Dialects from Tropical Islands: Caribbean Spanish in the United States 1138069752, 9781138069756

Table of contents :
Cover
Half Title
Series Page
Title
Copyright
Dedication
Contents
Acknowledgments
List of contributors
Introduction: Caribbean Spanish dialects in the United States: theoretical, empirical, and sociolinguistic perspectives
SECTION 1 Phonetics and phonology
1 Rhotic realizations of the Puerto Rican community in Western Massachusetts and Puerto Rico
2 Differences in the use of /l/ and /ɾ/ in two communities of Puerto Rican Spanish speakers in the United States
3 Laterals in contact: Miami-Cuban Spanish and English /l/
4 Se comen la [s] pero a veces son muy fisnos: observations on coda sibilant elision, retention, and insertion in popular Dominican(-American) Spanish
5 The sociolinguistic distribution of Puerto Rican Spanish /r/ in Grand Rapids, Michigan
SECTION 2 Morphology and syntax
6 Explaining pronominal subject placement variation across two generations of Caribbean Spanish speakers in New York City
7 The effect of person on the subject expression of Spanish heritage speakers
SECTION 3 Sociolinguistic perspectives
8 Evidence of creolized English grammar in the Spanish of Dominicanos on St. Thomas, US Virgin Islands
9 Puerto Rican evaluations of varieties of Spanish
10 Aquí no se cogen las guaguas: language and Puerto Rican identity in San Diego
11 Caribbean Spanish influenced by African American English: US Afro-Spanish language and the new US Caribeño identity
Index

Citation preview

i

Dialects from Tropical Islands

Dialects from Tropical Islands: Caribbean Spanish in the United States provides a comprehensive account of current research on Caribbean Spanish in the United States from different theoretical perspectives and linguistic areas. This edited volume highlights current scholarship and linguistic analyses in four major areas relative to Caribbean Spanish in the United States: phonological and phonetic variation, morphosyntactic approaches, sociolinguistic perspectives, and heritage-language acquisition. This volume will be of interest to linguists and philologists who specialize in Spanish, Caribbean Spanish, Spanish in the United States, or in Romance languages in general. Wilfredo Valentín-Márquez is an assistant professor of Spanish at Millersville University. Melvin González-Rivera is an associate professor of Spanish at the University of Puerto Rico, Mayagüez.

Routledge Studies in Hispanic and Lusophone Linguistics Series Editor: Dale Koike University of Texas at Austin

The Routledge Studies in Hispanic and Lusophone Linguistics series provides a showcase for the latest research on Spanish and Portuguese linguistics. It publishes select research monographs on various topics in the field, reflecting strands of current interest. Titles in the series: Current Research in Puerto Rican Linguistics Edited by Melvin González-Rivera Lexical Borrowing and Deborrowing in Spanish in New York City Towards a synthesis of the social correlates of lexical use and diffusion in immigrant contexts Rachel Varra Biculturalism and Spanish in Contact Sociolinguistic case studies Edited by Eva Núñez-Méndez Lusophone, Galician, and Hispanic Linguistics Bridging frames and traditions Edited by Gabriel Rei-Doval and Fernando Tejedo-Herrero Pragmatic Variation in Service Encounter Interactions across the Spanish-Speaking World Edited by J. César Félix-Brasdefer and María Elena Placencia Dialects from Tropical Islands Caribbean Spanish in the United States Edited by Wilfredo Valentín-Márquez & Melvin González-Rivera For more information about this series please visit: www.routledge.com/RoutledgeStudies-in-Hispanic-and-Lusophone-Linguistics/book-series/RSHLL

iii

Dialects from Tropical Islands Caribbean Spanish in the United States Edited by Wilfredo Valentín-Márquez and Melvin González-Rivera Series Editor: Dale Koike Spanish List Advisor: Javier Muñoz-Basols

First published 2020 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 © 2020 selection and editorial matter, Wilfredo Valentín-Márquez & Melvin González-Rivera; individual chapters, the contributors The right of Wilfredo Valentín-Márquez & Melvin González-Rivera to be identified as the authors 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 Names: Valentín-Márquez, Wilfredo, editor. | González-Rivera, Melvin, editor. Title: Dialects from tropical islands : Caribbean Spanish in the United States / edited by Wilfredo Valentín-Márquez & Melvin González-Rivera. Description: New York : Routledge, 2020. | Series: Routledge studies in Hispanic and Lusophone linguistics | Includes bibliographical references and index. Identifiers: LCCN 2019025778 (print) | LCCN 2019025779 (ebook) | ISBN 9781138069756 (hardback) | ISBN 9781315115443 (ebook) Subjects: LCSH: Spanish language—Dialects—Caribbean Area. | Spanish language—Dialects—United States. | Spanish language—Spoken Spanish—Caribbean Area. | Spanish language—Spoken Spanish— United States. Classification: LCC PC4838 .D53 2020 (print) | LCC PC4838 (ebook) | DDC 460.9/8611073—dc23 LC record available at https://lccn.loc.gov/2019025778 LC ebook record available at https://lccn.loc.gov/2019025779 ISBN: 978-1-138-06975-6 (hbk) ISBN: 978-1-315-11544-3 (ebk) Typeset in Times New Roman by Apex CoVantage, LLC

v

A mis hijos Edwin y Nelson, quienes me mantienen al día con las innovaciones del español puertorriqueño. —Wilfredo A Alma Simounet, con el mayor de los agradecimientos. —Melvin

vii

Contents

Acknowledgments List of contributors

ix x

Introduction: Caribbean Spanish dialects in the United States: theoretical, empirical, and sociolinguistic perspectives

1

WILFREDO VALENTÍN-MÁRQUEZ AND MELVIN GONZÁLEZ-RIVERA

SECTION 1

Phonetics and phonology 1 Rhotic realizations of the Puerto Rican community in Western Massachusetts and Puerto Rico

5

7

ALBA ARIAS

2 Differences in the use of /l/ and /ɾ / in two communities of Puerto Rican Spanish speakers in the United States

35

MICHELLE F. RAMOS-PELLICIA

3 Laterals in contact: Miami-Cuban Spanish and English /l/

54

BRANDON M. A. ROGERS AND SCOTT M. ALVORD

4 Se comen la [s] pero a veces son muy fisnos: observations on coda sibilant elision, retention, and insertion in popular Dominican(-American) Spanish

71

ALMEIDA JACQUELINE TORIBIO AND ARIS MORENO CLEMONS

5 The sociolinguistic distribution of Puerto Rican Spanish /r/ in Grand Rapids, Michigan WILFREDO VALENTÍN-MÁRQUEZ

88

viii Contents

viii

SECTION 2

Morphology and syntax 6

Explaining pronominal subject placement variation across two generations of Caribbean Spanish speakers in New York City

111

113

CAROLINA BARRERA-TOBÓN AND ROCÍO RAÑA RISSO

7

The effect of person on the subject expression of Spanish heritage speakers

128

ANA DE PRADA PÉREZ AND INMACULADA GÓMEZ SOLER

SECTION 3

Sociolinguistic perspectives 8

Evidence of creolized English grammar in the Spanish of Dominicanos on St. Thomas, US Virgin Islands

151

153

DANIEL S. D’ARPA

9

Puerto Rican evaluations of varieties of Spanish

166

EVA-MARÍA SUÁREZ BÜDENBENDER

10

Aquí no se cogen las guaguas: language and Puerto Rican identity in San Diego

184

ANA CELIA ZENTELLA

11

Caribbean Spanish influenced by African American English: US Afro-Spanish language and the new US Caribeño identity

201

TERESA SATTERFIELD AND JOSÉ R. BENKÍ JR.

Index

220

ix

Acknowledgments

We acknowledge with immense gratitude the contribution of the group of scholars who served as reviewers and offered pertinent recommendations to the chapter authors: Gabriela Alfaraz, Scott M. Alvord, Alba Arias, Phillip Carter, Daniel S. D’Arpa, Justin Davidson, Nydia Flores-Ferrán, Jonathan Holmquist, Tania Leal, Antonio Medina-Rivera, Jim Michnowicz, Lauren Miller, Julia Oliver Rajan, Michelle F. Ramos-Pellicia, and Israel Sanz-Sánchez. We are also thankful to the colleagues who submitted their work for publication in this volume.

Contributors

Scott M. Alvord, Brigham Young University Alba Arias, Roanoke College Carolina Barrera-Tobón, DePaul University José R. Benkí, Jr., University of Michigan and ForeSee Daniel S. D’Arpa, Mercer County Community College Inmaculada Gómez Soler, University of Memphis Aris Moreno Clemons, The University of Texas Ana de Prada Pérez, Maynooth University Michelle F. Ramos-Pellicia, California State University San Marcos Rocío Raña Risso, The CUNY Graduate Center Brandon M. A. Rogers, Ball State University Teresa Satterfield, University of Michigan Eva-María Suárez Büdenbender, Shepherd University Almeida Jacqueline Toribio, The University of Texas Wilfredo Valentín-Márquez, Millersville University Ana Celia Zentella, Professor Emerita, UC San Diego and Hunter College/CUNY

1

Introduction Caribbean Spanish dialects in the United States: theoretical, empirical, and sociolinguistic perspectives Wilfredo Valentín-Márquez and Melvin González-Rivera This volume aims to provide a comprehensive state of the art of current research on Caribbean Spanish in the United States from different theoretical perspectives and linguistic areas. Specifically, it highlights current scholarship and linguistic analyses on three major areas that correspond to the sections in which the book is organized: (1) phonological and phonetic variation, (2) morphosyntactic approaches, and (3) sociolinguistic perspectives. The term “Caribbean Spanish” refers here to a linguistic cluster of three dialects: the Spanish spoken in Cuba, Dominican Republic, and Puerto Rico by the 25 million inhabitants of those islands and the 7 million speakers in the mainland United States who identify themselves as Caribbean Hispanics according the U.S. Census 2010. We also consider the territories of Puerto Rico and the US Virgin Islands as part of the geographic scope of the term “the United States” in the title of this volume. Although each of the Caribbean Spanish varieties has its own distinctive characteristics, their grouping into a single category is based not only on geographic proximity but also on the structural features shared by them. In a global vision of the region, the term “Caribbean Spanish” may include the Spanish varieties of Panama, coastal Colombia, coastal Venezuela, and parts of coastal Central America and Mexico. However, due to the obvious constraints of a volume of this nature, we limit the term here to the native varieties of the Hispanic insular Caribbean exclusively: Cuba, Dominican Republic, and Puerto Rico. In tune with this particular criterion, we have the phrase “dialects from tropical islands” in the title, therefore excluding Panama, Colombia, Venezuela, Central America, and Mexico. Along with the scrutiny of dialectal peculiarities regarding the internal structure of the Spanish language, this collection includes studies that contextualize Caribbean Spanish in a social dimension, including a look into the implications that the linguistic projection of a particular Caribbean identity has for intra- and interethnic relations. Regarded among the most innovative varieties of the language, Caribbean Spanish dialects are of interest to linguistic research, and the conditions for language and dialect contact set in the United States make this group a fascinating object of study.

2 Wilfredo Valentín-Márquez et al.

2

In the first section, Phonological and phonetic variation, Chapter 1 examines variation in the production of the PRS trill /r/ in two cities of Western Massachusetts with large Puerto Rican populations, Holyoke and Springfield. Alba Arias studies which phonetic realizations are the most common and which factors might predict their use, paying special attention to the backed /r/ ([x], [χ]) and analyzing the transmission of the trill realization among first- and secondgeneration PRS speakers. In Chapter 2, Michelle F. Ramos-Pellicia compares the usage patterns of the variants of /ɾ / and /l/ among three generations of speakers in two Puerto Rican communities, one in the US East Coast and one in the US Midwest. She also explores whether the third generation’s variety displays phonological patterns comparable in frequency and environment to those found in the first and second generations in both communities. The results show that Puerto Ricans on the East Coast demonstrated a preference for lateralization in word-final position, whereas the preference for [ɾ ] declines across the three different generations of Midwest Puerto Ricans. The combination and interaction of linguistic (e.g., phonological environment, lexical context) and extralinguistic factors (e.g., level of education, exposure to and use of Spanish within the community, linguistic isolation, and negative prestige) result in the patterned use of the variants of /ɾ / and /l/ in both communities. In Chapter 3, Brandon M.A. Rogers and Scott M. Alvord study contactinduced change on Miami-Cuban Spanish and English phonology. They acoustically examine Miami-Cuban Spanish and English /l/. Miami Cubans from three immigrant groups were recorded in both Spanish and English, and tokens of /l/ from both Spanish and English were then extracted and measured for “lightness” or “darkness” by subtracting the F1 measurements from their corresponding F2 measurements. Results are indicative of Miami-Cuban English /l/ being less resistant to contact-induced change than its Spanish counterpart. In Chapter 4, Almeida Jacqueline Toribio and Aris Moreno Clemons contribute additional insight to the understanding of the unwarranted appearance of [s] in Dominican Spanish being conditioned by phonological and social factors. They examine ways in which [s]-maintenance and [s]-intrusion are sanctioned in everyday interactions and in public discourse. Their work draws on observations of the incidence of coda-[s] in the audio and visual landscapes of Dominican Republic and on its representation and evaluation on social media and press outlets in national and diasporic communities. In Chapter 5, Wilfredo Valentín-Márquez examines the distribution of the main variants of /r/ among Puerto Ricans in Grand Rapids, Michigan, a city where Puerto Ricans are a minority in the total population and also in the Hispanic community. Since opportunities to find Puerto Rican speakers whose social ties put them in more frequent interaction with speakers of other varieties are greater in a city where they constitute a smaller fraction of the Hispanic population, Grand Rapids provides a favorable scenario to study the effects of language and dialect contact on PRS. Hence, while the central part of the investigation is a quantitative analysis of the sociolinguistic variation of /r/ based on age and

3

Introduction 3

gender, the study also seeks to determine whether differences associated with the speakers’ maintenance of social ties within their national group provide explanatory insight into the distribution of /r/. In Section 2, Morphology and syntax, Chapter 6 reports on the placement of subject personal pronouns by first- and second-generation Spanish-speaking Caribbeans in New York City. Carolina Barrera-Tobón and Rocío Raña Risso observe that differences across and within generations in pronominal subject placement are conditioned by speakers’ English skills, socioeconomic status, and education. They examine whether these differences can be explained as a consequence of the differential input for acquisition of Spanish received by the first- and second-generation speakers—in other words, as evidence of the development of a dialect of Spanish in transition from a Caribbean variety to a US variety, with a word order more similar to that of English. In Chapter 7, Ana de Prada Pérez and Inmaculada Gómez Soler explore preferences in subject expression forms among Spanish heritage speakers enrolled in a public university in Florida. They examine the effects of language contact on subject expression in heritage speakers of different proficiencies (high vs. low) and varieties (Caribbean vs. non-Caribbean), contrasting data from two different grammatical persons (first yo “I” and third él/ella “he/she”) and three speech connectivity contexts. Finally, Section 3 includes chapters based on sociolinguistic perspectives. In Chapter 8, Daniel S. D’Arpa discusses how Spanish spoken by Dominicans on St. Thomas, US Virgin Islands, shows evidence of influence from Englishlanguage features that are consistent with Dominican Spanish in other contact environments and some new features that are emerging as the result of uniquely St. Thomas English Creole influences. Since Dominicans on St. Thomas report that they are not fully accepted by English-speaking St. Thomians, this paper considers the social aspirations of Dominicans and whether their language choices are explained by these social pressures. Additionally, it examines how Dominican youths make language choices that function as speech acts in which they mean to negotiate their identity on a spectrum between assimilating to St. Thomian culture and choosing solidarity with their Dominican heritage. In Chapter 9, Eva-María Suárez Büdenbender examines Puerto Ricans’ attitudes and perceptions toward their own dialect and other varieties of Spanish. She reports on largely positive views of PRS, although the effects of bilingualism with English and the increase of code-switching in younger speakers is of concern to many participants. Also, the informants appear to seek to distance themselves from other Caribbean varieties. Furthermore, the attribution of linguistic prestige seems to be guided by the perceived socioeconomic status and level of education of the speakers of these varieties. In Chapter 10, Ana Celia Zentella investigates the effects of linguistic and cultural swamping on the linguistic practices and identities of Puerto Ricans far from their primary island and mainland bases. She reports on the results of research in San Diego, where Puerto Ricans live 30 miles from the Mexican border and constitute less than 1% of the population. Interviews and vocabulary questionnaires

4 Wilfredo Valentín-Márquez et al.

4

probe the extent to which PRS in San Diego has changed as a result of contact with Mexican Spanish and the impact on Puerto Rican identity. Of particular interest are the familial, gender, economic, cultural, and political variables that make most continue to speak in traditional ways, while some switch to Mexican vocabulary and phonology completely or selectively. Despite experiencing some negative Mexican attitudes toward PRS, there was unanimous support for Mexicans, for the maintenance of Spanish in general, and for PRS in particular. However, Zentella argues that the loss of Spanish in the next generation militates against the formation of a PR-Mex koine and may be responsible for relaxing the Spanish requirement for Puerto Rican identity. In Chapter 11, Teresa Satterfield and José R. Benkí Jr. present a new US Spanish variety termed “US Afro-Spanish,” which likely originated in US East Coast communities with racially diverse Spanish-speaking Caribbean immigrants in intense contact with Anglophone African Americans. They explore this variety’s status as a Spanish ethnolect/in-group sociolect, documenting its phonetic and phonological bases through acoustic analyses and demonstrating its salient morphosyntactic features. The authors hypothesize that, as urban US-Caribbean Latino youth navigate intricate social networks, they reject sociocultural markers of “whiteness” to intentionally extend stigmatized “black” linguistic indices across a bilingual repertoire of English and Spanish, thereby constructing new identities as 21st-century American youth of color. Compared to the significant presence of Caribbean Hispanics in the United States, current research on this population—and crucially its ramifications to areas such as language contact, change, and education—has been somehow underexplored. New theoretical perspectives and analyses have been desperately called for given the increasing interest in this area of research among graduate students and scholars. The goal of this volume is to cover these existing lacunae on Caribbean Spanish in the United States by providing a state-of-the-art collection of articles from different perspectives and linguistic areas, maintaining a coherence among its contents in spite of the varied approaches. Because the book focuses on the native varieties of the Hispanic insular Caribbean spoken in the United States and in US territories, where interaction with speakers of English and of other Spanish dialects occurs, the common thread defining the cohesiveness of the chapters stems from a contact linguistics framework. We thank all the contributors and reviewers who helped us complete this project and are glad to offer an encompassing state-of-the-art review on Caribbean Spanish in the United States.

5

Section 1

Phonetics and phonology

7

1

Rhotic realizations of the Puerto Rican community in Western Massachusetts and Puerto Rico Alba Arias

1. Introduction The Spanish trill /r/ is described as having two or more brief occlusions between the tongue apex and the alveolar ridge (Martínez Celdrán, 1998). However, outside of the prescriptive description, experimental studies have shown an enormous amount of variation in the actual production of trills (Díaz-Campos, 2008; Willis, 2006, 2007; Willis & Bradley, 2008). This study focuses on language variation in the Puerto Rican diaspora in Western Massachusetts, a community where due to the unique legal situation between Puerto Rico and the United States, back-and-forth migration waves have taken place since 1950 (Center for Puerto Rican Studies, 2016). More concretely, this analysis provides a comprehensive description of onset trill in the cities of Holyoke and Springfield (Western Massachusetts) in order to identify differences between those trill realizations found in the diaspora and in Puerto Rico. Both linguistic and sociolinguistic factors are examined (Beaton, 2015; Delgado-Díaz & Galarza, 2015; Ramos-Pellicia, 2004; Valentín-Márquez, 2007). In this manner, the present study addresses larger questions in the field regarding social factors underlying language variation in diasporic communities.

