3b) 01 Thomas: is (.) wild (([velt])) cool (Fin) wild (Dut) 02 Laura: viileä is cool 03 Thomas: a:h ja: 04 Laura: villi (.) is wild (([velt])) wild (Fin) wild (Dut) In order for the reader to be able to decode the transcript accurately, I would recommend following the common practice of highlighting other-­/mixed-­language words with italics13 if there is a clear matrix language (such as English) from which these words or syllables can adequately be differentiated, and providing a translation and a gloss of the translated-­into language in the (unnumbered) line immediately below the turn. Furthermore, if the pronunciation of any item is relevant either interactionally for the participants (i.e., it is oriented to in the interaction or it causes subsequent comprehension issues); from an analytic perspective (e.g., it is unclear as to which language is used); or from the reader’s perspective (its written-­out form risks misleading the reader or it has various possible spellings that the reader might easily misinterpret), I would also recommend providing a phonetic transcription of the item, either in the transcription (as in 3b) or in the body text of the analysis (see e.g., Brouwer 2004). Transcribing accents and variation In talk that is predominantly in lingua franca English, variation can be expected. How important is it, then, to indicate these variations from standard English? Hepburn and Bolden (2017: 53–54) propose three alternatives for transcribing non-­native talk: A) attempting to “consistently capture the accented speech” e.g., by using eye dialect, B) transcribing “broadly, for the most part only slightly modifying standard orthography”, C) transcribing accents “only when aspects of the speaker’s pronunciation become an oriented to feature of interaction”. They suggest that the first is the most suitable for analysts who focus on pronunciation. As has already been argued in this paper, eye dialect is not the ideal option for tracking pronunciation particularities. I will further illustrate its arbitrariness

188  Kaisa S. Pietikäinen with the next extract (4) which has been transcribed using eye dialect (from Pietikäinen 2018): (4)

01 Carmen: b ud i have no-­idea(h), (.) uih what time. (0.2) i know-­↑is ear↓lier 02 like you: said,= 03 Kjetil: =>yea well,< (.) urr:: (.) she wrote?

In the first turn from a Mexican English speaker Carmen, there are three words that are not spelled in standard orthography (if intonation markers, lack of capitalisation, and hyphenation are not taken into consideration): “bud”, “idea(h)”, and “uih”. The first one, pronounced [bʌd] is a common American English assimilation of ‘but’ in front of a vowel and could therefore be transcribed using standard orthography, if the recommendations from the section entitled “Pronunciation particularities and eye dialect” are followed. The “(h)” after “idea” signifies aspiration; the word is uttered in a breathy manner. This presents no problems with deciphering the transcript, but neither does it add to the analysis – the breathy voice likely results from the word stress on the word-­final vowels. The third non-­standard word “uih” is pronounced [aɪ] and could possibly have been less equivocal had it been transcribed simply as “ai”. The most interesting variation in Carmen’s turn is, however, not readable in this eye dialect transcription at all. Namely, she pronounces the word “earlier” as [ˈɪərliər]; the syllable “ear” is pronounced similarly to the body part. Of course, the word could perhaps have been transcribed in eye dialect as “ierlier”, but this transcription does not solve the problem and may instead give the impression that the second vowel is [e] rather than [ə]. Another option would perhaps have been to add the letter “i” before “earlier” (“iearlier” or “i-­earlier”), however, this may have given the impression that some of the vowels are longer than others (or pronounced as a cut-­off in the latter version). Interactionally, this difference in pronunciation has little value – the husband displays no difficulty of understanding the word14 – but analytically, if one was interested in why Carmen is here using this pronunciation while she has previously in the recording used the standard pronunciation of the same word, this distinction would be an important one to make. As can be seen here, eye dialect does not offer this possibility. Similar accent-­related features prominently present in Carmen’s English that would not be visible in eye dialect are: ‘coming’, [ˈkɒmɪŋ] rather than [ˈkʌmɪŋ], and ‘talking’, [ˈtʌlkɪŋ] rather than [ˈtɒːking]. Besides variation in vowel quality, lingua franca English often comes with variation in consonant pronunciation. Typically, the Finnish-­ and Spanish-­speaking participants in my ELF data aspirate voiceless stops /p, t, k/ less prominently than in standard English, while phones that are less frequent in Finnish /g, d, b/ are oftentimes pronounced as hard consonants by Finnish speakers, whereas Dutch speakers’ /s/ is often of a softer quality than that of standard English. These example qualities would not be visible in eye dialect in any way. For these reasons, alternative A from Hepburn and Bolden (2017; outlined earlier) is not recommended

Notes on CA transcription  189 for ELF data. Instead, if detailed accounts of accents are vital for the analysis, the IPA is recommended. Otherwise I  suggest a combination of alternative B with slight modifications detailed earlier and specific detailing of pronunciation where it becomes essential from the perspective of the participant, the analyst, or the reader. In the next section, I propose a simpler alternative for specifying pronunciation or accent details than the IPA. An alternative to eye dialect Although endorsed by some (e.g., Bucholtz 2000), the IPA takes years to master. It is also very time-­consuming, and the resulting transcripts can only be correctly interpreted by trained phoneticians, as has been argued elsewhere (e.g., Haberland and Mortensen 2016: 586). However, there are other, simpler options for describing pronunciation accurately enough for studies that have their main focus on intersubjectivity in interaction (as opposed to detailed phoneme qualities), particularly those that focus on ELF. Like already mentioned, I do not recommend using phonetic symbols for producing whole transcriptions, only for disambiguation purposes for syllables or words that are ambiguous or where the pronunciation differs drastically from the standard orthography of the word in question, or where it causes disturbances in the interaction (i.e., it can be identified as a trouble source for repair work). It is not realistic to expect that those conversation analysts whose interests are not specifically in phonetics would be willing to use the IPA system to a very detailed level, especially as CA transcriptions are already extremely time-­ consuming to produce. Yet, learning a few extra phonetic symbols to more accurately transcribe ambiguous words seems feasible. I  suggest that analysts learn the most common symbols used in dictionaries, e.g., the Longman Pronunciation Dictionary. These symbols include the most common consonant sounds in the English language (examples from Received Pronunciation) /p, t, k, d, b, f, v, s, z, m, n, r, l, w, h, g (as in get), j (as in yes)/ and a few special symbols: /ʃ/ (sheep), /tʃ/ (chair), /ʒ/ (measure) /dʒ/ (John), /θ/ (thin), /ð/ (there), and /ŋ/ (ring). The symbols for short vowels are: /ɪ/ (bin), /e/ (Ben), /æ/ (bat), /ʊ/ (put), /ʌ/ (cut), /ɒ/ (pot), and /ə/ (tiger); long vowels (elongation marked with either ː or a colon); / iː/ (keep), /uː/ (move), /ɑː/ (car), /ɔː/ (horse), /ɜː/ (burn); and diphthongs: /aɪ, eɪ, ɪə, eə, ɔɪ, əʊ, aʊ, ʊə/ (see Wells 2001 for more information). Of course, these symbols are based on pronunciation standards, and English speakers’ actual pronunciations may appear as numerous variations of these. For example, the symbol /r/ represents the postalveolar approximant (IPA: /ɹ̠ /, often spelled /ɹ/) common to British/American English rather than the alveolar trill (IPA: /r/), the trilled r, which is more common to, e.g., speakers of Spanish, Finnish, and Irish English. Also, unlike in IPA, the aspiration of voiceless stops /p, t, k/ is not noted but is expected. If these factors are/become important for the analysis, they should be mentioned separately, or the use or the full IPA system should be considered. In the phonetic transcription system detailed earlier, word stress is also indicated with a straight apostrophe preceding the syllable(s) that is stressed. This should

190  Kaisa S. Pietikäinen cause no confusion with the marking of contractions if the phonetic transcription is provided as a transcriber comment in the transcript, i.e., in double brackets and square brackets (see extract 3b, line 01). The primary function of the transcript in a publication is to provide the reader a precise account of how “that” was said in “why that now” (Schegloff and Sacks 1973). The benefit of using phonetic symbols for disambiguation in transcripts is that they are a standard way of marking pronunciation which  – even if the actual phoneme quality may be a little varied – provides the reader with an easy standard tool for disambiguating pronunciation particularities. Hence, it offers a more accurate interpretation of the original data than the eye dialect does. Such a standard is useful where the full IPA cannot, for one reason or another, be used. The next section zooms out from the particularities of pronunciation and turns its focus on the analyst. It scrutinises possible caveats in analysing multilingual data from a practical perspective and suggests a routine for analysing problematic multilingual/lingua franca data sections.

Understanding one’s data: transcribing multilingual interaction Some textbooks (e.g., Hepburn and Bolden 2017) air a general recommendation that conversation analysts should work on languages of which they are native or “near-­native” speakers “since transcription is a vital part of the analytic process” (2017: 131). In the case of ELF, however, there are no native speakers, and when participants in the interactions can, theoretically speaking, switch into any language, the recommendation becomes impossible to follow. How can an analyst overcome this issue? Analysts are often rather lonely in their line of work: most transcription work is either done by the analyst him-­/herself. Only those who have sufficient funding have the opportunity to hire an assistant transcriber or employ a PhD student to do (a part of) the transcription work. However, PhD students and assistants often lack training in CA transcription, so the principal investigator is the one who needs to make sure the transcripts adequately represent the data. S/he must also make the ultimate decision on interpreting what has been said. Often analysts feel underequipped to analyse multilingual/lingua franca data if they are not fluent in the languages present in the data. This should not, in any way, be a reason for leaving multilingual data unanalysed. Multilingual interaction is greatly underrepresented in research on language and interaction, which has led to a monolingual bias both inside the research community as well as among laypeople. For this reason alone, we should attempt to seek solutions for including multilingual data in our work. In this section, I will describe some common difficulties that have to do with the limitations of the lonely analyst and suggest some practices to overcome these problems. I will start with an example. In 2018, I collected new interactional data from ELF families. These are families where the parents speak English as their main shared language but where neither parent is a L1 English speaker (Pietikäinen 2017). One

Notes on CA transcription  191 of the families had recorded discussions featuring English, Norwegian, and Mandarin Chinese, and as I do not understand Mandarin, I hired a student assistant to transcribe the Chinese sections into both Chinese characters and Pinyin and to translate these turns into English.15 My assistant did her work independently, and then we compared our transcripts. Interestingly, in the transcript of the first recording, my assistant’s transcription (6)16 differed quite drastically from my own (5): (5) Boy: o↑kay (.) ↑stop ma: (6) Boy: 我可以 stop 吗 wo ke yi stop ma Can I stop it? (7) Mother: 呃不可以 E Bu ke yi No, cannot. In the recording, the boy was examining the voice recorder with wonder. I had interpreted the trilingual boy’s utterance as telling his mother to stop the recording (5), but my assistant insisted that the boy was speaking (mainly) Mandarin and asking his mother whether he could stop the recorder (6). Considering the mother’s next turn (7) which denies this request and the fact that the boy seems to be closer to the recorder than his mother (his voice is louder), I decided to follow my assistant’s interpretation. What I want to convey with this example is that transcriptions are always subjective: what one hears as Yanny, is heard as Laurel by someone else.17 With multilingual data, the risk of confusion is even higher, because the possibilities for similar-­sounding combinations are increased. I therefore recommend analysts to use the next-­turn proof procedure as was used earlier: implications of earlier turns can commonly be concluded from how the interlocutors respond to the earlier turn.18 If the interactants are available and willing, the analyst can also ask for their interpretation or a basic transcription of uncertain sections. However, when using this method, a degree of caution should be maintained. Non-­linguists are often not trained to listen to their own voices and may find it difficult to trace exactly what was said and how. I therefore recommend the following four-­step protocol for unravelling unclear sections in multilingual data: 1

2

First listening to the section several times and making a “mock transcript” of all the syllables uttered and how they sound. A phonetic transcription may be helpful at this stage, although it is not always possible if the speech is very fast, covered by other noises, etc. Asking the speaker to produce a transcript of the section, emphasising that they should write each word exactly as heard, without omitting anything.

192  Kaisa S. Pietikäinen

3

4

A translation of the section is also needed. Showing the analyst’s mock transcriptions to the speakers is not recommended in order to avoid imposing one’s own interpretations on the speakers. Checking the speaker-­provided transcript against the mock transcript while listening to the original recording. Google Translate or other translation software that provides audio examples of pronunciations can also be helpful with deciphering word boundaries, pronunciation and meaning. If any syllables are left untranscribed or uncertain, the speaker should be consulted concerning these again. The help of a colleague who understands the language in question can also be solicited when specific linguistic questions are at hand, or when pragmatic meanings are unclear.

Using this methodology in my own work has resulted in interesting findings that I would likely have missed otherwise. For example, in transcribing new recordings from a Finnish-­German ELF couple, I noticed some words that I could not figure out on my own; see the bolded sections in (8) and (9): (8)

(9)

01 Henrik: an-­this ↑one is (.) lighte:r, (4.4) mm 02 Minna: yeah it’s much higher quality, (.) it was on (tarch-­zu) that (.) i bought it, 01 Henrik: 02 Minna: 03 Baby: 04 05 Minna: 06 Henrik:

SCHAU M[AL] [huo]maatsä

[mitä laa]tikossa on [hu:h  ]

(2.3) oo::↑oo::::::h (1.9) (tesh-­to::) dein eigenes (.) (teht-­t’) (.) telefo:n.

In (8), line 02, Minna is talking about ingredients for a cake, saying that she bought “it” on what sounded like tarchzu [ˈtɑʁʒu]. When I asked her about this word, she told me that she and her husband were talking about cocoa powder and using wordplay of the Finnish word tarjous (offer). There is quite an interesting story of translanguaging behind this specific word: After the Hungarian/German husband had moved to Finland, he had seen the word tarjous and thought it looked like French, wondering why Finnish shops would use this word in their windows. When he asked his wife about the word, she thought his French pronunciation of it sounded so hilarious that it became an insider joke between them. So, now they use the French pronunciation of the Finnish word when talking about products on offer; it has become a part of their “couple tongue” (Pietikäinen 2017). In (9) the same couple are opening a children’s toy phone package in front of their baby. In line 01, the husband exclaims “Look!” in German, while in line 02 the wife asks the child “do you notice what is in the box?” in Finnish. However, what happens in lines 05 and 06 is interesting. They both utter a strange word “tesh-­to::” /[ˈteʃˈtɔː] and “teht-­t’ ” [ˈteħt t] (“your own teht-­t’ telephone”, line 06), and even the speakers themselves could not recall what they said or what the word

Notes on CA transcription  193 might mean. The wife reckoned that it might have been something written on the box. I did an online search for the kind of a toy phone that they were discussing and found a picture of the product package. It says “Tech-­TOO” on the cover. Hence, the strange words in the recording were concluded to be the couple’s interpretations of what was written on the box. Although an analyst can never understand entirely fully what participants’ conversations are about – as we do not have access to their world and their thoughts – unclear sections and words can at times result in interesting findings if more effort is put into the transcription phase of the analysis especially regarding unclear words or utterances. Soliciting help from the speakers themselves may be helpful if it is in any way possible, and if the analyst is not fluent in the language(s) present in the data, naturally the help of a colleague or an assistant can be useful.

Concluding recommendations Although conversation analytic transcription conventions were initially developed based on native speaker data, the conventions have developed along with the recent increase of research on multilingual interaction (see e.g., Hepburn and Bolden 2017). However, consistent recommendations for transcribing ELF interaction for CA-­informed enquiry have hitherto not been made. This chapter has discussed conversation analytic transcription of lingua franca/multilingual interaction. It has argued that eye dialect is not ideal for disambiguating pronunciation particularities of lingua franca English. Instead, it proposes using only slightly modified standard orthography and common colloquial contractions (gonna, cos, etc.) as well as apostrophes for other contractions. Both weak and strong forms of words are recommended to be spelled in standard way, as stress marking (underlining) should give enough information to the reader on whether a word is uttered in its strong or weak form. If, however, the analyst notes that the speaker’s pronunciation influences the proceeding interaction, phonetic spelling of the trouble source can be used for clarification purposes. The same applies for analytically interesting words (see, e.g., the “earlier” example in (4)), or if there is a danger of confusing the reader. As mastering the whole range of the IPA system requires specialist knowledge in phonetics, it is here suggested that analysts could use a simplified system. The recommendations detailed earlier do not, however, necessarily apply to other languages than English. In languages where orthography follows pronunciation more closely than in English (such as Finnish, Spanish), eye dialect is less ambiguous than in English, and therefore less problematic. It is important to always perform the analysis from the original recording rather than from the transcript alone. However, transcripts are important notes of how, when and what has been said, and they can be used both for particularising insights as well as for gaining an overview of a larger amount of data. In publications, the primary aim for transcripts is to present the data to the reader and therefore their comprehensibility should be considered important. The present chapter proposed a method for detailing sections of multilingual data that are otherwise difficult for the analyst to understand. Although the common recommendation is that analysts

194  Kaisa S. Pietikäinen should know the languages they work with, the translingual nature of ELF does not always allow this, which is why alternative methods need to be considered. This chapter has attempted to contribute to the ongoing discussion concerning a good CA transcription practice particularly from the perspective of multilingual/lingua franca interaction. It does not attempt to be an exhaustive account of how transcription work of ELF materials should be conducted. Rather, it attempts to find solutions for common problems in the transcription phase and outline choices that analysts will have to make in this process. It is my hope that in forthcoming ELF-­CA research, analysts would consider the three perspectives of the researched, the researcher, and the reader, and elucidate the important methodological decisions they make in the process of transcription to a greater detail.

Notes 1 Acknowledgement: I would like to extend my gratitude to Kumiko Murata, Daisuke Kimura, and an anonymous reviewer, for providing their insightful feedback on an earlier version of this paper. This work was supported by the Alfred Kordelin Foundation. 2 I use the term multilingual in this paper to highlight the multitude of possible mediums of interaction the interlocutors may choose from – e.g., monolingual medium, bilingual medium, translingual medium (the “halfway-­between mode”; see Gafaranga and Torras 2001) – and the possibility of multiple languages speakers can draw from and refer to in lingua franca interaction. 3 In doing so, Preston (2000) actually uses the same argument as conversation analysts have previously used against IPA. In the 1974 version of the simplest systematics for turn-­taking paper, Sacks et al. (1974: 734) plead the following: Some linguists have objected to our use of modified English spelling – rather than, say, IPA symbols: the result, they claim, resembles a sort of funnypaper-­English, and could have derogatory connotations. Our reply is that we have simply tried to get as much of the actual sound as possible into our transcripts, while still making them accessible to linguistically unsophisticated readers; we certainly mean no disrespect to the parties cited. 4 With ‘phoneme’ I simply refer to a unit of sound that creates a distinguishable contrast with other sounds that interactants and analysts are able to detect. With this definition, I wish to resign from any juxtapositioning of native and non-­native accents that the term might otherwise suggest. 5 With ‘first-­order entextualisation’, Haberland and Mortensen (2016) refer to the original data (audio or video recordings). 6 I am here mostly referring to transcriptions of audio recordings, see e.g., ten Have (2007) and Jenks (2011). For video recordings, further notation requirements prevail, see e.g., Mondada (2014). 7 If possible, the original recordings should be made available for the readers, although often ethical considerations prevent this practice. 8 And possibly even a visualisation of intonation, see Hepburn and Bolden (2017: 155–158). 9 It is unclear what “ampitious” is supposed to signify, except that Jim rephrases it approximately as “sounds like you could go to sleep on the phone”. Maggie orients to it as a description of her “huffing and puffing” voice quality. 10 Such as in other three-­syllabic words ‘ambitious’, ‘factitious’, etc. 11 I acknowledge the problematicity of pronunciation standards especially in relation to lingua franca data, but since the readers of academic publications will rarely have

Notes on CA transcription  195

12

13 14 15 16 17 18

access to original recordings, they will need some tools that they can use as points of reference through which they can imagine a rough representation of the analysed interaction. Ja may be a Dutch affirmative, but Thomas’s turn in line may also be a version of Finnish ai jaa which is a change-­of-­state token that treats the previous turn as new information (Koivisto 2016). As the meaning of the turn remains roughly the same regardless of the switched-­into language, it is here not relevant (or possible) to label the language. Hepburn and Bolden (2017: 141) recommend italics or bold. They also use curly brackets in an example extract. Rather, the clarification question in line 03 indicates his uncertainty of whether the earlier arrival of a ferry is expressed by the wife herself or in the friend’s message that she had just been reading to him. For transcribing Mandarin Chinese, see Kobin H. Kendrick’s advice in Hepburn and Bolden (2017: 134–135). Unfortunately, these transcripts lack tone markers in Pinyin. See www.theguardian.com/technology/2018/may/16/yanny-­or-­laurel-­sound-­illusion­sets-­off-­ear-­splitting-­arguments. However, this is not entirely secure method either; apparently, in Chinese, the subject can be omitted from a clause, which is why the mother’s turn can be interpreted both as “No you can’t” and “No I can’t”.

References Brett, D. 2009. Eye dialect: Translating the untranslatable. In S. Cocco, M. Dell’Utri, S. Falchi and S.L. White (eds), Lost in translation: Testi e culture allo specchio. AnnalSS 6, 49–62. Available from: https://core.ac.uk/download/pdf/11689734.pdf [accessed 10 February 2019]. Brouwer, C.E. 2004. Doing pronunciation: A specific type of repair sequence. In R. Gardner and J. Wagner (eds), Second language conversations. London: Continuum, pp. 93–113. Bucholtz, M. 2000. The politics of transcription. Journal of Pragmatics 32(10), 1439–1465. Cogo, A. and M. Dewey. 2006. Efficiency in ELF communication: From pragmatic moves to lexico-­grammatical innovation. Nordic Journal of English Studies 5(2), 59–94. Firth, A. 2012. Conversation analysis and Lingua Franca. In C.A. Chapelle (ed), The encyclopedia of applied linguistics. Oxford: Wiley Blackwell, pp. 1043–1047. Gafaranga, J. and M.-­C. Torras. 2001. Language versus medium in the study of bilingual conversation. The International Journal of Biligualism 5(2), 195–219. Gumperz, J.J. and N.B. Berenz. 1993. Transcribing conversational exchanges. In J.A. Edwards and M.D. Lampert (eds), Talking data: Transcription and coding in discourse research. Hillsdale, NJ: Lawrence Erlbaum, pp. 91–121. Haberland, H. and J. Mortensen. 2016. Transcription as second-­order extentualization: The challenge of heteroglossia. In A. Capone and J.L. Mey (eds), Interdisciplinary studies in pragmatics, culture and society. Heidelberg: Springer, pp. 581–600. Hepburn, A. and G.B. Bolden. 2017. Transcribing for social research. London: Sage Publications. Jefferson, G. 1983. Issues in transcription of naturally-­occurring talk: Caricature versus capturing pronunciational particulars. Tilburg Papers in Language and Literature. 820.08.203. Tilburg University, pp. 1–12. Jefferson, G. 1996. A case of transcriptional stereotyping. Journal of Pragmatics 26, 159–170. Jefferson, G. 2004. Glossary of transcript symbols with an introduction. In G.H. Lerner (ed), Conversation analysis: Studies from the first generation. Amsterdam: John Benjamins, pp. 13–31.

196  Kaisa S. Pietikäinen Jenks, C.J. 2011. Transcribing talk and interaction. Amsterdam: John Benjamins. Kaur, J. 2016. Conversation analysis and ELF. In M.-­L. Pitzl and R. Osimk-­Teasdale (eds), English as a Lingua Franca: Perspectives and prospects. Berlin: de Gruyter Mouton, pp. 161–168. Koivisto, A. 2016. Receipting information as newsworthy vs. responding to redirection: Finnish news particles aijaa and aha(a). Journal of Pragmatics 104, 163–179. Kowal, S. and D.C. O’Connell. 2014. Transcription as a crucial step of data analysis. In U. Flick (ed), The Sage handbook of qualitative data analysis. Los Angeles: Sage Publications, pp. 64–78. Krapp, G.P. 1925. The English language in America (vol. 1). New York: Century. Matsumoto, Y. 2011. Successful ELF communications and implications for ELT: Sequential analysis of ELF pronunciation negotiation strategies. The Modern Language Journal 95, 97–114. McArthur, T. 2003. Eye dialect. In T. McArthur (ed), Concise Oxford companion to the English language (1st ed.). Online. Oxford: Oxford University Press. Available from: www.oxfordreference.com/view/10.1093/acref/9780192800619.001.0001/acref­9780192800619 [accessed 7 February 2019]. Mondada, L. 2014. Conventions for multimodal transcription. Available from: https:// mainly.sciencesconf.org/conference/mainly/pages/Mondada2013_conv_multimodality_copie.pdf [accessed 7 February 2019]. Ochs, E. 1979. Transcription as theory. In E. Ochs and B. Schieffelin (eds), Developmental pragmatics. New York: Academic Press, pp. 43–72. O’Neal, G. 2015a. ELF intelligibility: The vowel quality factor. Journal of English as a Lingua Franca 4(2), 347–358. O’Neal, G. 2015b. Segmental repair and interactional intelligibility: The relationship between consonant deletion, consonant insertion, and pronunciation intelligibility in English as a Lingua Franca in Japan. Journal of Pragmatics 85, 122–134. Pietikäinen, K.S. 2017. English as a Lingua Franca in intercultural relationships: Interaction, identity, and multilingual practices of ELF couples. Doctoral dissertation, University of Helsinki. Pietikäinen, K.S. 2018. The intercultural, translingual and multicontextual ELF: Implications for CA and potential empirical directions. Colloquium presentation at ELF11 – The 11th International Conference of English as a Lingua Franca. Kings College London, 4–7 July. Preston, D.R. 1985. The Li’l Abner syndrome: Written representations of speech. American Speech 60, 328–336. Preston, D.R. 2000. Mowr and mord bayud spellin’: Confessions of a sociolinguist. Journal of Sociolinguistics 4(4), 614–621. Sacks, H., E.A. Schegloff and G. Jefferson. 1974. A simplest systematics for the organization of turn-­taking for conversation. Language 50, 696–735. Schegloff, E.A. and H. Sacks. 1973. Opening up closings. Semiotica 8, 289–327. ten Have, P. 2007. Doing conversation analysis: A practical guide (2nd ed.). London: Sage Publications. Wagner, J. and R. Gardner. 2004. Introduction. In R. Gardner and J. Wagner (eds), Second language conversations. London: Continuum, pp. 1–17. Walpole, J.R. 1974. Eye dialect in fictional dialogue. College Composition and Communication 25(2), 191–196. Wells, J. 2001. IPA transcription systems for English. Available from: www.phon.ucl. ac.uk/home/wells/ipa-­english-­uni.htm [accessed 25 February 2019].

12 Accounting for asymmetries in ELF interactions Three disparate Asian settings Alan Thompson

Introduction Every day, people from differing cultural and linguistic backgrounds use English in interaction in a multitude (possibly billions) of settings  – some established over long terms, some cropping up only for a moment. The situational diversity “in terms of the constellation of users, the sociocultural background of speakers, their sociolinguistic repertoires, and their migration patterns” (Cogo 2012: 289) and speaker differences in “proficiency,  .  .  . motivation, and purpose” (House 2012: 173) have resulted in English language practices of great variation and unpredictability. This diversity is present on several levels. First, there is diversity in the backgrounds and repertoires of the individual participants. Compared to the assumptions inherent in much sociolinguistic terminology (“community”, “minority”, “language”, etc.), modern social reality is characterised by “superdiversity” (Vertovec 2007), where many individuals (and perhaps most participants in English as a lingua franca (ELF)) are decidedly mobile and interconnected. No two trajectories of experience (of social settings, of languages learned, or of education) are alike, which means that researchers must attend to the individual biographical dimensions of dialect and repertoire (Blommaert and Backus 2013). Second, each interaction group is made up of individuals with different ‘diversities’, as it were. As mobility has increased and interaction groups have become more and more transient (Hülmbauer 2009; Pitzl 2018; Thompson 2017), researchers observe more marked differences among individuals’ attitudes to communication (Darvin 2017) and their access to participant roles (Mežek 2018; Thompson 2016), in addition to the differences in linguistic and pragmatic repertoires. Third, there is diversity not only within but also across interaction groups. No two groups are alike, not only due to differences in how they are made up of individuals, but in terms of the purposes, goals, and sociolinguistic realities of the institutions and regions in which they are embedded (Björkman 2016, 2018; Kaur 2018). The first level of diversity is problem (and opportunity) for the individual, who must marshal his or her various social and linguistic repertoires and competencies so as to meet the needs of the present communication situation. The second level of diversity (of differences within the group) is a problem for groups, who must

198  Alan Thompson discern with what their interlocutors are familiar, find commonalities in repertoires, and co-­construct a makeshift set of norms to deal with the gaps. The third level of diversity (of differences across groups) is a problem for the researcher of ELF, who must be careful not to assume that characteristics or processes of communication manifested in one ELF interaction or communication situation will necessarily be present in others. Given this lack of uniformity, it is no surprise, then, that ELF interactions are said to be “characterised by continuously negotiated, hybrid ways of speaking” (Seidlhofer 2011: 4), and “constantly brought into being in each context of communication” (Canagarajah 2007: 91). ELF research (which was focused at first on relatively highly educated and uniform school and business settings in Europe) is perhaps still just beginning to deal with the stark contrasts in worldwide ELF situations – in socioeconomic level, education, profession, and exposure to international media. We might expect, then, with continuing investigation of the disparate situations that exist around the world and across social strata, further discoveries of widely diverging kinds of hybrid practices. And, given the unlikelihood of finding that participants in most situations are equally equipped and empowered, or that institutions and purposes are disinterested and unbiased, we should expect to find not only diversity, but also asymmetry, to be a core feature. Even before careful analysis of the situations and language practices in these settings, this casual observation of diversity, in and of itself, gives rise to two important sets of questions regarding how we theorise and delimit the ELF field of study. 1

2

How do diversities and asymmetries (within individual and within group) manifest themselves in ELF situations? Are isolated peculiar features of certain communication situations of interest when the objective is to say something meaningful about ELF communication in general? How do participants in ELF situations use strategies to deal with diversities and asymmetries in the interaction or communication situation? Can the effective use of these strategies be a measure by which performance or competence in ELF is judged? Can we say that there are “good” and “bad” ELF practices? Is there a minimal level of communicative competence in English, or a minimum threshold of communicative success, that is required in order for the interaction to be considered an instance of ELF?