2. Background 2.1 A sociolinguistic context: Puerto Rican diaspora The United States is home to the second largest Spanish-speaking population in the world: more than 46 million speakers (Instituto Cervantes, 2017). The vitality of US Spanish is due to the high immigration rates from several Spanish-speaking countries over decades, with most of the Latino population coming from Mexico (65% of all US Latinos), followed by Puerto Ricans (9%), Cubans (4%), and Dominicans (2.8%) (Pew Hispanic Center, 2012; Potowski, 2015). Interestingly, the Latino population in the state of Massachusetts does not reflect these numbers. Although Mexicans are the largest origin group in the United States, Puerto Ricans are the dominant group in Massachusetts, making up 46% of its total Latino population (Pew Hispanic Center, 2012). In urban

8 Alba Arias

8

centers in Western Massachusetts specifically, there has been a historical presence of Puerto Ricans, with migrations occurring since 1950. Puerto Ricans often migrated directly from the island to Western Massachusetts to work in tobacco fields, but others originally immigrated to New York City or Hartford, Connecticut, but moved north in search of employment in seasonal agriculture and blue-collar industries (Our Plural History, 2008). The Puerto Rican population in Massachusetts has increased over the past decades, reaching 266,125 habitants in 2010 (U.S. Census Bureau, 2010), or 9.6 % of the total population of Massachusetts. Two of the cities where Puerto Rican settlements were established are the cities of Holyoke and Springfield. Springfield, on the eastern bank of the Connecticut River, has an estimated population of 153,060 (U.S. Census Bureau, 2010). Holyoke, 8 miles north, has a population of 40,135 (U.S. Census Bureau, 2010). Holyoke has the largest population per capita of Puerto Ricans outside the island of Puerto Rico, while Springfield ranks 7th (U.S. Census Bureau, 2010). With the exception of the impressionistic work of Shouse de Vivas (1986), the variety of Spanish spoken in the urban centers of Holyoke and Springfield remains unexplored. This circumstance justifies the importance of analyzing Puerto Rican Spanish (PRS) in Massachusetts. In this study, we are concerned specifically with the phoneme /r/. 2.2 Phonetic description and distribution of /r/ According to traditional phonological descriptions, Spanish has two contrastive rhotic sounds, the vibrante simple “tap” /ɾ / and the vibrante múltiple “trill” /r/. The Spanish trill /r/, the linguistic variable of interest in this project, is described as having two or more brief occlusions between the tongue apex and the alveolar ridge for its normative realization (Hualde, 2005; Martínez Celdrán, 1998). Recasens and Pallarés (1999) point out that this rhotic presents a certain articulatory difficulty, requiring precise control of aperture and airflow with minimal deviation in oropharyngeal and subglottal pressure (Henriksen, 2014). This fact may explain why outside of the traditional description, experimental studies have shown an enormous amount of phonetic variation in the trill that differs in place of articulation (coronal, velar, and uvular), manner of articulation (approximants, fricatives, taps, flaps, and vocoids) and laryngeal setting (voiced, voiceless, and breathy voiced) (Blecua Falgueras, 2001; Bradley & Willis, 2012; Díaz-Campos, 2008; Henriksen & Willis, 2010; Henriksen, 2014; Willis, 2006, 2007; Willis & Bradley, 2008). Regarding distribution, the trill contrasts with the tap in intervocalic position, resulting in minimal pairs such as para “for” vs. parra “vine.” Outside of this context, taps and trills can be in complementary distribution: trills appear in word-initial position or after a heterosyllabic consonant (e.g., rosa “rose” and Israel “Israel”), while taps occur after a tautosyllabic consonant in an onset cluster or word-final position before a vowel-initial word. Taps and trills can also show free variation in word-medial position before a consonant and in word-final position before a consonant or a pause (Hualde, 2005). For the

9

Rhotics in Massachusetts and Puerto Rico 9

purpose of this study, trill variation will be analyzed in intervocalic position, where this segment is clearly contrastive, as well as in word-initial and after /n,l,s/ positions where the trill is also commonly produced. 2.3 Variation of /r/: Puerto Rican Spanish Variation in the production of the phonemic trill has served as a defining feature of Spanish dialectal variation in both Latin American and Peninsular varieties (Lipski, 1994). PRS is not an exception in terms of trill variation. In fact, as many as 11 realizations have been reported: [ɾ ], [r], [h], [hɾ ], [hr], [ɹ], [xr], [xɾ ], [x], [R], [χ] (Graml, 2009; Hammond, 2000; Hualde, 2005; Navarro Tomás, 1948; Valentín-Márquez, 2007). This variation is known to depend on linguistic factors (e.g., word position or stress) as well as sociolinguistic factors (e.g., whether speakers are from urban vs. rural regions of Puerto Rico or the prestige attributed to those realizations) (Graml, 2009; Medina-Rivera, 1997). Some studies that have analyzed the PRS trill were constrained to the island of Puerto Rico, analyzing not only the metropolitan area (San Juan) (López Morales, 1983; Matta de Fiol, 1981; Navarro Tomás, 1948) but also non-metropolitan regions (Caguas, Mayagüez, Cabo Rojo, etc.) (Delgado-Díaz & Galarza, 2015; Graml, 2009; MedinaRivera, 1997). Other studies, while not strictly focused on rhotics, analyzed the Puerto Rican trill in the US mainland (Lamboy, 2004; Valentín-Márquez, 2007), making also a further step comparing PRS in the diaspora with the one spoken on the island (Ramos-Pellicia, 2004; Valentín-Márquez, 2007). The backed /r/, which has different variants itself (e.g., [x], [χ]), is considered a salient feature of PRS, just as we find phenomena such as coda /s/ weakening [eh.tá] for estás or liquid neutralization [a.mól] for amor (Potowski, 2015). With respect to the socio-indexical meaning of backed articulations of /r/, members of the speech community have been shown to associate it with rural origin or low sociocultural level (Graml, 2009; Medina-Rivera, 1997). Other studies, however, reveal positive attitudes toward backed realizations, seen as a sign of Puerto Ricanness (Lamboy, 2004; Medina-Rivera, 1997). Beginning with Navarro Tomás’s study of the Spanish of Puerto Rico in 1948, there has been a rich body of sociolinguistic research focused on the backed /r/ in PRS. These studies typically involve production work showing that the realization can be predicted by both sociolinguistic and linguistic factors. Most studies find that the backed variant is more prevalent in intervocalic position, as well as when the rhotic is produced in the lexically stressed syllable of a word (Graml, 2009). Sociolinguistic factors affecting the use of the backed variant include gender (more frequent for males), origin (more frequent for rural speakers), and age (more frequent for middle-aged adults) (López Morales, 1983; Matta de Fiol, 1981). Moreover, the production of the backed variant has been shown to be more common in informal interviews than in more controlled experiments (Graml, 2009; Medina-Rivera, 1997). Of specific interest for the present study are the comparative studies that investigate the use of the trill in the Puerto Rican diaspora in the US mainland. Valentín-Márquez (2007) compared rhotics

10

Alba Arias

10

in the PRS spoken in Grand Rapids, Michigan, to the PRS spoken in Cabo Rojo, Puerto Rico, finding a similar distribution of backed /r/ in both communities, with the backed variant being more common in word-initial position and in unstressed syllables. Notably, this finding regarding stress contradicts the studies just mentioned (Graml, 2009), suggesting differences even among PRS varieties (in Puerto Rico or in the US mainland). Moreover, in contrast to prior studies, for the PRS spoken on the island (Graml, 2009), ValentínMárquez found no significant differences in gender. In the case of Cabo Rojo, the backed /r/ (frequency distribution of 16%) is more common among middle-aged speakers. For Grand Rapids, the same realization (15% frequency distribution) is produced significantly more among speakers with lower levels of education and when using terms related to Puerto Rican nationality (e.g., puertorriqueño, “Puerto Rican”). On the other hand, Ramos-Pellicia (2004) analyzed (in addition to other linguistic variables), rhotics in syllable-final and word-final positions across three generations of PRS speakers in Lorain, Ohio. She also explored this variation on the island of Puerto Rico to determine whether PRS in Lorain displays different or similar patterns in the production of rhotics in those specific positions. For speakers in both Lorain and Puerto Rico, lateralization, trill, retroflection, and deletion were reported. Results reveal that PRS speakers in Lorain and on the island follow a similar pattern in the production of /r- ɾ /. However, in Lorain, speakers produce more retroflection, which is almost never used on the island. An interesting finding is that the preference of the normative trill declines across the three generations in Lorain, with the first generation favoring normative /r/ and the third generation using more lateralization and more retroflex /r/ than the second and first generations.

3. Present study The present study contributes to the body of research on language use in language-contact situations, analyzing the transmission of the trill realization in two different Puerto Rican communities: the Western Massachusetts diaspora and Puerto Rico. Specifically, it pays special attention to the backed /r/ realization. Linguistic and sociolinguistic factors are examined to shed light on the potential trill variation and to show whether or not there are differences among the realizations found in Puerto Rico versus Western Massachusetts. For this diasporic setting, we focus on the cities of Holyoke and Springfield because of their large Puerto Rican populations. 3.1 Research questions and hypotheses RQ1.

H1.

Is there trill variation in the Puerto Rican community in Western Massachusetts? If so, does it mirror the trill variation found on the island? Specifically, is the backed /r/ among the allophones produced? Since previous studies have shown that there is variation in trill production on the island and in other diaspora settings (Graml, 2009; Hammond, 2000;

11

RQ2. H2.

Rhotics in Massachusetts and Puerto Rico

11

Valentín-Márquez, 2007), we hypothesize that there will also be trill variation in Western Massachusetts. Moreover, given that back-and-forth migration waves between Puerto Rico and Massachusetts have been in constant increase since 1950 (Center for Puerto Rican Studies, 2016), we predict that the same trill variants found in the aforementioned Western Massachusetts communities will occur on the island. Since Lamboy (2004) and ValentínMárquez (2007) reported the presence of backed /r/ in the diaspora, we predict that this variant will also be produced in Massachusetts. If we find similar trill variation, are the predictors (sociolinguistic and linguistic) of the use of /r/ variants similar in the two settings? Given the back-and-forth migration waves previously mentioned, the hypothesis is that the factors that might predict the use of /r/ variants will be similar in both settings. Specifically, we predict that linguistic and sociolinguistic predictors such as stress (Graml, 2009; Valentín-Márquez, 2007); word position (Graml, 2009; López Morales, 1983); generation (Ramos-Pellicia, 2004); gender (Matta de Fiol, 1981); age (Graml, 2009; Medina-Rivera, 1997); origin (Graml, 2009); and type of task (Medina-Rivera, 1997) affect trill variation.

4. Methodology 4.1 Materials During the first ten minutes, participants were engaged in an informal conversation about family life, school or work to solicit natural conversation, before the linguistic experiment. Afterward, three experimental production tasks were performed to answer the aforementioned research questions: a picture description task, a map task, and a reading task. The narration of the children’s picture book by Mercer Mayer, Frog, Where Are You? (1969), elicited multiple productions of the trill segment (/pé.ro/ dog, /rá.na/, frog). This task is common in the research on trill variation in Spanish (Henriksen & Willis, 2010; Willis, 2007; Willis & Bradley, 2008). In the map task, participants collaborated with the experimenter to reproduce a route shown on their map. Participants produced target sentences written on the map that contain the phonemic trill segment in intervocalic position (n = 3) (i.e., parroquia del pueblo, town’s parish), wordinitial position (n = 3) (i.e., rosas de la bahía, bay’s roses) and after /n,l,s/ (n = 3) (i.e., avión israelí, Israeli plane). In addition, participants read a total of 24 words (12 targets and 12 fillers) embedded in the frame sentence Diga____otra vez, with the same three conditions considered in the map task. Fillers and target words were produced in random order, and the speaker read a single repetition. As in the map task, the 24 words were checked with a PRS speaker from the community to confirm that all tokens are familiar in the Puerto Rican variety. The last activity was a written language background questionnaire to get all linguistic and extralinguistic information for each participant. The experiment took approximately 45 minutes for each participant.

12

Alba Arias

12

4.2 Participants Two groups of ten participants took part of the experimental task: one group of Puerto Rican speakers was recorded for the US mainland variety of PRS (Holyoke and Springfield) and the other group for the island variety (see Table 1.1). The group recorded in Massachusetts was controlled in terms of their origin on the island: metropolitan (San Juan, Bayamón) vs. non-metropolitan area (San Lorenzo, Florida, Salinas, Comerío). Moreover, participants were divided according to generation (first or second), considering Silva-Corvalán’s criteria (1994). The group recorded on the island was also controlled by origin: metropolitan (Bayamón, San Juan) vs. non-metropolitan areas (Ponce, Cayey, Caguas, Vega Baja, Humacao). Participants were paid for their participation from grant funds provided by The Cognitive Science Initiative of the University of Massachusetts, Amherst. 4.3 Equipment Participants were recorded in a quiet room or in a laboratory setting. US mainland data were collected between November and December 2015. Puerto Rican data were collected in April 2016. Recordings were made with a Samson Zoom H2 Handy Digital Recorder at a sampling rate of 44,100 Hz with a head-mounted microphone. Participants were told to talk how they would in a natural situation. Table 1.1 Demographic and linguistic information for each participant #

Generation

Origin

Place

Gender

Age

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Second First Second Second Second First First Second Second First First First First First First First First First First First

Comerío Salinas Bayamón San Lorenzo San Lorenzo San Juan Florida Salinas Humacao San Juan Ponce Bayamón San Juan Ponce San Juan San Juan Caguas Vega Baja Humacao San Juan

Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Puerto Rico Puerto Rico Puerto Rico Puerto Rico Puerto Rico Puerto Rico Puerto Rico Puerto Rico Puerto Rico Puerto Rico

Male Female Female Female Female Female Male Female Female Male Male Female Female Male Female Male Female Male Female Female

23 60 21 24 57 55 41 45 55 49 24 22 49 28 22 22 54 28 50 50

13

Rhotics in Massachusetts and Puerto Rico

13

4.4 Acoustic analysis and coding An average of 65 tokens were elicited from each participant. As a result, given that 20 speakers participated in the production task, a total of 1,292 phonemic /r/ were analyzed. Sound files were digitally transferred into WAV format and analyzed with the acoustic analysis software Praat (Boersma & Weenink, 2014). Three separate statistical analyses were carried out with the following dependent variables: Normative vs. non-normative /r/ (i.e., [r] vs. the remaining /r/ realizations found among the data) and backed vs. non-backed /r/. Following Beaton (2015), Delgado-Díaz and Galarza (2015), and Graml (2009), the linguistic factors that might predict different /r/ realizations are stress, word position (initial, intervocalic, after /n,l,s/), previous sound (/a,e,i,o,u/, consonant), and following sound (/a,e,i,o,u/). The sociolinguistic variables include generation (first, second), age (young, middle-aged and older adults), gender (male vs. female), task (picture description task, map task, and reading task), origin (metropolitan vs. non-metropolitan area) and age of arrival to continental United States (early, 20–30 years; middle, 30–50 years; late, 50–60 years).

5. Qualitative results: data description In the following section, figures of the most common realizations among the overall data analyzed are presented. Next, frequency distribution for both settings under study are detailed. 5.1 Phonetic realizations Figure 1.1 shows the spectrographic output for a word-medial normative trill, containing two apical occlusions and the corresponding amplitude reduction of the waveform. Normative trills were separated from approximated trills. This latter realization typically involves the approximation of two articulators but without enough precision in order to create the necessary turbulent airflow and therefore complete occlusions. Since this realization falls between vowels and fricatives, its spectrographic formant structure is not as precisely defined as that of the vowels that surround it. Figure 1.2 shows an example of a 0-occlusion approximated trill for perro “dog.” A backed phonemic trill is presented in Figure 1.3. Even though several realizations of backed /r/ have been reported (Graml, 2009), all single voiceless fricative realizations produced in the posterior oral cavity were grouped into the same category (without distinguishing between uvular vs. velar places of articulation). As can be observed in Figure 1.3, we find that a long period of voiceless frication and a reduction of amplitude of the waveform characterize this realization. Another frequently occurring variant in the corpus is a pre-aspirated /r/. It starts with pharyngeal friction and ends with one [hɾ ] or two apical occlusions [hr]. When it is voiced, it correlates with the pre-breathy voiced tap/trill

Figure 1.1 Normative trill in the utterance corriendo, or “running.”

Figure 1.2 Approximated trill in the utterance perro, or “dog.”

15

Rhotics in Massachusetts and Puerto Rico

15

Figure 1.3 Backed realization of /r/ [x] in the utterance carretera, or “road.”

[ɦɾ -ɦr] described by Willis (2006) for Dominican Spanish. For the present study, although most of pre-aspirated /r/ analyzed were voiced, we did not distinguish the aspiration in terms of voicing. Figure 1.4 shows an example of a pre-breathy segmental portion (appearance of the voicing bar and reduction in amplitude as well as formant structure) followed by multiple occlusions [ɦr]. 5.2 Frequency distribution The acoustic analysis revealed seven different variants (trill/approximated, trill/ backed, trill/pre-aspirated, trill/pre-aspirated, tap/tap/post-tap, and trill frication) not only on the island but also in the Puerto Rican community in Western Massachusetts. However, a more detailed analysis of the results (Figure 1.5) reveals that although in both settings the most commonly occurring trill variant is the normative articulation, [r], the frequency distribution of the remaining realizations is slightly different depending on the participant’s origin. While the second most frequent variant in Massachusetts is the approximated trill, the backed /r/ occupies this place in Puerto Rico. It should be emphasized that although backed realizations are more common in Puerto Rico (18%, n = 117) than in Massachusetts (16.53%, n = 104), this difference is minimal (1.47 %, n = 13), suggesting that the backed /r/ behaves nearly identically in both settings under study. As for the pre-aspirated or pre-breathy realizations, the percentages show that they are more common in Massachusetts than in Puerto Rico. Finally, post-tap and trill frication realizations are more common in Puerto Rico.

16

Alba Arias

16

Figure 1.4 Pre-breathy realization of /r/ [ɦr] in the utterance garrar, or “to grab.” Trill Approximated Trill Backed /r/ Pre-aspirated trill Tap Pre-aspirated tap Miscellaneous (Post-tap/trill frication) 0%

20% PR

40%

60%

MA

Figure 1.5 Frequency distribution of /r/ variation.

6. Quantitative results: statistical analysis To assess the role of our different linguistic and sociolinguistic variables in relation to the realization of /r/, data were analyzed using R (R Core Team, 2013) with the R packages lmerTest, lme4, and stats. We performed three series of mixed modal linear regression analysis with the realization of /r/ as a dependent variable and the speaker as a random effect. For the realization of /r/, we collapsed all realizations to be normative /r/ or backed /r/ vs. everything else. The first model analyzed data from both places under study (Western Massachusetts and the island of Puerto Rico); the second model analyzed data from Western Massachusetts,

17

Rhotics in Massachusetts and Puerto Rico

17

whereas the third one only included data from the island of Puerto Rico. We considered as linguistic variables stress, position, and previous and following sound. As sociolinguistic variables, we included generation, age, gender, origin, task, age of arrival, and place. Models and interactions were compared using ANOVA. 6.1 Results for Western Massachusetts and Puerto Rico As stated, in the first mixed effect analysis, we considered data from both places under study (Western Massachusetts and the island of Puerto Rico) to ascertain a general understanding about the predictors that affect the realization of /r/. In particular, we were interested to test whether or not place is indeed a significant predictor with relationship to the use of non-normative trills and specifically backed trills. We did not include generation and first age of arrival in this analysis, since they are not relevant to Puerto Rico’s data. When normative /r/ was selected as the dependent variable, the best-fit model included stress, age, origin, task, and place as fixed effects. Origin (p < 0.001) and task (p < 0.001) were significant predictors, while place, gender, and stress were not significant. Partially confirming our hypothesis, place does not affect the realization of normative trills, suggesting similar variation in Puerto Rico and in the diasporic setting of Western Massachusetts. Origin (p < 0.001) was a significant predictor of the presence of normative /r/ in the overall data (β = −2.10, SE = 0.59, z = −3.51, p < 0.001) (Figure 1.6).

Presence of Normative /r/ based on Origin 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 Metro area

Non-metro area

Origin Figure 1.6 Presence of Normative /r/ as a function of origin for the overall data.

18

Alba Arias

18

Speakers originally from non-metropolitan areas (β = −2.10, SE = 0.59, z = −3.51, p < 0.001) produce more non-normative realizations than speakers from the metropolitan area, in line with previous research on Puerto Rican Spanish (Graml, 2009; López Morales, 1983). Moreover, in consonance with our expectations, speakers produced significantly more normative realizations in the reading task than in the picture description task (β = 0.65, SE = 0.17, z = 3.67, p < 0.001)—that is to say, when the task involved more spontaneous speech, a common finding in sociolinguist research (Labov, 1973; RamosPellicia, 2004) (Figure 1.7). Although place was not a significant predictor in the mixed modal linear regression analysis performed, we also assessed the interaction of place and other predictors in relationship to the variation of normative /r/. In the best-fit model, two-factor interaction that included place was significant. Figure 1.8 shows that middle-aged speakers from Puerto Rico produced significantly more normative /r/ realizations than younger speakers from Massachusetts (β = −3.59, SE = 1.48, z = −2.42, p < 0.01). This finding might suggest that older speakers in Puerto Rico are more aware of prestigious realizations than younger speakers in the diaspora are. Another alternative explanation is that younger speakers, in fact, might produce non-normative realizations to sound more Puerto Rican, especially in a diasporic setting (Lamboy, 2004; MedinaRivera, 1997).