The first set of these questions can only be answered partially by any one study, and some tentative answers will be suggested by the data that are considered later in this chapter. Reasonably complete answers, however, might have to wait until the phenomenon of ELF is more surely situated among other inquiries into language in social context and understood as one facet of the long story of language contact and change, particularly with regard the processes that work upon and within languages of wider communication (see Thompson 2017). There have been many studies in the last two decades that touch on the second set of questions, regarding participants’ ELF strategies. The data to be presented

Asymmetries in ELF interactions  199 in this chapter offer some confirmation of and additions to other researchers’ findings and interpretations. Here, I will briefly review some of the contextual factors that have been found to influence communication behaviours in ELF situations, with the behaviours interpreted as purposeful or strategic. Perhaps the basic source of variation and asymmetry lies in the differences in experience and proficiency with respect both to the working language (English) and to the practice of intercultural communication. Meierkord (2002) showed how participants who saw themselves as less proficient, in a style she called “language stripped bare”, used laughter and pauses strategically to maintain their participation in interactions, while those who perceived themselves as more proficient also used similar strategies to avoid embarrassing their interlocutors. Another common source of variation and asymmetry in most social interactions (especially workplace communication and service encounters) is differences in status and in institutional role or membership category. Mežek (2018), for example, reports from a study of high-­stakes academic interactions (PhD oral defences), where, in the context of different roles and status levels, some speech acts are perceived as potential face threats (examiners need to be critical but not seem harsh; candidate needs to show expertise but not seem immodest). The most prominent behaviour response of the participants was laughter, which was explained as a strategy to build a non-­adversarial atmosphere in order to mitigate these threats while continuing to perform the potentially threatening acts. Pullin (2013) shows a contrasting kind of situation, of business meetings where the aim is to preserve (fictionally) equal roles and status despite apparent differences in expertise and competence. Thus, the speech acts of giving clear directives and of correcting others’ misunderstandings were accomplished with hedging behaviours (perhaps you know, sometimes, well we just, I think . . . I don’t make myself quite clear, etc.). Asymmetry also arises out of a mismatch between stated goals and social norms or expectations. Toomaneejinda and Harding (2018) show how, in ELF academic group discussion, there is an understood need to disagree, but that this competes with a social preference to avoid explicit displays of oppositional stance, often because of the participant’s knowledge of a mismatch in cultural norms, or a feeling that it is “not their place” to comment on these norms (i.e., in addition to the mismatch between task and norms, there is also a kind of cultural asymmetry, as it emerges out of the participants’ cultural memberships.). They identify three salient complex linguistic and interactional strategies: partial agreements with a shift in focus, indirect disagreement through complex turn-­management (e.g., waiting for others to initiate), and the use of gaze (e.g., rolling eyes and direct gaze as an invitation to another to intervene, because “it would [not] be acceptable to explicitly contradict” [2018: 325]). These findings show how asymmetries in contextual factors (proficiency, experience, status, role, goals) may contribute to the adoption of specific strategies and language practices by participants in communicative events, or interactions. Although these factors are present to some degree in all interactions, they may be compounded in ELF interactions by the variation in proficiency and experience among participants, as participants’ ways of accommodating to these sorts of

200  Alan Thompson contextual factors have likely been acquired in widely differing settings, according to their cultural and linguistic backgrounds. It is these accommodations – ELF participants’ responses to asymmetric contextual factors by linguistic and paralinguistic means – that are the object of study in this chapter. In the next section, I will briefly describe the three widely distinct ELF communication settings investigated, the data that were collected, and methods of analysis. In the following section (“Language practices”), I will describe the features of the interactions in their contexts – that is, the description will consist of an identification of some of the salient language practices and behaviours of the participants and the contextual factors of the interactions. This will be followed by a discussion of the impact that diversities and asymmetries in ELF situations have on the strategies and language practices of users as they work towards successful communication.

The settings, data, and methods Shared characteristics of the settings Obviously, the settings considered in this chapter make up only an extremely small sample from the larger set of global and diverse instances of ELF. (Although this is true of all field studies of ELF communication, it needs to be explicitly admitted here because inherent in the intent of this chapter is to offer a picture of diversity and asymmetry in ELF; in a paper as short as this, though, such a picture cannot be drawn but only suggested.) The initially recognised commonalities among the settings are 1) that the settings were located in Japan, 2) that a large proportion of the participants were from countries in south and east Asia, 3) that the interaction groups were established and recurring (over a period of weeks to months) but not long-­term, 4) that English was the language most often used, and 5) that there was (within each setting) a wide variety of linguistic and educational backgrounds. There are also, most likely, other commonalities unrecognised by the researcher. The characteristics that were not shared among the settings cannot be listed exhaustively, but important differences (and correlations of those differences with features of the interactions) will be noted in the description of language practices in the next section. Types of data collected The data consist of audio-­recordings, ethnographic notes, and interviews, gathered at the three settings (comprising English language users meeting in an institutional setting, and described in more detail later in this chapter and in Table 12.1) which will be labelled in this chapter as follows: 1 2 3

ICDC: post-­graduate interns at an inter-­governmental development agency UpLink: graduate and undergraduate students in an English-­medium Business Administration programme CommCare: healthcare professionals in an NGO-­operated training programme

Asymmetries in ELF interactions  201 Table 12.1  The three settings investigated, event types, and nationalities of participants ICDC Event types Nationalities

UpLink Event types Nationalities

10 post-­graduate interns at inter-­governmental development agency Meetings, work sessions of 2 and 3 participants East Eurasia: Japan (3), China (2), Philippines (1), Indonesia (1), Thailand (1), Myanmar (1) Central Eurasia: Jordan (1) 61 graduate and undergraduate students in English-­medium Business Administration programmes Discussions, presentations, and question-­answer sessions East Eurasia: Japan (7), Hong Kong (10), China (1), Malaysia (2) Central Eurasia: Kazakhstan (3), Saudi Arabia (1) West Eurasia: France (17), Germany (4), Portugal (3), Sweden (2), Norway (2), Finland (1), Lithuania (1) Africa: Senegal (1) North America: Mexico (3), Canada (2)

CommCare Event types Nationalities

South America: Colombia (1), Venezuela (1) 15 healthcare professionals at NGO training programme Discussions, role-­plays East Eurasia: Cambodia (1), Indonesia (1), Japan (3), Philippines (1), Thailand (2), Timor-­Leste Central Eurasia: Bangladesh (2), Nepal (2), Pakistan (1), Sri Lanka (1) North America: USA (1)

These interactions were naturally occurring; that is, they would have occurred with or without the interest of the researcher. At some interactions the researcher was present and was able to make ethnographic observations of the physical layout and the activities of participants as the interaction progressed, and these were used to annotate the transcriptions of the audio recordings. Also, after each recorded interaction, the researcher attempted to hold an interview with the participants within the next few days, in order to hear the participants’ accounts of 1) on the communicative success of the interaction(s), 2) on features of the situation that facilitated or impeded communication, and 3) on features of language practice (their own or others’) that stood out to them. These interviews were achieved with

202  Alan Thompson at least one of the participants for all recordings, so that partial reconstructions could be made of the non-­audio elements of the interactions. The first setting (ICDC) was the regional office of an inter-­governmental development agency, located in Japan, and staffed by research interns and administrative staff who were either Japanese or nationals of other countries in Asia (the office’s jurisdiction). The stated working language of the office was English, but since many of the research interns had completed post-­graduate degrees in Japan, Japanese was available as an option and often used. The interns were mostly on short terms of 2  years, so that the workplace culture could be described as medium-­term, one where situation-­specific practices and understandings could be expected to have developed. Audio recordings were made at meetings (by the researcher) and during one-­on-­one and small group conversations (by the participants, researcher not present). The second setting (UpLink) was an English-­medium business studies programme (MBA programme and pre-­MBA undergraduate programme) at a Japanese university. Most of the students were on one-­ or two-­semester exchanges from a home university, and a few were enrolled in a 2-­year MBA programme. Consequently, there was great variance in length of stay among the 61 students participating in the study, so that this setting could not necessarily be characterised as one in which medium-­term group interaction gives rise to shared understandings and practices. Data were collected from assigned discussions and presentations (including question-­answer sessions). The third setting (CommCare) was a 5-­week training programme for Asian community health managers held at a non-­governmental organisation in Japan. The programme consisted of morning and afternoon sessions where participants shared their experience and expertise, and were led by programme facilitators in discussions and role-­plays which addressed common problems in the organisation of community health care in developing countries. The participants were lodged in a dormitory on-­site so that their interactions with each other extended to cooperation in daily chores and the preparation of meals. The intensive contact meant that the setting could be characterised as an intact speech community, and as such it was reasonable to expect that some shared understandings and conventional practices (including language practices) might develop, in spite of the short-­lived nature of their interaction. The working language of the training programme was English, and while Japanese and other languages were occasionally used among compatriots or others with experience in the languages, the group interactions were mainly in English. The audio-­recordings were of a variety of interactions including discussions and role-­plays (including the preparation of role-­plays). Methods of analysis The interactions, once transcribed, were coded to mark the “interactional social order” (that is, the patterning of action and utterance turns, or the turn pairs and sequences) (ten Have 2007) constructed or maintained by the participants. The interviews were coded to mark contextual factors that the participants asserted or

Asymmetries in ELF interactions  203 to which they alluded, and linguistic features (in their own or another’s language practice) they noticed. The next task was the identification of salient turn structures and linguistic features, which was done according to the perceived norms of researcher and participants.1 Frequently occurring features or turn structures were further analysed functionally, according to systemic-­functional grammar categories of Halliday (1985), in order to elucidate how these features might relate to contextual factors mentioned by participants. In addition, in one of the settings (UpLink) some quantitative analysis of the patterning of frequently occurring modality markers (think, like, kind of, guess, etc.) was conducted after it was noticed that these were concentrated in the speech of certain participants.

Language practices The object of description is features in the language practice of participants in the three ELF settings, the most prominent of which are listed in this section. The list is necessarily a biased identification (or naming) of features, as it is those features that the researcher or participants noticed. However, the intent is to describe the features – without favour to any theoretical orientation, and without any valuation of the features – simply as deviations (relative to what is perceived as normal) in grammatical or discoursal patterns. Participants themselves, however, did often use value-­laden terms, distinguishing between 1) practices that they considered consistent with ‘standard native-­speaker usage’, 2) purposeful adjustments made to enhance or simplify communication, and 3) non-­standard usages, which they often called “mistakes”.2 Informed partly by participant accounts of the interactions and partly by contextual factors that were identified in other studies, I will attempt to explain the main features as strategic and purposeful adjustments when such interpretations are plausible. Action turns, pause turns, and laughter turns Two of the settings (ICDC and CommCare) were characterised by a sparse interaction order, showing several similarities to the “language stripped bare” characterisation of ELF (Meierkord 2002). Excerpt 1 (ICDC) is an example of an interaction that contained very few turns that could be coded as responses or acknowledgements.

Excerpt 1 (ICDC): An example of an action turn For each excerpt, the following information is given in a table above the transcription. Initial P R L

M/F male male female

Age 50–64 25–34 25–34

Nationality Indonesia Thailand China

Main language Indonesian Thai Mandarin

Role supervisor researcher researcher

204  Alan Thompson 1 2 3 4 5 6 7→ 8

P:

so Rong-­san, you may have another activity to, er, to manage under this MLA activity (.) which you, let us say, missed last year. yes. and I think (.) I will (.) I will ask again Li-­san to (.) to take care of this, China mission project. (4) ((taking notes)) ok’ now the next is about staff assignments’

R: P: L: P:

Here, P is explaining to R and L their responsibilities in overseeing international research projects. R responds to his supervisor’s (P’s) indirect command with a simple affirmative (line 4), while the only indication L gives that she has heard and acknowledged the request is to continue taking notes (line 7), a behaviour that P clearly interprets as acknowledgement, i.e., an action turn, as he continues to another topic in line 8. Excerpt 2 (ICDC), in which supervisor P asks about the researchers’ individual projects, shows a pause used as a topic boundary marking turn, and laughter as a turn that completes an adjacency pair.

Excerpt 2 (ICDC): An example of pause & laughter turns P M 1 2 3 4→ 5→ 6 7

male female P: SX: P: SS:: P: M:

50–64 25–34

Indonesia Japan

Indonesian Japanese

supervisor researcher

ok so (.) I think (.) then the only thing remaining works (.) is this (qqq) workshop. yeah. yeah, that’s (2.5) what about your individual projects. @@@ is it going well (.) uh’ Mami-­san’ no.

In line 4 there is a 2.5-­second pause, after which the speaker does not complete the thought but instead relies on the long pause (without other initiating turns) to signify that no more comment need be made and then moves on to another topic, i.e., the researchers’ individual projects. Although the transcription shows the utterances before and after the pause on the same line (by convention, as there is no intervening utterance), the line is probably best understood as three turns – the unfinished turn (yeah, that’s); the topic boundary turn; and the question turn (what about your individual projects.). This excerpt also illustrates how laughter may serve specific conversation management functions. At first glance, the role of laughter seems similar to that it had in Meierkord’s (2002) data, that of replacing the several forms of conventional backchanneling. Certainly, the laughter in line 5 might initially appear as a

Asymmetries in ELF interactions  205 backchannel showing nervousness at the prospect of having to answer the question. Yet, in the context of the interaction, it is also possible to treat the laughter in line 5 as a full-­fledged turn: its primary function is not to indicate an attitude (attention, nervousness) to the question in the preceding turn, but rather, to act as an attempt at a complete response to the question, an attempt which is seen to fail when, after the question is reformulated (line 6), a direct answer of “no” (line 7) is deemed unavoidable. As such, the laughter turn forms one half of a common adjacency pair (question-­answer). Excerpt 3 (ICDC) shows a laughter turn carrying a similar function. The supervisor P is expressing his opinion that a certain staff assignment is senseless, an opinion which his audience appears to share.

Excerpt 3 (ICDC): An example of a laughter turn P 1 2 3→ 4 5

male P: SS: P:

50–64

Indonesia

Indonesian

supervisor

so, how can you assign her (.) related with the human rights issue. @@@ so, please give me some suggestions (.) hhh how to deal with that.

The first turn in lines 1 and 2, given the shared knowledge of the interactants (i.e., that their working group does not research human rights issues, while a new member (“her”) has a specialisation in human rights), can be interpreted as a rhetorical question, i.e., a statement to the effect of “we are not able to assign her to a human rights issue.” In line 3, the other members of the meeting display their agreement with laughter, which, as in Excerpt 2, forms one half of an adjacency pair (rhetorical question-­ confirmation). That P understands that they have agreed is shown by how he begins his next turn, with a conjunction that presumes agreement on the antecedent (so, . . .). There are two complementary possible motivations (each complementing the other) for using laughter (see Excerpts 2 & 3), actions (Excerpt 1), and pauses (Excerpt 2) as meaningful conversation turns. First, although the communities are transient (lasting weeks to months) and the ranks, institutional roles, and membership categories of participants were not clearly defined, participants reported that they were aware that other members of the group had much higher or lower status levels “in their home countries”. Thus, it is likely that they thought more “appropriate” responses were called for, but in the absence of a specific set of interactional norms, only the minimal responses were available. Also, the laughter turns may have been motivated by a need to avoid threats to face, as in Mežek’s (2018) study. Second, even if some common set of international norms might apply, because of the participants’ variable experience in international situations, the “appropriate” expressions may not be equally retrievable to all participants. If participants were aware of these gaps and asymmetries, they may have been

206  Alan Thompson willing to let the sparse interaction order stand and simply assume intentions of comity on the part of the others. Simplification and repetition (accommodation) The discourse pattern of certain speakers was often composed of short clauses with regularly falling intonation (shown in the transcript by a period or full stop); the clauses were sometimes repetitions or variations of previous clauses, either what the speaker him/herself had just said or what his/her interlocutor had said. Excerpt 4, where one trainee at a community health workshop (S3) takes a leadership role in getting the group organised, is an example.

Excerpt 4 (CommCare): An example of repeated short clauses S2 S3 S4 S6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

female male male male S3: S2: S3: SX-­f: S6: S3: S6: S3: S4: S2: S3: S6: S2: S3: S6: SX-­f: S3: S6: S3:

25–34 50–64 25–34 35–49

Cambodia Pakistan Thailand Philippines

Khmer Urdu Thai Filipino

trainee trainee trainee trainee

are you CLEAR about this issue (.) I think er (.) about this (.) what we want to do. (3) advocacy xxx oh sorry. it’s not there. it’s not there. it’s not there. yeah. (.) this is our plan. (.) ok? we haven’t DONE this. oh we haVEN’T. no we have (.) not done this. this is our plan. NOT done? plan. we will (.) we want to do this. ah:: ok. ah:: ok? so I thought we have done this so that (.) ok. no. no. no. this is this this (.) we want to do it. ok. this is what (.) our objective. this is our task.

S3 simplifies his contributions by the use of three devices – simple clause structure, pauses, and repetition – seen in lines 1–2, 8–9, 11, 14, 20–21, and 23. Although his main interlocutor, S6, is also proficient and uses complex clauses (e.g., line 18) in his

Asymmetries in ELF interactions  207 contributions, S3 is consistent throughout the excerpt in using single clauses separated by pauses, in using simple clauses (except for the WHAT + clause construction in line 2), in avoiding complex noun phrases (by replacing them with deictic words and pronouns), and in restating his points in successive parallel clauses, sequenced so as to give progressively more clarity and insistence to his message, as shown in Table 12.2. Table 12.2  Structure of clauses in S3’s speech in Excerpt 4 Lines

Sequence of clauses

Function

1–2

are you CLEAR about this issue → what we want to do it’s not there → this is our plan → we haven’t done this → this is our plan → we want to do this we want to do it → this is . . . our objective → this is our task

Meaning of ‘issue’ clarified by second clause meaning of ‘plan’ reinforced by negatives in 1st and 3rd clauses, ‘want’ in 5th clause

8–11

21–23

1st clause and meanings of ‘objective’ and ‘task’ provide mutually reinforcement

As the speakers who employed this discourse pattern tended to be those taking leading or dominant roles in the interaction, a plausible interpretation is that this is a strategy used by more proficient speakers to accommodate to participants with lower proficiency, to improve the chances of communicative success. Ideational repetition in adjacent utterances Another characteristic of interaction order that was observed at the ICDC and CommCare settings was the high frequency of participants’ turns that were made up primarily of ideational meaning, especially of repetitions of the ideas contained in the previous utterances, and the low frequency of turns that were made up primarily of orientational meaning (cf. Halliday 1985).3 Ideational turns (or clauses) are those that, to some extent, directly express the ideas the speaker expresses or refers to, by reference that is nearly complete (e.g., “the Prime Minister of Japan wants to change the country’s constitution”), or partial and supported by pronouns and deixis (e.g., “he wants to change the constitution”). Purely orientational turns, on the other hand, refer to ideas (often in the previous utterances) without expressing them (e.g., “he does?”). As can be seen in the preceding examples, the operative difference, at least for the purposes of this study, is whether the clause’s main verb or predication expresses an idea (ideational), or whether it merely refers to an idea that is expressed elsewhere in the interaction (pure orientational). In the transcribed data, the overwhelming majority of clauses in the participants’ turns carried the ideational function, even to the extent that the ideas are repeated in the responses to these turns (instead of being referred to by auxiliary or modal verbs). Two examples of instances of such ideational repetition are shown here (Excerpts 5 and 6).

208  Alan Thompson In Excerpt 5, the participants are assigning themselves roles for a role-­play in community health services management.

Excerpt 5 (CommCare): An example of ideational exchanges S1 S4 1 2 3 4

female male S1: S4: SX-­m:

25–34 25–34

Japan Thailand

Japanese Thai

facilitator trainee

you’re not decided? decided. these are district officer. you need to be a support (.) to the n g o. xxxxx my n g o.

Here the first clause in line 2 expresses the ideational content of the first turn, but the new information, the orientational content “I am . . .” is only implied. This is in contrast to the practice of expressing only the orientational content, “Yes, I am” or “Yes, I have”. In Excerpt 6, where the participants are discussing the publication of proceedings from a conference, the ideational content “coming” (i.e., “being published”) is repeated for several turns (see lines 4–6), where the unmarked practice in standard usage is to use pro-­forms such as “it is”, “it is?”, etc.

Excerpt 6 (ICDC): An example of ideational exchanges P L M 1 2 3 4→ 5→ 6→

male female female P: L: M: P: P:

50–64 25–34 25–34

Indonesia China Japan

Indonesian Mandarin Japanese

supervisor researcher researcher

first of all, our project, yeah’ the proceeding is not yet coming. (xxx) it (x) coming. it’s coming? already? already coming. (2) ok.

Repetition, in and of itself, is hardly a feature of ELF. Mauranen notes that there are many uses for repetition in ELF and non-­ELF contexts alike: “ELF speech . . . uses repetition for gaining processing time, ascertaining comprehensibility, marking discourse boundaries, showing interactive alignment and affiliation, and structuring monologue” (Mauranen 2012: 206). What is possibly distinctive in the present data is that the repetition is predominantly of ideational elements in the interactions, rather than orientational, and that the choice is made to repeat the ideational elements from previous utterances (e.g., ‘decided’, ‘coming’) rather than to simply show orientation to those elements (e.g., by phrases such as ‘I am.’ and ‘it is.’). The ideational content of the repetitions would plausibly be a strategy to enhance the clarity of the message even if this means that the tone of the conversation is more abrupt and less talk is expended on maintenance of

Asymmetries in ELF interactions  209 the interpersonal relationship (e.g., with phrases of orientational meaning such as yeah, you’re right). Again, as seen in the discussion of action/pause/laughter turns earlier, this is a conversation structure that is, in the Hallidayan sense, sparse in orientational meaning – that is, the participants’ phrases convey mainly ideational meaning (what is happening) and little orientational meaning (how I/you are oriented towards what is happening). The participants have, consciously or unconsciously, selected this structure, which allows participants to neglect asymmetries among themselves, assume good intentions, and concentrate on the message. Distinct in-­group address practices In addition to examining the order in the interactions and the relationships among turns, word frequencies were also measured, treating the transcribed interaction data as a small corpus. One word that turned up as markedly more frequent than would be expected in globally compiled English corpora was the Japanese honorific address form san. According to the participants’ accounts in interviews and confirmed by the recorded interactions, the most common address form in two of the settings (ICDC and CommCare) was a name appended by san. In the UpLink setting, given names were the most common form, but name appended with san was very common among those more proficient in Japanese (which included the Hong Kong participants and about half of the French participants). The name used was determined by the preference of the addressee, and could variably be a family name, a given name, or a shortened form. However, counter to the participants’ claims of haphazard use, use of san was observed to be limited in a regular and perhaps unconscious way. At ICDC, members of one’s research team and other close working partners, both Japanese and non-­Japanese, were addressed or referred to by san, whereas for others English titles (predominantly Miss and Mr) were employed. At UpLink, the use of san was similarly observed to be more frequent among close-­knit groups such as, for example, when the discussion group was made up of those that were enrolled in the same programme. This tendency was usually not observable within brief segments of the interaction, except for Excerpt 7 (ICDC) where it shows quite clearly. In this excerpt, E, a research team leader from the Philippines, is speaking to members of his team, about a departing technical assistant (Punda), her replacement assistant (Pani), and a current assistant (Wu).

Excerpt 7 (ICDC): An example of in/out-­group address practices E 1 2 3 4 5

male E:

34–49

Philippines

Filipino

team leader

I think (.) Miss Punda is going to leave (.) and there is another person coming (.) Miss Pani (.) do you know him (.) Miss Pani (2) also an Indian (.) and she will be helping Ko-­san. And Mr Wu (.) who is now helping Ko-­san (.) is going to be working under Andy-­san.

210  Alan Thompson In the same turn, -­san was limited to peers (co-­researchers Ko and Andy, in lines 4 and 5) while English terms were used for technical assistants (Punda, Pani, and Wu in lines 1, 2, 3, and 4), showing a certain social distinction was being made. The use of san, then, at least in this working group, can be best described as an appellation used for a perceived in-­group, and within this in-­group it was applied irrespective of status, rank, nationality, or linguacultural background (e.g., Japanese secretary to Filipino group leader, and vice versa), reflecting perhaps a consensus of provisional equality. The boundary between the in-­group and out-­group was delineated in two ways: by hierarchy and by length of association. The director of the workplace was usually referred to as “the director” or as “Dr H”; his deputy was referred to as “Mr M”. This means of asserting provisional equality can clearly be useful in an environment when participants are anxious about the differences in rank among them (which they reported themselves to be). In this way, this distinct address practice might also be understood as a means of dealing with asymmetries with respect to linguacultural backgrounds, rank, etc., while simultaneously showing respect. Differential use of modality markers Another focus of the investigation into word frequencies was modal auxiliary verbs and other devices used to mark modality. When the interaction data from all settings was viewed in aggregate, the frequencies were, by and large, unremarkable. There are, however, many other means by which speakers of English express modality, or, in other words, their orientation (both epistemic and attitudinal) to ideas in the interaction or in the context. One of these means is the use of modality markers, such as think, like, kind of/kinda, maybe, etc., that formally resemble lexical ideational content, but are actually used to display the participants’ orientation to the idea (the ideational content of the utterance) either in terms of the status of their knowledge of the idea (epistemic) or in terms of their affective orientation to the idea (attitudinal). Excerpts 8 and 9 show how the modality marker (think in Excerpt 8) differs from the lexical item expressing ideational content (think in Excerpt 9, line 3).

Excerpt 8 (UpLink): Use of the modality marker think S2 female 17–24 Hong Kong Cantonese student 1

S2: er i think the summary here sums it up pretty well.

Excerpt 9 (UpLink): Use of the lexical item expressing ideational content think S8 male 17–24 France French student 1 2 3 4

S8:

it’s the same thing in the case their main objective is just to complete the mission that’s it (.) we should try to think long term more Asian way. we are going to work together or they are gonna have good relationship (.)

Asymmetries in ELF interactions  211 In the interaction data from the UpLink setting, out of 57 uses of the token think, 55 were orientational, and 2 were ideational, illustrating that this is a very common means of expressing modality. Similarly, the token like was used largely as an orientational marker, though with much use as well in construction of ideational content (of 74 uses, 51 were orientational, and 23 were ideational), as seen in Excerpt 10.

Excerpt 10 (UpLink): An example of the use of like as an orientational marker S2 female 17–24 Hong Kong Cantonese student 1 2 3

S2:

oh another thing i wrote down in my analysis Louis saw the company like family like company are really close where David doesn’t understand what he is concerned about

The aggregate data for the most common of this set of devices for expressing modality can be seen in Table 12.3. The investigation of modality markers in a subset of the data from the UpLink setting revealed an interesting correlation between their frequency and the role of the speaker in an interaction. Although all participants were equal in terms of their membership categories (e.g., students in an unstructured discussion group), there were a few participants who took the dominant role in some interactions and a more peripheral role in others. When participants took a more dominant role in an interaction, the markers of orientational modality (in Table  12.3) were more frequently used, relative to the total word count of that participant. When the same participant took a more peripheral role, these markers were less frequently used, again relative to the participant’s total contribution. To compare these speakers’ behaviour in conversations where they took dominant roles to that in conversations where they were more passive, each speaker in each discussion was given a unique identifying number (so that a given individual would be treated separately in different discussions). When modality markers were identified and counted for each speaker (manually weeding out the Table 12.3  Most frequent modality markers in the corpus subset Term

Count

Orientational uses

Relative frequency

think like kinda/kind of maybe know sure guess if probably

57 74 28 25 28 7 6 38 0

55 51 27 25 25 7 6 3 0

0.89% 0.83% 0.44% 0.41% 0.41% 0.11% 0.10% 0.05% 0%

212  Alan Thompson Table 12.4 Use of the modality markers kind of/kinda and like arranged by decreasing participation by speaker Speaker

Words

Turns

Average intervening turns

Kind of/kinda (relative freq.)

Like (relative freq.)

S2 (A) S11 S3 S6 (B) S8 S7 (A) S9 S1 (B) S4 S10 S5

1543 1137 1037 865 747 732 320 397 207 303 114

149 94 129 82 71 40 38 48 23 16 15

1.59 1.59 1.93 1.93 2.22 4.55 5.05 5.54 12.17 14.79 18.44

0.71% 0.36% 0.19% 1.27% 0.00% 0.00% 0.31% 0.00% 0.48% 0.33% 0.00%

1.88% 0.94% 0.29% 2.66% 0.94% 0.21% 0.37% 0.36% 0.48% 0.49% 0.00%

Note: ‘(A)’ and ‘(B)’ note the same individual in different interactions.

ideational uses), and compared to an independent variable measuring interaction dominance,4 a clear correlation was discovered, especially for two of the modality markers – kind of/kinda and like – as seen in Table 12.4. When we look more closely at pairs of discussion speakers who are actually the same individual in different discussions, the correlation is even more striking, and strongly suggestive of a dependent relation. In Table  12.4, one individual, A, is represented by the speaker identifying numbers S2 (dominant role) vs. S7 (peripheral role), and another, B, by S6 (dominant role) vs. S1 (peripheral role). If we focus on these two individuals, omitting the other rows in Table 12.4, we see a distinct contrast (Table 12.5). Table 12.5 Use of the modality markers kind of/kinda and like by two individuals, comparing dominant participation with peripheral participation Individual

Speaker

Words

Turns

Average intervening turns

Kind of/kinda (relative freq.)

Like (relative freq.)

A – dominant A – peripheral B – dominant B – peripheral

S2 S7 S6 S1

1543 732 865 397

149 40 82 48

1.59 4.55 1.93 5.54

0.71% 0.00% 1.27% 0.00%

1.88% 0.21% 2.66% 0.36%

For both individuals, A from Hong Kong and B from Japan, when dominant in a discussion (by the measures I have employed here), the marker kind of/kinda has a relative frequency higher than the overall frequencies of most modality markers (higher than all except think and like, see Table 12.3 earlier). Yet when only

Asymmetries in ELF interactions  213 peripherally involved in the discussion, this marker is not used at all by either individual. Similarly, and probably more significantly due to the higher overall frequencies, both individuals use the orientational modality marker like more than eight times as often (relative to total number of words spoken) when in a dominant discussion role than when in a peripheral role. Although the total number of words wherein these observations are made is small, the similarity in the patterns for these two markers (kind of and like) suggest that there may indeed be different language practices corresponding to the asymmetry in the roles taken up by conversation participants. Correlation of practices to situations The disparate observations of conversational strategies and grammatical-­ discoursal adjustments, described in the previous section, do not form a picture of English language practice in any one setting, but instead suggest the degree to which different ELF settings (situations and participants), and the asymmetries among them, may give rise to very different language practices: 1 2 3 4 5

use of minimal, non-­elaborated, functional turns in a sparse interaction order; simplified, short, repeated clauses; explicit repetition of ideational content from previous utterances; distinct address practices for in-­group and standard practices for out-­group; and use of certain subtle modality markers while taking a dominant role in an interaction.

The distinctly sparse interaction order was prevalent mainly in ICDC setting, perhaps influenced by the overall more serious tone and by the tendency (observed and reported in interviews) to switch to Japanese for greetings, partings, and “small talk” topics such as weekend activities or traffic. This asymmetric division of labour between languages may have led to a perception among participants that they could not engage in small talk during English interactions, which were to be work-­related, and, hence, that all non-­essential comment on work matters was unnecessary. One instance of switching to Japanese to make such a comment seems to be indicative of this attitude (Excerpt 11).

Excerpt 11 (ICDC) An example of switching to Japanese to make a non-­essential comment R T

male female

25–34 25–34

Thailand Japan

Thai Japanese

researcher researcher

1 R: they (.) they should have this type 2 T: yeah this type of information available for the ministry 3 really. dakedo taihen dakara ne. syou ga nai ne. @@@ (translation: but it’s difficult, right? so what can you do?)

214  Alan Thompson Explicit repetition was observed mainly at the ICDC and CommCare settings. These were the settings where the lowest English competence levels were reported by participants (talking of themselves or of others), and CommCare was also the setting where there was the widest variance between levels of formal education. The strategy of reinforcing the communication of ideas, together with the adjustment of repeating ideational content (“Decided?” – “Decided.”) could thus be interpreted as arising from real anxiety that ideas may not be transmitted otherwise. This anxiety would be higher at these two settings than at the business school setting (UpLink) for another reason as well: the stakes of the interactions at the workplace (ICDC) and in-­service training programme (CommCare) were generally greater than they were in the non-­assessed discussions at the business school. The distinct address practice using -­san was observed across all settings but particularly well defined and consistent at ICDC, and only extending to about one third of students at UpLink. The hypothesised purpose for the practice (to mark an in-­group) provides an explanation for this difference. At ICDC there was a clear in-­ group which was the research team making up most of the participants in the study; at CommCare the entire group of trainees and facilitators were the in-­group (there was no out-­group to be often referred to); and at UpLink there were only a few small groups of students for whom being at ease in Japanese was a group marker. Finally, the use of casual modality markers by dominant participants mainly in the UpLink setting, and not in the others, may be due to the perception (by researcher and participants) that there were participants who were native-­speakers or indistinguishable from native-­speakers in terms of proficiency. This may have contributed to an exo-­normative expectation that the most able participants would employ native-­speaker usage when taking a dominant role in the discussions. Additionally, there were no established procedures for meetings as there were at ICDC, so dominant participants might prefer to use expressions that were more subtle (e.g., kind of, like; as opposed to more explicit expressions such as I’ve decided that . . . or This is what we should do.) in their expression of decision-­ making authority and conversational control.

Discussion The first set of questions put forth in the introduction dealt with how isolated and situational features of communication can tell us anything about ELF in general. At first, it looks as if the practices in the array illustrated earlier are not reducible to common principles. Yet, in the summary of correlations of practices with settings I have just given, there is, at the end of each, a plausible (though inadequately supported) explanation of how the practice could be motivated by some specific features of the situation. Could it be that this greater dependence on situational and participant factors, and not on wider sociolinguistic or cultural factors, is the main characteristic of ELF communication? Hülmbauer seems to suggest this when she writes that the traditional notion of a stable ‘speech community’ . . . cannot apply to lingua franca contexts. . . . ELF users become members of situational speaking

Asymmetries in ELF interactions  215 communities: . . . with speaker constellations frequently being tied and untied anew for each emerging interaction. (Hülmbauer 2009: 325) The extent to which there are features that are common to all instances of ELF communication may be limited, indeed, to the extent that all ELF situations are alike, that is, to a vanishingly small extent. The second set of questions had to do with how participants respond to diversity and asymmetry in ELF situations and whether we can evaluate ELF practice – whether we might be justified in calling certain ELF practices “good” or “bad” ELF. Plausible motivation for each of the language practices (as illustrated earlier), even if the explanations are flawed, suggest an evaluative criterion, analogous to the evaluation of successful traits in biological evolution. Echoing Halliday’s claim that the form of language has been determined by the functions it has evolved to serve (1973: 7), the criterion is simply that a certain language practice is good if serves its purpose. Explicit repetition of ideational content, if it actually succeeds in lowering the chance of misunderstanding, is “good” ELF; that is, the accumulation of successes using this practice might lead to its wider adoption by speakers in ELF situations. Conversely, if a non-­standard practice that is communicatively successful is resisted or avoided – perhaps by a non-­native speaker of English under the influence of too much admiration for established norms, or by a native speaker of English who is ignorant of the benefits of certain non-­standard practices in ELF settings – this is “bad” ELF; happily, however, this sort of resistance of non-­standard features might lead to fewer communicative successes and eventually wane. The settings considered here have revealed asymmetries in various dimensions – in language proficiency and experience in intercultural communication; in status, membership category, and role; and in attitudes towards goals, norms, and social expectations. Any successful communication is the discursive accomplishment of a group (cf. Firth 1996), and the diversity and asymmetry within the group is one more challenge that must be met with effective language practices and communication strategies. As a group is unlikely to remain communicatively cohesive if its communications are not successful, it would not be surprising to find some convergence towards good practices within a group over the period of sustained contact. Of course, the practices and strategies observed here probably comprise no more than a small subset of the many ways that participants in ELF situations attempt to meet this challenge and to converge towards “good” ELF. What this sampling of disparate settings shows is that the irregular, asymmetric aspects of ELF situations may be among the most important factors when considering how this process towards effective communication evolves.

Appendix Transcription conventions

P, S1 SX SX-­f SS . ? ’ (.) (2.5) do MDI @@  . . .  xxx ((comment))

(single initial, or S followed by numeral) identifier for participant unknown participant unknown participant (f = female; m = male) all or many participants at once falling intonation rising intonation slight rise as in listing intonation short pause (up to .5 seconds) longer pause, with approximate timing in seconds accent on underlined word or syllable, extra prominence spoken as a sequence of letters (/em-­di-­ai/) Laughter overlapping speech unintelligible speech, with ‘x’ = one syllable extra-­textual information and comments

Notes 1 Heritage (1995) asserts that, for ‘pure’ conversation analysis, the consideration of talk-­ in-­interaction is sufficient, without consulting participants. This stance is based on the assumption that the researcher knows the cultural and linguistic norms that embed the interaction, and therefore can determine the function of each turn or linguistic feature. This assumption does not hold for the analysis of ELF interactions. 2 Although it is impossible to define exactly what ‘standard native-­speaker usage’ is, I do not hesitate to use such a phrase when I am referring to the emic categories that structure the participants’ perceptions and behaviour. This is partly following Seidlhofer’s (2011: 6) reasoning that the term ‘native speaker’ is not so much problematic in terms of what it denotes (an L1 speaker) but rather in terms of the privilege and authority that it (to many) connotes. It is also because participants repeatedly refer to a broad, tacit (though imprecise) understanding of ‘standard’ or ‘native-­speaker’ usage, and the existence of this tacit understanding is a social fact that potentially influences ELF users. 3 To speak of turns as having ideational and orientational meaning is a gross simplification of Halliday’s systemic-­functional grammar. Most turns simultaneously have ideational and orientational functions, as well as carrying a textual function. However,

Asymmetries in ELF interactions  217 this distinction, as it is described here – derived but not identical to Halliday’s distinction – captures a significant feature of the interactions. 4 Interaction dominance was measured by counting, for each participant, the number of turns that intervened between their subsequent turns, and taking an average (in Table 12.4, ‘average intervening turns’). In a dyadic conversation, for example, each participant would have an ‘average intervening turns’ value of 1; in a balanced three-­ person conversation, the value would be about 2.

References Björkman, B. 2016. English-­medium instruction and English as the Lingua Franca in higher education in Central and Northern Europe. In M.-­L. Pitzl and R. Osimk-­Teasdale and Ruth (eds), English as a Lingua Franca: Perspectives and prospects. Berlin: de Gruyter Mouton, pp. 57–68. Björkman, B. 2018. English as a Lingua Franca in spoken genres in the international university: Introduction. Journal of English as a Lingua Franca 7(2), 225–228. doi:10.1515/ jelf-­2018-­0019. Blommaert, J. and A. Backus. 2013. Superdiverse repertoires and the individual. In I. de Saint-­Georges and J.-­J. Weber (eds), Multilingualism and multimodality: Current challenges for educational studies (The future of education research). Rotterdam: Sense Publishers, pp. 11–32. Canagarajah, S. 2007. After disinvention: Possibilities for communication, community and competence. In S. Makoni and A. Pennycook (eds), Disinventing and reconstituting languages. Clevedon: Multilingual Matters, pp. 233–239. Cogo, A. 2012. ELF and super-­diversity: A case study of ELF multilingual practices from a business context. Journal of English as a Lingua Franca 1(2), 287–313. doi:10.1515/ jelf-­2012-­0020. Darvin, R. 2017. Social class and the inequality of English speakers in a globalized world. Journal of English as a Lingua Franca 6(2), 287–311. doi:10.1515/jelf-­2017-­0014. Firth, A. 1996. The discursive accomplishment of normality. On “Lingua Franca” English and conversation analysis. Journal of Pragmatics 26, 237–259. Halliday, M.A.K. 1973. Explorations in the functions of language. London: Edward Arnold. Halliday, M.A.K. 1985. An introduction to functional grammar. London: Edward Arnold. Have, P. ten. 2007. Doing conversation analysis. Thousand Oaks, CA: Sage Publications. Heritage, J. 1995. Conversation analysis: Methodological aspects. In U.M. Quasthoff (ed), Aspects of oral communication. Berlin: de Gruyter Mouton, pp. 391–418. House, J. 2012. English as a Lingua Franca and linguistic diversity. Journal of English as a Lingua Franca 1(1), 173–175. doi:10.1515/jelf-­2012-­0008. Hülmbauer, C. 2009. “We don’t take the right way: We just take the way that we think you will understand” – the shifting relationship between correctness and effectiveness in ELF. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle upon Tyne: Cambridge Scholars Publishing. Kaur, J. 2018. ELF in spoken genres in the international university: Of contextual factors and non-­linguistic resources. Journal of English as a Lingua Franca 7(2), 403–410. doi:10.1515/jelf-­2018-­0020. Mauranen, A. 2012. Exploring ELF: Academic English shaped by non-­native speakers. Cambridge: Cambridge University Press. Meierkord, C. 2002. “Language stripped bare” or “linguistic masala”? In K. Knapp and C. Meierkord (eds), Culture in Lingua Franca conversation. Frankfurt am Main: Peter Lang, pp. 109–133.

218  Alan Thompson Mežek, Š. 2018. Laughter and humour in high-­stakes academic ELF interactions: An analysis of laughter episodes in PhD defences/vivas. Journal of English as a Lingua Franca 7(2), 261–284. Pitzl, M.-­L. 2018. Transient International Groups (TIGs): Exploring the group and development dimension of ELF. Journal of English as a Lingua Franca 7(1), 25–58. doi:10.1515/jelf-­2018-­0002. Pullin, P. 2013. Achieving “comity”: The role of linguistic stance in Business English as a Lingua Franca (BELF) meetings. Journal of English as a Lingua Franca 2(1), 1–23. doi:10.1515/jelf-­2013-­0001. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Thompson, A. 2016. Modality practices among ELF users in academic discussions: Dominant and peripheral participants. In L. Lopriore and E. Grazzi (eds), Intercultural communication: New perspectives from ELF. Roma: TrE-­Press, pp. 473–484. Thompson, A. 2017. Diversity of users, settings, and practices: How are features selected into ELF practice? Journal of English as a Lingua Franca 6(2), 205–235. doi:10.1515/ jelf-­2017-­0011. Toomaneejinda, A. and L. Harding. 2018. Disagreement practices in ELF academic group discussion: Verbal, nonverbal and interactional strategies. Journal of English as a Lingua Franca 7(2), 307–332. doi:10.1515/jelf-­2018-­0016. Vertovec, S. 2007. Super-­diversity and its implications. Ethnic and Racial Studies 30(6), 1024–1054. doi:10.1080/01419870701599465.

Part IV

Narrative approaches in ELF research Exploring attitudes towards ELF

13 ‘Place-­reflexivity’ as an imaginary kaleidoscope to explore methodological issues in ELF research Masuko Miyahara Introduction As English as a lingua franca (ELF) research in Japan expands and increases (Murata 2016), there comes along with it a diverse range of approaches and methods used to conduct these studies. According to Murata (2016), there are basically three approaches most frequently used in current ELF research: corpus-­ based studies, conversation analytic studies, and narrative approaches. This paper focuses on the third approach, narratives. In particular, it discusses some of the emerging methodological issues involved in analyzing narrative research by using data collected from a study that offers a unique perspective on the understanding of the process of L2 identity construction and development in EMI (English as a Medium of Instruction) classrooms in higher education in Japan (Miyahara 2015), where English is a lingua franca. As ELF research diversifies both thematically and methodologically, it is essential that researchers develop a more critical eye to evaluate the various research approaches employed and to evaluate the appropriate methodology and method to explore their own research inquiry. Based on the understandings of researchers’ reflexivity, this paper attempts to argue for the importance for researchers to establish a space for critical and reflective thinking towards their research. More specifically, it will draw on the most recent discussions in qualitative research methodology that argue for a fuller account of the ‘where’ in the research encounter, that is, more focus on the research site where the study is conducted. In recent times, scholars have increasingly pointed out that the spatial contexts in which the ‘where of method’ matters have received less attention compared to the ‘how’ of qualitative research (i.e., Anderson et al. 2010; Cresswell 1996; Sin 2003). The social relationship in methodology of the research and the process of researcher reflexivity in terms of the positionality of the researcher and the participant have been well discussed. The chapter thus introduces a less documented perspective on reflexivity, the ‘where’ in the research. It will draw on the notion of place-­reflexivity as proposed by Swaminathan and Mulvihill (2019) by revisiting one of my previous research projects that explores the construction of L2 identity of Japanese learners at higher education in Japan. The chapter begins by presenting an overview of how ELF identity is conceptualized by referring to previous relevant studies conducted on the subject. It then

222  Masuko Miyahara offers a brief summary of how narrative inquiry is understood in the sample study (Miyahara 2015) used in this chapter. I then outline how researchers’ reflexivity could be advanced by drawing on the understandings of place-­reflexivity (Swaminathan and Mulvihill 2019) that brings the research site to the forefront rather than as a backdrop of the study. After providing a descriptive account of the sample study and employing the notion of place-­reflexivity in revisiting the original data, I attempt to illustrate that, with the shift of focus on the ‘where’ of qualitative research, researchers are now able to develop increased sensitivity and a more nuanced understanding towards the data. As is often pointed out, reflexivity should be reflected at all stages of the research process, but it is especially important at the analysis stage that place matters (Swaminathan and Mulvihill 2019). At this present age when a surge of diversity and plurality in ELF research can be anticipated, the chapter concludes by arguing the importance of serious contemplation on the methodological issues that surround ELF research. In this first book on research methodology in ELF research, it is most timely to bring the fundamental concerns on research methodology in narrative studies to centerstage.

Situating the scene ELF identity and the use of narrative approach Defining ELF as ‘any use of English for communication among speakers of different first languages for whom English is the communicative medium of choice, and often the only option’ (Seidlhofer 2011: 7), an increasing number of ELF studies have begun to acknowledge that ELF speakers’ identity construction is a relevant and unavoidable part of ELF communication (Baker 2009; Jenkins 2007; Nogami 2019). In these studies, following poststructuralist understanding, the identity of the second language learner is discursively constructed and conceptualized as multifaceted, fluid, dynamic, and emerging in interactions with others (i.e., Block 2007; Norton 2013; Miyahara 2015). Work on the construction of ELF identities using a narrative approach is exemplified well, for instance, in Virkkula and Nikula’s (2010) narrative study, which investigated Finnish students’ identity transformation from that of a language learner to language user during their study abroad in Germany. Another narrative study that explored similar transitions of university students’ identities was conducted by Iino and Murata (2016), but this time, in an EMI setting at a Japanese university. In both cases, the students underwent a change in their L2 identities: from confident English language learners who are conscious of their language in relation to the ‘native speaker’ competence, to language users who have acquired communicative capability (Widdowson 2016). The sample study employed in this chapter follows the trajectory of one of my participants, Megumi (a pseudonym, Miyahara 2015), who follows similar patterns in her identity construction as a language learner to a language user at a multilingual liberal arts college in Tokyo, the main research site. As exemplified in an increasing number of identity-­related ELF studies, narrative inquiry is now one of the prominent methodological approaches used (Iino

‘Place-­reflexivity’ as kaleidoscope  223 and Murata 2016; Kanno 2003; Nogami 2016; Miyahara 2017; Takino 2016) in trying to understand the implications of one’s identity in ELF interactions. Understanding the complexities involved in forging L2 identities required me to examine the experiences of the participants from their perspectives and to listen to their ‘voices’. This called for a methodology that would allow me to be sensitive to the learners’ (and users’, for that matter) account of their experiences as understood from their viewpoints. Narrative approach that focuses on the meaning making process of the participants, and where I could emphasize my role in co-­producing the narratives, appeared to be the best way to probe into the inner complexities of my research participants. In the sample study in this chapter, following Bruner (1990) and Clandinin (2007), narratives are understood as fundamentally stories of ‘experience’. In the experience-­centered approach, narratives are the means of human sense-­making where human beings create meaning from their experiences both individually and socially (Bruner 1990; Polkinghore 1995). Connelly and Clandinin (1990) have argued that narrative inquiry stems from an understanding of human experience in which people, both individually and socially, lead storied lives. Life is ‘storied’ in the way that people make sense of who they are (and others are) by interpreting their past in terms of their present lives and selves as well as their future lives and selves. Thus, narratives are not only about people describing their past experiences, but also about how individuals understand those experiences and how they ascribe meanings to those experiences (Barkhuizen 2013; Clandinin and Connelly 2000). Understanding the researcher’s reflexivity Researching learners’ identity naturally involves a certain degree of interpretative activity by the researcher. More than we would like to admit, researchers are thus inevitably as much a part of the study as the participants. How we take into account the effect of our presence on our research through researchers’ reflexivity is a critical issue. Although discussions of incorporating researchers’, reflexivity in the research process seem to be established as an important strategy (i.e., Finlay 2012; Mann 2016; Miyahara 2017, 2019; Prior 2016; Riessman 2008, 2012; Roulston 2010), there appears to be no common understanding towards the notion of reflexivity, and the concept is often confused with reflection. In fact, one of the reasons for my interest in the notion of ‘reflection’ emerged from this very question: what is the difference between ‘reflection’ and ‘reflexivity’? In this paper, following Finlay (2012), I  take reflection to mean ‘thinking about’ something after the event and reflexivity, in contrast, to mean an on-­going self-­awareness. That is, the former is more of a descriptive process, whereas, the latter, refers to how you position yourself in the research context and contemplate on how one’s own self could influence the actions one takes (self-­awareness), including their influence on the interactions with the participants. In fact, in general, reflexivity is comprised of mainly two dimensions: an understanding of researchers’ positionality by the researchers themselves and how that affects the overall research process and outcomes. The question is not whether one should engage in reflexivity, but

224  Masuko Miyahara how and to what degree, all of which would depend on the focus of the research (Finlay 2012). Reflexivity, understood as researchers’ positionality, is akin to turning the mirror on yourself. But some scholars have criticized self-­reflexivity because of its self-­indulgent nature (i.e., Guillemin and Gillam 2004). In my view, the important point is to always consider how the researcher’s personal disclosure does for qualitative analysis and how it contributes to understanding the phenomena at hand (Berger 2015). Furthermore, self-­reflexivity should not be just about the ‘self’, so to speak, although extending one’s reflexive gaze also helps to reinforce and furthers one’s awareness of the self. Researchers should not merely discuss how their personal identities shaped their narrative projects; they should also consider how these identities intersect with the institutional, cultural, and socio-­political context (Nagar and Geiger 2007). Furthermore, turning the mirror on one’s previous works could reveal the situatedness (the academic, theoretical, political, social, etc.) of our interpretations. For instance, depending on the researcher and their relationship with the participant, the same experience of a participant could be narrated quite differently. There may be discrepancies between one’s telling and another depending on when, who, and even where the talks were constructed, but these differences do not necessarily indicate that one is more fictive than the other. The different telling is considered to be merely another interpretation of the same individual’s narrative. A kaleidoscope is a metaphorical term often used to describe such notion of multiple ways of seeing: ‘You look and see one fascinating complex pattern, the light changes, you accidentally move, or deliberately shake the kaleidoscope, and you see – composed of the same elements – a somewhat different pattern’ (Stanley 1992: 178). Reflexivity can function as an imaginary kaleidoscope for us to employ in order to broaden our ways of thinking and understanding of our research. There appear to be various ways of working with reflexivity. Some are connected to the co-­construction of data; others are interested in the positioning of the researcher, the participant, and their ever-­changing relationship, and so forth. Still yet, others consider reflexivity in qualitative research ‘as a tool to measure legitimacy, validity, rigor’ as well as ‘to maintain one’s integrity’ (Kim 2016: 250). However, in this paper, I would like to depart from the norm of reflexivity that tends to put prominence on the social relationships, for instance, the positionality of the researcher or the researcher-­participant relationship to more emphasis of the context of the research site, and instead, ‘appreciate the multifaceted and the interactive nature of the respective research sites, and the role they play in the production of knowledge’ (Anderson et al. 2010: 592) – that is, move towards a methodological praxis that explicitly includes the researcher, researched, and the place of method rather than the conventional approach in the research that tends to focus on the dialogue between the researcher and the participants. As Tilley puts forth: [Place] is a medium rather than a container for action, something that is involved in the action and cannot be divorced from it. [It] is meaningfully

‘Place-­reflexivity’ as kaleidoscope  225 constituted in relation to human agency and activity. A humanized space forms both the medium and outcome of action, both constraining and enabling it. (1994: 10) I intend to pursue this goal by revisiting one of my previous studies on identity constructions of L2 learners from learners to users in an ELF context. In doing so, I draw heavily on the concept of place-­reflexivity as advocated by Swaminathan and Mulvihill (2019), which I turn to in the following section. Positioning place-­reflexivity Understanding place as comprised of two interrelated dimensions (Cresswell 1996; Sin 2003), the social and the geographical, the notion of place-­reflexivity conceptualized by Swaminathan and Mulvihill (2019) attempts to put emphasis on the rich, contextual factors that research sites represent. It specifically addresses the ‘where’ of qualitative research and intends to acknowledge the significance of the role place plays in the meaning-­making process. Their notion of place-­ reflexivity is a response to the generally accepted methodology which tends to obscure the relationship of the researcher and the participant to the place, the research location, or the research site in which research occurs. They question the prominence put on the social dimension of place, where the researcher’s ability to interpret and reflect on the power structures and relationships are at the center of discussion, and, instead, argue for the inclusion of a more broader understanding, where place can be considered as an active participant in the constitution of relationship, and, consequently, in the production of knowledge. In addition, it could be helpful to provide clarifications as to some of the terms used in this chapter, namely, place, site and context, at this point. According to scholars, the differences appears to represent a disciplinary focus. For example, site (Creswell 2014; Agnew and Duncan 2014) is often used in sociology, while, in education, context is used to contemplate variety of factors that includes place (Agee 2002). Apart from terminologies, it is, however, important to emphasize that place-­ reflexivity is not regarded as an alternative concept to replace the traditional understanding of reflexivity. Rather it is understood as one that seeks to complement the current dominant focus on the social dimension, where the relationship or the action and/or interactions of participants among themselves as well as the participants and researcher are the main areas of concern. The aim of such move is to enrich and contribute to the methodological complexities of qualitative research. Integrating the ideas of space, place and site by Soja (2010), Bhabha (2009) and Turner (1969), Swaminathan and Mulvihill understand ‘place’ as ‘spaces that are used in research by scholars; material as well as conceptual places for both long-­term and ephemeral time periods’ (2019: 8). It acts as a methodological tool that can assist researchers keep place at the center of research rather than merely as a backdrop: move sites from static, already there, background concept to an interactive dynamic space so that questions of power, ethics of respect and the

226  Masuko Miyahara intersections of place, globalization and indigenous ways of knowing can be culled from sites and fieldwork through a process of place-­reflexivity. (Swaminathan and Mulvihill 2019: 3) As a methodological tool to facilitate reflexivity that ‘goes beyond research subjectivity and interrogates the familiar’ (p.  9), they propose several aspects to place-­reflexivity in order to help researchers focus on the details in the field as data is gathered and analyzed by referring to the following six features (from Swaminathan and Mulvihill 2019: 9): a b

c d

e

f

Place-­reflexivity is an attention to sites that seeks to capture the physical location of the site and examine it in relation to other sites. It promotes reflection on the location and context and the wider environment. Place-­reflexivity means attending to the everyday flow of people through places. The ways in which the everyday habits of people become stable over time in particular places is part of site-­see through place-­reflexivity. Place-­ reflexivity also means paying attention to biographical discourses to highlight connections between people and places. Place-­reflexivity means thinking about resemblances or using the research imagination to deconstruct place from taken for granted ideas. Place-­reflexivity helps researchers to examine how places are organized and whether they promote or inhibit social interaction. It also means thinking about the voices which are less heard than others, who gets marginalized at what site. Place-­reflexivity involves self-­consciously observing and listening to how participants regard places. It means attending to the spectrum of emotions in certain places and learning from them. It also means the researcher attending to their own emotions and senses. Place-­reflexivity helps researchers with the awareness that stories that people tell in sites and about sites need to be noticed, acknowledged, and documented. It also means looking for liminal spaces where multiple stories can co-­exist.