Presence of Normative /r/ based on Task 1.00

Proportion

0.75

Realization Other

0.50

Normative /r/

0.25

0.00 Reading task

Map task

P. description task

Task Figure 1.7 Presence of normative /r/ as a function of task for the overall data.

19

Rhotics in Massachusetts and Puerto Rico

19

Presence of Normative /r/ Age:Place MA

PR

1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 Young

Middle-aged

Young

Middle-aged

Speaker Figure 1.8 Presence of normative /r/ as a function of age:place for the overall data.

When backed /r/ was selected as the dependent variable, the best-fit model included stress, age, origin, task, and place as fixed effects. Age (p < 0.001), origin (p < 0.001), and task (p < 0.01) were significant predictors, while place and stress were not significant. Once again, place does not affect the realization of /r/, in this case of backed trills, finding similar variation in Western Massachusetts and in Puerto Rico. Age, however, was a significant predictor (Figure 1.9), with middle-aged speakers producing significantly less backed /r/ than younger speakers produced (β = −4.38, SE = 1.33, z = −3.29, p < 0.001), in line with Medina-Rivera’s (1997) and Graml’s (2009) research on the island of Puerto Rico. Moreover, speakers from the non-metropolitan area produced significantly more backed /r/ realizations than speakers originally from the metropolitan area (β = 6.15, SE = 1.66, z = 3.68, p < 0.001) (Figure 1.10). Figure 1.11 shows that speakers produced significantly more backed realizations in the map task than in the picture description task (β = 0.65, SE = 0.24, z = 2.68, p < 0.001). We expected the opposite results, since the picture description task involved more spontaneous speech. We attribute this finding to the fact that the map task was conducted after the picture description task. Therefore, participants might have been less affected by the observer’s paradox (Labov,

Presence of Backed /r/ based on Age 1.00

Proportion

0.75

Realization Other

0.50

Backed /r/

0.25

0.00 Young

Middle-aged

Age Figure 1.9 Presence of backed /r/ as a function of age for the overall data.

Presence of Backed /r/ based on Origin 1.00

Proportion

0.75

Realization 0.50

Other Backed /r/

0.25

0.00 Metro area

Non-metro area

Origin Figure 1.10 Presence of backed /r/ as a function of origin for the overall data.

21

Rhotics in Massachusetts and Puerto Rico

21

Presence of Backed /r/ based on Task 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 Reading task

Map task

P. description task

Task Figure 1.11 Presence of backed /r/ as a function of task for the overall data.

1973). Finally, we conducted interactions between age, origin, and task, but none were significant when backed /r/ was selected as the dependent variable. After assessing the role of both linguistic and sociolinguistic predictors in relationship to our dependent variables for all phonemic /r/ analyzed, we divided the overall data into two different groups based on the place of recording: Western Massachusetts and Puerto Rico. That way, we would be able to compare the different factors that might affect trill variation in each setting. In addition to the variables considered in the previous models, generation and first age of arrival were also included in the Western Massachusetts data. In the following sections, findings for each group are reported. 6.2 Results for Western Massachusetts When we consider normative /r/ as the dependent variable, the best-fit model included only age, origin, and task. All of them were significant predictors, showing similar results for the presence of normative /r/ compared to those from the overall data. The correlations were not significant, though. Age was a significant predictor of the presence of normative /r/ in Western Massachusetts (β = 1.75, SE = 0.79, z = 2.2, p < 0.05). Young adults produce more non-normative realizations than middle-aged adults do (Figure 1.12). It suggests that speakers might use non-normative trills for nationalistic purposes

22

Alba Arias

22

Presence of Normative /r/ based on Age 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 Young

Middle-aged

Age Figure 1.12 Presence of normative /r/ as a function of age for Western Massachusetts.

or to sound more Puerto Rican (Lamboy, 2004), trying to distinguish themselves from other Spanish communities in Western Massachusetts, such as Dominicans, Salvadorians, and/or Mexicans. It is also likely that younger speakers have grown up in an English environment (e.g., at school or work). Therefore, they have not been exposed to standard Spanish norms that would help to suppress non-normative realizations. Further analysis that factors in networking, education, and/or socioeconomic variables might shed light on this finding. Origin was also a significant predictor, with speakers originally from a non-metropolitan area of the island producing significantly more non-normative realizations (β = −3.06, SE = 0.77, z = −3.94, p < 0.001) (Figure 1.13). Finally, the model also showed that task was a significant predictor of normative /r/ (Figure 1.14), with speakers producing more non-normative realizations when they were completing the picture description task than when they were performing the reading task (β = 1.31, SE = 0.35 z = 3.7, p < 0.001) or the map task (β = 1.36, SE = 0.34 z = 3.97, p < 0.001)—that is, when the task involved more spontaneous speech. On the other hand, when we consider backed /r/ as a dependent variable, age was the only significant predictor (β = −4.67, SE = 2.35, z = −1.98, p < 0.05), with similar results to the just-mentioned analysis: in Western Massachusetts, young adults produced more backed trills than middle-aged adults did (Figure 1.15). In contrast to results from the overall data, origin and task were not significant.

Presence of Normative /r/ based on Origin 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 Metro area

Non-metro area

Origin Figure 1.13 Presence of normative /r/ as a function of origin for Western Massachusetts.

Presence of Normative /r/ based on Task 1.00

Proportion

0.75

Realization Other

0.50

Normative /r/

0.25

0.00 Reading task

Map task

P. description task

Task Figure 1.14 Presence of normative /r/ as a function of task for Western Massachusetts.

24

Alba Arias

24

Presence of Normative /r/ based on Age 1.00

Proportion

0.75

Realization 0.50

Other Backed /r/

0.25

0.00 Young

Middle-aged

Age Figure 1.15 Presence of backed /r/ as a function of age for Western Massachusetts.

6.3 Results for Puerto Rico Analyzing data exclusively from Puerto Rico, when we consider normative /r/ as the dependent variable, the best-fit model included stress, origin, age and task, finding stress (p < 0.05), origin (p < 0.05) and task (p < 0.05) to be the significant predictors. Results for Origin and task showed the same pattern as in Western Massachusetts: as expected, metropolitan area speakers produced more normative /r/ than did participants originally from the non-metropolitan area (β = −1.65, SE = 0.76, z = −2.16, p < 0.05). Moreover, speakers produced more non-normative realizations when they were completing the picture description task than when they were performing the map task (β = −.58, SE = 0.22 z = −2.56, p < 0.05). In contrast to results from Western Massachusetts, stress was a significant predictor of the presence of normative /r/ in Puerto Rico (Figure 1.16), in line with Graml’s research (2009) on the island. The analysis revealed that when the syllable containing the rhotic sound is stressed, it favors a normative realization (β = 0.41, SE = 0.19, z = 2.13, p < 0.05). Also, stress was the first significant linguistic predictor among all the mixed-model analyses performed. Finally, when we considered backed /r/ as dependent variable, the best-fit model showed that origin (p < 0.001), Age (p < 0.01), task (p < 0.01), and following sound (p < 0.001) were significant predictors. In Puerto Rico, Young adults produced more backed trills than middle-aged adults did (β = −5.46, SE = 1.68,

25

Rhotics in Massachusetts and Puerto Rico

25

Presence of Normative /r/ based on Stress 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 Unstressed

Stressed

Stress Figure 1.16 Presence of normative /r/ as a function of stress for Puerto Rico.

z = −3.25, p < 0.01) and speakers from non-metropolitan areas produced more backed realizations than speakers from metropolitan areas did (β = 5.08, SE = 1.38, z = 3.67, p < 0.001). Similar results were found in Western Massachusetts. Task was also a significant predictor of backed /r/. However, against our expectations, speakers in Puerto Rico produced more backed realizations when they were completing the reading task than when they were performing the picture description task (β = 1.69, SE = 0.52, z = 3.21, p < 0.01). This result contradicts our findings on backed /r/ as a function of task for the overall data. In contrast to our results from Western Massachusetts, the mixed-effects model also showed that following vowel was a significant predictor of the presence of backed /r/ in Puerto Rico (Figure 1.17). When the rhotic segment is followed by /o/, a backed trill is predicted (β = 2.01, SE = 0.56, z = 3.57, p < 0.001) in comparison to when it is followed by /a/. This result could be explained due to coarticulation: the presence of a backed vowel may contribute to the implementation of the backed /r/.

7. Discussion The first research question asked whether there is trill variation in the Puerto Rican community in Western Massachusetts. In line with previous studies (Graml, 2009; Lamboy, 2004; Valentín-Márquez, 2007), the results confirm that there is such variation. In fact, we reported up to seven different variants: trill, approximated trill,

26

Alba Arias

26

Presence of Normative /r/ based on Following Sound 1.00

Proportion

0.75

Realization 0.50

Other Backed /r/

0.25

0.00 a

e

i

o

u

Following Sound Figure 1.17 Presence of backed /r/ as a function of following sound for Puerto Rico.

backed trill, pre-aspirated trill, pre-aspirated tap, tap, and a post-tap/trill frication variant. To our knowledge, all realizations found were previously reported in PRS (Navarro Tomás, 1948; Graml, 2009) with the exception of the post-tap/trill frication category, which was found in another Spanish variety (Mexican Spanish) in the Chicago diaspora (Henriksen, 2014). Confirming our hypothesis as well as previous studies focused on different communities in the US mainland and in Puerto Rico (Graml, 2009; Lamboy, 2004; Ma & Herasimchuck, 1972; Valentín-Márquez, 2007), the backed /r/ is also produced among our data. Qualitative results point out that the same set of phonetic realizations are found in the two settings under study (Puerto Rico and Massachusetts), in line with Ramos-Pellicia’s (2004) and Valentín-Márquez’s (2007) research, as well as with our hypothesis. Valentín-Márquez (2007) found the same five variants (two alveolar and three posterior realizations) in the interviews conducted in Puerto Rico (Cabo Rojo) than in those conducted in Grand Rapids (Michigan). Although Ramos-Pellicia (2004) analyzed different contexts, she showed the same variants regarding the variable (r) in Lorain, Ohio, and in Puerto Rico (lateralization, retroflection, deletion). Returning to our data, the frequency distribution of the realizations found is slightly different between the two communities under study. However, in both settings, the alveolar trill is the most common realization, a find that is in line with the results of other sociolinguistic

27

Rhotics in Massachusetts and Puerto Rico

27

studies focused on PRS (López Morales, 1983; Medina-Rivera, 1997; ValentínMárquez, 2007). Our statistical mixed-effects analysis for the overall data confirmed that place (Western Massachusetts vs. Puerto Rico) was not a significant predictor, not only for the production of normative /r/ but also for the use of backed /r/—which is in line with Valentín-Márquez’s research (2007), which found a similar distribution of backed /r/ in Cabo Rojo and Grand Rapids. As for the predictors of trill variation, we partially confirmed our second hypothesis. While there are differences in regard to the linguistic factors considered, the analysis also showed some similar sociolinguistic factors that predict not only the use of normative /r/ but also the use of backed /r/ in the diaspora and in Puerto Rico. This result might suggest that, on the one hand, due to the back-and-forth migration waves between Puerto Rico and Massachusetts, the main socio-indexical factors associated with trill realizations on the island are available in the diaspora. On the other hand, as explained later, individual speaker differences as well as other sociolinguistic variables not considered in the study (e.g., education, socioeconomic status) might cause the variation found for the backed /r/ analysis in particular. Among the overall data analysis (encompassing data from Western Massachusetts and Puerto Rico), origin was a significant predictor, reporting the same results for both normative /r/ and backed /r/: speakers originally from non-metropolitan areas produce more non-normative realizations and, specifically, more backed /r/ than speakers from the metropolitan area, in line with previous sociolinguistic research on PRS (López Morales, 1983; Matta de Fiol, 1981). Age was also a significant predictor, but only for the analysis that took backed /r/ as a dependent variable: young adults produced more backed realizations than middle-aged adults did. Interestingly, task was a significant factor: depending on the variable selected (normative /r/ or backed /r/), the model reported opposite patterns. As expected, overall, speakers significantly produced more normative trills in the reading task than in the picture description task. However, contrary to our expectations and to previous studies (Medina-Rivera, 1997; Ramos-Pellicia, 2004), overall, speakers significantly produced more backed realizations in the map task than in the picture description task. As stated before, we attribute this finding to the fact that the map task was the second exercise that the participants had to accomplish, so the subjects might have felt more confident with the experiment as well as with the researcher. Opposing our hypothesis, in addition to place (Western Massachusetts vs. Puerto Rico), gender was another non-significant factor. Following previous sociolinguistic research, we would expect women to favor more the production of the prestigious or standard variants than men do (López Morales, 1983; Matta de Fiol, 1981; Ramos-Pellicia, 2004; Valentín-Márquez, 2007). In the next two series of mixed-model analysis, we divided the overall data into two groups based on the place of recording: Western Massachusetts or Puerto Rico. Once again, two dependent variables were considered: normative /r/ and backed /r/. Results for the normative /r/ analysis revealed that task and origin were significant, showing the same pattern in Puerto Rico and Western Massachusetts.

28

Alba Arias

28

However, we also found different statistically significant factors in each setting. In Western Massachusetts, age affected the use of normative trills, while in Puerto Rico, the normative /r/ is constrained not only sociolinguistically (task, origin) but also linguistically (stress), as the normative trill was produced more in stressed syllables, in line with previous research on the island (Valentín-Márquez, 2007). Results for the backed /r/ analysis revealed that only age was a significant factor in both settings under study. As with the findings for the overall data, young adults (19–40 years old) produced more non-normative realizations than middle-aged adults did (41–60 years old), in line with Medina-Rivera’s (1997) and Graml’s (2009) research on the island. As stated before, young participants might see this realization as a sign of Puerto Ricanness (Medina-Rivera, 1997; Lamboy, 2004). For Western Massachusetts, PRS speakers might try to distinguish themselves from other Spanish-speaking communities in the area, such as Dominicans, Salvadorians, and/or Mexicans. It is also possible that not only in Massachusetts (diasporic setting) but also in Puerto Rico, those younger speakers have grown up in an English environment (school, work, friends), speaking Spanish only at home, in a family setting. Therefore, those speakers have not been exposed to other Spanish varieties, including the standard norms that would help them to avoid non-normative realizations: in this case, the stigmatized backed /r/. Other sociolinguistic variables, such as education, socioeconomic status, or social networks should be considered in further studies to shed light on this variation. Since language is essential to constructing and negotiating diaspora identities and relationships (Canagarajah & Silberstein, 2012), it will be necessary to analyze the effect of participants’ attitudes and linguistic practices as well as their relationship with (1) other members of their diasporic community, (2) members of other communities that coexist in the same diasporic setting, and (3) the island of Puerto Rico (travels to Puerto Rico/attitudes toward the Spanish spoken on the island). Thus, since diasporas are changing communities, it will be crucial to analyze participants’ responses related to language attitudes and networking in the diaspora and to the island of Puerto Rico, in order to address the possible role that bilingualism and language maintenance and change play on the rhotic variation. Furthermore, following previous studies (Lamboy, 2004; Valentín-Márquez, 2007), education and socioeconomic status should be included as sociolinguistic factors. Since less prestigious variants are associated with low education and povertylevel communities (Silva-Corvalán, 1994), we could expect that higher frequencies of backed /r/ are more frequent among PRS speakers with lower socioeconomic status and lower levels of education (Valentín-Márquez, 2007). In this manner, we would be able to confirm Valentín-Márquez’s regional distribution of backed /r/ in the United States, which would be conditioned by socioeconomic factors. Among the Massachusetts data, no other factors were significant for the backed /r/ analysis. However, results from Puerto Rico revealed that age, origin, task, and following sound were significant predictors. As expected, the model showed that the backed /r/ is produced more often when articulated before /o/, a finding attributed to coarticulation.

29

Rhotics in Massachusetts and Puerto Rico

29

Presence of Backed /r/ in MA (individual variation) 1.00

Proportion

0.75

Realization 0.50

Other Backed /r/

0.25

0.00 1

2

3

4

5

6

7

8

9

10

Speaker Figure 1.18 Presence of backed /r/. Individual speaker differences in Western Massachusetts.

The overall outcome is that even though there are some exceptions, such as following sound and stress (linguistic predictors), similar sociolinguistic patterns (principally origin, task, and age) predict trill variation in Western Massachusetts and in Puerto Rico. Interestingly, quantitative results showed more similarities for the normative /r/ analysis than for the backed /r/ analysis: while age is the only significant factor among the diaspora data, age, origin, task, and following sound were also significant in Puerto Rico. We attribute those differences to individual speaker differences. In fact, if we look closer at the data (Figures 1.18 and 1.19), mainly two participants are driving the variation backed vs. non-backed /r/ realizations (in both settings, Western Massachusetts and Puerto Rico). For Puerto Rico, those speakers are both young male adults from the non-metropolitan area (Ponce and Vega Baja), university students, and those proud of their Puerto Rican roots (as they mentioned in the natural conversation before the linguistic experiment). In the case of Western Massachusetts, speakers who produced backed /r/ realizations were also young adults originally from the non-metropolitan area on the island. Moreover, both were second-generation speakers. This result suggests that considering more participants would be crucial in order to corroborate differences in generation. Furthermore, as stated before, other sociolinguistic variables should be taken into account (education, socioeconomic status, networking) to shed light on this individual variation. On the other hand, for the presence of normative /r/, although there are some differences, Figures 1.20 and 1.21 show how the data are more balanced. With the

Presence of Backed /r/ in PR (individual variation) 1.00

Proportion

0.75

Realization 0.50

Other Backed /r/

0.25

0.00 11

12

13

14

15

16

17

18

19

20

Speaker Figure 1.19 Presence of backed /r/. Individual speaker differences in Puerto Rico.

Presence of Normative /r/ in MA (individual variation) 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 1

2

3

4

5

6

7

8

9

10

Speaker Figure 1.20 Presence of normative /r/. Individual speaker differences in Western Massachusetts.

31

Rhotics in Massachusetts and Puerto Rico

31

Presence of Normative /r/ in PR (individual variation) 1.00

Proportion

0.75

Realization 0.50

Other Normative /r/

0.25

0.00 11

12

13

14

15

16

17

18

19

20

Speaker Figure 1.21 Presence of normative /r/. Individual speaker differences in Puerto Rico.

exception of one speaker in Massachusetts, the participants produce both normative and non-normative /r/ realizations. In this regard, Ramos-Pellicia (2004) reports on her cross-generational research of rhotics in syllable-final and wordfinal positions among PRS speakers in Lorain (Ohio) that at least for the first generation, Lorain Puerto Ricans are aware of the negative prestige that [l] has on the island, so they use other variants, such as the retroflex trill, as an alternative to avoid the use of the lateralization. Similarly, we can attribute our findings (more variation when we take normative /r/ as dependent variable in comparison to backed /r/) to the fact that PRS speakers in Western Massachusetts and in Puerto Rico might be more conscious of the negative prestige of the backed /r/, associated with rural origin or lower socioeconomic status on the island (Navarro Tomás, 1948). Therefore, instead of producing it, they make more use of the other possible non-normative variants: approximated trill, pre-aspirated trill, pre-aspirated tap, tap, and a post-tap/trill frication variant.

8. Conclusion Rhotic variation has long been of interest to linguists, in Puerto Rico and in the diaspora, but as mentioned earlier, non-impressionistic research on the phonetic and phonological variation of PRS has focused on Western Massachusetts. Thus,

32

Alba Arias

32

this study is unique in that it studies acoustic data on onset /r/ in an area of the United States that is experiencing a rapid growth in its Puerto Rican population. To shed more light on the general research of how language variation plays out in this particular diasporic community, future studies might improve a production task that can be performed by both Puerto Rican speakers on the island and in Western Massachusetts, but they must carefully control for the origin of the speakers. Moreover, collecting data from third-generation speakers in Massachusetts would be crucial. That way, it would be possible to continue Ramos-Pellicia’s (2004) lines of research and analyze in depth whether or not the socio-indexical factors associated with more backed realizations are available to second- or third-generation speakers in this diasporic community. To conclude, in spite of its limitations (uneven speaker distribution, individual speaker differences, few participants, lack of networking analysis), this study adds to the growing body of literature on Spanish sociophonetics, providing a detailed analysis of trill variation for different generations of PRS speakers in Western Massachusetts and showing the similar and the different linguistic and sociolinguistic patterns that affect trill variation in comparison to those predictors of the island of Puerto Rico.

References Beaton, M.E. (2015). Coda liquid production and perception in Puerto Rican Spanish. Doctoral dissertation. Retrieved from https://etd.ohiolink.edu/ Blecua Falgueras, B. (2001). Las Vibrantes del Español: Manifestaciones Acústicas y Procesos Fonéticos. Doctoral dissertation. Retrieved from http://hdl.handle.net/ 10803/4859 Boersma, P., & Weenink, D. (2014). Praat: Doing phonetics by computer [Computer program]. Version 5.4.01. Retrieved from www.praat.org/ Bradley, T.G., & Willis, E.W. (2012). Rhotic variation and contrast in Veracruz Mexican Spanish. Estudios de Fonética Experimental, 21, 43–74. Canagarajah, S., & Silberstein, S. (2012). Diaspora identities and language. Journal of Language, Identity, and Education, 11, 81–84. Center for Puerto Rican Studies. (2016). The story of U.S. Puerto Ricans: Part four. Retrieved from http://centropr.hunter.cuny.edu/education/puerto-rican-studies/storyus-puerto-ricans-part-four/ Delgado-Díaz, G., & Galarza, I. (2015). ¿Qué comiste [x]amón? A Closer Look at the Neutralization of /h/ and Posterior /r/ in Puerto Rican Spanish. In E.W. Willis, P. Martín-Butragueño, & E. Herrera (Eds.), Selected proceedings of the 6th conference on laboratory approaches to romance phonology (pp. 70–83). Somerville, MA: Cascadilla Proceedings Project. Díaz-Campos, M. (2008). Variable production of the trill in spontaneous speech: Sociolinguistic implications. In L. Colantoni & J. Steele (Eds.), Selected proceedings of the 3rd conference on laboratory approaches to Spanish phonology (pp. 47–58). Somerville, MA: Cascadilla Proceedings Project. Graml, C. (2009). Puerto, RICO en variación: Variation socio-phonétique et son auto- et hétérosurveillance par les locuteurs—le cas de la vélarisation du /r/ en espagnol portoricain. Doctoral dissertation. Retrieved from www.theses.fr/2009PA100056

33

Rhotics in Massachusetts and Puerto Rico

33

Hammond, R. (2000). The phonetic realizations of /r/ in Spanish: A psychoacoustic analysis. In H. Campos (Ed.), Papers from the 3rd Hispanic linguistics symposium: Hispanic linguistics at the turn of the millennium (pp. 80–10). Somerville, MA: Cascadilla. Henriksen, N. (2014). Sociophonetic analysis of phonemic trill variation in two subvarieties of Peninsular Spanish. Journal of Linguistic Geography, 2, 1–21. Henriksen, N., & Willis, E. (2010). Acoustic characterization of phonemic trill production in Jerezano Andalusian Spanish. In M. Ortega-Llebaria (Ed.), Selected proceedings of the 4th conference on laboratory approaches to Spanish phonology (pp. 115–127). Somerville, MA: Cascadilla Proceedings Project. Hualde, J.I. (2005). The sounds of Spanish. Cambridge: Cambridge University Press. Instituto Cervantes. (2017). El español: una lengua viva. Informe 2017. Retrieved from https://cvc.cervantes.es/lengua/espanol_lengua_viva/pdf/espanol_lengua_viva_2017. pdf Labov, W. (1973). Sociolinguistic patterns. Philadelphia, PA: University of Pennsylvania Press. Lamboy, E. M. (2004). Caribbean Spanish in the metropolis: Spanish language among Cubans, Dominicans, and Puerto Ricans in the New York City Area. New York: Routledge. Lipski, J. (1994). Latin American Spanish. New York: Longman Publishing. López Morales, H. (1983). Estratificación social del español de San Juan de Puerto Rico. México: Universidad Nacional Autónoma de México. Ma, R., & Herasimchuck, E. (1972). The linguistic dimensions of a bilingual neighborhood. In J. Fishman, R. Cooper, & R. Ma (Eds.), Bilingualism in the Barrio. Language Science Monographs, 7 (pp. 347–464). Bloomington, IN: Indiana University Press, Research Center for the Language Sciences. Martínez Celdrán, E. (1998). Análisis espectrográfico de los sonidos de habla. Barcelona: Ariel. Matta de Fiol, E. (1981). La rr velar en el español hablado en San Juan: estudio de actitud lingüística. Boletín de la Academia Puertorriqueña de la Lengua Española, 2, 2–24. Mayer, M. (1969). Frog, where are you? New York: Dial Press. Medina-Rivera, A. (1997). Variación Fonológica y Estilística en el Español de Puerto Rico. Unpublished doctoral dissertation, University of Southern California. Navarro Tomás, T. (1948). El español de Puerto Rico: Contribución a la geografía lingüística hispanoamericana. Río Piedras: Editorial de la Universidad de Puerto Rico. Our Plural History, Springfield MA. (2008). Puerto Rican communities in the valley. Retrieved from http://ourpluralhistory.stcc.edu/recentarrivals/puertoricans.html/ Pew Hispanic Center. (2012). Characteristics of the 60 largest metropolitan areas by Hispanic population. Retrieved from www.pewhispanic.org/2012/09/19/characteristics-ofthe-60-largest-metropolitan-areas-by-hispanic-population/ Potowski, K. (2015). Ethnolinguistic identities and ideologies among Mexicans, Puerto Ricans, and “MexiRicans” in Chicago. In R. Márquez-Reiter & L. Martín Rojo (Eds.), A sociolinguistics of diaspora: Latino practices, identities and ideologies. New York: Routledge. Ramos-Pellicia, M.F. (2004). Language contact and dialect contact: Cross-generational phonological changes in a Puerto Rican community in the Midwest of the United States. Electronic doctoral dissertation. Retrieved from https://etd.ohiolink.edu/ R Core Team. (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Retrieved from www.r-project.org

34

Alba Arias

34

Recasens, D., & Pallarés, M.D. (1999). A study of /r/ and /rr/ in the light of the ‘DAC’ coarticulation model. Journal of Phonetics, 27, 143–169. Shouse de Vivas, D. (1986). El uso de [l] variante de ‘r’ en el habla de Puerto Rico. In Actas del V Congreso Internacional de la Asociación de Lingüística y Filología de la América Latina (pp. 622–631). Caracas: Universidad Central de Venezuela. Silva-Corvalán, C. (1994). Language contact and change: Spanish in Los Angeles. Oxford: Oxford University Press. United States. (2010). U.S. Census Bureau. Retrieved from www.census.gov Valentín-Márquez, W. (2007). Doing being Boricua: perceptions of national identity and the sociolinguistic distribution of liquid variables in Puerto Rican Spanish. Available from ProQuest database (3287649). Willis, E.W. (2006). Trill variation in Dominican Spanish: An acoustic examination and comparative analysis. In N. Sagarra & A.J. Toribio (Eds.), Selected proceedings of the 9th Hispanic linguistics symposium (pp. 121–131). Somerville, MA: Cascadilla Proceedings Project. Willis, E.W. (2007). An acoustic study of the “pre-aspirated trill” in Narrative Cibaeño Dominican Spanish. Journal of the International Phonetic Association, 37, 33–49. Willis, E.W., & Bradley, T.G. (2008). Contrast maintenance of taps and trills in Dominican Spanish: Data and analysis. In L. Colantoni & J. Steele (Eds.), Selected proceedings of the 3rd conference on laboratory approaches to Spanish phonology (pp. 87–100). Somerville, MA: Cascadilla Proceedings Project.

35

2

Differences in the use of /l/ and /ɾ / in two communities of Puerto Rican Spanish speakers in the United States Michelle F. Ramos-Pellicia

1. Introduction In the linguistic literature, it has been reported that the Spanish of Puerto Rico, like the Caribbean Spanish dialects of Dominican Republic and Cuba, has [l] and deletion as /ɾ / variants. Álvarez-Nazario (1972, 1990, 1991), Guitart (2004), Navarro Tomás (1948), Ramos-Pellicia (2004), and Valentín-Márquez (2007) note that /ɾ / is often pronounced as [l] in syllable-final position or word-final position, so that, for example, “hablar” (“to speak”) is pronounced [a.blál]. Retroflection has not been documented as a variant of /ɾ / in Island PRS, but it has been reported in the speech of Puerto Rican speakers in the US Midwest (Ramos-Pellicia, 2007). Rhotacization of /l/ occurs less frequently than lateralization; but when observed, it occurs in word-final position (Valentín-Márquez, 2007), and it is found with less frequency in three different generations of PRS speakers in the United States (Ramos-Pellicia & Ayala, 2008). Given what we know about PRS on the island of Puerto Rico and in the United States, I speculate that the frequency rates of the use of [ɾ ] and [l] of Puerto Rican speakers who live in the United States will vary depending on exposure to PRS from the island and contact with American English. Likewise, the third generation will exhibit patterns of usage that vary as a result of interactions with speakers of the first and second generation as well as with American English speakers and may present patterns of attrition and higher rates of retroflection of /ɾ / as a result of the contact with English.

2. Objectives There are several objectives for this project. First of all, I want to compare the frequency patterns of the use of variants for the phonemes /ɾ / and /l/ among three generations of speakers in two Puerto Rican communities in the United States. Second, I am interested in exploring whether the third generation’s variety of these communities displays phonological patterns comparable in frequency and environment to those found in the first and second generations in both communities. Furthermore, in one of the two communities, data demonstrating usage of a retroflex /ɾ / was collected among speakers of different

36

Michelle F. Ramos-Pellicia

36

generations. Given the occurrence of a retroflex /ɾ / among those speakers, I consider further evidence against attributing the occurrence of the retroflex /ɾ / as an alternative for /ɾ / to American English influence.

3. Hypotheses The two Puerto Rican communities described here are close-knit. They do not maintain regular contact with the PRS spoken in Puerto Rico. However, speakers of all generations are exposed to this linguistic variety. The contact is the result of their continuous interactions with other speakers who still use this linguistic variety within their close-knit communities. These linguistic interactions cannot be avoided; they are totally necessary because these are individuals who are either family members or friends. Thus, the interactions happen for a variety of reasons. While they do not maintain regular contact with PRS, they do not travel frequently to the island of Puerto Rico; do not call as often and use Spanish; do not interact with more speakers of PRS; or do not have the constant flow of other PRS speakers who visit from other similar communities located elsewhere in the United States who do speak PRS. This is important because PRS has maintained certain characteristics that distinguish it from, for instance, Mexican Spanish. Thus, even when the speakers of these two communities are not in continuous contact with their corresponding ancestral variety as a result of their geographical location in the United States, I predict that both communities will make the same use of their ancestral variety. I base this prediction on the notion that this linguistic variety for these speakers at the core of their identity. I propose that even when all factors are equally maintained, the size of a community will make a difference. If a variety of a language has a lack of input from the ancestral variety but maintains the same type of input from the diasporic variety and has more exposure to the regional dominant language, I predict that the community size will play a role in the use of certain variations. The prediction will hold true despite geographical distance between the communities that speak the same variety of the language. In other words, in the two communities in question, one located on the East Coast and the other in the Midwest, the same results will not be found for /ɾ / and /l/. Due to the difficulty in finding speakers in a small community of speakers of a minority language, as well as other patterns that result from American English influence, the smaller sized community will not have a clearer model to reproduce, and thus, the results will differ from the ancestral variety or even from the model used by speakers in the same community. On the other hand, if a minority linguistic community is larger because of the possibility of having more contacts with speakers of the same minority language in the region, there will be a greater tendency to hold the patterns found in the ancestral variety. Thus, I put forth a proposal somewhat similar to that of Van Coetsem’s phonological borrowings, but in which the roles of the “source language” and the “recipient language” are reversed (Van Coetsem, 1988).

37

Use of /l/ and /ɾ / in US PRS speakers

37

Given this context, I consider the following three objectives for this work with its possible hypothesis. 1

2

3

With respect to the frequency patterns of use for the phonemes /ɾ / and /l/ among three generations of speakers in two Puerto Rican communities in the United States, I hypothesize that a pattern more similar to that of the ancestral variety will be found in the first generation. The pattern will decrease in frequency from the first generation to the second generation, with the third generation having little to no resemblance with the ancestral variety. The third generation’s variety of these communities might display phonological patterns that are not comparable in frequency and environment to those found in the first and second generations in both communities. It might be that the third generation will present lower frequency rates of the standard variant of /ɾ /, as well as few cases of lateralization and deletion as a result of little to no contact with the ancestral variety and more contact with American English. Furthermore, a retroflex /ɾ / was collected among speakers of different generations in the Lorain Puerto Rican community. Given the occurrence of a retroflex /ɾ / among those speakers, I consider further evidence that could attribute the occurrence of the retroflex /ɾ / as an alternative for the lateralization of /ɾ /. This retroflex /ɾ / could be the result of American English influence.

4. The two language communities The two language communities of Puerto Ricans that I considered for this study are located in different areas in the United States. Access to both Latino communities, their size differences, their similarities in socioeconomic status and educational levels, and their differences in terms of rural development vs. urban development made these interesting groups to study when considering the maintenance of and change in a PRS linguistic variety. My interest in comparing these two Puerto Rican communities is the result of my curiosity in whether community size, location, and an urban setting vs. a rural setting play a role in maintaining Spanish in general and of their linguistic variety in particular. For the present work, I am interested in comparing the two communities to see if intergenerational transmission, gender, and position in the words play a role in the maintenance of the ancestral variants for /ɾ /. One of the Puerto Rican communities is located in Hazlet, a town approximately 35 miles south of Jersey City, New Jersey. According to the 2010 U.S. Census Bureau, Hazlet’s total population is 20,3334. Of the total population, 1,601, or 7.9%, are Latinos, and 806 (4%), or almost half of the Latinos, are Puerto Ricans. Along with the Puerto Ricans in Hazlet, there are Mexicans (141, or 0.7%) and Cubans (129, or 0.6%), and 752, or 2.6%, of Latinos that belong to other ethnic groups. Hazlet has a small Puerto Rican community. It is located in an industrial urban area of New Jersey. The data were the result

38

Michelle F. Ramos-Pellicia

38

of a convenience sample made possible by the contacts that a student collaborator had in the area. Unfortunately, the information about the Puerto Rican community is limited. The Puerto Rican community of Hazlet can be described as close and dense and having highly interconnected networks. The other community of Puerto Ricans that I studied is located in Lorain, Ohio. Because of connections I have in the community and because of its long and welldocumented history, this Puerto Rican community seemed like a natural choice for my data collection. Lorain is 30 miles to the west of Cleveland, Ohio. According to the 2016 U.S. Census Bureau, Lorain is inhabited by 63,714 people, of whom 28.3% (18,044) are Latino. Of this 28.3%, 21.8% (or 13,896) are Puerto Ricans; 4.8% (or 3,032) are Mexicans; 0.3% (or 173) are Cuban, and 1.5% (or 943) belong to different Latino groups. The Puerto Rican community in Lorain, Ohio, was established in the late 1940s. At that time, the steel company sought a labor force of US citizens who were hard working, would accept low wages, and could pass as “American.” Another company concern was to replace Mexican workers at risk of deportation by the Immigration and Naturalization Service. The Immigration and Naturalization Service threatened a raid and the expulsion of all nonAmerican workers, a group into which many Mexicans fell. It was imperative, then, to look for American replacements (Valdés, 1991). The steel mill company hired Puerto Ricans from the inland areas of the island of Puerto Rico to replace the Mexican labor. The Lorain Puerto Rican community is well established and has open and highly interconnected networks.

5. Methods The data were collected from recordings of 63 speakers. Three speakers belong to one of the three generational groups in Hazlet, New Jersey. The remaining 60 speakers were members of the Puerto Rican community in Lorain, Ohio. Of them, 20 speakers belong to each of the three generational groups and were evenly divided by generation and gender—that is, ten individuals per cell. The difference in size was done to replicate the proportion of Latinos in Hazlet, New Jersey, and in Lorain, Ohio: 4% of Puerto Ricans in Hazlet vs. 21.8% of Puerto Ricans in Lorain. At the time of data collection, I divided the speakers according to their “generation”. For the purposes of my research, I followed a definition for “generation” that differs from lay knowledge—that is, not in the same sense as age and parentage but rather according to the same criteria that Silva-Corvalán (1989, 1994) followed in her study of the Mexican American community in Los Angeles: 1 2

The first generation consists of people who were born in Puerto Rico and who moved to the United States after age 12. The second, of those who were born in the United States and whose parents came from Puerto Rico.

39 3

Use of /l/ and /ɾ / in US PRS speakers

39

The third, of those people who were born in the United States, whose parents were also born in the United States, but whose grandparents were born in Puerto Rico.

The main goal of these interviews was to obtain both the formal and the most casual speech from the participants, as well as their intuitions, judgments, and information about language choice. With this in mind, I followed the sociolinguistic paradigm of interviewing techniques for data collection developed by Labov (1966, 1972) and the participant observation techniques first developed by Blom and Gumperz (1972) and later adapted by Labov and his colleagues (Labov, Cohen, Robins, & Lewis, 1968). For all the groups, I followed the same steps—that is, we started with the strategy that will elicit the more formal speech then transition from different speech styles until I could elicit a more informal and relaxed speech. First, I asked the participants to read a series of word tables (three in total) and a short paragraph. The word tables were read twice so that there were two instances of each word, in case the first word recorded was not clear enough. Immediately after the reading of the paragraph, I engaged the participants in a short informal conversation dealing with different topics: hobbies, school experiences, and the like. Our conversations also included more targeted questions intertwined with the more casual and informal interactions. These targeted questions dealt with the specifics of the studies in terms of language use, opinions about language use, attitudes about language, and culture. These conversations were casual enough so that even the participants’ friends and relatives would also sometimes join us in our interaction. These interactions usually lasted for an hour, though some were as short as 45 minutes, and others were longer, the longest ones being 105 minutes. As mentioned in previous paragraphs, the participants read word lists and paragraphs and answered a variety of questions aimed to encourage natural conversation. The word lists included lexical items such as “televisor,” “yerba,” “horno,” “martes,” “hierva,” “singular,” “urna,” and “normal,” among other words that were mixed with words that contained “r” in onset position or did not contain “r” at all. I also prepared a list of questions for the informal conversation in Spanish, such as the following: 1 2 3 4 5 6 7 8 9

Would you prefer that people speak Spanish here, and why? Are any members of your family more likely to speak English than others? Do older people use more English? At what times? With whom? Do younger people use more English? At what times? With whom? Do you use English with older/younger family members? At what times? Are any members of your family more likely to speak Spanish than others are? Do older people use more Spanish? At what times? With whom? Do younger people use more Spanish? At what times? With whom? Do you use Spanish with older/younger family members? At what times?

40

Michelle F. Ramos-Pellicia

40

Even when these questions are listed here in English, they were all asked and tailored to a conversation in PRS. I used GOLDVARB X for Windows (Sankoff & Rand, 1999) to calculate the probabilistic effect of linguistic and extralinguistic factors to the application of the variable. Internal factors (e.g., grammatical category, aspects of the phonological environment) and external factors (e.g., generation) were considered in the coding of the data: 1 2 3 4 5 6 7 8

the phonological positioning within the word: word final or syllable final/ word internal the context according to preceding segment: [a], [e], [i], [o], [u] the phonological context according to following segment: bilabial, alveolar, velar consonants, pause the type of word according to the number of syllables (i.e., word length): monosyllabic or polysyllabic the type of word according to grammatical category: noun, adjective, verb, verb infinitives, adverb the speech style: reading, talking the gender: male, female, nonbinary the generational group according to age of arrival to the United States: first, second, third

Navarro Tomás (1948) and Álvarez-Nazario (1972, 1990, 1991) have pointed out that the lateralization of /ɾ / is found more often in syllable-final position in Island PRS; thus, the factor group phonetic environment included syllable-final, word-internal position and syllable-final, word-final position. Furthermore, deletion as a variant of /ɾ / has been reported in word and syllable-final positions, while the Costa Rican retroflex variant is reported in syllable-final position and especially phrase-final position (Lipski, 1994). These occurrences justify the need to include the type of phonological environment according to position within the word as an internal factor. The extralinguistic factors that were considered were speech style, gender, and generation. Generation plays a role in lateralization, as documented in Lamboy (2004) and Ramos-Pellicia (2004, 2007, 2009), as does length of residence in the United States. I included gender because previous studies of PRS, both on the island and in the United States, have documented that gender plays a role in the maintenance of the standard variant (Terrell, 1980; Ramos-Pellicia, 2004, 2007, 2009) and other times in the preference for lateralization (Poplack, 1979; Figueroa & Hislope, 1999; Holmquist, 2003, 2004). Socioeconomic status and level of education did not yield any reliable tendencies. The participants from both communities are working class and have limited education. GOLDVARB X for Windows (Sankoff & Rand, 1999) calculated the probabilistic effect of each described factor on the application of the variable. If a factor has a value higher than 0.500, then the factor favors the application of the variable. When a factor has a value lower than 0.500, then the factor disfavors the

41

Use of /l/ and /ɾ / in US PRS speakers

41

application. A value of exactly 0.500 means that the factor shows no effect on the rule.