Swaminathan and Mulvihill (2019) emphasize that we should not reflect on what we observe and base our understandings on the assumptions we hold about our research, but we should contemplate and deliberate on how we come to know what we think we know. It is in this line of thought that I later turn to re-­examine my data and findings from the sample study. Place-­reflexivity is particularly apt for the purpose of this chapter, firstly, because of its relevance in the analysis stage of research as the authors claim: It is in the analysis of qualitative data that we suggest place-­reflexivity can bring greater awareness to the ways in which place matters in qualitative research. A place-­reflexivity would need to not only take into account social and relational networks (e.g. social dimension of place) but in addition

‘Place-­reflexivity’ as kaleidoscope  227 analyze place for its material and environmental characteristics thus moving sites into the foreground of qualitative thinking. (Swaminathan and Mulvihill 2019: 3; the current author’s emphasis in italics) I believe this is an important point to make since at the data collection stage, many researchers do in fact think seriously about where the data is to be collected. For example, using interviews, researchers would be most conscious about the place or the location where the interview could take place in order to create the most comfortable atmosphere for the participants and, hence, hope to build rapport with the participants and to be able to elicit rich stories for the research. Secondly, I also regard place-­reflexivity as a particularly relevant concept in revisiting my sample study that explores the transition in my participant’s subject position from a language learner to that of a legitimate English user. Although studies, including my own, often conclude that ELF users’ identity tend to emerge and flourish as they are exposed to a variety of ELF encounters, the details of the ELF contexts that the participants operate in are not sufficiently explored or rather not explicitly presented. The focus on the peculiarities in different ELF contexts that make up the environment that enables learners to construct new identities or transform identities is obscured or taken for granted in many cases. In the sample study that I  draw on in this chapter, I  refer to one of my participants, Megumi, and attempt to examine the details of ELF encounters that eventually enable her to construct a new identity as a language user and how the materials and the environmental characteristics of the different sites had influenced in the creation of the new subject positions.

The sample study Drawing on one of my previous works, I would now like to critically reflect on the story of Megumi’s identity formation in different ELF contexts, and to illustrate how the discussions of place-­reflexivity so far have provided additional insights to the original study. In doing so, I will first give a brief outline of this research by presenting a descriptive account about the researcher, the research site, the participant, the methodological approach including data collection methods, the analytical process, and the reflexive strategies used in the sample study. An overview of the sample study Narrative data for this chapter is drawn from a study that aimed to shed light on the understanding of process of L2-­related identity construction and development among Japanese English learners at the tertiary level (Miyahara 2015). The particular feature of this study was its attempt to integrate the socially and the psychologically oriented perspective on L2 identity formation. The research contributed in highlighting the instrumental agency of individuals in responding to and acting upon the social environment and in developing, maintaining and/ or reconstructing their desired identities as L2 users. It offered unique insights

228  Masuko Miyahara into the role of experience, emotions, social and environmental affordances, and individuals’ responses to these, in shaping their personal orientations to English and the transformation of their self-­perceptions as English learners to users as they envision themselves in an international community of English-­speakers (or as Iino and Murata would say ‘from EFL learners to users’) (2016: 111). The original study helped to show the intricate analysis of past, present, and future dimensions of individuals’ L2-­related experiences and trajectories and how these dimensions are intertwined through the process of narrative construction as participants relate their thoughts, and the researcher represents and interprets their stories (for details, refer to Miyahara 2015). The researcher In line with Canagarajah (1996: 324), in any research, ‘the subjectivity of the researchers  – with their complex values, ideologies, and experiences  – shapes the research activities and findings’, I believe it is crucial that researchers write themselves into the research by providing a brief account of relevant aspects of the researcher that appear to be pertinent to the study. Thus, what follows is an account of the aspects of my trajectory as a learner, practitioner, and a researcher that are particularly relevant in the shaping of the research project. My interest in identity and language learning is largely rooted in my educational experiences in and out of Japan. I  was born in Japan but spent most of my formative years overseas. After graduating from a high school in the United States, I returned to Japan to attend a college in Tokyo, which happens to be the research site of the sample study. My educational experience as such is what some academics like Kanno (2003) would call the ‘returnee’ experience. One apt phrase that would characterize my language learning experiences over the years would be riding on a virtual roller coaster of ups and downs in language learning experiences and, henceforth, undergoing identity shifts and periods of mixed emotions towards myself as a language learner and user and, later, as a practitioner and researcher. The research site and the participants The research site, a private university located in the suburbs of Tokyo, aims to build a global community where a diversity of people from various ethnic and religious backgrounds gathers together. This institution is also famous for its bilingual identity, and, in Japan, the graduates and alumni are regarded by the general public as being fluent and well versed in the English-­speaking language and culture. English is used on a daily basis as a means of communication not only in classes, but also in the daily lives of the students and faculty alike. The university has several college-­wide courses that are required components for all students, and the English Language Program (ELP), that was later renamed ‘English for Liberal Arts’ (ELA), is one of them. Students, for whom English is a second language, must study English intensively for the first 2 years. The main focus

‘Place-­reflexivity’ as kaleidoscope  229 of the program is designed to teach English for academic purposes with a focus on critical thinking. The curriculum is further complemented by an optional study abroad program referred to as the Study English Abroad (SEA) Program. First-­ and second-­year students can take part in a 6-­week program during the summer break at various universities located in the United States, the United Kingdom, Canada, New Zealand, and Australia (at the time of the research). A total of six students, volunteers, in their first year of their 2-­year English language curriculum participated in the study. For the purpose of this chapter, I will focus on one of them, Megumi (a pseudonym), who exemplified most clearly the identity changes from a language learner to that of a language user. Data collection The method employed in this particular study is what is generally characterized as autobiographical narratives. I used a narrative interview strategy based on a series of semi-­structured questions to generate data in that the researcher did not have a list of questions but rather a range of topics to be covered (Block 2006). The language in which talks were conducted is related to the jointly constructed nature of the interview process in narrative studies, and thus, the participants were given choices, but none opted for English, and thus, the language used in all interviews was Japanese. Five sets of interviews were conducted over a period of 1 year, each kept within an hour and a half, mainly for practical reasons (e.g., time to conduct these interviews was limited for the participants). The narrative data was supplemented by other sources such as weekly journals, audio recordings or group discussions, and weekly self-­reports during the 6-­week study-­abroad programs for those who participated in the SEA Program. The talks were audio-­taped and transcribed in their entirety using a simplified transcription style. In terms of practicality, translations from Japanese to English were prepared for selected sections during the course of the analysis. The transcripts were translated by the researcher and were reviewed and cross-­checked by a bilingual colleague. Analytical model: incorporating the researcher’s reflexivity The analytical approach taken in the original study situates its philosophical approach to its analysis within the constructivist perspective on narratives (Gubrium and Holstein 2002; Kasper and Prior 2015; Mann 2011), where the focus is more on the how questions as opposed to the what questions. The ‘what’ questions include ‘what happened’ and ‘what were the experiences people had’, while the ‘how’ questions include ‘how do the participants position themselves while telling the stories?’ and ‘how does the interpersonal and/or social relationship shape the making of the stories?’ (Frost 2011). Obviously, any ‘analytical model’ will oversimplify the empirical practices of narrative analysis. I  do not intend to prescribe or endorse a certain model, but the Six Step Analysis Model (refer to Miyahara 2015 for a detailed account of the model) represents my attempt to take into account the content (what) and form (how) as well as the context

230  Masuko Miyahara where context can range from the micro-­local to the macro-­global. The purpose in providing the analytical framework is to offer a guide to how the analysis was carried out, and also, to give transparency to the process of attempting to offer some insightful order to the multiple accounts of human experiences that my participants brought to the research site. The precise framework used will naturally depend on the enquiry, but in any narrative analysis, the reflexive involvement of the researcher at various stages of the analysis and representation of the data is inherent (Miyahara 2017).

Place-­reflexivity in action: Megumi’s stories In this section, I offer an example of how reflexivity was practiced by thinking more consciously about how different sites contributed in promoting a deeper understanding of how learners adopted new identities as language learners to language users as they moved from their educational contexts in Japan to those abroad. Informed by place-­reflexivity, I ventured out to re-­examine the original data in order to explore the different ways learners experienced their learning by focusing on the changes in their identities from learners to users. Practicing place-­ reflexivity enabled me to understand the diverse ways in which place could be thought of more critically and purposefully, sometimes confirming or reinforcing the findings from the original study. At other times, it expanded and enhanced my thinking of the subject matter. For this chapter, I focus on one of my participants, Megumi, and her narratives to illustrate how, by engaging in place-­reflexivity, I was able to gather additional insights that allowed me to see how ELF identity was formed from a more detailed and nuanced perspective. In employing the six areas of place-­reflexivity presented earlier, I found that some features could occur simultaneously or that some could be more prominent than others, making it thus increasingly difficult to distinguish which feature is at play at a given time and space. Also, certain features do not exhibit their characteristics in a linear fashion, but they can overlap or co-­exist or are liable to reoccur. Therefore, my discussions of applying the features of place-­reflexivity to Megumi’s narratives will be holistic in the sense that they will be considered as a group of interrelated factors. Megumi in her new environment of a liberal arts college Megumi is a ‘jun-­Japa’, which is a jargon and an emic term used in the environment of this bilingual college that refers to L1 Japanese students who were educated in the Japanese educational system. The use of this term distinguishes them from other group members such as kikoku (returnee) or seputen (September entry students who come from educational institutions overseas or the international schools in Japan) or ryugakusei (international students who often have native-­like proficiency of English or are experienced ELF users). These terms were created and used by the students themselves since the inception of the university almost 70 years ago (see Iino and Murata 2013, 2016, for the use of similar terms in a different institution in Japan).

‘Place-­reflexivity’ as kaleidoscope  231 Although Megumi had been attending English conversation school since kindergarten and, therefore, was ‘exposed’ to English from an early age, English was basically a formal school subject associated with exams that had little functional purposes. Megumi in this context consistently refers to the need to learn English for exams and also emphasizes its instrumental potential for her future career:

Extract 1 If we study English, it would be useful to pass the entrance exams. English will also open up a lot of doors to different kinds of careers. So, I know it is important to study and get good grades for English. (Miyahara 2015: 113) Megumi was influenced by the official discourse that emphasized the importance of a school subject, which generated a strong sense of obligation to learn English. Megumi’s recollection of learning English at the secondary level, the pedagogical approach was typical, with an emphasis on language structure. Teachers would use government approved textbooks that followed a curriculum specified by the government. Many hours were allocated to language forms, including grammar and phrasal expressions, which were usually based on the ‘standard’ native speaker norms. Rote memory and pattern practices were the dominant instructional styles, coupled with another commonly used pedagogical method, the translation method. Megumi explained that ‘teachers used Japanese to translate everything: it was exam-­oriented English’. In this context, Megumi was positioned as ‘language learner’ of English with little awareness of viewing it as a ‘tool’ for communication as exemplified in one of her initial interview extracts: ‘To me, English was a language that gaijin (meaning ‘foreigners’ in Japanese) used, and not a tool to use to communicate with people’. Megumi’s views towards English and English language learning started to go through a major transformation in her new academic environment at university. Practicing place-­reflexivity, let us now pay more attention to the details of this site and contemplate on what and why this site is considered as ‘bilingual’. As Megumi states in Extract 2, there is indeed a sense of diversity especially in terms of the make-­up of the students: a mix of nationality, ethnicity, language, and culture. Even among the Japanese students, there are the returnees or students who have been educated in the international school system in Japan. Similarly, the faculty come from all over the world, from a diverse range of professional backgrounds with non-­Japanese faculty teaching Japanese culture or history. In such an environment, the medium of communication and/or instruction is mainly in English. Both English and Japanese can be heard all over the campus. There are a lot more opportunities for encounters of different values and ways of thinking compared to a more monolingual educational atmosphere. Two months after first being immersed in the English-­medium university environment, Megumi, notices the differences about her new learning context.

232  Masuko Miyahara

Extract 2 With overseas students, returnees, and people who are good at speaking English here on campus, you get a feeling that ‘English’ isn’t just in the textbooks. It’s real! You can see people using it. This was quite new to me. And, it’s fun to use English to communicate your ideas. At high school, the questions we were asked in our English classes were very simple; you could respond to them by just saying yes or no. But now, I have to think, and try to put my opinions into one coherent argument. It’s challenging, but interesting. (Miyahara 2015: 113) In my original study, I  used a very holistic descriptor, ‘bilingual environment’ in an attempt to emphasize the uniqueness and globalized nature of this institution. However, instead of presenting such a cursory view, foregrounding the site enabled me to think and interrogate the material and environmental characteristics a site offers, and the implications of the relationship between people and the research site, where interactions among people take place, that reveals fresh insights that had gone previously unnoticed. In addition, from Extract 2 we can see that Megumi also notices the change in the pedagogical approach and the changes made in her interactions with her teacher and classmates as a result of it. In fact, although the students in these first-­year ELP classes are mainly jun-­Japa students, English is the sole medium of instruction. As noted earlier, there is a focus on nurturing critical thinking as well as the accumulation of the English knowledge in these classes. The pedagogical approach is mainly in the form of group discussions where the students are given a topic to discuss and then presented in a form of presentations, which is subsequently consolidated in an academic paper. Compared to a lecture-­driven type of approach that is a more common form of teaching used in many language classes at the Japanese tertiary level, there is a lot more interaction with the teacher as well as their classmates in the classroom. However, in one of Megumi’s personal email exchanges with me, she ‘confesses’ that activities in this ‘new style of learning’ (learner centered) sometimes marginalized people. Those who are proficient in English gets to speak out more, while the less able ones keep silent, and aren’t given much chance to voice their thoughts. I wish the teacher could do something about this. (email from Megumi in York to the researcher) Megumi’s comments bring me to discuss one of the features of place-­reflexivity that conceptualizes the relationship that people have to places as constituting ‘a sense of place’. These are subjective emotions or feelings that represent the interactions of the affective dimensions between human and the research site. Referring to Relph (1976), Swaminathan and Mulvihill explicate that positive emotions can lead to a sense of belonging, while, on the other hand, negative dispositions can provoke a sense of isolation or a sense of being an outsider.

‘Place-­reflexivity’ as kaleidoscope  233 This is an interesting phenomenon to note when we consider how Megumi’s conceptualization of English expanded to include not only the Anglophone norms but also different varieties of ‘Englishes’ by observing how English was used by overseas students on campus as seen in the following extract, Extract 3. In a way, English became more personalized and she now felt ‘closer’ to the language because she was required to use it in a meaningful way:

Extract 3 I felt for the first time I have likings for English. It wasn’t anything formal. Neither was it difficult. It was something that was used in everyday life and something you use to communicate with people. I  felt an affinity towards English and also, I realized that people had different accents, and was okay. (Miyahara 2015: 120) Informed by the place-­reflexivity framework, we notice that, at the more macro-­ level of the institution, the campus atmosphere contributed in creating a sense of ‘internationality, multicultural’ community for Megumi, but, at the micro-­level of the classroom, her interlocutors were mainly L1 Japanese speakers with whom she was required to communicate in English. This illustrates how the material environment of the place assisted in creating this new view towards English as real and meaningful, and fun to use. It also exemplifies how the place interacted with the participant to create a new awareness towards English. That is, immersed in the English-­medium university environment, Megumi was now able to observe how English is used not only with native English but also with ELF speakers alike. This, in turn, helped her to generate a view of herself as an English user and envision herself functioning in an ELF community, which influenced the way she related to the world. This suggests how the place with its various material resources could interact to create new subject position or identity. With Megumi’s positive views and affinity towards English as well as her increasing self-­awareness as a user, it came as no surprise to find out that she had applied for the 6-­week study abroad program at the University of York in the UK. Megumi explains her decision as follows:

Extract 4 Now that I  feel comfortable with the language and have less fear about communicating with people who speak a different language, I  wanted to go abroad. That is, not just with the family on holidays with Japanese tour guides, but on my own to study. (Miyahara 2015: 122) Shifting to the scene on a different platform, namely, in the study abroad context, I will examine the findings of the original study by incorporating some of the components of place-­reflexivity in order to engage in reflexivity more intentionally.

234  Masuko Miyahara The study abroad program The summer program overseas, known by the acronym SEA Program, is a program for first-­and second-­year students that is designed to improve students’ language proficiency (rather than studying the subject content) while experiencing different cultures at their designated locations. All the programs at the sites have been developed in collaboration with the university, and they are credit-­bearing courses so that students are able to fulfill a portion of their ELP unit requirements during their 6-­week program. In the first interview upon her return from the study-­abroad program in York, UK, Megumi described her experience as ‘great’! She was not sure if it had a substantial effect in developing her linguistic performance, but the rewarding and enriching experience of interacting with people from different cultures was a great discovery for her as illustrated in this extract:

Extract 5 I really appreciated the talks with friends I made there. Especially with one particular Spanish student. We keep in touch through emails. I found out that although my English is not perfect, people will try to understand you. They don’t usually ridicule you because of your English. By asking them to repeat things, and rephrasing things, you can make yourself understood. You don’t have to be perfect. It’s okay to make mistakes. (Miyahara 2015: 122) At York University, the classes were generally kept small with about ten to fifteen students per class. As in most study abroad programs, the make-­up of the students was diverse in terms of nationality, ethnicity, and culture. The classes were also streamed by a placement test, and classes were organized so that, in the mornings, they were oriented towards building up students’ knowledge of English and practicing the four skills; in the afternoons, the approach was more communicative, open-­ended, and project-­based. Planning a weekend trip to London was one such project, which would require the students to work in groups to make the necessary travel arrangements and plan sightseeing trips using on-­line resources. Upon their return from such trips, students were asked to make a presentation of their preparations prior to their trip and to report on the trip upon their return. Not surprisingly, such hands-­on projects were very appealing to the students, as Megumi explains here:

Extract 6 These afternoon projects were fun. I also liked to do volunteer work in the stores in the high streets. We rotated stores every afternoon for about a week. This was another highlight of my stay in York. Yes, you were communicating! It was real. It was great! (Miyahara 2015: 123)

‘Place-­reflexivity’ as kaleidoscope  235 These were the positive aspects of Megumi’s narratives of her studies in York, which had propelled her to take more initiatives towards her learning. For instance, as she was interested in acting, she would search websites and seek information on the most sought-­after shows in that area. Arranging for tickets, finding directions to the theatre on her own were all enjoyable ‘tasks’ for Megumi. As she noted in one of her weekly email exchanges with me, ‘I am learning and using English. It feels great!’. On the other hand, Megumi encountered several disappointments and challenges while at York. Contrary to what Megumi had expected, the students in her class were L2 English learners and users. This is obvious since the program is targeted for international students, but it appears that Megumi thought she would be engaging with L1 English speakers (as shown in her email): I had expected to myself to be talking to native English speakers, but all my classmates were from different parts of the world. Although this was nice, but it would have been nice if I could have interacted more with the British from my own generation. (email to the author from York) However, Megumi could get a glimpse of the British people and culture through her middle-­class host family from an Italian heritage. While enjoying herself with her new acquired identity as an ELF user communicating with other ELF users in her class, Megumi had at the same time, the desire to interact with native English speakers. The other difficulty was related to the variety of interactional patterns in class group discussion. In Japan, her classroom interactions were mainly with her Japanese counterparts; in comparison, at York University, she was immersed in an ELF environment where English served as the medium of communicative transactions and interactions. In Megumi’s class at York, students came from a wide range of places each with their different cultural profile such as Kuwait, China, Spain, Italy, and Germany. One-­third of the students in Megumi’s class were from Kuwait, but Megumi found these students ‘quite vocal sometimes even to the point of being obnoxious’ (from Megumi’s weekly journal during her study abroad). Her comments in Extract 7 illustrate her reservations:

Extract 7 I had a hard time with these students from Kuwait and China (in my class). In group discussions, they were so loud. The Chinese would try to stop the Kuwaitis sometimes. But it didn’t help much. They just kept going. And even if I had something to say, by the time it was my turn I would have forgotten what to say; or, many times, by then, the discussion would have headed in a different direction. . . . Although the teacher noticed it and tried to facilitate things. Their English was not much different from ours, but they spoke more fluently. Their behavior puzzled me. . . . (Miyahara 2015: 123)

236  Masuko Miyahara One component to place-­reflexivity is to critically approach the site by contemplating how places are organized and how they could promote or impede social interactions as well as to contemplate on what this means in terms of the relationship of the participant to the site. The interview extracts earlier brought to the surface how Megumi faces difficulties in acknowledging herself as a language user. Back in Japan, she experiences for the first time what it is like to see herself more of a language user rather than imbued on identity as a language learner. However, even in this environment, the dominant language and culture is Japanese, using English mainly with her Japanese cohort. In contrast, at York, Megumi is genuinely interacting in the ELF environment. She is not a bystander anymore, but a legitimate ELF communicator actively functioning in ELF transactions. As we have observed, she had experienced both the positive and negative emotional forces of developing identity as a language user. As her words towards the end of Extract 7 indicates, we find her wondering about the reasons why these students from the Middle East exhibit, from her point of view, such forceful behavior. However, this to me, suggests that Megumi is trying to understand her interlocutors by showing respect for cultural diversity, and not merely as a criticism towards unfamiliar culture. Some afterthoughts on practicing place-­reflexivity In terms of the various features of place-­reflexivity, in my previous study, I noticed that, although my discussions had covered, more or less, most of these areas, it was done unconsciously and without much heed and awareness towards how place could be considered as an active agent in the research encounter. In activating the rather passive category of the research site and paying more attention to the material resources that place has to offer could provide deeper insights for the research. The need for more explicit focus of place is crucial. For instance, upon entering college, its multilingual and academic environment helped Megumi, a jun-­Japa, become aware of herself as a language user. The physical characteristics of the campus such as its location away from the central Tokyo, to the suburbs, the greenery and the vast space was very much unlike the university campus in the center of the city and often gave the impression of ‘foreignness’. Such environmental characteristics coupled with the international atmosphere where students would not only converse in Japanese, but in English in different varieties have enabled Megumi to observe first-­hand, ‘English in action’. In addition, her classes and tutorials were conducted all in English. The majority of the faculty were non-­Japanese, both English and non-­native English speakers, who used English as the common language for communication. The pedagogical approach was based on the ideas of liberal arts education where critical thinking is at the base. Such material and environmental characteristics appeared to facilitate Megumi’s awareness of herself as an English learner transforming herself to additional identity of a user. In my original study, the construction of Megumi’s identity from learner to user was based on the grounds of ‘exposure’ to the bilingual environment of the research site. However, incorporating several aspects of place-­reflexivity has enabled me to see in a more explicit manner the details of the site, and the significance and meaning place has in containing diverse types of interrelationship among different users of English.

‘Place-­reflexivity’ as kaleidoscope  237 Megumi’s study abroad experiences at York consolidated her identity as language user. Through both the positive and negative experiences at this site, she consolidates here identity as language user. She develops understandings that the Anglophone form of English is not the only norm and that there could be varieties in English; linguistic deficiencies are regarded not as negative, although she also simultaneously shows her desire for communicating with native English speakers. In addition, she also encounters negative disposition in the process of forming her language user identity, but such negative experiences led to her awareness to develop a respect towards different cultures as implicated from her interview extract about her classroom interactions with the Arab students. She also experiences a sense of ownership towards English as she realizes the potentials of her communicative capabilities (Widdowson 2016) to make English her ‘own’ by having the opportunities to use her English in ‘genuine’ ELF contexts. In my original study, the developments towards English ownership was visualized as a more linear process progressing from one to the other: that is, from language learner, to language user, and then, eventually, building a sense of ownership towards the language. Employing the place-­reflexivity framework and also an ELF perspective helped me to look the phenomena form more nuanced and multidimensional perspectives, and now, I conceptualize the process to be represented as a tri-­level concentric circle where the three components do not progress in a linear manner, but could sometimes going back and forth (depending on the context), eventually building up on one another as seen in Figure 13.1.

宒宺宱宨宵家宫宬害季宷宲宺室宵宧家季 宷宫宨季宯室宱宪宸室宪宨季

宏室宱宪宸室宪宨季宸家宨宵

宏室宱宪宸室宪宨季宯宨室宵宱宨宵季

Figure 13.1 Conceptualization of language learner, to user, and developing a sense of ownership

238  Masuko Miyahara Place-­reflexivity has illustrated that the research practice is not merely a dialogue between the researcher and the researched, but should be conceived as a ‘trialogue’ that includes the researcher, the researched, and the research site where the methodology is being practiced. It is also crucial that we put prominence on the interrelationship between these three actors and what they produce. Place could include unique arrangements of people, landscape, buildings, etc. or the non-­human actors as well. Place in this context ascribes agency to non-­human actors, and in some cases, they are as important as human actors (Hinchcliffe et al. 2015). The important point is to grant agency to place and to acknowledge its influence in the production of knowledge.

Final remarks and limitations Overall, place-­reflexivity has offered me a new lens in which to ponder reflexively on my research. However, one of the difficulties of incorporating place-­reflexivity to my original study was how to conceptualize place in creating virtual spaces or liminal spaces such as the imagined communities (Kanno and Norton 2003) that learners could aspire (Dörnyei 2009; Miyahara 2015). How does the various aspects of place interact with the participant, the researcher and the context in creating such a virtual space? This is certainly another area that needs further exploration. In the meantime, as ELF research firmly establishes itself in the field of applied linguistics, approaches to ELF research will certainly diversify and increase as research inquiries expand. Particularly, studies that attempt to explore ELF users’ voice, attitudes, and identities should critically examine their methodology and methods and take heed to be mindful of the researchers’ involvement in the entire research process. It is my hope that this chapter may serve as a useful guideline to illuminate such an approach, contribute in delineating the issues at stake, and make researchers aware of the challenges that they may encounter.

References Agee, J. 2002. Winks upon winks: Multiple lenses on setting in qualitative educational research. Qualitative Studies in Education 15, 569–585. Agnew, J.A. and J.S. Duncan (eds). 2014. The power of place (RLE social & cultural geography): Bringing together geographical and sociological imaginations. London: Routledge. Anderson, J., P. Adey and P. Bevan. 2010. Positioning place: Polygic approaches to research methodology. Qualitative Research 10, 589–604. Baker, W. 2009. Language, culture and identity through English as a Lingua Franca in Asia: Note from the field. The Linguistic Journal 4, 8–35. Barkhuizen, G. 2013. Introduction: Narrative research in applied linguistics. In G. Barkhuizen (ed), Narrative research in applied linguistics. Cambridge: Cambridge University Press, pp. 1–16. Berger, R. 2015. Now I see it, now I don’t: Researcher’s position and reflexivity in qualitative research. Qualitative Research 15(2), 219–234. Bhabba, H.K. 2009. In the cave of making: Thoughts on third space. In K. Ikas and G. Wagner (eds), Communicating in the third space. New York: Routledge, ix–xiv.