6. Phoneme /ɾ / in Spanish The phoneme /ɾ / in syllable-final and word-final positions covers several variants, including a tap [ɾ ] produced with the tongue touching the alveolar ridge once, in free variation with a trilled or multiple [r] articulated on the same region with the tongue touching the alveolar ridge multiple times. Other variants are [l] and deletion, which both occur in word-final and syllable-final position. In certain dialects, such as Dominican Spanish, /ɾ / vocalization is found as another variant. Interestingly, an /ɾ / sound like American English [ɻ ], produced with the tongue’s apex going backward without touching the roof of the mouth, has been found in several Central American Spanish dialects. The retroflex [ɻ ] articulation is not found in other varieties of Spanish. IPRS

IPRS

IPRS

IPRS

LPRS

Gloss

[be.ˈβer] [fa.ˈβor] [a.ˈpar.te] [ˈlar.go]

[be.ˈβeɾ ] [fa.ˈβoɾ ] [a.ˈpaɾ .te] [ˈlaɾ .go]

[be.ˈβel] [fa.ˈβol] [a.ˈpal.te] [ˈlal.go]

[be.ˈβeø] [fa.ˈβoø] [a.ˈpaø.te] [ˈlaø.go]

[be.ˈβeɻ ] [fa.ˈβoɻ ] [a.ˈpaɻ .te] [ˈlaɻ .go]

“drink” “favor” “apart” “long”

Many researchers of Spanish in the US Southwest have argued that the retroflex [ɻ ] is a result of American English influence. Ornstein (1974) claims that the retroflex [ɻ ] is due to American English interference and explains that this phoneme is pronounced more like the corresponding American English sound than the corresponding one found in most other Spanish dialects. Furthermore, Lastra de Suárez (1975) also attributes to American English interference the substitution of a retroflex [ɻ ] for Spanish /ɾ / or /r/ in this variety. Sánchez (1973) states that the retroflex [ɻ ] found in the informal speech of Texas Mexican American speakers is a common instance of AE phonological interference in this dialect (e.g., [karne] [kaRne] [R] = (retroflex)) (Standard Spanish “carne”, “meat”). Cassano (1973, 1977) argues that New Mexico Spanish speakers substitute an AE retroflex for the Spanish flap /ɾ /. In the Spanish of New Mexico, Alonso (1930) heard an intervocalic and syllable-final /ɾ / similar to the AE /ɾ / with less retroflection. Espinosa (1930) provides a similar description of New Mexico Spanish. Se oye en posición intervocálica o final de sílaba. Se parece un poco a la r final del inglés hablado en el oeste de los Estados Unidos, pero la lengua no se arquea tanto hacia arriba y hacia atrás como en el sonido angloamericano. (141) It is heard in intervocalic or syllable-final position. It is somewhat similar to r in final position in the English spoken in the west of the United

42

Michelle F. Ramos-Pellicia

42

States, but the tongue does not arch as far upward and back as in the English sound. (The translation is mine.) All these varieties of Mexican American Spanish are in contact with American English. As well as Mexican American Spanish, other varieties of Spanish have a retroflex as a variant of /ɾ / in coda position. The retroflex variant is found in the central areas of Costa Rica. The retroflex occurs in syllable-final and especially phrase-final position. Speakers of Costa Rican Spanish extend the environments where the retroflex appears to those where trilled /ɾ / is produced in fast speech (Lipski, 1994). Belizean Spanish and Mexican Spanish as spoken in the Yucatan Peninsula are other Central American Spanish varieties that realize /ɾ / as a retroflex [ɻ ] (Hagerty, 1996). According to Lipski (1994), the retroflex [ɻ ] is more common among Belizean Spanish speakers than among the Spanish speakers of the Yucatan. The retroflection of /ɾ / is an innovation in these Spanish dialects when compared to other dialects of Spanish. The type of phonological environment according to the position within the word was considered because lateralization and deletion as variants of /ɾ / have been reported in word-final and syllable-final positions. Additionally, this internal factor was considered because the retroflex variant found in the central areas of Costa Rica is documented in syllable-final position and especially phrase-final position (Lipski, 1994).

7. Review of research on Puerto Rican Spanish /ɾ / When considering the use and distribution of /ɾ / and /l/ in PRS, keep in mind that we might be talking about different tendencies in Island PRS and in Mainland PRS. 7.1 Island Puerto Rican Spanish /ɾ / In this section, I will first consider the studies of PRS as spoken in Puerto Rico as the ancestral variety of the speakers in Hazlet and Lorain. After I consider the different studies of Island PRS, I will turn to the review of PRS as spoken on the mainland. Navarro Tomás (1948) in his seminal study of PRS documented the use of /ɾ /. He observed that the maintenance of /ɾ / was characteristic of the speakers in the highlands in the northwest, while the lateralization of /ɾ / was found in the northeast region and neutralization was more engrained in the San Juan and nearby regions. Navarro Tomás (1948) also reported that lateralization was used sporadically in the upper educated levels of Puerto Rican society. Island PRS, like the Caribbean Spanish dialects of Dominican Republic and Cuba, has [l] and deletion as /ɾ / variants. Navarro Tomás (1948) and ÁlvarezNazario (1972, 1990, 1991), in their studies of Island PRS, note that /ɾ / is often pronounced as [l] in syllable-final position or word-final position, so that,

43

Use of /l/ and /ɾ / in US PRS speakers

43

for example, “hablar”, “to speak” is pronounced [a.ˈblal]. Retroflection has not been documented as a variant of /ɾ / in Island PRS. On the other hand, Terrell (1980) observed that Puerto Rican women showed a higher tendency to the use of the standard /ɾ /, with younger adults and older speakers producing a lateralization of /ɾ / more frequently than do middle-aged speakers. Later on, López Morales (1983) found that San Juan male residents produced the lateralization of /ɾ / more often than female residents. Contrary to Terrell (1980), López Morales’s (1983) work observed that lateralization is more often found among speakers from San Juan who were either born in San Juan or moved to the San Juan region before age six. These speakers were more likely to prefer lateralization when compared to speakers with rural origins who moved to San Juan in their teenage or adult years. Lateralization was reported among younger speakers with some education, ranging from high school to some college work (Prosper-Sánchez, 1995) in the northwest. Prosper-Sánchez (1995) reported that older speakers in the same area with no high school or college education did not show as high a frequency of use of lateralization. In the western highlands, Holmquist (2003, 2004) documented that the youngest women used lateralization with higher frequencies, presenting counter evidence to the principle that women in Western societies show a higher tendency to the use of the standard forms (Labov, 1990). While with a low frequency, the younger speakers show a preference for lateralization, but not the older speakers. In a control group from the southeast and northwest of Puerto Rico, RamosPellicia (2004, 2007, 2009) found the lateralization of /ɾ / in word-final position, as previously described in the literature (Navarro Tomás, 1948; Álvarez-Nazario, 1972, 1990, 1991). More formal style (i.e., reading) yielded more cases of [ɾ ] among Island PRS speakers, while talking did not favor the occurrence of [ɾ ], instead favoring [l] (70%, N = 495). Female speakers of Island PRS favored the use of [ɾ ] more than the male speakers, a result that is not surprising and that has been described in the linguistic literature: women tend to favor the use of the prestigious variants as long as they are exposed to them. The Puerto Rican males were almost evenly divided in their usage of [ɾ ] (46%, N = 476) and [l] (49%, N = 504). Both age groups were evenly divided between [ɾ ] and [l]. The older Puerto Ricans preferred [ɾ ] (55%, N = 555) over [l] (39%, N = 393), just as the younger group, with 54% for [ɾ ] and 43% for [l]. Lateralization was also shown to be favored among men, young adults, and speakers with rural origin in the eastern central region of Puerto Rico (MedinaRivera, 1997). The same author (Medina-Rivera, 2014) in a more recent study of the frequency of /ɾ / as a lateral in written documents proposed that El fenómeno de la lateralización en Puerto Rico va más allá de lo que los estudios muestran, ya que solo toman en consideración los registros orales. Las nuevas redes sociales de comunicación le brindan al sociolingüista

44

Michelle F. Ramos-Pellicia

44

otras fuentes para examinar el lenguaje y un mayor número de registros. Hasta cierto punto, los hablantes puertorriqueños utilizan formas no estándar como el cambio de “r” por “l” como una manera de poner en evidencia un fenómeno que va adquiriendo un sentido de identidad nacional. (336–337) The lateralization phenomenon in Puerto Rico goes further than what studies have demonstrated because these studies take into consideration only the oral registers. The new social networks of communication provide the sociolinguist other sources to examine language and a larger number of registers. Up to a certain point, the Puerto Rican speakers use non-standard forms like the change from “r” to “l” in a way that makes evident a phenomenon that is acquiring a sense of national identity. (The translation is mine.) Now that we have considered what has been documented of Island PRS, it can be concluded that there is a clear tendency for preferring the lateralization of /ɾ / regardless of region but that gender, age, and rural origin are extralinguistic characteristics that play a role in the frequency of its production. 7.2 US Puerto Rican Spanish /ɾ / Several studies have concentrated on the production of /ɾ / in PRS on the mainland, among them Decker (1952), Ma and Herasimchuck (1972), Shouse de Vivas (1978), Poplack (1979), Figueroa and Hislope (1999), Lamboy (2004) Valentín-Márquez (2007), and Ramos-Pellicia (2004, 2007, 2009). In an earlier study of PRS in Lorain, Ohio, Decker (1952) mentions several /ɾ / variants. In this study, Decker found [r] and [l] as allophones of /ɾ / and an intermediate variant: Also may be found a variety of l and r which bears articulatory characteristics of each, the variety being an allophone of the two phonemes l and r. One can best approximate this articulation by forming the phoneme l with the posterior and lateral areas of the tongue while allowing the tip of the tongue to articulate an alveolar relaxed or lightly flapped r. (116) R-colored vowels are most common in Southern American English (SAE) words like bear, fur, and perfect; in these words, the vowel and the r are a single sound and are written as such in IPA (International Phonetic Alphabet) and SAE phonetic transcription. This r-coloring was found in 10% of all responses, but never without the addition of a true alveolar or prealveolar r. Furthermore, Decker (1952) mentions that [ɾ ] was articulated on the alveolar ridge and that “a markedly relaxed r (similar to SAE)” occurs after a vowel, “so that the first element properly belongs to the preceding r-colored vowel”

45

Use of /l/ and /ɾ / in US PRS speakers

45

(Decker, 1952, p. 116). The description that Decker (1952) provides of this [ɾ ] variant refers to a tap with a preceding r-colored vowel, not a retroflex. Decker (1952) also observes that lambdacisms—“equalization” (in his own words, “neutralization”) of /l/ and /ɾ /—and /ɾ / vocalization are found with less frequency in Lorain PRS than in previous studies of PRS. Ma and Herasimchuck (1972) documented the use of lateralization among Puerto Rican speakers in New Jersey. The researchers reported that those with a lower educational level, low socioeconomic status, and rural origin and who have lived longer in the United States showed a preference for lateralization. Such an occurrence of /ɾ / was found to be more frequent among younger adults and middle-aged speakers. Poplack (1979), in her study of a working-class community of Puerto Ricans in Philadelphia, found limited lateralization. The cases of lateralization were more frequent among speakers from the southwest of the island, women, and participants with lower educational levels. Contrary to Poplack (1979), Shouse de Vivas (1978) found that speakers who came from the northeast of the island and lived in Springfield, MA, presented more lateralization. Similar to previous studies, Shouse de Vivas (1978) also found that factors such as age, education, and years of living in the United States influenced the preference for lateralization. In other words, younger speakers, participants with limited education, and speakers who have lived the longest in the United States presented a tendency that favors lateralization more. Ramos-Pellicia (2004, 2007, 2009) studies reported that men and women presented similar frequencies of the lateralization of /ɾ / in Lorain, Ohio. The researcher also found that lateralization was associated with second-generation immigrants more than with speakers who from the first and third generations. On the other hand, Lamboy (2004) compared first- and second-generation Puerto Rican speakers in New York City. Lateralization was found in 25% of occurrences, but Lamboy found no significant differences between the two generational groups. Figueroa and Hislope (1999) found that lateralization occurred five times more frequently in male speakers than in female speakers. Figueroa and Hislope (1999) also reported “other” variants (e.g., semivocalization, the mixed variant, deletion, gemination, and the English alveolar flap) and suggested that PRS in Indiana is undergoing change in its phonetic inventory. In summary, it is clear that there is an inversely proportional correlation between lateralization and education. The more education a speaker has, the lesser the tendency to lateralize. There is also a direct relation between years of living in the United States and lateralization. The longer a speaker of PRS has lived in the United States, the greater the chance for lateralization of /ɾ /. On the other hand, there is no direct relation between age, gender, region of origin, and the lateralization of /ɾ /. In the data for this work, however, I will consider generation and gender as the only extralinguistic factors. These extralinguistic factors showed to play a role in

46

Michelle F. Ramos-Pellicia

46

the lateralization of /ɾ /, contrary to the studies on PRS in the United States already discussed. With respect to socioeconomic status and level of education, all community members who participated in the study belong to the working class and did not attain a level higher than a primary school education or a high school diploma.

8. Results While very few cases of rhotacism were heard during the interviews, many cases of lateralization were collected from the Puerto Ricans from Hazlet. In wordinternal position, words like [pwél.to], “puerto”; [o.pol.tu.ni.ðád], “oportunidad”; and [re.kwél.do], “recuerdo”, were collected. Words with lateralization in word-final position were [vja.hál], “viajar”; [a.bl ál], “hablar”; and [eh.tu. ðjál], “estudiar”. Table 2.1 presents the results of the tokens for PRS in Hazlet. It shows the number of tokens collected and analyzed. The number of tokens is small because of the sampling size. Of the total number of tokens, 18, there were 14 cases, or 13% of the standard variant—with 87% of cases, or 94 tokens, of lateralization and no cases of deletion or retroflection. I only found seven cases of rhotacism in the conversations with the PRS speakers in Hazlet. Table 2.2 shows that it was the first-generation speaker who produced all the cases of rhotacism perhaps because of the proximity to the ancestral variety of Island PRS. Rhotacism was found in words like [kwár] “cual”; [er.kó.ko] “el coco”. Lateralization occurred the most in front of alveolar consonants: [d], [n], [s], [t]—that is, “recue[l]do”, “impo[l]tante”. The /ɾ / at the end of verbs in the infinitive was lateralized the most—that is, “trabaja[l]”, “pensa[l]”. Lateralization through assimilation occurred in third-generation speech—that is, “cria[l]lo”, “aprende[l]lo”. Various words were also pronounced with the standard phoneme as well—that is, “pue[ɾ ]to”, “aprende[ɾ ]”—especially in the third generation. In the data, /ɾ / was lateralized in word-internal position (81.2%), while wordfinal position represented 97.4% of lateralization. Lateralization of /ɾ / occurred more after [a] (94.4%) than after [e] (87.7) or [o] (66.7%). Similarly, /ɾ / was Table 2.1 Number of tokens of /r/ variants for PRS speakers in Hazlet, NJ Group

Number of tokens analyzed

[r]

[l]

[ɻ ]

Ø

PRS, Hazlet, NJ

108

14 / 13%

94 / 87%

0

0

Table 2.2 Number of /l/ > [r] tokens for Puerto Rican speakers in Hazlet, NJ Position

First Generation

Second Generation

Third Generation

Rhotacism in word-final position (only) PRS, Hazlet, NJ

7

0

3

47

Use of /l/ and /ɾ / in US PRS speakers

47

lateralized more when it was followed by a pause (93.3%) than when followed by an alveolar consonant (80%). In the lexical category, infinitives presented the highest percentage for lateralization (97.1%), followed by verbs (92.3%), nouns (82.1%), and the factor group comprised by adjectives and adverbs (68.4%). Before a statistical analysis could be done, I had to either collapse several factors or eliminate several factor groups, because they presented knockouts of singletons that prevented a binominal analysis with VARBRUL. Among the factor groups eliminated were following phonological context and word length. In the preceding phonological environment, I had to eliminate [i] due to a lack of variation. Similarly, for grammatical function, the factors adjective and adverb were collapsed because of singletons in the data. Once these factor groups and/or factors were eliminated or collapsed, the statistical analysis yielded a significance of 0.009. The VARBRUL statistics, as seen in Table 2.3, demonstrate that /ɾ / is more likely to be lateralized in word-final position (0.785), with word-internal position (0.325) disfavoring the application of the rule. The vowel in the preceding environment was also demonstrated to play a role in /ɾ / lateralization. As Table 2.3 demonstrates, [a] favors /ɾ / lateralization (0.582), but [e] (0.483) and [o] (0.370) disfavor /ɾ / becoming [l]. Infinitives (0.683) in particular and verbs (0.519) in general favored the application of lateralization. Nouns (0.454) and both adverbs and adjectives (0.285) disfavor the application of the rule. Few cases of rhotacism were found in the conversations among the first- and thirdgeneration speakers. Because there was not a significant display of rhotacism, a statistical analysis was not performed in any of the two data sets. Similarly, because of the limited data from Hazlet, a statistical analysis was not possible with respect to the contribution of factors to the lateralization of /ɾ / in Hazlet. Third-generation speakers with more exposure to Spanish followed the same pattern of use for the /ɾ / variables found in Island PRS: they preferred [l] over [ɾ ]. Among the first-generation speakers, the retroflex /ɾ / is related to their exposure to American English through its monolinguals in the community and their Table 2.3 Contribution of factors to the lateralization of /r/ in Lorain PRS Factor Position Word final Word internal Preceding Environment [a] [e] [o] Lexical Category Infinitives Verbs Nouns Adverbs and adjectives Significance

Probability Value 0.785 0.325 0.582 0.483 0.370 0.683 0.519 0.454 0.285 0.009

48

Michelle F. Ramos-Pellicia

48

families: speakers of the second and third generations who are bilingual. The retroflex in the first generation has several sources: American English from the American English speakers in the community; second and third generations’ American English; and first-generation speakers with some contact with American English. For the second and third generations of Puerto Ricans in Lorain, on the other hand, American English influence is behind the production of the retroflex when reading in Spanish. These speakers are recurring to the use of this alternative for /ɾ / because they have received most of their education in American English and read mostly, if not exclusively, in this language. No cases of the retroflex were observed—or, rather, heard—among the participants in Hazlet, New Jersey. Given what I know of the PRS in Lorain, Ohio, I was expecting to hear a retroflex as a variant for /ɾ / in Hazlet, New Jersey, as well. As in Lorain, Ohio, Hazlet’s participants also have extended direct contact with American English speakers, even when the contact with first-generation speakers with some contact with American English is not as extensive. No cases of rhotacism were collected from the Puerto Ricans in Lorain, Ohio, so a statistical analysis could not be performed. Puerto Rican participants in Lorain follow a similar pattern of usage of [ɾ ] as that reported in the studies of PRS on the island of (Navarro Tomás, 1948; Álvarez-Nazario, 1972, 1990, 1991). As shown in Table 2. 4, speakers of PRS in Lorain favored [ɾ ] in syllable-final position more than in word-final position, where [l] was used more. Lorain Puerto Ricans preferred [ɾ ] in their more formal style, but not in their conversations, where [l] was preferred. Thus, they maintain the use of the more conservative variable in their most monitored form of speech, just as the Puerto Ricans in Puerto Rico do, while preferring to use [l] in their conversations (see Table 2.4). Puerto Rican speakers in Lorain are deleting [ɾ ] in conversations, while choosing to make use of a retroflex [ɻ ] when they are asked to read in Spanish.