‘Place-­reflexivity’ as kaleidoscope  239 Block, D. 2006. Identity in applied linguistics. Where are we? In T. Omoniyi and G. White (eds), The sociolinguistics of identity. London: Continuum, pp. 34–49. Block, D. 2007. Second language identities. London: Continuum. Bruner, J. 1990. Acts of meaning. Cambridge, MA: Harvard University Press. Canagarajah, S. 1996. From critical research practice to critical research reporting. TESOL Quarterly 30(2), 321–331. Clandinin, D.J. 2007. Handbook of narrative inquiry. Thousand Oaks, CA: Sage Publications. Clandinin, D.J. and F.M. Connelly. 2000. Narrative inquiry: Experience and story in qualitative research. San Francisco, CA: Jossey-­Bass. Connelly, F.M., and Clandinin, D.J. 1990. Stories of experience and narrative inquiry. Educational Researcher 19(5), 2–14. Creswell, J.W. 2014. A concise introduction to mixed methods research. Thousand Oaks, CA: Sage publications. Creswell, T. 1996. In place/out of place: Geography, ideology and transgression. Minneapolis, MN: University of Minnesota Press. Dörnyei, Z. 2009. The L2 motivational self system. In Z. Dörnyei and E. Ushioda (eds), Motivation, language identity and the L2 self. New York: Multilingual Matters, pp. 9–24. Finlay, L. 2012. Five lenses for the reflexive interviewer. In J. Gubrium, J. Holstein, A. Marvasti and K. McKinney (eds), The Sage handbook of interview research: The complexity of the craft. Thousand Oaks, CA: Sage Publications, pp. 317–332. Frost, N. 2011. Qualitative research methods in psychology. London: Open University Press. Gubrium, J.F. and J.A. Holstein. 2002. Handbook of interview research: Context and method. London: Sage Publications. Guillemin, M., and Gillam, L. 2004. Ethics, reflexivity, and “ethically important moments” in research. Qualitative Inquiry 10(2), 261–280. Hinchcliffe, S., M. Kearnes, M. Degen and S. Whatmore. 2015. Urban wild things: A cosmopolitical experiment. Environment and Planning D 23, 643–658. Iino, M. and K. Murata. 2013. We are Jun-­Japa: Dynamics of ELF communication in an English medium academic context. Waseda Working Papers in ELF 2, 84–100. Iino, M. and K. Murata. 2016. Dynamics of ELF communication in an English – medium academic context in Japan: From EFL learners to ELF users. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts. Oxford: Oxford University Press, pp. 111–131. Jenkins, J. 2007. English as a Lingua Franca: Attitude and identity. Oxford: Oxford University Press. Kanno, Y. 2003. Negotiating bilingual and bicultural identities: Japanese returnees betwixt two worlds. Mahwah, NJ: Lawrence Erlbaum. Kanno, Y. and B. Norton. 2003. Imagined community and educational possibilities: Introduction. Journal of Language, Identity and Education 2(4), 241–249. Kasper, G. and M. Prior. 2015. Analyzing storytelling in TESOL interview research. TESOL Quarterly 49(2), 226–255. Kim, J. 2016. Understanding narrative inquiry. Thousand Oaks, CA: Sage Publications. Mann, S. 2011. A  critical review of qualitative interview in applied linguistics. Applied Linguistics 32(1), 6–24. Mann, S. 2016. The research interview: Reflexive practice and reflexivity in research process. Basingstoke: Palgrave Macmillan. Miyahara, M. 2015. Emerging self-­identities and emotions in foreign language learning: A narrative oriented approach. Bristol: Multilingual Matters.

240  Masuko Miyahara Miyahara, M. 2017. Using a narrative approach in ELF research: Developing a reflexive framework. Waseda Working Papers in ELF 6, 91–109. Miyahara, M. 2019. Methodological diversity in emotion studies: Reflexivity and identities. Journal for the Psychology of Language Learning 1, 83–105. Murata, K. 2016. Introduction: ELF research, corpora and different approaches to data. Waseda Working Papers in ELF 5, 1–12. Nagar, R. and S. Geiger. 2007. Reflexivity and positionality in feminist fieldwork revisited. In A. Tickell, E. Sheppard, J. Peck and T. Barnes (eds), Politics and practice in economic geography. London: Sage Publications, pp. 267–278. Nogami, Y. 2016. A narrative approach to ELF-­focused research: Diary study for investigating identity construction of Japanese ELF speaker. Waseda Working Papers 5, 136–152. Nogami, Y. 2019. Identity and pragmatic language use among East Asian ELF speakers and its implications for English-­medium education. In K. Murata (ed), English-­medium instruction from an English as a Lingua Franca perspective. Abingdon, Oxon: Routledge, pp. 176–198. Norton, B. 2013. Identity and language learning: Extending the conversation (2nd ed.). Bristol: Multilingual Matters. Polkinghore, D.E. 1995. Narrative configuration in qualitative analysis. International Journal of Qualitative Studies in Education 8(1), 8–25. Prior, M. 2016. Emotions and discourse in L2 narrative research. Bristol: Multilingual Matters. Relph, E. 1976. Place and placeness. London: Pion. Riessman, C. 2008. Narrative methods for human sciences. Newbury Park, CA: Sage Publications. Riessman, C. 2012. Analysis of personal narratives. In J. Gubrium, J. Holstein, A. Marvasti and K. McKinney (eds), The Sage handbook of interview research: The complexity of the craft. Thousand Oaks, CA: Sage Publications, pp. 367–380. Roulston, K. 2010. Reflective researcher: Learning to interview in the social sciences. Thousand Oaks, CA: Sage Publications. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press. Sin, C.H. 2003. Interviewing in “Place”: The socio-­spatial construction of interview data. Area 35(3), 305–312. Soja, E.W. 2010. Seeking spatial justice. Minneapolis, MN: University of Minnesota Press. Stanley, L. 1992. The auto/biographical, vol. 1: The theory and practice of feminist auto/ biography. Manchester: Manchester University Press. Swaminathan, R. and T. Mulvihill. 2019. Learning to “site-­see”: Place reflexivity as a methodological tool for qualitative researchers. International Journal of Qualitative Studies in Education 32(8), 1–16. Takino, M. 2016. Using narrative inquiry in ELF Research: Exploring BELF users’ perspective. Waseda Working Papers 5, 117–135. Tilley, C. 1994. A phenomenology of landscape: Places, paths and monuments. Oxford: Berg. Turner, V. 1969. The ritual processes: Structure and anti-­structure. London: Routledge and Kegan Paul. Virkkula, T. and T. Nikula. 2010. Identity construction in ELF contexts: A case study of Finnish engineering students working in Germany. International Journal of Applied Linguistics 20(2), 251–273. Widdowson, H. 2016. Competence and capability: Rethinking the subject English. In K. Murata (ed), Exploring ELF in Japanese academic and business contexts. London: Routledge, pp. 213–223.

14 Understanding the ELF phenomenon through narrative inquiry A diary study on identities of Japanese ELF users1 Yoko Nogami Introduction Since Seidlhofer et al. (2006: 21) called for more ‘qualitative studies with a strong ethnographic element’, research investigating beyond English as a lingua franca (ELF) linguistic description is on the increase, facilitating our understanding about ELF speakers themselves, the environment they are in, and sociocultural factors that may influence ELF communication. For instance, a narrative-­based research study by Baker (2015) demonstrated the fluidity and complexity of culture in ELF communication, and Kalocsai (2014) revealed how ELF plays a role as part of social practice in the development of a specific community of practice based on an ethnographic, narrative-­driven study. In particular, a narrative approach has been taken up for the investigation of identity issues of ELF speakers (e.g., Baker 2009, 2011, 2015; Guido 2019; Jenkins 2007; Kalocsai 2014; Ng and Dodge 2015; Nogami 2020; Seidlhofer 2011). Due to the increased numbers of the findings of ELF research with a narrative approach, we now appreciate that ELF is used not purely for communicative purposes, but that it needs to be treated just as communication in any languages through which speakers develop new identities and negotiate different identities in a globalised society. Recent and rather rapid developments in ELF research have revealed the linguistic nature of ELF, particularly in lexis, lexicogrammar, pronunciation, and pragmatics (see Jenkins, Baker, and Dewey 2018 for a comprehensive introduction to the field of ELF study). The wide range of linguistic features of ELF is often characterised as a manifestation of ELF speakers’ creativity (e.g., Cogo 2009; Kalocsai 2009; Pitzl 2009, 2012; Seidlhofer 2009) and flexibility, which form part of their accommodating and collaborative behaviours during interactions. These linguistic features are also said to be in use in order to ensure mutual understanding among interactants (e.g., Kaur 2009, this volume; Pullin Stark 2009) and sometimes to signal the speaker’s cultural identity. For example, strategies such as code-­switching (Cogo 2009; Kalocsai 2011, 2014; Klimpfinger 2009), using idioms and metaphors (Fiedler 2011; Pitzl 2009), and integrating first language (L1) communicative norms into ELF conversation are linked to a connection with the speaker’s cultural identity (Pölzl and Seidlhofer 2006). We now know quite well about what ELF speakers do linguistically and what these linguistic features may

242  Yoko Nogami function as, but we need to explore more where ELF speakers are situated and why they perform linguistically as discussed in descriptive ELF research through exploring the process of ELF users’ identity construction in order to deepen our understanding further on ELF phenomena. In this light, this chapter will present how narrative inquiries have enriched our insights to ELF communication by reviewing previous studies as well as drawing examples from my own research study, which employed longitudinal diary observation as a narrative method for investigating ELF users’ identity. The chapter is composed of two parts. In the first part, a review of several ELF-­focused, qualitative studies discussing issues relating to ELF speakers’ identity will be presented. In the second part, findings of a case study in which I investigated identities of Japanese ELF users will be presented.

Understanding the ELF phenomenon through a narrative approach Narrative inquiry is a discipline within the field of qualitative research. It relies on the actual words spoken or written by individuals (e.g., diaries, field notes, interviews, and autobiography), by focusing on the stories of lived experience and reconstruction of the individual’s experience as told by the individuals (Clandinin and Connelly 2000; Clandinin 2007; Etherington 2007). Narrative inquiry is a research method to systematically gather and analyse the narrative data (i.e., the stories), but also the phenomenon itself under study that represents people’s stories as told by individuals (Etherington 2007). The emphasis in the narrative approach is on both how the stories are narrated as well as what stories are told. Narrative inquiry considers the way people create subjective meaning in their lives and the manner a sense of self and identity are negotiated as the stories unfold (Clandinin and Connelly 2000; Clandinin 2007). ELF-­focused research with a qualitative approach expands our insights on ELF communication from ELF speakers’ individual standpoints in the social context. This section will review previous ELF-­related research that employed a narrative approach, which facilitates our understandings of ELF as social phenomenon in a certain emerging community and/or TIG (Transient International Groups, see Pitzl 2018) and the underlying process of ELF speakers’ identity construction. Culture, language, and identity in intercultural communication through ELF One important concept when exploring ELF communication is the notion of culture and its relationship with language. According to the poststructuralist perspective, culture is neither a pre-­conceptualised nor defined notion, but something constructed and reconstructed during the course of social interactions (Baker 2009, 2015). In terms of intercultural communication through ELF, in which people with diverse lingua-­cultural backgrounds participate, we can no longer view language and culture as being directly linked, or based solely at a ‘national’ level:

Understanding through narrative inquiry  243 for instance, the English language as a representation of US/UK culture (Baker 2015). Baker also maintains that ELF communication should not be regarded as being neutral from any form of culture (Baker 2009, see also Sung 2014). Thai ELF speakers in Baker’s (2009) ethnographic research study told of their ambivalent attitudes towards English language use and culture in their intercultural communication through ELF. They often valued local varieties of English and did not want to give away their L1 or Thai identity in their ELF interaction. Meanwhile, they set limits on how much they follow native speakers of English (NSE) norms and changed their behaviour to fit their own context when speaking English (Baker 2009). This indicates that local and/or national identities can be a salient feature of ELF communication. ELF speakers occasionally desire to retain and express their local-­and/or national-­based identities (Baker 2009; Sung 2014). Similar findings are also reported elsewhere: ELF speakers desire to preserve L1 identity, for example, through accent (Baker 2009; Jenkins 2007; Sung 2014) and pragmatic language use (Nogami 2019; Sung 2014), instead of wishing to be identified with NSE when using ELF. ELF speakers’ English-­speaking identity operates alongside or in addition to their L1/first culture (C1) identity, especially in order to fulfil needs for self-­identification through ELF. Baker further states that ELF gives speakers a means to express themselves in a way that is more open and free from particular cultural constraints (Baker 2009, see also Nogami 2011; Sung 2014). Furthermore, Baker argues that ELF speakers wish to belong to a wider global community, which is not directly relevant to any particular lingua-­culture, where they can express their multicultural and multilingual-­selves (Baker 2009: 29). This multicultural-­self represents hybridity, which is sometimes local, and sometimes global. ELF speakers can take a situationally specific ‘third place’ (Kramsch 1993), and they can move in-­between local, national, and global cultures (Baker 2009: 11). In sum, Baker (2009: 18) supports that ELF embodies the meanings of its users’ surrounding context and culture with reference to the complexities of global and local use of English. This perspective of culture and identity that represents a dynamic, fluid, and sometimes contradictory nature, is important in recognising how ELF users communicate ‘in diverse and transitory’ ELF situations (Baker 2009: 9). However, such flexible notion of culture and diversity of ELF communication is sometimes not appreciated fully by English as a foreign language (EFL) teachers. Their identity as an EFL teacher and an ELF user seems to appear contradictory due to the culture surrounding English language teaching, which constrains their attitude toward ELF, to which I now turn. EFL teacher identity and attitude toward ELF Conflicting and ambivalent attitudes toward the English language is ubiquitous, especially among EFL teachers, and they are reported to show negative attitudes toward the legitimacy of ELF (e.g., Jenkins 2007; Pedrazzini and Nava 2011; Phan 2008). Views of EFL teachers on the language often reflect their professional role, which is guided by the needs and desires of their students, the students’ parents, as

244  Yoko Nogami well as the national curriculum, which traditionally encourages the acquisition of Standard English (Seidlhofer 2018). They may also be influenced by local expectations of a teacher’s role, for instance as a role model of social morality (Phan 2008). Their role as EFL teachers inevitably creates inconsistency in their attitude toward ELF and friction in their identity between as an EFL teacher and an ELF user. The most extensive ELF-­focused study on teachers’ identities is one conducted by Jenkins (2007) on accent-­related identities. Seventeen EFL teachers of various L1 backgrounds were interviewed about their personal history and experience. Their narratives, which revealed their aspirations to reach so-­called NSE-­level proficiency, demonstrated their desire of being acknowledged as competent professionals. Some participants showed appreciation of different varieties of English and referred to their own accent as part of their L1/C1 identity. Others displayed awareness of different varieties of English both in the inner and outer circles but did not seem to enjoy contemplating different varieties as possible emerging models other than the ‘standard NSE model’. In a similar vein, the participants disclosed that it was acceptable for others to have their own accents in ELF settings but not for themselves, which demonstrates their conflicting attitude toward ELF. Whereas, slightly more positive attitudes were reported when they spoke as an ELF speaker and not as an EFL teacher. That is, for some, retaining their own L1-­influenced accent was important for their identity. Others said they disliked their own accents but still liked to retain them for their identification. However, they again expressed a strong desire to have an NSE accent, which indicates there was more of a tolerance of, rather than a fondness for, L1-­influenced accents (Jenkins 2007: 186–189). Similar findings, but with more positive attitudes toward ELF accents were reported in a smaller scale study conducted by Pedrazzini and Nava (2011) on teachers’ identities in Europe. The participants’ narratives reveal that having their own L1-­influenced accents signals what they have in common as ELF speakers, which creates a shared commonality (Nogami 2019) and a shared sense of community (Pedrazzini and Nava 2011). Identity continues to play a critical role in English speakers’ attitudes toward their accents and thus toward ELF in the future (Jenkins 2007), especially because accents or prosodic features are the most obvious and recognizable aspect of speakers’ linguistic features. It is important for EFL teachers along with learners to be exposed more to different varieties of English as well as ELF communication, which can lead to better understanding of diversity of ELF communication and how speakers identify themselves in English-­speaking settings (Jenkins 2007; Iino and Murata 2016). A narrative approach to ELF-­focused research helps make sense of the complexity and fluidity of ELF users’ lives, in particular, with a view to the manner of identity negotiation in various social contexts in which ELF users are set. Transition of identities from a learner to an ELF user Less like EFL teacher identity, ELF users’ identities tend to flourish as they are exposed to intercultural encounters through ELF. In Virkkula and Nikula’s research (2010), the Finnish students’ narratives demonstrated changes of their identities before and after their ELF-­mediated cross-­cultural experiences during

Understanding through narrative inquiry  245 study abroad. Before going abroad, the students focused on the deficiency in their English language skills by comparing themselves to NSE norms of correctness. However, after completing a study abroad programme in Germany, where the medium of communication was English, their concerns for correctness of forms diminished, and their views changed to a celebration of success in English communication for daily survival. In sum, the study reveals that the English learners gain a new mindset for identity-­building through ELF encounters, which actually enables the participants to construct new identities as users of ELF rather than as unsuccessful learners of English (Virkkula and Nikula 2010: 268–270). Phan’s (2009) study provides a detailed account of identity construction among Asian students through ELF encounters.2 The ethnographical study on eight Asian international students studying in a Thai university reveals the process of them taking ownership of English for their own sake. The participants’ narratives disclose how fruitful ELF encounters fostered their sense of achievement as successful English language learners and intercultural communicators as well as their legitimacy as ELF speakers. Through their experience in Thailand, they felt that they were in control of English, not controlled or colonised by English whilst they adapted their English to suit their local needs. It also allowed them to realise that they are multilingual people who speak proficient English, as well as mediators of their L1/C1. Phan’s study depicts such processes of the students developing a sense of ownership of ELF. The participants acknowledged the English language as a provider of opportunities for intercultural communication and as a gateway to get to know others in the globalised world. Through exposure to successful intercultural communication through ELF, ELF users’ identities can thrive while building a sense of global citizenship, a sense of belonging to a local community, and an appreciation of their identities as multilingual/multicultural people (Kalocsai 2014; Nogami 2011; Phan 2009; Virkkula and Nikula 2010). Again, a narrative inquiry represents people’s stories of their lives and experiences as told by them, revealing, for example, a state of transition of identification as from learners to ELF users and an emergence of sense of belonging to a local community as well as the wider global world. Thus, a narrative-­driven ELF study helps us understand the intricate process of ELF users’ identity construction and the multiplex nature of identities. Communities of practice and ELF In ELF communication where language and culture are not inexorably connected (Baker 2009; Sung 2014), the sociolinguistic notion of a speech community, which embodies uniformity and regularity, becomes less relevant (Seidlhofer 2011). However, ELF speakers often create a form of culture or ‘space’ through ELF that can be physically (Kalocsai 2014) as well as virtually and globally networked3 (Yashima 2009). Employing the social theory of Communities of Practice (CoP)4 (Lave and Wenger 1991), some ELF researchers have investigated the manner in which ELF is used to index and construct membership in ELF-­related communities, such as Erasmus exchange students in Central Europe (Kalocsai 2014), business managers in Germany (Ehrenreich 2009), and English language teachers in Italy (Pedrazzini and Nava 2011).

246  Yoko Nogami Kalocsai’s (2014) study examined the manner in which her participants created a community called the ‘Erasmus family’, which they belonged to and identified with. She observed the way in which the participants developed their own practices, through mutual engagement (Wenger 1998) such as parties and trips, to build and maintain a friendship-­oriented community. This construction and maintenance of the community collectively became their shared goal, i.e., joint enterprise (Wenger 1998). Embarking on the joint enterprise, Kalocsai’s participants developed various linguistic practices through ELF such as repetition, code-­switching, and use of humour to build solidarity and rapport among themselves. These linguistic practices were created jointly and exploited collectively among themselves and became their shared repertoire (Wenger 1998) in the community of practice. The Communities of Practice approaches to ELF, such as Kalocsai’s (2014) ethnographic study, enable descriptive ELF studies and social theory to be merged. Linguistic practices among ELF speakers can be connected to the understanding of what linguistic practices mean to ELF speakers and negotiation of their identities. More research of this kind that combines the interactional approach to linguistic natures of ELF and the narrative approach to ELF speakers’ identity and the environment in which they are situated will be indispensable for the further development of ELF research (see also Miyahara, this volume). ELF-­focused, narrative research reveals ELF users’ perceptions and attitudes toward ELF from their own perspectives, leading to better understanding of culture and identity in ELF settings (e.g., Baker 2009); the dynamic and fluid nature of ELF speakers’ identities (e.g., Jenkins 2007; Phan 2008); and their belonging to and formation of an ELF-­resourced community (e.g., Kalocsai 2014). Through their stories as told by ELF users, narrative inquiry facilitates capturing complexity, subtlety, and profundity of the ways in which ELF users organise and obtain meaning from their lived experiences so that we can learn from it. In the next section, my research study employing a diary method will be presented as another example of ELF-­focused narrative research. The study explored first-­hand intercultural experiences of Japanese ELF users and the process of their identity construction. A diary study provides avenues to explore detailed descriptions of an ELF user’s immediate intercultural experiences and their own reflection on the encounters. The experiences can include not only a face-­to-­face interaction but also a virtual interaction such as the one through email exchanges and messaging on social media. A diary study allows access to potentially insignificant incidents that research participants come across which may be difficult to be inquired about at the later stage by other qualitative methods such as interviewing.

A diary study on understanding identities of Japanese ELF users as a form of narrative inquiry The study The study was conducted to investigate how Japanese ELF users construct and negotiate their identities through their experiences in intercultural communication. The longitudinal diary study was one of the qualitative methods that was

Understanding through narrative inquiry  247 implemented in the author’s original investigation (Nogami 2011), and was designed to explore in-­depth insights of the identities of the research participants based on their narratives about the daily experiences and opportunities in communicating through ELF. I was interested in hearing the participants’ own words on how they perceived their own thoughts and feelings relating to their day-­to-­day intercultural encounters through ELF. Diary observation as a qualitative research method Alaszewski (2006: 1) defines a diary as ‘a document created by an individual who has maintained a regular, personal and contemporaneous record’. A number of applied linguistics studies used diaries as a means to investigate the process of second or foreign language learning from an emic perspective and as a source for learners’ perspectives on their experiences of language learning and use (e.g., Bailey 1983; Jackson 2008; Norton Peirce 1993). Such diary studies enable us to reflect on 1) the diarists’ language learning experience both in the classroom and outside the classroom (e.g., workplace, home, and social life); 2) their thoughts, feelings, and perceptions concerning different situations of English language use with any speakers of the language; and 3) the opportunities they have to interact with various speakers of the language in daily life (Norton Peirce 1993: 37–42). A diary study can be used to acquire an individual’s stories of their lived experiences as narratives. It can be particularly useful for studying the impact of social structures on an individual and how that relates to one’s identity construction (Bartlett and Milligan 2015). Narrative inquiry attaches importance to context and pursues understanding phenomena, such as ELF communication, through the meanings that people ascribe to them. A diary study allows space for a diarist to record a more detailed, temporal experience, thus it often grants researchers a deeper understanding of an individual’s actions, experiences, thoughts, and emotion around a particular topic. Accordingly, in the case of the current research, it can capture the process of ELF users’ identity construction in a more detailed manner. Follow-­up email interviews further assist exploring the related issues in more depth, which prove useful for capturing the meaning and weight that an individual attach to different events and experiences in their lives (Bartlett and Milligan 2015; Corti 1993). Participants and method The participants of the original diary study were nine Japanese users of English, who all reported that they used English regularly outside of formal classrooms. Three were living in Britain (all female, in their mid-­20s to mid-­30s), and six were living in Japan (five females and one male, early 20s to mid-­30s) at the beginning of the diary study.5 The diarists were asked to record occurrences, actions, and events that were conducted in English and that surprised them or struck them as unusual and to observe critically such incidents. In particular, they were asked to investigate any opportunity of interaction in English with any English speakers and also to reflect

248  Yoko Nogami critically on their engagement in the interaction. They were also asked to write their feelings and thoughts on this matter.6 The diarists sent me their diary entries each time as email attachments. Most of the time, the participants were inquired further to clarify or expand on the issues that had been raised in the diary entries. I often asked questions such as ‘Could you elaborate more on what you felt in this event?’ and ‘Could you clarify what you wrote?’. The file with my questions and comments was sent back to the original diarist by email, and then the diarist responded to my questions. This email exchange procedure continued throughout the course of the diary study. This kind of follow-­up interview about diary entries can be one of the most reliable methods of obtaining essential research information, which enables us to clarify and enrich our data (Corti 1993). Regular email correspondence also enables the maintenance of recurrent message exchanges and can thus be beneficial not only for accessing more insightful information but also for encouraging and motivating the participants to continue their dialogues with the researcher. Moreover, regular exchanges of messages can result in the establishment of positive interpersonal relationships between the researcher and the participants. The diarists were asked to complete a diary entry at least once a week for the duration of 6 months. Generally, I  received their diary entries weekly or fortnightly. The study lasted for 6 to 8 months, depending on the respective participants’ frequency of diary entries. At the end of the diary study, I  gathered 130 diary entries in total for analysis, with an individual’s number of diary entries ranging from 12 to 20. The data was analysed qualitatively using an interpretive approach (Miles and Huberman 1994) through two lenses: a cross-­sectional study and a case study. Due to the limited space of this chapter, only one case study will be presented in this chapter.

A case study and discussion In this section, findings and discussion from a case study of Hana (pseudonym) will be presented. It includes a short biography and narratives from Hana’s diary accounts that represent the most salient and robust matter for her identities relating to ELF communication and how she sees and positions herself as an ELF user in the situated contexts. Hana’s case study was chosen because it appeared relevant to various aspects which have recently gathered much attention in ELF research, such as ELF communication in a business context, participation in a virtual community through ELF use, and negotiation of identity. Hana’s story: imagined community of ELF users in a business context Hana is in her late 20s and works in an office of an international food trading company in Japan. She has been a keen English language learner especially throughout and after her undergraduate studies. She belonged to the English Speaking Society at university and had regular meetings with a university English language

Understanding through narrative inquiry  249 instructor along with like-­minded friends outside formal curricula to practice English. During her undergraduate studies, she participated in a 3-­month English language-­learning programme at a British university as a member of a group sent from her university. A year later, she went to an American university for 10 months as an exchange student, which she thoroughly enjoyed. It was thus natural for her to look for a job where she could employ her English language skills after graduating from university. When the diary study started, she had already worked for the international food trading company for about 5 years. Her diary entries mainly depicted her daily email exchanges in English used as a lingua franca and her occasional face-­to-­face interactions through ELF with business associates from all over the world. Through her diary entries, we can witness her learning trajectory of the pragmatic aspects of language use through ELF in a business context. It involves her realisation of and resistance to others’ pragmatic language use, in particular, for building a rapport, and her effort to create opportunities for shared practice with others, indicating her participation in an imagined community (Anderson 2006; Kanno and Norton 2003) of business ELF users. In Excerpt 1, she shows her resistant feelings to the message she received from a business partner because it sounded business-­like.

Excerpt 17 I received a reply from the Italian man (50s) with whom I  exchange messages about twice a month. Normally, he would start a message with a greeting such as ‘Hello. How are you?’, but [this time] he got down to business straight away, and wrote ‘Attached is the document you required’. I  wondered why. Perhaps he was busy, but from the recipient’s point of view, it felt cold. Because I haven’t actually met him in person, I didn’t feel friendly toward him. (Hana’s Diary 7: 2009–07–23) Hana felt uncomfortable after receiving such an unsociable, business-­like message and perceived that she was not acknowledged as a human being with emotional feelings. In response to receiving such an email, she continued to describe how she would normally respond to such a message.

Excerpt 2 Before getting down to business, for example, when there is something trivial to say like ‘How was your weekend? I hope you had a good weekend,’ when it’s after the weekend, the message shows the writer’s personality, and then I can feel a sense of closeness. When I receive such a [friendly] message, I  tend to prioritise replying to it rather than waiting until later. (Hana’s Diary 7: 2009–07–23)

250  Yoko Nogami Even though it is her job to exchange emails for business matters, she became resistant to a message that sounded too methodical. When she received such a message, her need to be acknowledged or recognised, which is a fundamental aspect of identity construction (West 1992: 21), was not satisfied. Conversely, in the following diary entry (Excerpt 3), she wrote about an occasion when she received a message from another business associate. This writer included something interpersonal in the message, which reconfirmed her appreciation of a cordial message that reflects the writer’s personality in order to build interpersonal relationship among business associates.