Table 2.4 VARBRUL weights, application values and input and weights for the variant [r] in Lorain PRS (Total of tokens analyzed: 4,291) Factor Group

Factor

Phonetic Environment Word final Syllable final Style Reading Talking Gender Male speakers Female speakers Generation First Second Third Input: 0.511

Weight

App/Total

Input & Weight

0.419 0.595

0.43 0.61

0.43 0.61

0.603 0.245

0.61 0.26

0.61 0.25

0.470 0.527

0.49 0.54

0.48 0.54

0.590 0.466 0.351

0.58 0.49 0.39

0.60 0.48 0.36

49

Use of /l/ and /ɾ / in US PRS speakers

49

Table 2.4 also shows that male and female Puerto Ricans in Lorain are closely divided in their usage of [ɾ ] and [l], with the female speakers slightly preferring [ɾ ] over [l], while the male speakers present a slight difference between their usage of [l] and [ɾ ]. The preference to use [ɾ ] declines across the three different generations in Lorain, as the first generation favors [ɾ ] usage, while the second generation is barely divided between [ɾ ] and [l]—but not so for the third-generation speakers, who prefer [l]. A retroflex [ɻ ] was found among third-generation speakers (10%, N = 78) more than among second-generation speakers (2%, N = 33). First-generation speakers also used retroflex [ɻ ] (3%, N = 52). The retroflex variant could not be included in the statistical analysis, because of its low frequency. Lorain Puerto Ricans have maintained the same patterns of usage for (ɾ ) as documented in the literature on PRS in Puerto Rico. In other words, both groups (Lorain and Puerto Rico) use [l] in informal contexts and also use deletion as an alternative for /ɾ /. The gross distribution with regard to percentages, style, and phonetic environment are essentially the same: the numbers are not identical, but the trend is. On the other hand, Lorain Puerto Ricans are making use of a retroflex [ɻ ] when they are asked to read in Spanish. A likely reason for this is American English influence because speakers of PRS in Lorain are more likely to read in American English and have received most all of their education in this language. On the other hand, PRS speakers in Hazlet, while not producing enough tokens to make a statistical analysis possible, showed that they have a preference for lateralization, regardless of gender, style, or generation. These results differ from what is documented about Island PRS and PRS in the United States.

9. Conclusion Among the Puerto Ricans in Lorain, lateralization is more likely to occur in the infinitive form of the verbs. This is the impersonal verbal form that ends in /ɾ /, which speakers show a preference for in its lateral variant. One of the thematic vowels for the infinitives in Spanish is [a]. Of all the vowels preceding the infinitive form ending in [l], the data show that this is the vowel that favors lateralization the most. This is not surprising given that there are many infinitives that include this sound in particular, and many of these examples were collected in the recordings. The third generation’s pronunciation is influenced not only by the Spanish of the first generation but also by their formal education in Spanish, which resulted in more standardized pronunciation. The combination and interaction of linguistic (e.g., phonological environment, lexical context) and extralinguistic factors (e.g., level of education, exposure to American English, use of Spanish within the community, linguistic isolation, and negative prestige) result in the variables for /ɾ / and /l/ used by both communities of Puerto Rican speakers. The Puerto Ricans in Lorain demonstrated a preference for lateralization in word-final position (0.785), and although some cases of retroflex /ɾ / were found, the frequency was not

50

Michelle F. Ramos-Pellicia

50

enough to run a statistical analysis. The vowel in the preceding environment was also demonstrated to play a role in /ɾ / lateralization. Rhotacization occurred in the Spanish of the first and third generations, not the second, but with values that were considered statistically insignificant. Among the Midwest Puerto Ricans, the preference to use [ɾ ] declines across the three different generations: first generation (0.590), second generation (0.466), and third generation (0.3510). These speakers also use a retroflex as an alternative for /ɾ / when reading in Spanish. No cases of rhotacization were found in this community of speakers. As far as the use of /ɾ / and /l/ in the community of Puerto Ricans in Hazlet is concerned, the findings in this study were indeed congruent with the observed pattern of use for lateralization in PRS. However, while the second- and thirdgeneration speakers are well educated and the first-generation speakers had minimal education, lateralization still occurred with about the same frequency. Although this group has been fairly isolated linguistically, there is evidence of maintenance in this dialect, but with different trends. On the other hand, PRS speakers in Hazlet, while not producing enough tokens to make a statistical analysis possible, showed that they have a preference for lateralization, regardless of gender, style, or generation. These results differ from what is documented about PRS in general. For instance, in Island PRS, Navarro Tomás (1948) reported a sporadic use of the lateral variant among higher-educated speakers in Puerto Rico. The same phenomenon was reported among younger speakers with education that ranged from high school to some college work, while older speakers with no high school or college education did not show a high frequency of the lateralization of /ɾ / (Prosper-Sánchez, 1995). Further, the lateralization of /ɾ / is reported among PRS speakers in the United States who have limited education (Ma & Herasimchuck, 1972; Shouse de Vivas, 1978; Poplack, 1979). While the results are inconclusive, there are implications for the study of PRS as spoken in the United States. Contrary to what could be thought about this variety of Spanish in the United States, there is not a clear trend of how speakers use their language variety. Factors that undoubtedly play a role are exposure to American English and language transmission through generations. The more exposure to American English and the longer down the line a speaker is among the generational groups in the United States, the more they will produce a Spanish that will differ from their ancestral variety. Furthermore, other factors need to be considered. In these communities, there is no regular contact with their ancestral variety and other varieties of PRS in the United States. Additionally, there is a lack of input from the standard PRS variety. These three factors, as attested in the data discussed in this work, will have implications for the transformation of the many varieties of PRS found in the United States into individual varieties with differing tendencies. As time passes, these could change enough to pass from consideration as PRS varieties to Caribbean Spanish varieties. Even more so, as time progresses, these varieties could become completely different Spanish varieties spoken in the United States.

51

Use of /l/ and /ɾ / in US PRS speakers

51

References Alonso, A. (1930). Estudios sobre el español de Nuevo Méjico. Buenos Aires: Universidad de Buenos Aires. Álvarez-Nazario, M. (1972). La herencia lingüística de Canarias en Puerto Rico. San Juan: Instituto de Cultura Puertorriqueña. Álvarez-Nazario, M. (1990). El habla campesina del país. Río Piedras, PR: Editorial Universitaria. Álvarez-Nazario, M. (1991). Historia de la lengua española en Puerto Rico. San Juan: Academia Puertorriqueña de la Lengua Española. Blom, J.-P., & Gumperz, J.J. (1972). Social meaning in linguistic structures: Code switching in Norway. In J.J. Gumperz & D. Hymes (Eds.), Directions in sociolinguistics (pp. 400–434). New York: Holt, Rinehart and Winston. Cassano, P.V. (1973). The influence of American English on the phonology of American Spanish. Orbis, 22(1), 201–214. Cassano, P.V. (1977). Problems in language borrowing and lending exemplified by American Spanish phonology. Orbis, 36(1), 149–163. Decker, B.D. (1952). Phonology of the Puerto Rican Spanish of Lorain, Ohio: A study in the environmental displacement of a dialect. Unpublished M.A. thesis, The Ohio State University, Columbus, OH. Espinosa, A. (1930). Estudios sobre el español de Nuevo Méjico. Parte I. Buenos Aires: Biblioteca de Dialectología Hispanoamericana. Figueroa, N., & Hislope, K. (1999). A study of syllable final /r/ neutralization in Puerto Rican Spanish. Romance Languages Annual, 10, 563–569. Guitart, J.M. (2004). Sonido y sentido: Teoría y práctica de la pronunciación de lespañol con audio CD. Washington, DC: Georgetown University Press. Hagerty, T. (1996). The influence of English on the Spanish language of Belize. In M.D. Phillips (Ed.), Belize: Selected proceedings from the second interdisciplinary conference (pp. 131–142). Lanham, MD: University Press of America. Holmquist, J. (2003). Coffee farmers, social integration and five phonological features: Regional socio-dialectology in west-central Puerto Rico. In L. Sayahi (Ed.), Selected proceedings of the first workshop on Spanish sociolinguistics (pp. 70–76) Sommerville, MA: Cascadilla Proceedings Project. Holmquist, J. (2004). Social stratification in women’s speech in Rural Puerto Rico: A study of five phonological features. Paper presented at the Second International Workshop on Spanish Sociolinguistics (WSS2), SUNY at Albany, March 25–26. Labov, W. (1966). The social stratification of English in New York City. Washington, DC: Center for Applied Linguistics. Labov, W. (1972). Sociolinguistic patterns. Philadelphia, PA: University of Pennsylvania Press. Labov, W. (1990). The intersection of sex and social class in the course of linguistic change. Language Variation and Change, 2, 205–254. Labov, W., Cohen, P., Robins, C., & Lewis, J. (1968). A study of the non-standard English of Negro and Puerto Rican speakers in New York City. Report on Co-operative Research Project 3288. Washington, DC: Office of Education. Lamboy, E.M. (2004). Caribbean Spanish in the metropolis: Spanish language among Cubans, Dominicans, and Puerto Ricans in the New York City area. New York: Routledge. Lastra de Suárez, Y. (1975). El habla y la educación de los niños de origen mexicano en Los Ángeles. In E. Hernández-Chávez, A. Cohen, & A. Beltramo (Eds.), El lenguaje de

52

Michelle F. Ramos-Pellicia

52

los chicanos: Regional and social characteristics of language used by Mexican Americans (pp. 61–69). Arlington, TX: Center for Applied Linguistics. Lipski, J.M. (1994). Latin American Spanish. New York: Longman Linguistics Library. López Morales, H. (1983). Estratificación Social del Español de San Juan de Puerto Rico. México: Universidad Nacional Autónoma de México. Ma, R., & Herasimchuck, E. (1972). The linguistic dimensions of a bilingual neighborhood. In J. Fishman, R. Cooper, & R. Ma (Eds.), Bilingualism in the Barrio (pp. 347–464). Bloomington, IN: Indiana University Press. Medina-Rivera, A. (1997). Variación Fonológica y Estilística en el Español de Puerto Rico. Doctoral dissertation, University of Southern California, LA, California. Medina-Rivera, A. (2014). Lateralización de /ɾ / implosiva: la conciencia fonológica y sus manifestaciones en el español puertorriqueño oral y escrito. In A. Enrique-Arias, M.J. Gutiérrez, A. Landa, & F. Ocampo (Eds.), Perspectives in the study of Spanish language variation. Papers in Honor of Carmen Silva-Corvalán (pp. 321–340). Santiago de Compostela: Universidad de Santiago de Compostela. Navarro Tomás, T. (1948). El español en Puerto Rico: Contribución a la Geografía Lingüística Hispanoamericana. Río Piedras, PR: Universidad de Puerto Rico. Ornstein, J. (1974). Mexican American sociolinguistics: A well-kept scholarly and public secret. In B. Hoffer & J. Orenstein (Eds.), Sociolinguistics in the Southwest (pp. 91–121). San Antonio, TX: Trinity University. Poplack, S. (1979). Function and process in a variable phonology. Doctoral dissertation, University of Pennsylvania, Philadelphia, PA. Prosper-Sánchez, G.D. (1995). Neutralización homofonética de líquidas a final de sílaba: Aspectos sociolingüísticos en el español de Puerto Rico. Doctoral dissertation, University of Massachusetts, Amherst, MA. Ramos-Pellicia, M.F. (2004). Language contact and dialect contact: Cross-generational phonological changes in a Puerto Rican community in the Midwest of the United States. Doctoral dissertation, The Ohio State University, Columbus, OH. Ramos-Pellicia, M.F. (2007). Lorain Puerto Rican Spanish and ‘r’ in three generations. In J. Holmquist, A. Lorenzino, & L. Sayahi (Eds.), Selected proceedings of the third workshop on Spanish sociolinguistics (pp. 53–60). Somerville, MA: Cascadilla Proceedings Project. www.lingref.com, document #1526. Ramos-Pellicia, M.F. (2009). Language contact and dialect contact: Cross-generational phonological variation in a Puerto Rican community in the Midwest of the United States. Saarbrücken, Germany: VDM Verlag. Ramos-Pellicia, M.F., & Ayala, M. (2008). Phonological variations of /l/ and /ɾ / in three generations of Puerto Rican Spanish speakers. Paper presented at Hispanic Linguistics Symposium 2008, Université Laval, Québec, Canada. Sánchez, R. (1973). Nuestra circunstancia lingüística. Voices. Readings from El Grito. A journal of Mexican American thought 1967–1973 (pp. 420–449). Berkeley, CA: Quinto Sol Publications. Sankoff, D., & Rand, D. (1999). GOLDVARB X for Windows (version 2.0). Université de Montréal. Shouse de Vivas, D. (1978). El uso de [l] variante de ‘r’ en el habla de Puerto Rico. In Actas del V Congreso Internacional de la Asociación de Lingüística y Filología de la América Latina, ALFAL (pp. 622–631). Caracas: Universidad Central de Venezuela. Silva-Corvalán, C. (1989). Sociolingüística. Teoría y análisis. Madrid: Editorial Alhambra, S.A.

53

Use of /l/ and /ɾ / in US PRS speakers

53

Silva-Corvalán, C. (1994). Language contact and change: Spanish in Los Angeles. Oxford: Clarendon Press. Terrell, T.D. (1980). Current trends in the investigation of Cuban and Puerto Rican phonology. In J. Amastae & L. Elías-Olivares (Eds.), Spanish in the US: Sociolinguistic aspects (pp. 47–70). Cambridge: Cambridge University Press. U.S. Census Bureau. (2010). American fact finder, Hispanic or Latino by type. Data Profiles. Retrieved from https://factfinder.census.gov/faces/tableservices/jsf/pages/pro ductview.xhtml?src=CF U.S. Census Bureau. (2016). American community survey, 2016 ACS 5-year estimates data profiles. Retrieved from https://data.census.gov/cedsci/results/tables?q=puerto% 20ricans%20in%20lorain,%20ohio&g=1600000US3944856:0400000US72&tab=AC SProfile5Y2016.DP05&ps=app*page@1$app*from@RESULTS_ALL Valdés, D.N. (1991). Al norte: Agricultural workers in the Great Lakes Region, 1917–1970. Austin, TX: University of Texas. Valentín-Márquez, W. (2007). Doing being Boricua: Perceptions of national identity and the sociolinguistic distribution of liquid variables in Puerto Rican Spanish. Doctoral dissertation, University of Michigan, Ann Arbor, MI. Van Coetsem, F. (1988). Loan phonology and the two transfer types in language contact. In Publications in language sciences, 27. Providence, RI: Foris Publications.

3

Laterals in contact Miami-Cuban Spanish and English /l/ Brandon M. A. Rogers and Scott M. Alvord

1. Introduction As a bilingual city, Miami provides an ideal laboratory for studying the effects of language contact (e.g., López Morales, 2003; Lynch, 1999). Because of the rapid increase of Cubans in the United States during the past 60 or more years, Miami became the first city of more than 2,000,000 inhabitants whose population was a Hispanic majority (López Morales, 2003; Carter & Lynch, 2015). According to the Pew Hispanic Center, in 2014, 66% of the population of Miami-Dade County was of Hispanic origin (1,762,598 out of 2,662,874) (www.pewhispanic. org/ states/county/12086/). Likewise, it is considered “the most bilingual city in all of the Americas” (369) and “arguably the most dialectally diverse Spanishspeaking city in the world” (Carter & Lynch, 2015, p. 370) The Spanish and English spoken in Miami have shown effects of the years of language contact, particularly in the Miami-Cuban community (e.g., Boswell, 2000; Lynch, 1999, 2000; Carter & Lynch, 2015; Alfaraz, 2002; 2014; Alvord & Rogers, 2014; Rogers & Alvord, 2017). Most language-contact studies in Miami have focused on the influence of English on this variety of Spanish, particularly on the morphosyntax and lexicon (e.g., Varela, 1974; Otheguy & García, 1988; Lynch, 2000, among others). More and more studies have begun to examine the phonology and prosody of Miami-Cuban Spanish. For example, Varela (1992), studying Cuban Americans in general, documented the widespread reduction in the vowel systems of her participants in Miami and attributed this to English influence on their vowel phonology. Alvord (2010) analyzed Miami-Cuban absolute interrogative patterns in three immigrant groups and posits that these patterns in Miami-Cuban speech could be influenced not only by English but also by the various other Spanish varieties spoken in Miami. Alvord and Rogers (2014) performed an acoustic analysis of over 9,000 Miami-Cuban Spanish vowel productions from three different immigrant generations. Their results show that contrary to Varela’s (1992) findings, the Miami-Cuban Spanish vowel system is resistant to contact-induced change from English, even in the more-English-dominant immigrant groups. With regard to the effect of Spanish on Miami-Cuban English, there are few studies. Carter (2013) hypothesizes that there are several areas where Miami

55

Laterals in contact

55

English is perceived as differing from general American English. He states that “Miami English is a systematic, rule-governed variety of English with subtle structural influence from Spanish. It is spoken by native English speakers, mostly second-, third, and fourth-generation Latinos who learn it as their firstlanguage variety” (para. 5). According to Carter, some of these Spanish influences are lexical and phonetic. With regard to the phonetic influences, he states that there are some vowel differences between Miami English and general American English, along with “stronger-sounding consonants” (para. 6), specifically mentioning /l/ and /r/ as salient examples. There have been a limited number of studies on Miami Latino English. A notable number of these studies have discussed and analyzed the sociolinguistic implications and perceptions of English in Miami and the relationship it shares with Spanish (e.g., Carter & Lynch, 2015; Carter & Callesano, 2014; Carter, López, & Sims, 2014). Only two known acoustic studies on Miami English have been conducted. Labov, Ash, and Boberg (2006) analyzed the speech of a middle-aged white woman, and based on the results, they placed Miami English in the southeastern region on their linguistic atlas of the United States. In a more extensive acoustic study on Miami-Cuban English and Spanish vowels, Rogers and Alvord (2017) analyzed over 4,600 Miami-Cuban English vowel token from three immigrant groups. While their similar 2014 study found that the Miami-Cuban Spanish vowel system was resistant to Englishinduced change, their study on Miami-Cuban English showed that the MiamiEnglish vowel system is less resistant to change influenced by Spanish. Their results show a backward shift of the English front lax vowel /æ/ into the traditional low central space occupied by Spanish /a/, resulting in pronunciations such as [‘fa.ʃ n] for the word fashion. These results empirically confirm the anecdotal notion put forth by Carter (2013) that the Miami-English vowel system shows evidence of Spanish influence. Based on the anecdotal observations of researchers such as Carter (2013) and the empirical results recently put forth by Rogers and Alvord (2017), the potential effects of Spanish on Miami English merit investigation, given that most studies have been overwhelmingly unidirectional in their analyses of the contact relationship between Spanish and English in Miami. The current study is a preliminary acoustic analysis within in a laboratory phonology framework of Miami-Cuban Spanish and English /l/ from ten bilingual Miami Cubans belonging to three immigrant generations (Silva-Corvalán, 1994). The results offer an initial perspective regarding the possible influences that both languages have had on aspects of each other’s segmental phonology. e

2. Review of the literature While the use of two different allophonic variants, specifically [ɫ] and [l], to describe the different articulatory and phonetic properties of lateral productions is a common way to describe lateral variation in a number of previous studies,