Excerpt 3 After the holiday, there were many emails from overseas in the inbox. Most of the messages were just about a business matter, or they contained only a simple greeting such as ‘Hope you are doing well’. However, the message from a business associate (40s) from New Zealand included questions like ‘How was your Silver Week Holiday?’ and news about his family. Although my workload was heavy after the holiday, I prioritised replying to him, dedicating much time. After all, I thought it is important that an email message should reflect the writer’s personality. (Hana’s Diary 10: 2009–09–24) When a writer showed his/her personhood in the message, Hana was grateful and felt acknowledged. Then she invested more effort in replying to the message for the sake of a smooth business transaction as well as for building better interpersonal relationship, which in turn gave her sense of being acknowledged. In another diary entry, Hana admitted that she herself used to write emails on the job without adding interpersonal feelings to them. Through the experiences described in her diary entries, she gained a chance to reflect on her own practice too. She eventually felt resistant toward her own as well as others’ existing email writing practices that solely exchanged business information. Then she changed her own writing habits to build better relationships with her business associates. This recognition consequently had a positive influence on her identity as a business ELF user. About a month later, she wrote in another diary entry:

Excerpt 4 I heard from a Thai lady for the first time in a while. I have showed her around Kyoto before [when she came to Japan], so I  had a rather close business relationship with her. But recently, I  have only been sending bureaucratic-­ sounding messages. Then, the responses to my messages have seldom come from herself but from her assistants, and the messages were CCed to her. [Today,] after so long, I sent her a message that included some seasonal greetings like Japanese scarlet-­tinged leaves, and wrote about the restaurant we went together before. Soon, she herself wrote a reply to me right away. Our

Understanding through narrative inquiry  251 relationship had been fading recently, but I felt a sense of closeness with her again after having this small talk. She became more supportive in terms of business transactions too. After all, I recognised again that interpersonal relationships get better or worse with the presence or absence of a little [sociable] conversation. (Hana’s Diary 12: 2009–11–13) This diary entry depicts that investment in one’s linguistic practice is an investment in a learner’s own identity and future self (Norton 2000), which is also applicable to ELF interaction. After reconfirming the value of exchanging friendly messages with business associates, Hana put it in practice herself, meaning, she invested more effort in writing emails with a personal touch. Her investment was paid off as she received the positive reaction from the recipient. Her satisfaction derived from her investment, which gave her cause for achieving a positive sense of self. As shown in her narratives, Hana’s negotiation of her identity as an ELF user was characterised by the process of her resistance to and investment in a certain pragmatic language use in order to foster healthy business and interpersonal relationships in a business context. Hana’s case study depicted the process of learning and practice of the pragmatic language use through email exchanges in the global business world. In particular, she became aware of the significance of more elaborate pragmatic language use in order to build better interpersonal relationships with business associates abroad who she may not necessarily have face-­to-­face encounters. Hana’s stories revealed her learning trajectory of ELF use through everyday business email exchanges and her participation in an imagined community (Anderson 2006; Kanno and Norton 2003). There was no tangible community of ELF users that Hana had access to, but instead there was a virtual and global community that was connected by business email exchanges. Like in the Communities of Practices (Lave and Wenger 1991), people in imagined communities mutually engage in a process of learning by sharing information and experiences. In other words, members of the imagined community learn from each other by co-­constructing knowledge (i.e., joint enterprise). Subsequently, members ‘develop a shared repertoire of resources’: experiences, stories, tools, ways of addressing recurring problems, in short, a shared practice (Wenger 2006: Introduction section). Through her participation in the imagined community of business ELF users, Hana recognised the importance of email messages that facilitate meaningful interpersonal relationships, in particular for her improved self-­identification, and brought the practice into negotiation with the other members (Excerpts 1, 2, and 3). Then, Hana engaged herself and other members in the action of exchanging business emails as well as building better relationships (Excerpt 4). In turn, the person with whom she exchanged emails responded positively to Hana’s practice, and they seemed to be co-­constructing and re-­developing the shared practice of business email exchange for rapport building. Those practices can be seen as a resource in creating a group and as an indicator of belonging to a group (Kalocsai 2014). In short, Hana’s narrative demonstrated her active participation and

252  Yoko Nogami belonging to the imagined community of global business ELF users amidst learning and negotiating shared practice of pragmatic language use. She learned how to utilise her available linguistic resources to achieve what she hoped for, and thus empowered herself even more to act as an ELF user (Iino and Murata 2016). Thus, the current research has demonstrated that narrative inquiry employing a diary study enables a researcher to access a diarist’s temporal experience, their thoughts and emotions around a particular experience. Through an individual’s stories as told by them, a diary study felicitates deeper understanding of the contexts the diarist is situated and how they make sense of themselves and their experience in their lives.

Conclusion As have been presented in this chapter, ELF-­focused, narrative research largely provides the theoretical ground for interpreting ELF and culture, as well as insights into the multifaceted nature of ELF speakers’ identities ranging from EFL teachers’ negative attitudes toward ELF and to the positive influence of ELF. In particular, in the narrative research study I  conducted, Hana’s storytelling has also contributed to the understanding of a process of learning, a development of practice and negotiation of identity through her active participation in an imagined community of a business ELF users while engaging in ELF business communication. In other words, a narrative approach enables us to reach something perhaps linguistically trivial but incredibly significant in terms of ELF users’ self-­ identification in the context of sociocultural environments in which they are.8 As one of narrative approach research studies, a diary study can be one of the practical methods to obtain an individual’s detailed and profound narratives based on their immediate interactional experiences which researchers are unable to access directly. ELF research with an interactional approach informed us what ELF speakers do linguistically during interaction, and the narrative approach reveals the thinking processes behind the speakers’ linguistic practices. For developing more informed perspectives on the link between the linguistic nature of ELF and speakers’ expressions of identity through ELF, further investigation is required from an emic perspective, seen by the participants themselves. Additionally, more mixed-­ method research needs to be conducted in order to better understand both what ELF speakers do and why they do it. A study combining both interactional and narrative approaches would be able to complement the findings held so far, as Kalocsai (2014) asserts. To conclude this chapter, the advantages of choosing a diary method as a narrative approach research should be addressed. It can be beneficial for both researchers and participants for different reasons. First, a diary study provides the researcher with access to a rich source of data. A  diary method is a good alternative for interviewing when eliciting events that are easily forgotten and that are difficult to be remembered precisely by the person concerned. Diary entries also can be a propitious information source on the participants’ behaviour

Understanding through narrative inquiry  253 and experiences on a day-­to-­day basis as well as the participants’ reflection of the event they experience. As Hana’s case study demonstrated, her diary entries depicted the process of her realisation of the importance of the pragmatic aspects of ELF use and the negotiation of her identity based on her day-­to-­day activities through ELF, which may not have been significant enough to be remembered in the future. Also, diaries are suitable for collecting information that may be too sensitive to divulge face-­to-­face and which involves groups or activities that are hard to reach.9 In particular, access to ELF communication data in business environments is likely to prove difficult, thus a diary study may be a good alternative method to collect data if the focus of the investigation is on ELF users in a business context. Lastly, diary writing can be beneficial for the participants themselves to become aware of their use of English and issues surrounding it in the communities they participate in. In other words, the diary writing activities become an opportunity of awareness raising for the diarists like Hana’s final comment10 (Excerpt 5) shows.

Excerpt 5 Something that has changed since I started writing diaries is that I became aware of writing heartfelt messages that reflect my personality. Up until now, because I wanted to save time and because I worried about making mistakes, I tended to write about necessary business matters only and as minimally as possible. But after all, in order to build strong interpersonal relationships, I again recognised the importance of messages that reflect the writer’s personality. . . . I became aware of using humour. The more that there are linguistic, national, and cultural distances between others and myself, the more I tend to write something trivial. That’s a characteristic of mine as a user of English.  .  .  . In particular, when using English I  often exchange messages with somebody who is in a very different environment than me. So, it [i.e., using humour] is necessary for the other person to feel comfortable even to a small degree. (Hana’s reflective essay: 2010–01–13) Writing a diary is not an easy task for many individuals; however, it helps to increase mindfulness of what goes on during engagement in ELF interactions and to stimulate further language learning/using and identity construction. When appropriate instructions and regular encouragement are provided, diary writing can be a valuable experience for ELF users and second language learners/users. This is especially so when people experience ELF intercultural encounters, such as studying abroad (Jackson 2008) because of the self-­reflective and self-­analytic elements of diary writing. Thus, reflecting on and observing experiences by writing diaries could help ELF users to examine English language use in relation to larger communities and ‘find spaces for the enhancement of human possibility’ (Norton 2000: 153).

254  Yoko Nogami

Notes 1 This article is a modified version of my paper published in Waseda Working Papers in ELF (English as a Lingua Franca) Vol. 5 (pp. 136–152). The case study of Hana also appears in Nogami (2011). 2 Phan (2008) does not use the term ELF in her study; however, the context is an ELF setting in which her participants from different Asian backgrounds lived and studied while using English as a means for intercultural communication. 3 This is linked to Anderson’s Imagined Communities (2006, 1st edition 1983), which refers to “groups of people, not immediately tangible and accessible, with whom we connect through the power of the imagination” (Kanno and Norton 2003: 241). 4 The term Communities of Practice (CoP) refers to “groups of people who share a concern or a passion for something they do and learn how to do it better as they interact regularly” (Wenger 2006: Definition section). A number of ELF scholars (e.g., Ehrenreich 2009; Kalocsai 2014) incorporated the concept of the CoP into their ELF research with caution. Meanwhile several ELF scholars pointed out that an ELF interactional context is often more fluid than in “communities” in Wenger’s sense because ELF speakers do not necessarily engage in a process of learning regularly by co-­constructing knowledge or develop a shared repertoire of resources for a common goal (Jenkins 2015: 61–62), but rather, ELF takes place in “transient groups” (i.e, Transient International Groups) than in communities (Pitzl 2018). 5 Some participants travelled abroad during the study. 6 They were allowed to write their diary in either Japanese or English, and code-­ switching was permitted to enable them to express themselves in a more comfortable manner. 7 All the excerpts in this chapter were translated from the original Japanese by the author. The translation was validated by a bilingual colleague for accuracy and interpretation. 8 Only Hana’s case study was presented in this paper due to space limitations; however, other diarists in the larger study also wrote of stories and experiences that revealed sociocultural values and perceptions among ELF users that influenced their identity construction (Nogami 2011). 9 Some drawbacks of using diaries as a research method should also be addressed. Diary observation as a data collection method is time-­consuming in terms of both collecting and analysing data because the researcher needs to re-­read the entries multiple times to find regular patterns. It also heavily relies on the willingness of the diarists to keep the diaries regularly and truthfully. However, the richness of the data and a well-­built relationship with the diarists can compensate for such drawbacks. 10 Upon my request, several diarists agreed to write a reflective essay at the end of the diary study, reporting on how writing diaries made them realise things that they were never aware of, and how they thought such realisations have changed their perspective on their English use.

References Alaszewski, A. 2006. Using diaries for social research. London: Sage. Anderson, B. 2006. Imagined communities: Reflections of the origins and spread of nationalism (rev. ed.). London: Verso. Bailey, K.M. 1983. Competitiveness and anxiety in adult second language learning: Looking at and through the diary studies. In H.W. Seliger and M.H. Long (eds), Classroom oriented research in second language acquisition. Rowley, MA: Newbury House, pp. 67–102. Baker, W. 2009. Language, culture and identity through English as a Lingua Franca in Asia: Note from the field. The Linguistics Journal 4, 8–35.

Understanding through narrative inquiry  255 Baker, W. 2011. Culture and identity through ELF in Asia: Fact or fiction? In A. Archibald, A. Cogo and J. Jenkins (eds), Latest trends in ELF research. Newcastle: Cambridge Scholars Publishing, pp. 32–52. Baker, W. 2015. Culture and identity through English as a Lingua Franca: Rethinking concepts and goals in intercultural communication. Berlin: De Gruyter Mouton. Bartlett, R. and C. Milligan. 2015. What is diary method? London: Bloomsbury Academic. Clandinin, D.J. 2007. Handbook of narrative inquiry: Mapping a methodology. London: Sage. Clandinin, D.J. and F.M. Connelly. 2000. Narrative inquiry: Experience and story in qualitative research. Jossey-­Bass: San Francisco. Cogo, A. 2009. Accommodating difference in ELF conversations: A  study of pragmatic strategies. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle: Cambridge Scholars Publishing, pp. 254–273. Corti, L. 1993. Using diaries in social research. Social Research Update 2. Guildford: University of Surrey. Available from: http://sru.soc.surrey.ac.uk/SRU2.html [accessed 14 April 2011]. Ehrenreich, S. 2009. English as a Lingua Franca in multinational corporations – Exploring business communities of practice. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle: Cambridge Scholars Publishing, pp. 126–151. Etherington, K. 2007. Ethical research in reflexive relationships. Qualitative Inquiry 13(5), 599–616. Fiedler, S. 2011. English as a Lingua Franca – A native-­culture-­free code? Language of communication vs. language of identification. Apples – Journal of Applied Language Studies 5(3), 79–97. Guido, M.G. 2019. English as a Lingua Franca in migrants’ trauma narratives. London: Palgrave Macmillan. Iino, M. and K. Murata. 2016. Dynamics of ELF communication in an English-­medium academic context in Japan: From EFL learners to ELF users. In K. Murata (ed.), Exploring ELF in Japanese academic and business contexts: Conceptualization, research and pedagogic implications. London: Routledge, pp. 111–131. Jackson, J. 2008. Language, identity, and study abroad: Sociocultural perspectives. London: Equinox. Jenkins, J. 2007. English as a Lingua Franca: Attitude and identity. Oxford: Oxford University Press. Jenkins, J. 2015. Repositioning English and multilingualism in English as a Lingua Franca. Englishes in Practice 2(3), 49–85. Jenkins, J., W. Baker and M. Dewey. 2018. The Routledge handbook of English as a Lingua Franca. London: Routledge. Kalocsai, K. 2009. Erasmus exchange students: A  behind-­the-­scenes view into an ELF community of practice. Apples – Journal of Applied Language Studies 3(1), 25–49. Kalocsai, K. 2011. The show of interpersonal involvement and the building of rapport in an ELF community of practice. In A. Archibald, A. Cogo and J. Jenkins (eds), Latest trends in ELF research. Newcastle: Cambridge Scholars Publishing, pp. 113–137. Kalocsai, K. 2014. Communities of practice and English as a Lingua Franca: A study of Erasmus students in a Central European context. Berlin: De Gruyter Mouton. Kanno, Y. and B. Norton. 2003. Imagined communities and educational possibilities: Introduction. Journal of Language, Identity, and Education 2(4), 241–249. Kaur, J. 2009. Pre-­empting problems of understanding in English as a Lingua Franca. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle: Cambridge Scholars Publishing, pp. 107–123.

256  Yoko Nogami Klimpfinger, T. 2009. ‘She’s mixing the two languages together’ – Forms and functions of code-­switching in English as a Lingua Franca. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle: Cambridge Scholars Publishing, pp. 348–371. Kramsch, C. 1993. Context and culture in language teaching. Oxford: Oxford University Press. Lave, J. and E. Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. Miles, M.B. and A.M. Huberman. 1994. Qualitative data analysis (2nd ed.). London: Sage Publications. Ng, P.C.L. and P.S-­W. Dodge. 2015. Situating English as a Lingua Franca in context: Narratives from Japanese and Chinese Classrooms. Intercultural Communication Studies 24(3), 50–63. Nogami, Y. 2011. Japanese L2 English users’ second language identities and pragmatic use in relations of power and culture. Unpublished PhD thesis, University of Essex, U.K. Nogami, Y. 2019. Identity and pragmatic language use among East Asian ELF speakers and its implications for English-­medium education. In K. Murata (ed), English-­medium instruction from an English as a Lingua Franca perspective: Exploring the higher education context. Oxon: Routledge, pp. 176–193. Nogami, Y. 2020. Study abroad, identity, and attitude toward the English language. In M. Konakahara and K. Tsuchiya (eds), English as a Lingua Franca in Japan: Towards multilingual practices. London: Palgrave Macmillan, pp. 157–181. Norton Peirce, B. 1993. Language learning, social identity, and immigrant women. Unpublished PhD dissertation, Ontario Institute for Studies in Education, University of Toronto. Norton, B. 2000. Identity and language learning. Harlow: Pearson Education. Pedrazzini, L. and A. Nava. 2011. Researching ELF Identity: A  Study with non-­native English teachers. In A. Archibald, A. Cogo and J. Jenkins (eds), Latest trends in ELF research. Newcastle: Cambridge Scholars Publishing, pp. 269–284. Phan, L.H. 2008. Teaching English as an international language: Identity, resistance and negotiation. Clevedon: Multilingual Matters. Phan, L.H. 2009. English as an international language: International students and identity formation. Journal of Language and Intercultural Communication 9(3), 201–214. Pitzl, M-­L. 2009. ‘We should not wake up any dogs’: Idiom and metaphor in ELF. In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle: Cambridge Scholars Publishing, pp. 298–322. Pitzl, M-­L. 2012. Creativity meets convention: Idiom variation and re-­metaphorization in ELF. Journal of English as a Lingua Franca 1(1), 27–55. Pitzl, M-­L. 2018. Transient International Groups (TIGs): Exploring the group and development dimension of ELF. Journal of English as a Lingua Franca 7(1), 25–58. Pölzl, U. and B. Seidlhofer. 2006. In and on their own terms: The “habitat factor” in English as a lingua franca interactions. International Journal of the Sociology of Language 177, 151–176. Pullin Stark, P. 2009. No joke – This is serious!’ Power, solidarity and humour in business English as a Lingua Franca (BELF). In A. Mauranen and E. Ranta (eds), English as a Lingua Franca: Studies and findings. Newcastle: Cambridge Scholars Publishing, pp. 152–177. Seidlhofer, B. 2009. Accommodation and the idiom principle in English as a Lingua Franca. Intercultural Pragmatics 6(2), 195–215. Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford: Oxford University Press.

Understanding through narrative inquiry  257 Seidlhofer, B. 2018. Standard English and the dynamics of ELF variation. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 85–100. Seidlhofer, B., A. Breiteneder and M-­L. Pitzl. 2006. English as a Lingua Franca in Europe. Annual Review of Applied Linguistics 26, 1–34. Sung, C.C.M. 2014. Global, local or glocal? Identities of L2 learners in English as a lingua franca communication. Language, Culture and Curriculum 27(1), 43–57. Virkkula, T. and T. Nikula. 2010. Identity construction in ELF contexts: A case study of Finnish engineering students working in Germany. International Journal of Applied Linguistics 20(2), 251–273. Wenger, E. 1998. Communities of practice: Learning, meaning, and identity. Cambridge: Cambridge University Press. Wenger, E. 2006. Communities of practice: A brief introduction. Available from: http:// www.ewenger.com/theory/ [accessed 15 April 2011]. West, C. 1992. A matter of life and death. October 61(summer), 20–23. Yashima, T. 2009. International posture and the ideal L2 self in the Japanese EFL context. In Z. Dörnyei and E. Ushioda (eds), Motivation, language identity and the L2 self. Clevedon: Multilingual Matters, pp. 144–163.

15 Rigour in ELF language attitude research An example of a conversational interview study Tomokazu Ishikawa As might be expected of so abstract and serviceable a term, [attitude] has come to signify many things to many writers, with the inevitable result that its meaning is somewhat indefinite and its scientific status called into question. (Allport 1935: 798)

Introduction Attitude research in social psychology of language started as early as the 1930s (e.g., Pear 1931), and language attitudes have been a key concept in sociolinguistics since Labov’s (1966) seminal work. Within the relatively new research field of English as a Lingua Franca (ELF), there has been a large amount of work investigating language attitudes, especially since Jenkins’s (2007) global study on English teachers’ attitudes towards English accents. Despite the long research tradition (see e.g., Giles and Billings 2004) and the intensive research interest of many ELF researchers (see e.g., Baird and Baird 2018), language attitudes seem to lack a widely agreed definition to this date. In the first place, it is not clear whether it is necessary or beneficial for ELF researchers with different research focuses to have one single definition of language attitudes. In this regard, what Allport (1935) observed more than 80 years ago (see earlier) could still be the case. For the sake of scientific integrity, ELF researchers investigating ‘language attitude’ ought to characterise what it is and how it is being researched in their own studies. Regarding the question of the what, the controversy or difficulty surrounding the characterisation of language attitudes is especially observable in conceptualising their durability: how far they are stable or emergent (Ishikawa and Morán Panero 2016). A closely related issue is what may be called ‘non-­attitudes’ (Converse 1970), which may lead to ‘no opinion’ (Krosnick et al. 2002). ELF research has shifted away from associating language with geographical communities in the face of unprecedented global mobility and multilingualism (Jenkins 2015, 2017), to which the traditional concept of ‘native’ English speakers (NESs) bears a very limited relevance. Such amorphous linguistic interaction might not invite any attitudes from research participants.

Rigour in ELF language attitude research  259 Regarding the question of the how, it is dubious whether language attitude data belongs solely to participants. In particular, “interviews involve the co-­construction of experiential reality” (Holstein and Gubrium 2011: 161; see also Miyahara, this volume) or the co-­construction of knowledge (Talmy and Richards 2011). Obviously, the researcher-­interviewer is likely to contribute some knowledge of ELF to facilitate discussion in ELF-­related interviews, and what participants have to say is embedded in dynamic interaction. It is therefore challenging to analyse an individual’s language attitudes, no matter what they may be defined. The present chapter introduces one possible example of answering the ‘what’ and ‘how’ questions with a more focus on the ‘how’. The chapter starts by addressing the ‘what’ question in relation to the durability and ontology of language attitudes. It then discusses the ‘how’ question with empirical data and concludes by suggesting the importance of conceptual and methodological rigour in language attitude research. For this purpose, the author draws on his own conversational interview study (Ishikawa 2017a), which explores the attitudes of Japanese university students towards their own English, particularly the attitudes of those who would like to use English for global communication. This study is part of a larger research conducted as a doctoral project: A study of Japanese university students’ attitudes towards their English (Ishikawa 2016a; see also Ishikawa 2016b, 2017b).

The ‘what’ question: conceptualising and redefining language attitudes To characterise language attitudes, Garrett (2010) premises that language attitudes are evaluative dispositions to a language-­related psychological target. This premise well summarises three key denominators thus far proposed mainly in social psychology. First, attitudes are evaluative, in other words, entailing a varying degree of favourableness or disfavourableness (e.g., Petty et al. 1997). Second, attitudes are directed to a psychological object and, in the case of language attitudes, a linguistic phenomenon at any level, including language topics (e.g., Garrett 2010). Finally, as being dispositions formed through experience (e.g., Campbell 1963), attitudes cannot be directly observed but can be inferred from relevant responses including verbal accounts (e.g., Perloff 2017). These responses are not necessarily the direct reflection of attitudes owing to various situational factors (e.g., Eagly and Chaiken 2007). The author’s research regards everyone as a folk, whether linguists or non-­linguists, and attempts to include non-­specialist views in line with Niedzielski and Preston’s (1999/2003) sociolinguistic approach called folk linguistics. Assuming that people bring their various evaluative concepts to everyday life, folk linguistics investigates their accounts to determine “the underlying beliefs and belief systems” (Niedzielski and Preston 2009: 357) or “a reservoir of beliefs and concepts” (Preston 2010: 9). One may argue that this approach coincides with the aforementioned characterisation of attitudes if ‘beliefs’ are reinterpreted as ‘attitudes’. This is all the more so because beliefs in attitude research have often been identified with cognitive responses or

260  Tomokazu Ishikawa accounts triggered by attitudes rather than what underlies these responses or accounts themselves (e.g., Rosenberg and Hovland 1960; Bassili and Brown 2005/2014; cf. Fives and Buehl 2012). Accordingly, the present study identifies Garrett’s (2010) evaluative dispositions with an underlying reservoir of evaluative dispositional concepts in folk linguistics, but without employing the term ‘beliefs’. As dispositions, “the standard view of attitudes” (Banaji and Heiphetz 2010: 357) has long assumed that attitudes are stable to a varying degree rather than emergent from scratch. In contrast, constructionist scholars view them “as temporarily constructed judgments” (Wilson and Hodges 1992: 38). They have investigated the situatedness of evaluative practices around linguistic and other issues. However, it appears that these scholars either disregard an individual’s dispositional concepts on which his or her mental processing may be based (e.g., Hyrkstedt and Kalaja 1998) or focus considerably more on processing itself (e.g., Cargile et al. 1994). Even so, constructionist scholars (e.g., Potter and Wetherell 1987) provide notable insights. After all, the same psychological target may be interpreted differently through different representations of the object in question (e.g., Potter 1996). At the same time, an individual’s evaluation is inseparable from both ideological beliefs and indexical meanings surrounding a society (e.g., Coupland 2007). In turn, an individual’s dispositions may exist only in relation to the socially available representations of any given phenomenon. This leads to a nuanced understanding of the issue of ‘non-­attitudes’. That is, it may be impossible to detect any stable attitudes not only towards an unfamiliar linguistic phenomenon but also towards unfamiliar ways of representing a familiar linguistic phenomenon. It may be argued that the stability and emergence of language attitudes represent two sides of the same coin, if one admits the possibility that even highly stable dispositional concepts are processed and expressed in a relatively unstable manner. Empirically, it is possible that some individuals’ psychological processing manifests itself across their discursive accounts to the extent of being inferable as among dispositional concepts (Ishikawa 2016a). Instead, it is also possible that various and variable evaluative patterns feature empirical data to the extent of signalling a lack of evidence of any stable dispositions (Morán Panero 2016). Researchers should therefore refrain from taking it for granted that participants always have stable evaluative repertoires of any given psychological object. Based on this examination of language attitudes, the author’s research (re)defines them as being identified with a reservoir of evaluative dispositional concepts, directed to a linguistic phenomenon as interpreted through social experience, and underlying observable responses which are constructed situationally.

The ‘how’ question: exploring participants’ language attitudes This section introduces the conversational interview study as part of the author’s research on language attitudes in the field of ELF. To the effect of justifying his methodological choice, the section starts with a range of possible methods in language attitude research in general.

Rigour in ELF language attitude research  261 Approaches to methods in language attitude research There seem to be three approaches to methods in language attitude research: (1) societal treatment, (2) the indirect approach, and (3) the direct approach (e.g., Garrett et al. 2003). Societal treatment studies investigate how languages or language varieties and their users are viewed in a given society (e.g., Landry and Bourhis 1997). Diverse studies can fall into this category, such as analysis of language policy documents; consumer advertisements (e.g., television commercials); and what is called linguistic landscape, which “refers to the visibility and salience of languages on public and commercial signs in a given territory or region” (ibid.: 23). While useful for contextualising a given social sphere, societal treatment is obviously not necessarily synonymous with individuals’ attitudes. Other approaches should be combined together if the attitudes are the main target of research, whether within or outside the field of ELF. The indirect approach of language attitude studies refers to both the matched guise technique or MGT (e.g., Lambert et al. 1960) and its modified version called the verbal guise technique or VGT (e.g., Gallois and Callan 1981; Huygens and Vaughan 1983). This approach is indirect because respondents are typically made to believe that they are evaluating people who speak like an audio-­recorded sample instead of a language variety. Introduced by Lambert et al. (1960), the classic design of the MGT involves both presenting the same audio text in various languages or language varieties recorded by one or more speakers who produce plural versions under ‘guises’ and providing a questionnaire with rating scales (e.g., Dragojevic and Giles 2014). As such, the MGT inevitably resorts to the stereotypes associated with linguistic cues and presumes that a speech sample represents the members of a certain speech community. This experimental nature does not seem to accord with ELF research which “targets, and seeks to comprehend in situ, English communication across geographical boundaries” (Ishikawa 2016c: 129). The later version, the VGT, employs actual speakers of each language or language variety of a national or regional speech community (e.g., Chan 2016). Even if speech factors other than languages or language varieties (e.g., voice quality) are controlled carefully, the VGT similarly does not match the nature of global communication as “dynamic, pluralistic manifestations of linguistic resources in an international setting” (Ishikawa 2015: 39). In the first place, empirically, there is no such thing as an ‘ELF variety’ (e.g., Baker and Jenkins 2015). Having probably been the most dominantly employed, the direct approach of language attitude studies mainly refers to questionnaires and interviews, both of which involve “the asking of direct questions about language evaluation, preference etc.” (Garrett et al. 2003: 16). While closed-­response items, typically rating scales, have been an integral part in these questionnaires (e.g., Woolard and Gahng 1990; Coupland and Bishop 2007), they are likely to confine the scope of enquiry to predetermined evaluation categories which, in turn, may be interpreted variously among respondents. The reification and simplification of closed-­response items thus cannot be reconciled with heightened complexity and adaptability in lingua franca communication (e.g., Mortensen 2013).

262  Tomokazu Ishikawa Unlike closed-­response items, perceptual dialectology (e.g., Preston 2018) provides some openness since “[r]espondents may be asked to label maps with where different dialects are spoken, or to rate various areas (e.g., each state in the US) on how ‘correct’ and ‘pleasant’ the language spoken there is” (Lindemann 2005: 189). However, this map task is no less stereotypical than the MGT. Also, the respondents are constrained to answer geographically, which again is not compatible with communication scenarios which transcend geographical boundaries. Meanwhile, a questionnaire may comprise open-­ended items, so long as literate respondents feel its theme to be very relevant to them (Jenkins 2014). The main limitation of an open-­ended questionnaire is a likely low response rate (e.g., Dörnyei and Taguchi 2010). More precisely, the respondents would be more of those who have high interest and willingness to participate in research, to which due notice should be given for data interpretation. Like an open-­ended questionnaire, conversational interviews would assist in heeding what may be the crux of ELF research, that is, “the importance of viewing language from multiple dimensions in which its contextual embodiment is crucial” (Baird et al. 2014: 181). Conversational interviews form a contrast to pragmatic, business-­like transactions in the following three respects (see Eggins and Slade 1997/2004: 19–20). First, the flow of the interviews is unpredictable and  only vaguely around a research focus. Second, the interviews are informal and humorous so that interviewees can easily initiate topics and overlap or interrupt the interviewer. Finally, the interviews are relatively long, possibly exceeding an hour and thus contributing to eliciting contextually rich data. An example of conversational interview study In “an attempt to secure an in-­depth understanding of the phenomenon in question” (Denzin and Lincoln 2011: 5), the author’s investigation of language attitudes employed both an open-­ended questionnaire (Ishikawa 2016b) and conversational interviews (Ishikawa 2017b), along with a review of published research on societal treatment in the Japanese context. Again, the focus here is narrowed down to the conversational interview study. Methodological overview The study targeted only the undergraduates who had received Japanese education as well as had an interest and investment in global communication through English.1 In total, 18 first language Japanese users participated from diverse disciplines at 2 national and 6 private Japanese universities. The researcher held a once-­only interview with each of the 18 undergraduates in Japanese in Japan. There was no prepared interview question or prompt to minimise the imposition of a priori categories and propositions on participants. All the interviews were conducted in a casual manner in the hope that conversations between (quasi-­)equals would provide maximum freedom for both sides to explore any topic in depth. As is usual with this conversational method, each interview was relatively long, amounting to 20 hours and 11 minutes in total.

Rigour in ELF language attitude research  263 All the data was transcribed and analysed through the combination of qualitative content analysis and Eggins and Slade’s (1997/2004) speech functions analysis framework.2 Qualitative content analysis seeks to capture “the underlying deeper meaning of the data” (Dörnyei 2007: 246) by coding what is in the data, grouping the codes under overarching categories, and interpreting the relationships between the categories (e.g., Schreier 2012). Meanwhile, all the codes were derived only from a participant’s conversational moves which in Eggins and Slade’s (1997/2004) framework (described next) represented his or her stance and concerns rather than the researcher’s. As a tool to analyse the functions of conversational moves, the framework seems applicable to both original and translated data. Admittedly, the whole analytical procedure was inevitably influenced by the research focus and interest (e.g., Miles et al. 2014), even though there was no predetermined code or category. Eggins and Slade’s (1997/2004) framework Eggins and Slade’s (1997/2004) speech functions analysis framework has been developed from Halliday’s (1984, 1994) systemic functional description of dialogue, and consists of three broad types of conversational moves: opening, continuing, and reacting. Opening moves “function to initiate talk around a proposition” (Eggins and Slade 1997/2004: 194). They comprise both attending moves which secure the attention of an intended interlocutor and initiating moves which include giving or demanding either factual information or opinions. Continuing and reacting moves are achieved, respectively, by the current or another speaker in interaction, both functioning to “keep negotiating the same proposition” (Eggins and Slade 1997/2004: 195). Continuing moves comprise monitoring, prolonging, and appending moves. Monitoring moves make sure that an interactional situation is maintained. Examples include checking whether an interlocutor is following or not and inviting him or her to take a turn. Prolonging moves provide further information as the expansion of a prior move. Appending moves occur when a speaker engages in this expansion after losing and regaining the turn. Reacting moves are further classified into responses and rejoinders, both of which are either supportive or confrontive. While responses “move the exchange towards completion”, rejoinders “in some way prolong the exchange” (ibid.: 200). With regard to responses, supporting responses comprise developing, engaging, registering, and replying moves. Among them, engaging moves are simple agreement with attending moves. Developing moves occur when another speaker engages in the expansion of a prior move, but does so with a view to completing the exchange. Registering moves encourage a prior speaker to take the next turn, not providing any expansion (e.g., “Ah, that lady.”), while replying moves just indicate the willingness to accept what a prior speaker says (e.g., “Oh, yes.”). On the other hand, confronting responses comprise disengaging and replying moves. The former refuses to take part in the exchange, for instance, by keeping silent. The latter includes a contradiction, disavowal,

264  Tomokazu Ishikawa disagreement, or withholding response (e.g., “I don’t know.”) in the direction of completing the exchange. With regard to rejoinders, supporting rejoinders comprise tracking moves and subsequent responses. Tracking moves are subdivided into checking, confirming, clarifying, and probing moves. The first three types occur when another speaker seeks: (1) the re-­explanation of at least some of a prior move (i.e., checking); (2)  the verification of what he or she indicates having heard (i.e., confirming); or (3) additional information to understand a prior move (i.e., clarifying). And the last type occurs when another speaker engages in the expansion of a prior move, but does so in order to seek the confirmation of his or her expansion (i.e., probing). Subsequent responses may also be supporting by resolving, repairing, or acquiescing with prior moves. On the other hand, confronting rejoinders comprise challenging moves and subsequent responses. Challenging moves are subdivided into detaching, rebounding, and countering moves. Detaching moves seek to bring the current discussion up to an end. Rebounding moves question a prior move in terms of its “relevance, legitimacy or veracity” (Eggins and Slade 1997/2004: 212). And countering moves express “an alternative, counter-­position or counter-­interpretation” (ibid.). Subsequent responses may also be confronting by having prior moves unresolved, refuted, or re-­challenged. All these different speech function moves are summarised in Table 15.1. In practice, the author’s coding in qualitative content analysis targeted an interviewee’s turns which substantially contributed to discussion. These are the following ten moves in Eggins and Slade’s (1997/2004) framework described earlier:

Table 15.1  Speech function moves

OPENING MOVES -­attending -­initiating REACTING MOVES Responses: Supportive -­developing -­engaging -­registering -­replying Rejoinders: Supportive Tracking & Subsequent responses -­checking -­confirming -­clarifying -­probing

CONTINUING MOVES -­monitoring -­prolonging -­appending Responses: Confrontive -­disengaging -­replying Rejoinders: Confrontive Challenging & Subsequent responses -­detaching -­rebounding -­countering

Source: Based on Eggins and Slade’s (1997/2004: 191–213).

Rigour in ELF language attitude research  265 initiating moves as opening moves, prolonging and appending moves as continuing moves, developing moves as supporting responses, disengaging and replying moves as confronting responses, probing moves as supporting rejoinders, and detaching, rebounding, and countering moves as confronting rejoinders. It should be mentioned, however, that there was no disengaging or detaching move detectable in the collected data. Given that prolonging, appending, developing, and probing moves expand a prior move in their different ways, it follows that analysis was applied to an interviewee’s initiating moves, variously occurring expansion, confronting replying responses, and rebounding and countering rejoinders. Findings on ELF awareness The findings are manifold, largely corroborating the open-­ended questionnaire study: participants’ prevalently negative attitudes towards their English as dispositional concepts, coupled with the doxic experience3 within Japanese society that probably hindered grounding these attitudes on their own critical thinking as well as valuing their English communication ability with non-­Japanese (Ishikawa 2016b). At the same time, the interviews served as virtually the first opportunities for the participants to reflect on their English from ELF perspectives.4 The English quotations and interview extracts in this chapter are the researcher’s translations from Japanese. Admittedly, owing to the differences in Japanese and English sentence structures, where pauses and overlapping talk are cited in the English extracts is an approximation of the original Japanese version. Transcription conventions are provided in the Appendix. During the interviews, all the 18 participants reviewed their global communication experience and came to know ELF perspectives. While the researcher did not try to explain the perspectives themselves, he frequently posed questions, such as “Whose English was more understandable to you?” and “How would you define communicative success in those situations?” and occasionally gave a summary of a perceived contradiction in what he had heard. Based on the 18 interviewees’ initiating moves, variously occurring expansion, confronting replying responses and rejoinders in Eggins and Slade’s (1997/2004) speech functions analysis framework, qualitative content analysis resulted in producing ELF perspectives as one main category, together with the following three subcategories: ELF PERSPECTIVES • Entire approval of ELF perspectives • Partial approval of ELF perspectives • Non-­understanding of ELF perspectives Indeed, Eggins and Slade’s (1997/2004) framework not only lends itself to identifying what deserves to be processed through qualitative content analysis, but also provides assistance in understanding an interviewee’s utterances in an interactional context, as seen later with two extracts.

266  Tomokazu Ishikawa Thirteen participants approved of ELF perspectives entirely and quite readily. There were two commonalities across what they had to say. One commonality was that they had scarcely given serious consideration to ELF perspectives, which they tended to wonder why they had not recognised before. As one example, immediately after recalling that she “did feel the desire to sound like North American or British people (アメリカとかイギリスとかの人っぽくしゃべ りたいとは思ってました)”, Interviewee 6 burst into laughter and clapped her hands together loudly. The combination of the past tense in this quotation, her explosive laughter, and hand clapping appears to suggest that she no longer regarded her previous desire as reasonable. She added that a native-­centred view “has probably been imprinted on my mind since an early age (多分ちっ ちゃいころからの刷り込みがある)”. Given the strong words “imprinted on my mind (刷り込みがある)”, she had never had an opportunity to consider any alternative views. As another example, Interviewee 7 criticised her previous view of English as in Extract 1. While the researcher initiated the topic of an obsession with ‘correctness’ at the expense of communication, this topic was first suggested by the interviewee well before the start of the extract, in relation to exchange students from overseas.

Extract 1 (R: Researcher, I7: Interviewee 7)   1 R:   you say “deviant” or “mistaken” a lot [today   2 I7: [hmm   3 R: I was wondering what made you say [so   4 I7: [hmm   5 R: presumably deviant from the so-­called standard usage of grammar and   6 vocabulary in the United States or the United Kingdom.   7 [deviant pronunciation   8 I7: [hmmm­mmm. ahhhh   9 R: well it is deviant and thus [not good 10 I7: [deviant. not good 11 R: and 12 I7: and= 13 R: =shameful in the international [arena 14 I7: [hmm. hmm= 15 R: =er 16 I7: I’ve been thinking so haven’t I 17 R: but. actually we can’t communicate if obsessed with [that 18 I7: [hmmmm。pro[bably 19 having such an obsession itself is 20 R: [is that 21 I7: shameful isn’t it. @@@@ 22 R: I see 23 I7: phew

Rigour in ELF language attitude research  267 24 25 26 27 28 29 30 31 32 33 34 35

R: having such an obsession may in [fact be shameful I7: [indeed somehow it’s shameful R: is that [now I7: [that is all prejudice R: wh. when did you get to think so I7: well quite recently R: @[@@ I7: [yeah. just now. @@[@ R: [now I7: I didn’t realise that R: I [see I7:  [probably。before having this interview opportunity

Extract 1: Original Japanese version (R: Researcher, I7: Interviewee 7)   1 R: 「はずれている」とか「間違っている」ってのは今日いっぱい出て[来て   2 I7: [んー   3 R: それなんなんだろうなーって考えている[と   4 I7: [んー   5 R: やっぱりアメリカかイギリスのいわゆる標準的な文法や単語の使い方か  6 らはずれている.   7 [発音が違う   8 I7: [んーーー. はーー   9 R: でそのはずれていることでイコール[良くないし 10 I7: [はずれている. 良くない 11 R: し 12 I7: し= 13 R: =国際舞台で恥ず[かしい 14 I7: [んー. んー= 15 R: =ん 16 I7: と思ってたんでしょうね 17 R: でも. 実際そこにこだわってばかりだとコミュニケーションが[出来ない 18 I7: [んーー。 19 多[分こだわっているっていうことも 20 R: [てこと 21 I7: 恥ずかしいんでしょうね. @@@@ 22 R: なるほど 23 I7: ふうん 24 R: こだわっていることも実は[恥ずかしいのかもしれない 25 I7:  [もうなんか恥ずかしいことだと 26 R: それは[今 27 I7: [偏見のかたまりです 28 R: な. それはいつ思うようになったの

268  Tomokazu Ishikawa 29 30 31 32 33

I7: それはまあ最近ですね R: @[@@ I7:   [そう. だから今. @@[@ R: [今 I7: 気づかなかったんですよ 34 R: なる[ほど 35 I7: [多分。こんなこうゆう機会があるまで After producing a registering move (i.e., “hmmmm” in l.18), Interviewee 7 started seeing the obsession with ‘correctness’ as “shameful” (l.21). Remarkably, after the checking and resolving sequence (ll.24–25), she left the researcher’s probing (l.26) unresolved and appended fairly strong words “all prejudice” (l.27), thereby criticising her previous NES-­centred, normative way of thinking about English. She then acknowledged that she had probably never looked at English from ELF perspectives (ll.29, 31, 33, 35), and her louder, high-­ pitched voice (l.33) might have implied how novel they were. Sometime after the extract, when the researcher wondered whether she would revert to her previous view of English in the future, she made a disagreeing reply overlappingly: “No, I’ll no longer think that way (いや もう思わないと思います)”. This overlapping remark probably connoted how agreeable and influential ELF perspectives were. The other commonality across the aforementioned 13 participants was that they perceived ELF perspectives important and valuable in their different ways while tending to show positive feelings overtly. For instance, Interviewee 4 argued that, with ELF perspectives, “we’ll be able to step towards our interactants psychologically (精神的になんか相手に歩み寄ることが出来る)” and “approach their real selves (本当の自分たちに近づける)”. He did not wait for the researcher’s turn but went on to say “I  see (なるほど)” in a louder voice, suggesting how pleasantly convincing ELF perspectives were. To mention another example, in an enthusiastic tone of voice, Interviewee 8 asked the researcher to lecture on how to facilitate global communication for her colleagues at the university international social club. Likewise, Interviewee 9 insisted in a serious tone of voice that the researcher aim to “do break down Japanese people’s awfully contradictory understanding of English (実際すっごい矛盾がかった英語の 特に日本人の英語 への理解ってゆうのを取り壊[す])” (for this contradictory understanding, see Ishikawa 2017b). To give one more example, Interviewee 5 discussed ELF perspectives in relation to human development.

Extract 2 (R: Researcher, I5: Interviewee 5)   1 I5: uh. making what you say easier for your interactants to understand is   2 perhaps not limited to English   3 R: mmm

Rigour in ELF language attitude research  269   4 I5: isn’t it. it connotes consideration for your interactants. It’s like. when   5 people with different positions in society. or rather. people with   6 different upbringings   7 R: mmm=   8 I5: =people with different backgrounds. people with different first languages   9 talk with each other. Whether we can give our sympathy to those 10 differences. in communication really. has something to do with our 11 development as human beings [I think 12 R:                  [ah. I really agree. I almost feel like 13 that’s why we have English edu[cation 14 I5: [hmm 15 R: but now it seems like our education is taking a different course of action= 16 I5: =oh dear. interesting. ah. oh dear really interesting

Extract 2: Original Japanese version (R: Researcher, I5: Interviewee 5)   1 I5: あの. 相手に分かりやすいように話すってゆうのは  2 英語をちょっと超えたものかも   3 R: んー   4 I5: しれませんよね. 相手に対する思いやりだったり. そのこう.  5 身分 社会の身分が違う.  6 ま違うか. 育ちが違う人たち   7 R: んー=   8 I5: =育ってきた環境が違う人同士が. 言葉の違う人同士が話す時に.   9 そういう相手のところにシンパシーを持ちながら. こうコミュニ 10 ケーションできるかってのは非常に。その人間的にも成長出来る 11 ことにつながるもの[だと思う 12 R: [あー. すごい同感. だから英語教育が 13 ある気も[するんですよ 14 I5: [んー 15 R: でもそうじゃない方向に今の教育が行っているようで= 16 I5: =あっいや. 面白い. あー. あっいやっ本当に面白いですね What Interviewee 5 stressed in this extract was the importance of having “sympathy” (l.9) for those with different social, linguistic, and cultural identities in terms of human development. Thus, by the words “perhaps not limited to English” (l.2), he might have meant that effective global communication was not a purely linguistic issue, but a matter of social and cultural sensitivity. When the researcher briefly discussed Japanese English education (l.15), Interviewee 5 made one set of supporting replies in a rather excited tone (l.16). These replies appeared to reflect his keen interest in continuing to ponder on the interrelationship of ELF perspectives and human development. In fact, as final comments, he told the researcher with pleasant laughter that “I’ll probably can’t sleep tonight (今日は多分眠れないですね)”.

270  Tomokazu Ishikawa Three participants partially approved of ELF perspectives. Interviewee 1 acknowledged diversity in English but invariably saw North American NESs as ideal. This was probably due to the North American NES-­dominant approach to English with which he was so familiar. Also, however much they liked ELF perspectives, both Interviewees 3 and 10 claimed that how to use English with people from diverse backgrounds was not relevant to the English classroom, probably reflecting the strong normative approach to English which they had experienced in their own English education. Only two participants did not approve of ELF perspectives. Interviewee 11 initiated the topic of non-­NESs’ English and added: “I don’t see it as English (そ れは英語じゃないですね). It’s something peculiarly personal (なんか個人的 に変わっちゃったなんかみたいな)”. She had hardly any English communication experience outside private eikaiwa (i.e., English conversation) institutions, which tend to commodify white NESs (e.g., Kubota 2011). Also, Interviewee 12, who was then an applied linguistics student, suddenly told the researcher towards the end of the interview that “I’ve really disliked Jennifer Jenkins’s so-­called Lingua Franca Core (ジェニファー・ジェンキンズさんのリンガ・フランカ・ コアってゆうのがすごく嫌だったんですよ)”. He misunderstood it as arguing for “the importance of reduction or compromise (リデュースってゆった その歩み寄るってゆうことが大事だって)”. After all, university lecturers and publications do not always understand ELF research, thereby risking misinforming students (e.g., Ishikawa 2015). Discussion The present study offers empirical evidence on how easily and in what way Japanese university students appreciate ELF perspectives. It may be misleading, however, to conclude that many interviewees’ attitudes have completely changed. What the current study reveals is that they were unlikely to have any previous attitudes towards the unfamiliar representation of their English (i.e., from ELF perspectives) and that they were highly likely to develop positive attitudes towards it. In other words, even if their stable attitudes were negative towards the familiar representations of their English in the society, ELF perspectives provided a new way of viewing it, which was easy for them to appreciate. Thus, ELF awareness has a high potential to develop Japanese students’ positive attitudes towards their English as a global language, even if they “have been obsessed (強迫観念に陥っ てしまう)” (Respondent 1 in Ishikawa 2016b) with an NES-­dominant, normative approach to English. Seeing that these positive attitudes are more likely to encourage them to engage actively in English interaction and value their communication ability, future research and effort should be devoted to raising ELF awareness in Japanese and other similar contexts.

Conclusion At a broader level, language attitude research in the field of ELF seems to favour a study of contextualised, non-­experimental, and discursive accounts and analyse

Rigour in ELF language attitude research  271 them in the light of how they are expressed. At a more specific level, it would be fair to say that it is the remit of researchers to characterise language attitudes for their own studies. This is particularly because different ontological perspectives may lead to different choices and practices of research methods and analytical frameworks, as well as different interpretations of the data collected. The author’s conceptualisation of language attitudes, chosen research method of conversational interviews, and way of using Eggins and Slade’s (1997/2004) framework with qualitative content analysis are not claimed to be generalisable to language attitude research in the field of ELF. Rather, with his conversational interview study as one example, it should be argued that academic rigour requires researchers in this well-­studied area “to contemplate what and how it is being researched in the name of language attitudes as well as why possible alternatives are dispensable in any given study” (Ishikawa and Morán Panero 2016: 99).

Appendix Transcription conventions (adapted from VOICE Project 2007)

[ overlapping or interrupted speech = latched utterance xxx. the end of a perceivable block of utterance xxx。 significant pause @ laughter and a laughter-­like sound, approximating the syllable number (e.g., ha ha ha = @@@) xxx etc. spoken with specified features

Notes 1 The pilot study, with which the researcher found no major problem, was conducted at a UK university. 2 The examples of the previous application of Eggins and Slade’s (1997/2004) framework include Leung (2012) and Jenkins (2014). 3 The term doxic refers to the taken-­for-­granted ‘reality’ of everyday life which appears, and is perceived to be, “self-­evident and undisputed” (Bourdieu 1977: 164) in a social sphere. 4 Those who are less familiar with ELF awareness and the current thinking in the field of ELF could refer to Ishikawa and Jenkins (2019) and Ishikawa (2020, in press).

References Allport, G.W. 1935. Attitudes. In C. Murchison (ed), A handbook of social psychology. Worcester, MA: Clark University Press, pp. 798–844. Baird, R. and M. Baird. 2018. English as a Lingua Franca: Changing “attitudes”. In J. Jenkins, W. Baker and M. Dewey (eds), The Routledge handbook of English as a Lingua Franca. London: Routledge, pp. 531–543. Baird, R., W. Baker and M. Kitazawa [Baird]. 2014. The complexity of ELF. Journal of English as a Lingua Franca 3(1), 171–196. Baker, W. and J. Jenkins. 2015. Criticising ELF. Journal of English as a Lingua Franca 4(1), 191–198. Banaji, M.R. and L. Heiphetz. 2010. Attitudes. In S.T. Fiske, D.T. Gilbert and G. Lindzey (eds), Handbook of social psychology (5th ed.). Hoboken, NJ: John Wiley & Sons, pp. 353–393. Bassili, J.N. and R.D. Brown. 2005/2014. Implicit and explicit attitudes: Research, challenges, and theory. In D. Albarracín, B.T. Johnson and M.P. Zanna (eds), The handbook of attitudes. London: Psychology Press, pp. 543–574.

Rigour in ELF language attitude research  273 Bourdieu, P. 1977. Outline of a theory of practice (Trans. R. Nice). Cambridge: Cambridge University Press. Campbell, D.T. 1963. Social attitudes and other acquired behavioral dispositions. In S. Koch (ed), Psychology: A study of a science, Investigations of man as socius: Their place in psychology and the social sciences. London: McGraw-­Hill, pp. 94–172. Cargile, A.C., H. Giles, E.B. Ryan and J.J. Bradac. 1994. Language attitudes as a social process: A conceptual model and new directions. Language & Communication 14(3), 211–236. Chan, J.Y.H. 2016. A multi-­perspective investigation of attitudes towards English accents in Hong Kong: Implications for pronunciation teaching. TESOL Quarterly 50(2), 285–313. Converse, P.E. 1970. Attitudes and non-­attitudes: Continuation of a dialogue. In E.R. Tufte (ed), The quantitative analysis of social problems. London: Addison-­Wesley, pp. 168–189. Coupland, N. 2007. Style: Language variation and identity. Cambridge: Cambridge University Press. Coupland, N. and H. Bishop. 2007. Ideologised values for British accents. Journal of Sociolinguistics 11(1), 74–93. Denzin, N.K. and Y.S. Lincoln. 2011. Introduction: The discipline and practice of qualitative research. In N.K. Denzin and Y.S. Lincoln (eds), The Sage handbook of qualitative research (4th ed.). London: Sage Publications, pp. 1–19. Dörnyei, Z. 2007. Research methods in applied linguistics. Oxford: Oxford University Press. Dörnyei, Z. and T. Taguchi. 2010. Questionnaires in second language research: Construction, administration, and processing (2nd ed.). London: Routledge. Dragojevic, M. and H. Giles. 2014. The reference frame effect: An intergroup perspective on language attitudes. Human Communication Research 40(1), 91–111. Eagly, A.H. and S. Chaiken. 2007. The advantages of an inclusive definition of attitude. Social Cognition 25(5), 582–602. Eggins, S. and D. Slade. 1997/2004. Analysing casual conversation. London: Equinox. Fives, H. and M.M. Buehl. 2012. Spring cleaning for the “messy” construct of teachers’ beliefs: What are they? Which have been examined? What can they tell us? In K.R. Harris, S. Graham and T. Urdan (eds), APA educational psychology handbook (volume 2): Individual differences and cultural and contextual factors. Washington, DC: American Psychological Association, pp. 471–499. Gallois, C. and V.J. Callan. 1981. Personality impressions elicited by accented English speech. Journal of Cross-­Cultural Psychology 12(3), 347–359. Garrett, P. 2010. Attitudes to language. Cambridge: Cambridge University Press. Garrett, P., N. Coupland and A. Williams. 2003. Investigating language attitudes: Social meanings of dialect, ethnicity and performance. Cardiff: University of Wales Press. Giles, H. and A.C. Billings. 2004. Assessing language attitudes: Speaker evaluation studies. In A. Davies and C. Elder (eds), The handbook of applied linguistics. Oxford: Wiley Blackwell, pp. 187–209. Halliday, M.A.K. 1984. Language as code and language as behaviour: A systemic-­ functional interpretation of the nature and ontogenesis of dialogue. In R.P. Fawcett, M.A.K. Halliday, S.M. Lamb and A. Makkai (eds), The semiotics of culture and language (volume 1): Language as social semiotic. London: Frances Pinter, pp. 3–35. Halliday, M.A.K. 1994. An introduction to functional grammar (2nd ed.). London: Edward Arnold. Holstein, J.A. and J.F. Gubrium. 2011. Animating interview narratives. In D. Silverman (ed), Qualitative research (3rd ed.). London: Sage Publications, pp. 149–167.

274  Tomokazu Ishikawa Huygens, I. and G.M. Vaughan. 1983. Language attitudes, ethnicity and social class in New Zealand. Journal of Multilingual and Multicultural Development 4(2–3), 207–223. Hyrkstedt, I. and P. Kalaja. 1998. Attitudes toward English and its functions in Finland: A discourse-­analytic study. World Englishes 17(3), 345–357. Ishikawa, T. 2015. Academic rigour in criticising English as a Lingua Franca. Englishes in Practice 2(2), 39–48. Ishikawa, T. 2016a. A study of Japanese university students’ attitudes towards their English. PhD thesis, University of Southampton, Southampton. Ishikawa, T. 2016b. Japanese university students’ attitudes towards their English: Open-­ ended email questionnaire study. In N. Tsantila, J. Mandalios and M. Ilkos (eds), ELF: Pedagogical and interdisciplinary perspectives. Athens: Deree – The American College of Greece, pp. 87–94. Ishikawa, T. 2016c. World Englishes and English as a Lingua Franca: Conceptualising the legitimacy of Asian people’s English. Asian Englishes 18(2), 129–140. Ishikawa, T. 2017a. Japanese university students’ attitudes towards their English: Conversational interview study. In K. Murata and M. Konakahara (eds), Waseda working papers in ELF (vol. 6). Tokyo: Waseda ELF Research Group, pp. 110–125. Ishikawa, T. 2017b. Japanese university students’ attitudes towards their English and the possibility of ELF awareness. Journal of English as a Lingua Franca 6(2), 237–263. Ishikawa, T. 2020, in press. EMF awareness in L1-­shared classrooms. ELT Journal. Ishikawa, T. and J. Jenkins. 2019. What is ELF? Introductory questions and answers for ELT professionals. Center for English as a Lingua Franca Journal 5, 1–10. Ishikawa, T. and S. Morán Panero. 2016. Exploring language attitudes in ELF research: Contrasting approaches in conversation. Englishes in Practice 3(4), 74–109. Jenkins, J. 2007. English as a Lingua Franca: Attitude and identity. Oxford: Oxford University Press. Jenkins, J. 2014. English as a Lingua Franca in the international university: The politics of academic English language policy. London: Routledge. Jenkins, J. 2015. Repositioning English and multilingualism in English as a Lingua Franca. Englishes in Practice 2(3), 49–85. Jenkins, J. 2017. ELF and WE: Competing or complementing paradigms? In E.L. Low and A. Pakir (eds), World Englishes: Rethinking paradigms. London: Routledge, pp. 12–28. Krosnick, J.A., A.L. Holbrook, M.K. Berent, R.T. Carson, W.M. Hanemann, R.J. Kopp, R.C. Mitchell, S. Presser, P.A. Ruud, V.K. Smith, W.R. Moody, M.C. Green and M. Conaway. 2002. The impact of “no opinion” response options on data quality: Non-­attitude reduction or an invitation to satisfice? Public Opinion Quarterly 66(3), 371–403. Kubota, R. 2011. Learning a foreign language as leisure and consumption: Enjoyment, desire, and the business of eikaiwa. International Journal of Bilingual Education and Bilingualism 14(4), 473–488. Labov, W. 1966. The social stratification of English in New York City. Washington, DC: Center for Applied Linguistics. Lambert, W.E., R.C. Hodgson, R.C. Gardner and S. Fillenbaum. 1960. Evaluational reactions to spoken languages. Journal of Abnormal and Social Psychology 60(1), 44–51. Landry, R. and R.Y. Bourhis. 1997. Linguistic landscape and ethnolinguistic vitality: An empirical study. Journal of Language and Social Psychology 16(1), 23–49. Leung, C. 2012. English as an additional language policy-­rendered theory and classroom interaction. In S. Gardner and M. Martin-­Jones (eds), Multilingualism, discourse and ethnography. London: Routledge, pp. 222–240. Lindemann, S. 2005. Who speaks “broken English”? US undergraduates’ perceptions of non-­native English. International Journal of Applied Linguistics 15(2), 187–212.