56

Brandon M. A. Rogers and Scott M. Alvord

56

the process of darkening, or velarization, is generally considered to be scalar (Recasens & Farnetani, 1994). This chapter treats this phenomenon as part of a continuum as well, even though some previous studies may treat it as a binary process. 2.1 English and Spanish laterals While the current chapter views lateral “lightness” and “darkness” as a phonetic continuum, it must be mentioned that not all researchers have described lateral variation in English as gradient. In most studies, the light [l] is said to be produced in syllable-initial position, with the dark, or velarized [ɫ], in syllablefinal and sometimes nuclear position (e.g., Yuan & Liberman, 2009). While some studies consider these phones to be phonologically distinct, other researchers have called into question both the phonological status and the phonetic implementation of both variants (e.g., Sproat & Fujimura, 1993; Oxley, Buckingham, Roussel, & Daniloff, 2006; Yuan & Liberman, 2009). Several studies have argued that the differences between [l] and [ɫ] are allophonic and not phonemic. For example, Sproat and Fujimura (1993), EpsyWilson (1992), and Huffman (1997) all present evidence that the phonetic differences between [l] and [ɫ] are due to gestural and timing differences and argue that they are not discreet categories. Additionally, Oxley et al. (2006) found that the variation between light and dark laterals is gradient rather than binary. Crosslinguistically, the light/dark dichotomy has been analyzed in other languages, such as German and different varieties of Catalan (e.g., Recasens & Farnetani, 1994; Recasens, 2004; Recasens & Espinosa, 2005), and the results support the notion that there is no categorical difference between different variants of light and dark laterals. Rather, the differences are gradient and depend on different phonetic factors. Cross-dialectally, English has been shown to have a notable amount of lateral variation as well. For example, Thomas (2007) states that in African American English (AAE) /l/ is frequently either deleted or vocalized in different phonetic contexts (e.g., [‘fio] or [‘fi ] as variant pronunciations of “feel”). Lateral deletion and vocalization have also been attributed to Southern American English (SAE) by Caffee (1940). Van Hofwegen (2010) examined word-initial /l/ in AAE, documenting a gradual darkening from one generation of speakers to the next, possibly due to increased contact and exposure to other majority dialects of American English. She also posits that lighter /l/s may have been a feature of earlier varieties of AAE. Hall-Lew and Fix (2012) demonstrated that /l/ has differing levels of social capital among English speakers, with greater levels of vocalization being associated with non-white ethnicities. In North Wales, Morris (2017) showed variation of light and dark /l/ among bilingual speakers of Welsh and English, especially among female speakers, and that dark /l/ may have social significance among the specific speakers whom he analyzed. Researchers have documented lateral variation that patterns in accordance with social factors in contact varieties of English in the United States as well. Van Hofwegen (2009) indicates that male speakers of Chicano English in South Texas e

57

Laterals in contact 57

tend to produce darker laterals than their female counterparts. Likewise, laterals show evidence of being a marker of generational identity among speakers of Chicano English. Van Hofwegen (2009) indicates that speakers belonging to earlier generations tend to produce darker /l/s, more similar to “Anglo English” dialects, while younger speakers show a tendency to produce lighter laterals more associated with a Latino/Spanish-speaking ethnic identity. Slomason and Newman (2004) found similar results in New York, demonstrating that lighter /l/ production was associated with those who identified themselves as more Latino. With respect to Spanish, traditionally only one light lateral has been documented by a number of researchers (e.g., Navarro Tomás, 1918; Quilis, 1999; Recasens & Espinosa, 2005). Quilis (1999) states that the prototypical Spanish lateral is produced when the tongue makes contact with some area of the upper central region of the mouth, be it the alveolar ridge, the teeth, or the region between the palate and the alveolar ridge, allowing air to leave the mouth along one or both edges of the tongue. More specifically, he states that the phonemic realization of /l/ in Spanish is a linguoalveolar realization that appears in any position relative to a word and/or syllable (i.e., word-initial, word-internal, syllable-initial, and syllable-final positions). He further notes that in certain contexts, it assimilates its apical point of articulation to that of the following segment. The literature on Spanish laterals in contact with other languages is relatively sparse. However, laterals have proven to be of interest in the study of CatalanSpanish bilingualism. Simonet (2008, 2010) studied the lateral productions of Spanish-Catalan bilinguals in Majorca and found that both Spanish-dominant and Catalan-dominant speakers produced second-language (L2) laterals that were different from those in their L1. However, despite, these differences, their L2 productions still had more in common with their L1 productions than they did with those of native productions of their L2. Thus, while Catalandominant speakers may have produced lighter /l/s in their L2 Spanish, those productions still more closely resembled their traditional dark Catalan /l/s. Likewise, those Spanish-dominant bilinguals in the study produced darker laterals in their Catalan than in their Spanish, but their Catalan /l/s were still lighter than those of the Catalan-dominant bilinguals. 2.2 Lateral variation in Cuban Spanish One of the defining characteristics of Cuban Spanish is the variation of laterals. Alonso (1953), for example, reports articulatory assimilation to following consonants, which leads to the complete deletion of /l/. In some cases, this leads to gemination or lengthening of the consonant (algo!’a.g;o], el número![e.’n:u.mer.o]). Alonso (1953) states that in cases of liquid neutralization in syllable-final position, the resulting sound is neither /r/ nor /l/, but rather an intermediate sound that at times sounds more rhotic in nature and at others sounds more like /l/. Lipski (2004, 2008) also reports liquid neutralization in Cuban Spanish and the Spanish of Cubans residing in the United States. He states that of the liquids, /l/ is more resistant to rhotacism than /r/ is to lambdacism, a tendency

58

Brandon M. A. Rogers and Scott M. Alvord

58

also reported by Alfaraz (2007). He also confirms the tendency of /l/ to assimilate to following consonants and for the following consonant to geminate in both Cuban and Cuban-American Spanish. Most researchers agree that /l/ tends to weaken in post-nuclear position in all Caribbean varieties of Spanish, including Cuban (e.g.,Lipski, 2004; 2008; Almendros, 1958; Dohotaru, 2002; Vaquero, 1996). In Miami-Cuban Spanish, Alfaraz (2007) found that liquid assimilation has spread from older speakers to the younger generations. She found that men tend to assimilate /r/ more and that there was no significant difference between men and women for /l/ assimilation. She argues that this is evidence that /l/ assimilation is not as stigmatized or salient as /r/ assimilation. The current study continues the examination of laterals in contact in both the English as well as the Spanish spoken by the Miami-Cuban community. It also follows up on Carter and Lynch’s (2015) and Carter’s (2013) assertions that the phonology of English is affected by contact with Spanish, particularly with liquids. The specific focus of the present study is on the production of the lateral segment /l/ in Miami-Cuban Spanish and English and uses acoustic measurements to determine whether there is evidence in either language of influence from the other. The following research questions guided the study. RQ1: Is there evidence of crosslinguistic influence in Miami-Cuban Spanish and English /l/ across immigrant groups? RQ2: Given that gender patterning has been shown in different situations of language change, how do Miami-Cuban Spanish and English /l/ productions vary according to gender? RQ3: How do Miami-Cuban Spanish and English /l/ productions correlate with the linguistic factors of word position and following segment?

3. Methodology 3.1 Participants Ten Miami-Cuban bilinguals, fie male and five female, were recorded performing a reading task in both languages. Additionally, they each completed a background and language use questionnaire. Speakers were divided into three immigrant groups in accordance with the criteria used by Silva-Corvalán (1994). Participants in the first immigrant group (G1) (n = 2) were born in Cuba and arrived on US soil after age 11. The second immigrant group (G2) (n = 5) comprised individuals who were either born in Cuba and migrated to the United States before age six or were born in the United States to at least one parent from the first group. The third immigrant group (G3) (n = 3) comprised individuals who had at least one parent belonging to the second group. Sociolinguistic interviews were also performed, but it was determined that not all interviews had enough tokens of /l/ to obtain a reliable sample size from the spontaneous speech of all three groups. Therefore, only the reading task data was used. While spontaneous speech is preferable when studying sociolinguistic variation, more formal tasks, such as reading tasks, are

59

Laterals in contact 59

also advantageous because researchers can more or less control the number of tokens of interest. Future studies should include spontaneous speech as well. 3.2 Instruments Participants read a short story titled “El Sobrín” in Spanish1 and “Who Did Patrick’s Homework,” by Carol Moore, in English. Participants were recorded with a Marantz PMD 671 digital recorder at a recording rate of 44.1 kHz. All readable tokens of English and Spanish /l/ in syllable-initial and syllable-final positions were marked by using Praat (Boersma & Weenink, 2012). Because of the variation that Cuban Spanish /l/ undergoes, deleted and assimilated instances of Spanish /l/ were not included. Likewise, tokens of /r/ that underwent lambdacism were also excluded. Once all eligible tokens were marked, a script was run that extracted all tokens and measured each token’s F1 and F2 values in Hz at their midpoint (Recasens & Espinosa, 2005; Simonet, 2010). All F1 and F2 values were then normalized by using the Bark scale to account for individual variation and physiological differences. In all, 1,409 Spanish tokens and 765 English tokens were extracted, for a total of 2,174 tokens. To quantify the relative lightness or darkness of each token, the normalized F1 values were subtracted from their normalized F2 counterparts (e.g., Recasens & Espinosa, 2005; Simonet, 2010). As shown by Sproat and Fujimura (1993), the tongue dorsum lowers more when articulating [ɫ] than [l], resulting in lower F2 values (see also Epsy-Wilson, 1992). Therefore, differences between F1 and F2 values for darker productions are expected to be smaller than those of lighter productions of /l/. The advantage of this method of measurement is that it considers the gradience of the overall lightness and darkness of laterals shown in previous research and thus is able to give a more detailed picture of the acoustic behaviors of the speakers’ /l/ productions. These differences can be seen in Figures 3.1 and 3.2. While these examples are taken from differing phonetic contexts, they are intended to illustrate typical spectrographic differences

Figure 3.1 Example of acoustic characteristics of light [l] in Spanish word “familia” as produced by Speaker 1013.

60

Brandon M. A. Rogers and Scott M. Alvord

60

Figure 3.2 Example of acoustic characteristics of [ł] in English word “still” as produced by Speaker 1015.

between [ɫ] and [l] in a clear fashion. Figure 3.1 shows the spectrogram of the Spanish word “familia” (family). The large space between the F1 and F2 in the [l] are indicative of a non-velarized, or light, production. Figure 3.2, a spectrogram of the English word “still,” illustrates the small difference between F1 and F2 in the syllable-final dark [ɫ]. Finally, separate mixed-model random-effects analyses, with the variable of “speaker” included as a random factor, were run for both the English and Spanish /l/ data by using the syntax window in the SPSS (statistics software). In both cases, the dependent variable was the F2-F1 measurement. The independent variables were both sociolinguistic and linguistic. In Spanish, tokens were marked for syllable position and for coarticulation with the following segment (Sproat & Fujimura, 1993). Following segments were categorized based on whether they were a non-back vowel (/i, e, a/), a back vowel (/u, o/), a pause, or a consonant whose articulation involves upward dorsal movement (e.g., /g/ /k/). All other following segments were marked as “other.” English /l/s were marked for the same linguistic variables and the additional variable of whether the lateral was syllabic. The extralinguistic variables investigated for both languages were gender and immigrant group.

4. Results 4.1 Spanish The only significant main effect according to the Spanish data was for following segment (F(4, 1396.166) = 7.292, p < 0.0001). There were no significant main effects according to immigrant group (F(2, 3.980) = 0.279, p = 0.771), gender (F(1,3.978) = 5.190, p = 0.085), or syllable position (F(2, 1396.731) = 0.718, p = 0.488). Table 3.1 shows the results of the overall analysis of Spanish /l/

61

Laterals in contact 61

Table 3.1 Overall results for a mixed-model analysis of Spanish /l/ Variable

F

Sig

Immigrant group Gender Following segment Position Generation * gender

0.279 5.190 7.292 0.718 0.746

N.S. N.S. 0.000 N.S N.S.

Note: F(1,179.028) = 358.277, p < 0.05

Table 3.2 F2-F1 means for Spanish /l/ by following segment Following segment

F2-F1 mean (Bark) of following lateral

Standard deviation

Non-dorsal consonant Pause Non-back vowel Back vowel Dorsal consonant

7.379 7.023 7.006 6.715 6.364

0.390 0.431 0.365 0.366 0.443

Table 3.3 Post hoc results for significant following segment differences Following segment pair

F2-F1 mean difference (Bark)

Standard deviation

Sig

Post hoc

Non-back and back V Dorsal and non-dorsal C

0.292 −1.015

0.097 0.242

0.027 0.000

Back < non-back Dorsal < non-dorsal

Note: F(4, 1396.166) = 7.292, p < 0.0001

(F(1,179.028) = 358.277, p < 0.05). Table 3.2 illustrates the overall F2-F1 Bark means of /l/ when followed by each of the following segment categories. The means in Table 3.2 show that /l/ was the lightest when followed by a non-dorsal consonant and the most velarized—that is, darkest—when a dorsal consonant followed it. Bonferroni post hoc pairwise tests for following segment indicated that there were significant differences for F2-F1 means between /l/ before non-back vowel (/i, e, a/) versus /l/ before a back vowel (/u, o/) (p = 0.027), and for when /l/ was followed by a dorsal consonant versus a non-dorsal consonant (p < 0.0001). Table 3.3 illustrates these results. As can be predicted, when a back vowel followed Spanish laterals, they were darker than when followed by a non-back vowel due to the coarticulatory effects of the raised dorsum. Likewise, when /l/ preceded a dorsal consonant, it was generally more velarized than when a non-dorsal consonant followed it. These results point to coarticulation being an important factor in the production of Spanish /l/. In other words, coarticulation with back vowels resulted in a greater amount of tongue retraction when speakers articulated /l/, resulting in lower F2 values and darker laterals, while coarticulation with dorsal consonants led to lower F1 and F2 values, resulting in darker productions of /l/ as well.

62

Brandon M. A. Rogers and Scott M. Alvord

62

4.2 English The English analysis indicated that there were several more significant main effects than observed in the Spanish data. There were significant main effects for the immigrant group (F(2, 3.990) = 16.829, p = 0.011), gender (F(1, 3.985) = 12.786, p = 0.023), following segment (4, 749.492) = 2.478, p = 0.043), and word position (F(1, 749.641) = 132.164, p < 0.0001). There was also a significant interaction for immigrant group and gender (F(2, 3.991) = 8.989, p = 0.033). There was no significant main effects for whether English /l/ was nuclear (F(1, 749.977) = 0.019, p = 0.889). Table 3.4 shows the results of the English data. For the immigrant group, F2-F1 means gradually decreased with each subsequent group, suggesting a gradual darkening of English /l/ as Spanish proficiency possibly decreased. However, post hoc tests (F(2,3.990) = 16.829, p < 0.05) revealed that the only significant differences between immigrant groups were between the first group and the latter two. There were no significant differences between groups 2 and 3. Table 3.5 shows the overall means of each immigrant group and the results of the pairwise immigrant group comparisons are illustrated in Table 3.6. Table 3.4 Overall English results Variable

F

Sig

Immigrant group Gender Following segment Position Nuclear Immigrant group * gender

16.829 12.786 2.478 132.164 0.019 8.989

0.011 0.023 0.043 0.000 N.S. 0.033

Note: F(1, 8.097) = 941.382, p < 0.0001

Table 3.5 F2-F1 means by immigrant group Immigrant group

F2-F1 mean (Bark)

Standard deviation

1 2 3

6.188 4.549 4.188

.290 .201 .255

Table 3.6 Pairwise comparisons of immigrant groups Immigrant group pair

F2-F1 mean difference (Bark)

Standard deviation

Sig

Post hoc

1 and 2 1 and 3 2 and 3

1.638 2.000 0.362

0.330 0.366 0.299

0.023 0.017 0.880

2 0.05, Z2p = 0.029

F(2,16) = 0.816, p > 0.05, Z2p = 0.093

NH

F(2,28) = 2.559, p > 0.05, Z2p = 0.155

F(1,14) = 3.561, p > 0.05, Z2p = 0.203

F(2,28) = 1.922, p > 0.05, Z2p = 0.121

NL

F(2,22) = 1.878, p > 0.05, Z2p = 0.146

F(1,11) = 0.082, p > 0.05, Z2p = 0.007

F(2,22) = 2.938, p > 0.05, Z2p = 0.211

We created the materials controlling for variables reported in the variationist literature to affect production. We manipulated three language-internal variables: subject form (null/overt pronominal and overt lexical), grammatical person (1sg and 3sg, cf. Bentivoglio, 1987[1990]; Davidson, 1996; Enríquez, 1984; Morales, 1997; Otheguy et al., 2007; Silva-Corvalán, 1982, 1994; Travis, 2007), and speech connectivity (connect) (same referent, same TAM; same referent, different TAM; and different referent, different TAM: cf. Otheguy et al., 2007). Table 7.2 lists the variables manipulated in the AJT and their associated constraints, and it provides an example for each one. The combination of these different constraints resulted in six conditions: Condition Condition Condition Condition Condition Condition

1: 2: 3: 4: 5: 6:

1sg, 1sg, 1sg, 3sg, 3sg, 3sg,

same referent, same TAM same referent, different TAM different referent same referent, same TAM same referent, different TAM different referent, different TAM

We controlled for the following variables: clause type (all clauses were main clauses: cf. Lozano, 2008; Margaza & Bel, 2006; Morales, 1997; Otheguy et al., 2007; Silva-Corvalán, 1994), verb form ambiguity (all verbs were non-ambiguous: cf. Bayley & Pease-Alvarez, 1996, 1997; Silva-Corvalán, 1994; Travis, 2005), animacy (all referents were animate: cf. Lozano, 2008), and verb type (all verbs were external activities:4 cf. Bentivoglio, 1987[1990]; Enríquez, 1984; Otheguy & Zentella, 2012; Silva-Corvalán, 1994; Travis, 2007). Example (3) is a token from condition 4 (3sg, same referent, same TAM). We expected (a) to be the highest scored option since the context and the following sentence refer to the same person: María.

137

The effect of person on subject expression

137

(3) Yo soy enfermera. María también lo es y tiene un horario complicado. 1 2 3

Trabaja de noche los fines de semana. 1 Ella trabaja de noche los fines de semana. 1 María trabaja de noche los fines de semana. 1

2 2 2

3 3 3

4 4 4

4. Results The data were submitted to two statistical analyses, one for each grammatical person, using SPSS version 24, with the sentence rating as the dependent variable and two linguistic independent variables (subject form and connect). Two separate analyses were necessary given that the dependent variable only had two levels (null vs. overt pronominal subjects) in 1sg, while it had three levels (null vs. overt pronominal vs. lexical subjects) in 3sg. An initial repeatedmeasures ANOVA revealed a main effect of variety in 1sg (F(1, 43) = 4.290, p = 0.044, Z2p = 0.091), a marginal main effect of variety in 3sg (F(1, 64) = 3.578, p = 0.063, Z2p = 0.053) and interactions with the extralinguistic independent variable proficiency. In particular, there was a subject form by connect by proficiency interaction, F(2, 86) = 4.932, p = 0.009, Z2p = 0.103, in 1sg, and in 3sg, F(4, 256) = 8.482, p = 0.000, Z2p = 0.117. To better understand the grammars of each of the groups, separate repeated-measures ANOVAs were performed on each group: Caribbean HSs of higher proficiency (CH) (N = 11), Caribbean HSs of lower proficiency (CL) (N = 9), non-Caribbean HSs of higher proficiency (NH) (N = 15), and non-Caribbean HSs of lower proficiency (NL) (N = 12). 4.1 First-person singular subjects As seen in the Figures 7.1–7.4, the ratings for both null and overt pronominal subjects fell within the acceptance range (above 2.50 since we used a 1–4 scale) and were similar in all of the conditions for all of the speaker groups. The ratings for null subjects across speech connectivity (connect) contexts seemed to be rather similar, except for contexts with a different referent where the CH group’s rating was slightly lower. There did not seem to be differences across the speaker groups in the ratings of specific conditions or the patterns across conditions. The results from the 2 (subject form: null vs. pronominal) by 3 (connect: same referent and same TAM, same referent and different TAM, different referent) repeated-measures ANOVA further confirmed these trends. There was no main effect for subject form (CH: F(1,10) = 4.683, p > 0.05, Z2p = 0.319; CL: F(1,8) = 0.237, p > 0.05, Z2p = 0.029; NH: F(1,14) = 3.561, p > 0.05, Z2p = 0.203; NL: F(1,11) = 0.082, p > 0.05, Z2p = 0.007), connect (CH: F(2,20) = 2.200, p > 0.05, Z2p = 0.180; CL: F(2,16) = 0.962, p > 0.05, Z2p = 0.107; NH: F(2,28) = 2.559, p > 0.05, Z2p = 0.155; NL: F(2,22) = 1.878, p > 0.05, Z2p = 0.146), or a subject form by connect interaction (CH: F(2,20) = 2.848, p > 0.05, Z2p = 0.222; CL: F

138

Ana de Prada Pérez and Inmaculada Gómez Soler

138

Caribbean-Higher 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.31

3.45

3.07

Pronominal

3.48

3.42

3.42

Figure 7.1 Mean ratings for Caribbean-Higher group for first person

Caribbean-Lower 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.21

3.08

3.21

Pronominal

3.35

3.22

3.28

Figure 7.2 Mean ratings for Caribbean-Lower group for first person

(2,16) = 0.816, p > 0.05, Z2p = 0.093; NH: F(2,28) = 1.922, p > 0.05, Z2p = 0.121; NL: F(2,22) = 2.938, p > 0.05, Z2p = 0.211) for any of the groups. Thus, participants rated null and overt pronominal subjects similarly, rated sentences in different contexts (with different speech connectivity) similarly, and, most importantly, the lack of an interaction indicates that participants did not rate certain subject forms higher in one context than in another.