Rigour in ELF language attitude research  275 Miles, M.B., A.M. Huberman and J. Saldaña. 2014. Qualitative data analysis: A methods sourcebook (3rd ed.). London: Sage Publications. Morán Panero, S. 2016. Orienting to the spread of English as an international Lingua Franca: Voices from the Spanish-­speaking world. PhD thesis, University of Southampton, Southampton. Mortensen, J. 2013. Notes on English used as a Lingua Franca as an object of study. Journal of English as a Lingua Franca 2(1), 25–46. Niedzielski, N.A. and D.R. Preston. 1999/2003. Folk linguistics. Berlin: de Gruyter Mouton. Niedzielski, N.A. and D.R. Preston. 2009. Folk linguistics. In N. Coupland and A. Jaworski (eds), The new sociolinguistics reader. Basingstoke: Palgrave Macmillan, pp. 356–373. Pear, T.H. 1931. Voice and personality. London: Chapman and Hall. Perloff, R.M. 2017. The dynamics of persuasion: Communication and attitudes in the 21st century (6th ed.). London: Routledge. Petty, R.E., D.T. Wegener and L.R. Fabrigar. 1997. Attitudes and attitude change. Annual Review of Psychology 48, 609–647. Potter, J. 1996. Attitudes, social representations and discursive psychology. In M. Wetherell (ed), Identities, groups and social issues. London: Sage Publications, pp. 119–173. Potter, J. and M. Wetherell. 1987. Discourse and social psychology: Beyond attitudes and behaviour. London: Sage Publications. Preston, D.R. 2010. Variation in language regard. In P. Gilles, J. Scharloth and E. Ziegler (eds), Variatio delectat: Empirische Evidenzen und theoretische Passungen sprachlicher Variation. Frankfurt am Main: Peter Lang, pp. 7–27. Preston, D.R. 2018. Perceptual dialectology. In C. Boberg, J. Nerbonne and D. Watt (eds), The handbook of dialectology. Oxford: John Wiley & Sons, pp. 177–203. Rosenberg, M.J. and C.I. Hovland. 1960. Cognitive, affective, and behavioral components of attitudes. In M.J. Rosenberg, C.I. Hovland, W.J. McGuire, R.P. Abelson and J.W. Brehm (eds), Attitude organization and change: An analysis of consistency among attitude components. New Haven: Yale University Press, pp. 1–14. Schreier, M. 2012. Qualitative content analysis in practice. London: Sage Publications. Talmy, S. and K. Richards. 2011. Theorizing qualitative research interviews in applied linguistics. Applied Linguistics 32(1), 1–5. VOICE Project. 2007. VOICE transcription conventions [2.1]: Mark-­ up conventions. Available from: www.univie.ac.at/voice/page/transcription_general_information [accessed 12 September 2019]. Wilson, T.D. and S.D. Hodges. 1992. Attitudes as temporary constructions. In L.L. Martin and A. Tesser (eds), The construction of social judgments. Hove: Lawrence Erlbaum, pp. 37–65. Woolard, K.A. and T.-­J. Gahng. 1990. Changing language policies and attitudes in autonomous Catalonia. Language in Society 19(3), 311–330.

Index

Note: Page numbers in italic indicate a figure and page numbers in bold indicate a table on the corresponding page. Page numbers followed by ‘n’ indicate a note. abnormalities 21 academic achievement 4 academic writing 4, 9; anticipated/ hypothetical reader 64; approximation and fixing 69, 74 – 75; awareness 61; blog posts 59; bottom-up interaction 65; CARS (Create a Research Space) pattern 66; cognitive analyses 69 – 75; complexification 71 – 75; contentbased textual organisation 67 – 68; contrast 73 – 74; corpus data 58; data 58 – 60; digitalisation 63 – 64; editing cycles 57; ELF in academia 61 – 62; evaluative reports and research blogs 59; globalisation and mobility 61 – 62; imposed and spontaneous norms 65 – 66; ‘Inner Circle’ country 59; internationalisation 57, 61 – 63; language activity 58; language processing 69; language regulation 66; local and regional practice 57; macro-social perspective 60 – 64; methods 60 – 61; micro-social perspective 64 – 68; morpohology 71; PhD examiner 59; prepositions 72; processing and entrenchment 69 – 75; qualitative analysis 60 – 61; quality control methods 63; regularisation 69 – 71; research publishing 63 – 64; SciELF texts 59 – 60; second-language use (SLU) corpus 59, 64, 69; simplification 69 – 71; singular and plural noun 73; social interaction 60, 64 – 65; standard languages 66 – 67; subcorpora 59; text norms and effectiveness 64 – 68; text patterning 66 – 67; top-down interaction

64 – 65; traditional and digital forms 57 – 58; translocal text norms 68; Written English as a Lingua Franca in Academic Settings (WrELFA corpus) 59 accent-related identities 244 accents and variation 187 – 189 accessible resource 126 accommodation strategy 43, 119 – 120, 206 – 207, 207, 241 action turns 203 – 204 acts of identity 32 affiliative and mutually supportive behaviors 173 Agha, A. 65 Airey, J. 90 airy voice quality 184 – 185 Alaszewski, A. 247 Allport, G.W. 258 amorphous linguistic interaction 258 anonymization 103, 122n8 anticipated/hypothetical reader 64 applied linguistics 2, 3, 7 – 8, 26, 29 approximation 8, 22 – 23, 25, 69, 74 – 75 articles 137 Asian Corpus of English (ACE) 3, 5, 7, 10, 29 – 30; accessible resource 126; ACE online Homepage 130, 131; archiving of recorded data 129; articles 137; browse corpus interface 130, 131; common verbs 133, 134, 137; concordances 132, 132; data collection sites 127; demonstrative ‘this’ 139; ELF tendencies 133, 140; English as communicative medium of choice 126; listening to sound recordings 131–132, 131; participants for data collection 128;

Index  277 prepositions 139 – 140; question tag 138; recording and data transcription tools 128; relative pronouns interchangeability 136 – 137; searching online  132, 132; speech events in corpus 129; spoken corpus 126 – 127; third person singular 133 – 136, 134; transcription conventions 129 – 130; types of settings of recorded data 128 – 129; uncountable nouns as plural 137 – 138; verb tense 136; and VOICE data 133 – 140; Voicescribe 128 attitude research 258 – 260, 261 – 262; see also conversational interview study asylum seekers and immigration 7, 31 – 32 asymmetries: action turns 203 – 204; analysis methods 202 – 203; casual modality markers 214; contextual factors 199 – 200; cultural asymmetry 199; data collection 200 – 202, 201; distinct in-group address practices 209 – 210; diversity 197 – 198, 215; experience and proficiency 199; explicit repetition 213 – 215; features/ turn structures 203, 216n1; ideational repetition in adjacent utterances 207 – 209, 216n3; interactional social order 202; intercultural communication 199; language stripped bare 199, 203; laughter turn 199, 205 – 206; modality markers 210 – 213, 211, 212; pause and laughter turn 199, 204 – 205; practices to situations correlation 213 – 214; repeated short clauses 206 – 207, 207; settings, shared characteristics of 200, 215; simplification and repetition (accommodation) 206 – 207, 207; situational diversity 197, 214; social interactions 199 awareness 29, 61 back-stage practices 99 Baker, W. 241, 243 BELF see English as a lingua franca in business contexts (BELF) Bhabha, H.K. 225 Biber, D. 82, 87 bilateral constellation 106, 107 bilingualism 40 Bolden, G.B. 186 – 188 bottom-up interaction 65 Brett, D. 182 browse corpus interface 130, 131

Brumfit, C. 26, 29 Bruner, J. 223 Bucholtz, M. 184 business-related EMI setting 5 – 6 Bysouth, D.K. 30 Canagarajah, S. 41, 228 categories 21 casual modality markers 214 Clandinin, D.J. 223 clarity of expression 166 – 167 closed-response items 261 – 262 code-switching 43, 44, 98 – 100, 100 – 101, 167 – 168 cognitive perspective 8 Cogo, A. 3, 7, 8, 38, 42, 43, 99, 100, 105, 116, 164 coinages, use of 41 – 42 collaborative behaviours 241 collaborative writing 63 colloquial/dialectal language use 181 combined repetition 166 – 167 commitment formulaic sequences (FS) 147 Common European Framework of Reference (CEFR) 34 – 35 common repertoire of resources 49 common verbs 133, 134, 137 communication at global level 1 communication use 24 – 25, 26 – 27 communicative capability 32, 222, 237 communicative function 21 community 29, 31 community of practice (CoP) 6, 13, 47, 241, 245 – 246, 254n4 compatibility 162 – 164 competence 29, 31, 35, 163 complexification 71 – 75 compound construction 88, 89 Connelly, F.M. 223 content-based textual organisation 67 – 68 contrast 73 – 74 convergence 117 – 120 conversational interview study: appending moves 263; continuing moves 263; correctness 266 – 268; direct approach of language attitude 262; dispositional concept 265; doxic experience 265, 272n3; Eggins and Slade’s (1997/2004) framework 263 – 265, 264; human development 268 – 269; methodological overview 262 – 263; monitoring moves 263; NES-dominant approach 270; positive attitudes 270; prolonging moves

278 Index 263; qualitative content analysis 263; reacting moves 263; responses and rejoinders 263, 264; social and cultural sensitivity 269 ; supporting responses 263; tracking moves and subsequent responses 264 conversation analysis (CA) 2, 5 – 6, 8 – 10; affiliative and mutually supportive behaviors 173; code-switching 167 – 168; collaboration and cooperation 172 – 173; combined repetition/clarity of expression 166 – 167; communicative success and effectiveness 164; compatibility 162 – 164; concept 161; diverse linguacultural speaker 164; ELF interaction 161 – 162; ELF use 164; emic approach 179; gaps 169, 171 – 172; implications 172 – 174; interactional management of turntaking 169 – 172; international trading negotiations 161; intersubjectivity in talk 164 – 169; language competence 163; as methodology and analytical tool 164, 173 – 174; mutual understanding 168 – 169; non-native speaker interaction 163; overlapping talk 169 – 171; phonological intelligibility 165; preemptive practices 166; pronunciation 165 – 166; recipient design 173; reprehensive talk 173; sequential organization of talk 162, 165; small talk conversations 169; social action 162; turn-by-turn analysis 166; see also transcription co-operativeness 30 corpus annotation 9, 101 – 102, 122n5 corpus-based analysis 1 – 2, 5, 8 corpus compilation 97 – 98 corpus linguistic approach 7, 10 covert and overt resources 8, 23, 38; codeswitching 43, 44; cognitive nature 41, 43, 52n2; common repertoire of resources 49; community of practice 47; covert creativity 47 – 48; idiomatic expressions 42; language choice 50; language contact 41, 47 – 48; language separation/transformation 42 – 43, 47 – 49; methodological observations 44 – 45; monolingual ideologies 50; multilingual nature 41, 43, 45 – 46; multilingual turn 48 – 50; phonological covert resources 47 – 48; pragmatic covert resources 46 – 47; prepositional

nature 46 – 47; pronunciation 47 – 48; repertoire 41 – 42; translanguaging 43 – 44, 48 – 49; use of coinages 41 – 42 Create a Research Space (CARS) model 66, 84 cultural asymmetry 199 deficit approach 39 demonstrative ‘this’ 139 descriptive analyses 1, 3 descriptive linguistics 24 Dewey, M. 42 diachronic changes in research articles: academic English 81 – 82, 90 – 91; compound construction 88, 89; corpus word count, introduction 84 – 85, 85, 93n5; Create a Research Space (CARS) model 84; disciplinary differences 89 – 90; engagement and informal features 82, 89, 93n2; full-sentence construction 88; genre writing 90 – 91; introductory section 83 – 88; language editing 90, 93n7; learning deficiencies 82; Li and Ge research method 83; medicine and economics 83; methods 83 – 84; move structure 85; nominal group construction 88; noun phrases 87; novice-user corpora 82; physical and online services 82, 93n3; sentence structure and length 86 – 88, 87; small corpus analysis 82; structural and grammatical changes 82; titles 83 – 84, 88, 88 – 89; verb tenses and modals 86, 86 dialectology 24 diary 247 – 248, 252 – 253, 254n9 digitalization/digital publishing 63 – 64 direct approach of language attitude 261 – 262 disciplinary acceptability 24 disciplinary standard 9 dispositional concept 265, 259 – 260 diversity 197 – 198, 215 doxic experience 265, 272n3 durability of language 258 Eggins, S. 263 – 265, 271 Eggins and Slade’s (1997/2004) speech functions analysis framework 263 – 265, 264 Ehrenreich, S. 99, 114 emergent (multilingual) practices 119 – 120; see also English as a lingua franca in business contexts (BELF) emic perspective 5 – 6

Index  279 engagement formulaic sequences (FS) 147 engagement and informal features 82, 89, 93n2 English for Academic Purposes (EAP) 9, 81 English Language Program (ELP) 228 English language teaching (ELT) 2 English for Liberal Arts (ELA) 228 English as a Lingua Franca in Academic Settings (ELFA) 3, 5 English as a lingua franca in business contexts (BELF): accommodation 119 – 120; back-stage practices 99; bilateral constellation 106, 107; codeswitching 98 – 100; convergence 120; corpus compilation 97 – 98; data 105 – 106; divergent practice of language choice 120; emergent (multilingual) practices 119 – 120; front-stage realizations 99; group characterization and interaction profile 106 – 109, 107, 108; ideologies of diversity 99; interaction and participation profile 108, 108 – 109; micro-diachronic approach 98, 109 – 119; non-*English speech 100 – 105; predictability and generalizability 121, 122n19; situational multilingual practices 97 – 98; translanguaging 98 – 100; utterances 106 – 108, 107, 108, 122n13; VOICE 99, 100 English as a Medium of Instruction (EMI) 221 English for Specific Purposes (ESP) 9, 81 – 82 English user 233, 236 – 237 Erasmus family 245, 246 ethnomethodology 162 evaluation formulaic sequences (FS) 147 exclusive switches 112, 113 – 116 experience-centered approach 223 explicit repetition 213 – 215 eye dialect transcription 11 – 12; airy voice quality 184 – 185; colloquial/dialectal language use 181; commentary in analysis 185, 194n11; dots 185; firstorder entextualization 182, 194n5; in literary works 181; orthographic alternations 180, 183; pejorative stereotypes 181; plosiveness 184; pronunciation 179 – 180; pronunciation particularities 183 – 186; representability 180; sounds/phonemes 181, 194n4;

sound-symbol combinations risks 182; spelling 181 – 182; syllable 185 – 186; symbols 183; working transcripts 183 – 184 favourableness/disfavourableness 259 Finlay, L. 223 first-order entextualization 182, 194n5 Firth, A. 161, 163, 164 fixing 8, 69, 74 – 75 Flowerdew, L. 82 fluid language practices 40 folk linguistics 259 – 260 foreign word (FW) 101, 101 – 102, 122n5 formulaicity of spoken communication 5 formulaic sequences (FS) 10, 11, 143, 144, 146; automatic identification 153 – 154; classification of interpersonal functions 147; commitment FSs 147; congruency with frequency-based identification 151 – 154, 152; data 145; distribution across disciplines 147 – 148, 148; distribution of functional categories 148, 148 – 149; emotion/cognition 151; engagement FSs 147; evaluation FSs 147; formal variations 152 – 154; frequency and distribution 147 – 151; frequency-based approach 143, 144; grammatical irregularity/semantic opacity 146; hesitation in spoken discourse 153 – 154; identification criteria 146 – 147; IDIOM Search 151 – 152, 152; interpersonal functions 144 – 145; interruption by another formulaic unit 154; modality FSs 147; natural language use 143; open-choice principle 143, 149, 153, 154; politeness formulae 149 – 150; prefabricated sequences 143, 146; questions/invitation for comments 150; situation/register/ genre-specific formula 146 – 147; underlying frame 146, 154; word counts 145, 145 four-step protocol 191 – 193 Franzmann, A. 57, 62 front-stage practices 99 full-sentence construction 88 gaps: in conversation 169, 171 – 172; in transcription 118 – 119, 119 García, O. 40, 43 Gardner, R. 183 Garrett, P. 259, 260

280 Index gate-keeping encounters 31 Ge, G.-C. 82, 83 Graddoll, D. 61, 64 grammatical irregularity/semantic opacity 146 Granovetter, M. 69 Gray, B. 82, 87 global competition 4 globalisation and mobility 1, 61 – 62 Guido, M.G. 32 Haberland, H. 182 Halliday, M.A.K. 10, 203, 263 Hallidayan systemic functional grammar (SFG) 3, 6, 10, 12 Harding, L. 199 Have, P. ten 184 Hepburn, A. 186 – 188 high-stakes encounters 26 – 27, 29, 30 – 31, 32; Common European Framework of Reference (CEFR) 34 – 35; communication 29, 32; community and competence 31, 35; conformity 34; co-operativeness 30; domains 29 – 30; ELF interactions 30 – 31; encounters 31, 34; idiomatic patterns 34; international publishing 34; intervention 29; invariability 33; Language Analysis for the Determination of Origin (LADO) 32 – 34; monolingualism 33, 35; multilingualism 33; power differential 29; transcripts of speech 31 House, J. 98 Hülmbauer, C. 41, 214 Hutchby, I. 164 Hyland, K. 82 – 84, 89, 150 Hynninen, N. 8, 66 ideational repetition 207 – 209, 216n3 identity 241, 243, 252 ideologies of diversity 99 idiomatic expressions 34, 42, 143 Iino, M. 222, 228 imagined community of ELF users 248 – 252 imposed and spontaneous norms 65 – 66 inclusive use of multilingual elements 112, 122n16 in-group address practices 209 – 210 interactional social order 202 interactions types 2, 5 – 6, 11 – 12 interim stages of development 23 internationalisation 57, 61 – 63

international publishing 4, 34, 57 International Phonetic Alphabet (IPA) 12, 180, 189 – 190 international trading negotiations 161; see also conversation analysis (CA) interpersonal formulaic sequences (FSs) 11; see also formulaic sequences (FS) intersubjectivity in talk 164 – 169 intervention 29 introduction corpus word count 84 – 85, 85, 93n5 Ishikawa, T. 12, 13, 258 Jefferson, G. 179 – 181, 183 – 186 Jeffersonian system 179 Jenkins, J. 4, 24, 44, 98, 244, 258 Jenks, C.J. 30, 173 Jiang, F. 82, 84, 89 Kalocsai, K. 241, 246, 252 Kanno, Y. 228 Kappa, K. 172, 173 Kaur, J. 5, 6, 11, 161, 164, 166, 167 Kirkpatrick, A. 30 Kjellmer, G. 143 Klimpfinger, T. 43 Konakahara, M. 169 – 171 Kowal, S. 181 Krapp, G.P. 182 Kuteeva, M. 90 L1/LN tags function 110 – 111, 111, 116, 117 Labov, W. 258 Laitinen, M. 64 Lambert, W.E. 261 Language Analysis for the Determination of Origin (LADO) 32 – 34 language: attitudes 258; contact 39, 41; editing 90, 93n7; ideologies 41; of globalization 38, 41; of international communication 38; learner 230 – 231, 236; policy and planning 2; regulation 66; stripped bare 199, 203; see also attitude research laughter turn 199, 205 – 206 Le Page, R. 32 lexico-grammatical aspects of ELF interaction 5, 10 Liddicoat, A.J. 162 Li, L.-J. 82, 83 lingua franca communication and users attitude 1, 13; see also English as a

Index  281 lingua franca in business contexts (BELF) linguistic features 241 – 242 linguistic forms 21 – 24 linguistic resources 39 Li Wei 40, 43 longitudinal diary study 13 Mabe, M. 89 macro-social perspective 8 Martinez, R. 146 Maswana, S. 4, 9, 11, 81, 83 matched guise technique (MGT) 261 Matsumoto, Y. 165, 166 Mauranen, A. 4, 5, 8, 10, 22 – 24, 57, 64, 126, 154, 164 McArthur, T. 181 meaning-making process 225 Meierkord, C. 169, 171, 199, 204 Mežek, S. 199, 205 micro-diachronic analysis 9, 98; exclusive and intransparent L1 side sequences 115 – 116; exclusive and transparent L1 side sequences 113 – 115; exclusive switches 112, 116; gaps in transcription 118 – 119, 119; inclusive use of multilingual elements 112, 122n16; L1/ LN tags function/content 110 – 111, 111, 116, 117; L1 tags in side sequences 117 – 118, 118; *languages and speakers 109 – 111, 110, 111; main meeting phase 113 – 116; non-*English elements (L1/LN tags) 109 – 110, 110; opening phase of meeting 111 – 112; situational multilingual etiquette 115 micro-social perspective 8 Miyahara, M. 12, 13, 221 mock phonetic spelling 181 – 182 modality formulaic sequences (FS) 147 modality markers 210 – 213, 211, 212, 217n4 monolingualism 33, 35, 39, 40 Mortensen, J. 182 move structure 85 multilingualism 23 – 24, 33, 38, 43; practices 5, 6, 8, 9; resources 9 multilingual turn 39 – 40, 48 – 50 Mulvihill, T. 221, 225, 226, 232 Murata, K. 1, 221, 222, 228 narrative approach 2, 6 – 8, 12 – 13, 221 – 223, 242; accent-related identities 244; accommodation 241; collaborative

behaviours 241; community of practice 241, 245 – 246, 254n4; culture and 241, 242 – 243; diary observation 247 – 248, 252 – 253, 254n9; EFL teacher identity and attitude 243 – 244; Erasmus family 245, 246; ethnographic research 243; identity of ELF speakers 241, 243; imagined community of ELF users 248 – 252; interactional and narrative approaches 252; learner to ELF user 244 – 245; linguistic features 241 – 242; NSE-level proficiency 244; participants and method 247 – 248; practice and negotiation of identity 252; qualitative research method 242, 247; social morality 244; social practice 241; study 246 – 247; see also place-reflexivity Nava, A. 244 Negretti, R. 90 NES-dominant approach 270 next-turn proof procedure 190 – 191 Niedzielski, N.A. 259 Nikula, T. 222, 244 Nogami, Y. 12, 13, 241 nominal group construction 88 non-attitudes 258, 260 non-*English elements (L1/LN tags) 109 – 111, 110, 111, 116, 117 – 118, 117, 118 non-*English speech 100 – 103, 104, 105, 122n5, 122n7, 122n8; see also English as a lingua franca in business contexts (BELF) non-standard spelling 182 noun phrases 87 novice-user corpora 82 NSE-level proficiency 244 O’Connell, D.C. 181 open-access and digitalisation 8 open-choice principle 143, 149, 153, 154 open-ended items 262 open publishing 64 orthographic alternations 180, 183 overlapping talk 169 – 171 overt interaction 8, 23; see also covert and overt resources parole/performance phenomenon 21 pause and laughter turn 199, 204 – 205 pedagogical approach 232 Pedrazzini, L. 244 pejorative stereotypes 181

282 Index perceptual dialectology 262 Phan, L.H. 245 phonemes 181, 194n4 phonetic transcription system 189 – 190 phonological covert resources 47 phonological intelligibility 165 Pietikäinen, K. 5, 6, 11, 167, 168, 171, 172, 179 Pitzl, M.-L. 5, 9, 10, 11, 76, 97, 98 place-reflexivity 12 – 13; agency to place 238; analysis stage of research 226 – 227; analytical model 229 – 230; autobiographical narratives 229; bilingual environment 231 – 232; communicative capability 222, 237; conceptualization 237, 237; cultural diversity 236; data collection 229; ELF identity 222 – 223; English user 233, 236 – 237; experiencecentered approach 223; features 226; interrelated dimensions 225; language learner 230 – 231, 236; limitations 238; meaning-making process 225; methodological complexities 225; narrative approach 221 – 223; pedagogical approach 232; positioning 225 – 227; power structures and relationships 225; reflection vs. reflexivity 223; relevance 226 – 227; researcher reflexivity 223 – 225; research site and participants 228 – 229; sample study overview 227 – 228; selfawareness as user 233; self-reflexivity 224; Six Step Analysis Model 229 – 230; social relationship 221, 224, 225, 231 – 232; space, place and site idea 225; storied lives 223; Study English Abroad (SEA) Program 229, 234 – 236; subjectivity of researcher 228; subject position transition 227 plosiveness 184 PluS project 33 power differential 29 practices to situations 213 – 214 pragmatic meaning 24 – 25 pragmatic covert resources 47 predictability and generalizability 121, 122n19 pre-emptive practices 166 prefabricated sequences 143, 146 prepositions 72, 139 – 140 Preston, D.R. 180, 259 processing and entrenchment 69 – 75

pronunciation 165 – 166 pronunciation particularities 183 – 186 Pullin, P. 199 quality control methods 63 question tag 138 real-world solution 3, 7 – 8 Receptive multilingualism 52n2 recipient design 173 reflection vs. reflexivity 223 regularisation 69 – 71 regularities 21 relative pronouns, interchangeability of 136 – 137 Relph, E. 232 repeated short clauses 206 – 207, 207 repertoire 39 – 41 reprehensive verbal behaviour 30, 173 representability 180 researcher reflexivity 223 – 225 research: publishing 63 – 64; science 57 Roman Jakobson factors 24 Sacks, H. 162, 169 – 171 Sadoshima, S. 4 Schegloff, E.A. 163 Schmitt, N. 146 Schneider, E. 23, 144 SciELF texts 59 – 60 second language acquisition (SLA) 2 second-language use (SLU) 59, 64, 69 Seedhouse, P. 163 Seidlhofer, B. 1, 3, 7, 23, 29, 126, 161, 241 self-awareness as user 233 self-reflexivity 224 semi-phonetic spelling 182 sentence structure and length 86 – 88, 87 sequential organization of talk 162, 165 Sidnell, J. 163, 165 Sinclair, J.M. 64, 143 similects 22 – 23 simplification 69 – 71, 206 – 207, 207 singular and plural noun 73 situational diversity 197, 214 situational multilingual practices 97 – 98, 115; see also English as a lingua franca in business contexts (BELF) situational variability and regularities in ELF speech 57 situation/register/genre-specific formula 146 – 147 Six Step Analysis Model 229 – 230

Index  283 Slade, D. 263 – 265, 271 small corpus analysis 82 small talk conversations 169 Snow, C.E. 87 social action 162 social interaction 60, 64 – 65 social morality 244 social practice 241 social relationship 221, 224, 225, 231 – 232 societal treatment to attitude 261 sociocultural approach 6 socio-linguistics 29, 258 Soja, E.W. 225 Soler, V. 84 spoken corpus research 5, 9 spontaneous language regulation 8 standard language ideology 4, 66 – 67 storied lives 223 stories 242 structural and grammatical changes 82 Study English Abroad (SEA) Program 229, 234 – 236 superdiversity 197 Swales, J.M. 66, 84 Swaminathan, R. 221, 225, 226, 232 syllable 185 – 186 Systemic Functional Linguistics (SFL) 147 Tabouret-Keller, A. 32 talk-in-interaction 11 – 12 teacher identity and attitude 243 – 244 text vs. discourse 21 text norms and effectiveness 64 – 68 third person singular 133 – 136, 134 Thompson, A. 5, 6, 11, 12, 197 Tilley, C. 224 titles 83 – 84, 88, 88 – 89 Toomaneejinda, A. 199 top-down interaction 64 – 65 transcription conventions 45, 52, 129 – 130, 175, 186, 216, 272 transcription 11 – 12, 100 – 103; accents and variation 187 – 189; four-step protocol 191 – 193; International Phonetic Alphabet (IPA) 180, 189 – 190; Jeffersonian system 179; Latin script 186; matrix language 187; multilingual interaction 190 – 194; next-turn proof procedure 190 – 191; phonetic system 189 – 190; pronunciation 179 – 180; recommendations 193 – 194; see also eye dialect transcription

translanguaging 2, 8; bilingualism 40; codeswitching 43, 44; cognitive repertoires 43; communication 40 – 41; countability 44; English as a lingua franca in business contexts (BELF) 98 – 100; fluid language practices 40; language ideologies 41; linguistic hierarchies 40; linguistic resources 43 – 44; monolingual ideologies 40, 50; multilingualism 39, 48 – 50; professional community 48 – 49; repertoire 39 – 41; social justice perspective 40; transcription conventions 52; transformative perspective 40 Transient International Groups (TIG) 10 Tribble, C. 81 turn-by-turn analysis 166 Turner, J. 4 Turner, V. 225 uncountable nouns as plural 137 – 138 underlying frame 146, 154 usage-based research 3, 5, 7, 21 – 22 use-based research 3, 5, 7, 21 – 22 users identity construction 2 utterances 106 – 108, 107, 108, 122n13 variety 21, 23 verbal accounts 259 verbal guise technique (VGT) 261 verb tenses and modals 86, 86, 136 Vetchinnikova, S. 69 Vettorel, P. 43 Vienna Oxford International Corpus of English (VOICE) 3, 5, 7, 9, 10, 29, 43, 99, 100, 126, 133 – 140 Virkkula, T. 222, 244 Voicescribe tool 128 Wagner, J. 183 Walkinshaw, I. 30 Walpole, J.R. 182 Wang, L. 5, 9, 10, 126 Wang, Y. 143 Widdowson, H. 21, 122n2; communicative capability 32 working transcripts 183 – 184 written corpora research 4, 8 – 9 Written English as a Lingua Franca in Academic Settings (WrELFA corpus) 4, 8, 59; see also academic writing