139

The effect of person on subject expression

139

Non Caribbean-Higher 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.43

3.59

3.38

Pronominal

3.56

3.66

3.60

Figure 7.3 Mean ratings for Non Caribbean-Higher group for first person

Non Caribbean-Lower 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.55

3.56

3.45

Pronominal

3.43

3.58

3.45

Figure 7.4 Mean ratings for Non Caribbean-Lower group for first person

Overall, these results indicate that these groups of HSs rate null and overt subjects similarly across different conditions, thus evidencing the weakening of the pragmatic conditioning, as instantiated in the variables switch reference and TAM continuity of the null vs. overt pronominal subjects. Nonetheless, there is a possible task effect, a point that we return to in the discussion.

140

Ana de Prada Pérez and Inmaculada Gómez Soler

140

Caribbean-Higher 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.06

3.11

2.75

Pronominal

3.50

3.50

3.43

Lexical

3.52

3.64

3.63

Figure 7.5 Mean ratings for Caribbean-Higher group for third person

4.2 Third-person singular subjects The data for 3sg subjects also exhibited mean ratings above the acceptance range (see Figures 7.5–7.8). All groups, except the NL speaker group, rated overt subjects (both pronominal and lexical) higher than null subjects. Null subjects were rated lower in clauses with a different referent for speakers in the two higher proficiency groups. At higher proficiency levels, there did not seem to be differences between speakers of Caribbean and non-Caribbean varieties, while at lower proficiencies, more differences were attested. In particular, NL did not exhibit preferences for any subject form across any condition. The statistical analyses revealed that for both higher proficiency groups there was a main effect for subject form (F(2,20) = 14.057, p = 0.00, Z2p = 0.584, for CH, and F(2,28) = 23.379, p = 0.00, Z2p = 0.625, for NH), no main effect for connect (F(2,20) = 2.752, p > 0.05, Z2p = 0.216, for CH, and F(2,28) = 23.075, p = 0.00, Z2p = 0.622, for NH), and subject form by connect interaction (F(4,40) = 2.935, p = 0.03, Z2p = 0.227, for CH, and F(4,92) = 0.866, p > 0.05, Z2p = 0.036 for NH). For the CL group, there was a main effect for subject form F(2,34) = 20.604, p = 0.00, Z2p = 0.548, connect, F(2,34) = 0.816, p > 0.05, Z2p = 0.046, and a subject form by connect interaction, F(4,68) = 3.669, p = 0.01, Z2p = 0.178. The remaining three groups rated overt subjects significantly higher than null subjects, particularly in contexts with a different referent. Lastly, for NL, there was a main effect for subject form, F(2,46) = 7.773,

141

The effect of person on subject expression

141

Caribbean-Lower 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

2.50

2.68

2.57

Pronominal

3.33

3.26

3.29

Lexical

3.60

3.57

3.46

Figure 7.6 Mean ratings for Caribbean-Lower group for third person

Non Caribbean-Higher 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.03

3.23

2.68

Pronominal

3.62

3.67

3.53

Lexical

3.62

3.68

3.62

Figure 7.7 Mean ratings for Non Caribbean-Higher group for third person

p = 0.001, Z2p = 0.253, where null and overt pronominal subjects were rated significantly lower than overt lexical subjects, no main effect for connect, F(2,46) = 0.949, p > 0.05, Z2p = 0.040, and no subject form by connect interaction, F(4,92) = 0.866, p > 0.05, Z2p = 0.036. Thus, regarding subject form, all groups except for

142

Ana de Prada Pérez and Inmaculada Gómez Soler

142

Non Caribbean-Lower 4.00

AJT rating

3.50 3.00 2.50 2.00 1.50 1.00

Same reference & same TAM

Same reference & different TAM

Different referent

Null

3.28

3.28

3.32

Pronominal

3.42

3.38

3.43

Lexical

3.61

3.52

3.65

Figure 7.8 Mean ratings for Non Caribbean-Lower group for third person

NL rated overt pronominal subjects significantly higher than they rated null subjects. With respect to connect, only NH rated sentences with a different referent lower, which was probably due to the subject form by connect interaction. Lastly, all groups except for NL had a subject form by connect interaction, which indicated that the participants rated null subjects significantly lower in contexts of different reference than in context of same reference. Overall, the data for 3sg subjects seemed to indicate a general preference for overt pronominal subjects, as they were rated significantly higher than null and overt lexical subjects. This preference, however, did not constitute a loss of differentiation for contexts: there remained a distinction across contexts with different speech connectivity. Null subjects were rated significantly lower in contexts with a different referent. The NL group, however, seemed to have a weakened sensitivity to this pragmatic factor.

5. Discussion In this section we return to our RQs in light of the results. Our aim was to explore speaker preferences with respect to subject form across different discourse connectivity contexts in two different grammatical persons. Previous literature on SPE in Spanish HSs in the United States had returned conflicting results regarding the increased use of overt pronominal subjects and the loss of the pragmatic contrast between null and overt pronominal subjects. With the observation that previous literature had not been consistent in the grammatical person represented in

143

The effect of person on subject expression

143

their tasks/corpora, in our first RQ, we explored the role of person in this study by comparing the results of items with 1sg referents and those with 3sg referents. We anticipated a higher rating for overt pronominal subjects compared to null subjects, in the third person, since the previous literature that reported a contact effect included 3sg subjects. In contrast, we anticipated no contact effects in the rating of overt vs. null pronominal subjects in the 1sg items. The results partially confirmed these predictions. With 3sg subjects, most groups rated overt pronominal subjects higher than they rated null subjects across conditions. With 1sg subjects, the effect of language contact seems subtler given that there is no preference for either subject form. Given the task type and the lack of a monolingual group, though, contact effects cannot be fully discussed. Our assumption, based on previous literature, is that monolingual speakers would prefer null subjects. The task also had an effect, as expected, since HS have been shown to struggle with AJTs, written tasks, and metalinguistic tasks (Montrul, 2005; Montrul & Perpiñán, 2011). Some of these speakers were also recorded participating in a sociolinguistic interview (Prada Pérez, forthcoming), where there is evidence of more contact effects in 3sg than in 1sg. The data from the interview, however, did not show a preference for overt subjects, as they still used null subjects more frequently than they used overt subjects. Even in the absence of monolingual data, the results are suggestive of a stronger contact effect in 3sg vs. 1sg forms. Importantly, Spanish lacks a person morpheme in 3sg. Additionally, the difference between 1sg and 3sg in null-subject languages is not as surprising given languages like Hebrew, where null subjects are only permitted in 1sg and 2sg subjects. Artstein (1999) puts forth an optimality theory analysis of these facts, based on Aissen’s (1998) hierarchies for person and animacy (first/ second person > proper noun 3sg > human 3sg > animate 3sg > inanimate 3sg) and adding a hierarchy for subjects (null > overt). Defining markedness as depending on the misalignment between two hierarchies (i.e., an element ranked high in one hierarchy that is ranked low in another one), he establishes that there may be languages that only allow null subjects in first and second person, while the opposite—that is, languages that allow null subjects in third person but not in first or second—does not occur. According to this account, it is possible that Spanish in contact with English is not moving toward an English grammar where all subjects have to be overt but where the influence starts occurring where the null subjects are more marked. Since this paper included only animate 1sg and 3sg, it is expected that crosslinguistic influence, or simplification, first appears in 3sg. Thus, while the data are not compatible with a system like the one in Hebrew, given that speakers still accept null subjects in 3sg, the differential crosslinguistic effects according to person are not unexpected. Regarding the second RQ, on the pragmatic contrast associated with null and overt pronominal subjects, we anticipated a weakening of pragmatic contrast, as previous research evidences either a use of the overt pronominal subject in contexts of topic continuation or a weakening of the effect size of the variable (or a lower range of the factor group). Our results were largely consistent with these

144

Ana de Prada Pérez and Inmaculada Gómez Soler

144

predictions. In the first person, there were no subject form * connect interactions for any of the groups. In the third person, the pattern was the same across all three contexts: overt forms were rated higher than null subjects. Only the more-advanced groups (CH, NH) showed an interaction: null subjects were rated lower in different referent contexts than in the other two contexts with the same referent. Although previous literature indicated that 1sg subjects favored overt subjects, while 3sg subjects did not (Prada Pérez, 2009; SilvaCorvalán, 1994), in Otheguy and Zentella’s (2012) data, 3sg also favored overt subjects, even more than 1sg in mainland speakers. This was true even in their group with less contact with English (newcomers). Thus, it seems that more contact effects are present in 3sg. There could be several explanations for the difference in vulnerability across people. The deictic function of the 1sg vs. the referential function of the 3sg may have an effect on bilingual Spanish. While a higher use of the 1sg can be explained as due to the egocentric nature of speakers (Silva-Corvalán, 1994), in 3sg, it may serve a disambiguating function, as more than one referent may be available in the context. Additionally, while 1sg verb forms have a person and number morpheme in Spanish, 3sg verb forms do not. Therefore, there are significant differences between these people that may explain the differences in contact effects. Future research may be able to identify which of these differences better accounts for this outcome. RQs 3 and 4 inquired about the effect of two extralinguistic factors: proficiency and Spanish variety. In the first person, there did not seem to be any effects across proficiency, although evidence consistent with a contact effect exists in the similar rating of null and overt pronominal subjects across conditions, an effect that cannot be confirmed without data from speakers without contact with English. To the best of our knowledge, the previous literature focusing on first-person SPE did not contrast groups of different proficiencies and varieties. Therefore, we cannot contrast our data with theirs in this grammatical person. For those focusing on third-person only (only in the generative tradition) or on several grammatical persons, there seems to be a contact effect (Montrul, 2004; Otheguy & Zentella, 2012). In this paper, instead of comparing groups with more and less contact with English based on their experiences, we used an independent measure of Spanish proficiency. The degree of contact with English is determined in previous research based on the language background reported by participants, such as their generation. In our context, some thirdgeneration HSs were more proficient in Spanish than some second-generation HSs, as other factors seemed to affect their proficiency (e.g., time spent with Spanish-speaking grandmothers, neighborhood in Miami where they grew up, etc.). Thus, proficiency was a more explanatory factor in our analysis. Even though proficiency and contact with English are not necessarily correlated, we can compare previous literature, where contact with English was included as a variable, and our study, where we included proficiency instead. These studies conclude that those speakers with more contact with English exhibited an increase in overt pronominal expression and also some changes in the factors affecting the distribution—switch reference, in particular. Our results indicated

145

The effect of person on subject expression

145

that lower proficiency HSs were not sensitive to the interaction between subject form and discourse connectivity, while more-advanced HSs were sensitive to it. The results were also consistent with an increased acceptance of overt pronominal subjects overall, since they were rated significantly higher than null subjects in all four speaker groups. In sum, in our data, the variable proficiency revealed some differences between the two proficiency levels. Proficiency has not been widely examined in previous studies. It is possible that these previous studies have not found the variable proficiency as determining due to the predetermined groupings based on the results of the DELE-based proficiency measure, whereas we used a median split. Although using the median split analysis makes comparisons across studies more difficult, it makes comparisons across groups in contexts like ours possible. Additionally, this practice makes the test more viable. As previously pointed out, the test is in written format, based on a specific variety of Spanish (Peninsular) that is largely absent from HSs input and based on prescriptive rules. Not assuming preconceived levels and including participants with literacy in Spanish may have resulted in more realistic participant groupings. We acknowledge, nonetheless, that better measures of proficiency for Spanish HSs in the United States are still needed. Additionally, our group of participants is not representative of all types of HSs as these are college-age, second- or third-generation bilinguals who have taken Spanish courses at a university. This may explain the scarcity of low proficiency speakers, who may have been placed in the secondlanguage learner track instead, and the scarcity of advanced speakers, who may be first generation and excluded from the analysis because of the differences in the background (e.g., rarely college students and generally older in age). With respect to speaker region, differences across varieties are well attested in the literature for studies focusing on second-person singular (2sg) (Cameron, 1995; Lipski, 1994) or those including a variety of people (Otheguy & Zentella, 2012; Shin & Otheguy, 2013). It is not clear, though, whether the higher number of overt pronominal subjects in Caribbean Spanish applies across people. Otheguy and Zentella (2012) showed that more differences existed between Caribbean and mainland speakers among recent arrivals (newcomers) than among those raised in NYC (NYR). With respect to 1sg and 3sg, in Caribbean Spanish, 1sg favored overt pronominal subjects, while 3sg neither favored nor disfavored them. In Mainland Spanish, in contrast, both people favored overt pronominal subjects, with 3sg favoring them more than 1sg. Within NYR, Otheguy and Zentella (2012) showed evidence of convergence, where both regions favored overt pronominal subjects with 1sg and 3sg. Similarly, our participants showed evidence of convergence: both Caribbean and non-Caribbean speakers rated 3sg pronominal subjects higher than nulls and 1sg pronominal subjects similarly to nulls. Otheguy and Zentella (2012) reported a difference: for Caribbean speakers, 1sg favored the overt form more than 3sg, and the opposite trend was attested for Mainland Spanish speakers. In contrast, our Caribbean and non-Caribbean participants did not differ in which of the two people rated pronouns higher. It is possible, however, that subtle differences existed in production that were not identifiable in this task.

146

Ana de Prada Pérez and Inmaculada Gómez Soler

146

The motivation for the study was to shed some light to the conflicting results in the antecedent literature—in particular, on the different approaches to the study of SPE in bilingual Spanish. As stated earlier, there are methodological and conceptual differences between generative and variationist approaches, particularly the nature of SPE (variable vs. regulated by switch reference only), the grammatical persons included in the analysis (only 3sg, only 1sg, or several persons) and the tasks used (AJTs in generative studies on second-language acquisition vs. oral narratives vs. sociolinguistic interviews). The results clarified some of the conflicting results encountered in the previous literature. Our results indicated that these conflicting results were likely due to the grammatical person included in the analysis, as we reported stronger contact effects in 3sg than in 1sg. This study also showed that an AJT designed to take into account the results of previous variationist work was a viable task to examine SPE. It seemed, however, that some of the trends were exaggerated. In Prada Pérez (forthcoming), a subset of these participants participated in a sociolinguistic interview and did not seem to have lost sensitivity to the discourse factors regulating the distribution. Additionally, they did not use overt pronominal subjects at a higher rate than null subjects, a result that could be expected given the results here.

6. Conclusion This paper contributes to current discussions in the fields of language contact and bilingualism by presenting new SPE data to clarify some contradictory results in the previous literature. In particular, our findings revealed that the locus of crosslinguistic influence in SPE lies mainly within 3sg (as opposed to 1sg). These results are compatible with both variationist and generative approaches to bilingual SPE. Additionally, this study benefited methodologically from the advancements made in both approaches. Our study is not within limitations; specifically, the lack of assurance that participants are paying attention to the context makes examining syntactic reflexes of discourse features unreliable. Ideally, this study needs to be complemented by oral data in which further variables can be analyzed thoroughly (e.g., semantic verb type and verb form ambiguity). This type of data can further corroborate our findings about the effect of grammatical person in HS subject expression.

Notes 1 In the theoretical literature, some authors have argued that the unmarked form is the null subject (under the avoid the pronoun principle: Chomsky, 1981 and, more recently, Sheehan, 2006), and as pointed out by an anonymous reviewer, it could be argued that the null requires the speaker to produce less, thus being less marked. 2 In recent years, however, generative theoretical work has incorporated variation into their analysis, especially as it pertains to syntactic distributions depending on pragmatic aspects (e.g., Richards, 2008). Generative SLA research on this topic, in contrast,

147

The effect of person on subject expression

147

has reported the results as if no variation is present with respect to the distribution of null and overt subjects in Spanish. 3 Not all participants reported the city of origin. Thus, in the case of Colombian and Mexican origin, we are grouping the speakers with non-Caribbean speakers, which may be problematic. 4 We follow the verb classification in Otheguy and Zentella (2012), who adapted it from Enríquez (1984). Their guidelines for identifying external activity verbs state, “Suponen alguna actividad, sea física, social o comportamiento, ya sea en movimiento, reposo o pleno desarrollo” (Otheguy & Zentella, 2012, p. 266).

References Aissen, J. (1998). Markedness and subject choice in optimality theory. Ms. UC Santa Cruz. Alonso-Ovalle, L., Clifton, C., Frazier, L., & Férnandez-Solera, S. (2005). Null vs. overt pronouns and the topic—focus articulation in Spanish. Journal of Italian Linguistics, 14, 151–169. Amaral, P.M., & Schwenter, S.A. (2005). Contrast and the (non-) occurrence of subject pronouns. In D. Eddington (Ed.), Selected proceedings of the 7th Hispanic linguistics symposium (pp. 116–127). Somerville, MA: Cascadilla Proceedings Project. Artstein, R. (1999). Person, animacy and null subjects. In T. Cambier-Langeveld, A. Lipták, M. Redford, & E.J. van der Torre (Eds.), Proceedings of console VII, SOLE, Leiden (pp. 1–15). Leiden: Universiteit Leiden. Retreived from http://artstein.org/pub lications/person.pdf Bayley, R., & Pease-Alvarez, L. (1996). Null and expressed subject pronoun variation in Mexican-descent children’s Spanish. In J. Arnold, R. Blake, & B. Davidson (Eds.), Sociolinguistic variation: Data, theory, and analysis (pp. 85–99). Stanford, CA: Center for the Study of Language and Information. Bayley, R., & Pease-Alvarez, L. (1997). Null pronoun variation in Mexican-descent children’s narrative discourse. Language Variation and Change, 9(3), 349–371. Bentivoglio, P. (1987[1980]). Los sujetos pronominales de primera persona en el habla de Caracas. Caracas: Universidad Central de Venezuela. Cameron, R. (1995). The scope and limits of switch reference as a constraint on pronominal subject expression. Hispanic Linguistics, 6/7, 1–27. Carvalho, A.M., & Besset, R.M. (2015). Subject pronoun expression in Spanish in contact with Portuguese. In A. Carvalho, R. Orozco, & N.L. Shin (Eds.), Subject pronoun expression in Spanish: A cross-dialectal perspective (pp. 143–168). Washington, DC: Georgetown University Press. Carvalho, A.M., Orozco, R., & Shin, N.L. (Eds.). (2015). Subject pronoun expression in Spanish: A cross-dialectal perspective. Washington, DC: Georgetown University Press. Casanova Seuma, L. (1999). El sujeto en catalán coloquial. Revista Española de Lingüística, 29(1), 105–131. Chomsky, N. (1981). Lectures on government and binding. Dordrecht, Netherlands: Foris Publications Holland. Clements, C., & Domínguez, L. (2017). Reexamining the acquisition of null subject pronouns in a second language: Focus on referential and pragmatic constraints. Linguistic Approaches to Bilingualism, 7(1), 33–62. Cuza, A. (2013). Cross-linguistic influence at the syntax proper: Interrogative subject-verb inversion in heritage Spanish. International Journal of Bilingualism, 17(1), 71–96.

148

Ana de Prada Pérez and Inmaculada Gómez Soler

148

Davidson, B. (1996). ‘Pragmatic weight’ and Spanish subject pronouns: The pragmatic and discourse uses of ‘tú’and ‘yo’in spoken Madrid Spanish. Journal of Pragmatics, 26(4), 543–565. Dumont, J. (2006). Full NPs as subjects. In N. Sagarra & A.J. Toribio (Eds.), Selected proceedings of the 9th Hispanic linguistics symposium (pp. 286–296). Somerville, MA: Cascadilla. Enríquez, E. (1984). El pronombre personal sujeto en la lengua española hablada en Madrid. Madrid: Consejo Superior de Investigaciones Científicas. Erker, D., & Guy, G. (2013). The role of lexical frequency in syntactic variability: Variable subject personal pronoun expression in Spanish. Language, 88(3), 526–557. Erker, D., & Otheguy, R. (2016). Contact and coherence: Dialectal leveling and structural convergence in NYC Spanish. Lingua, 172/173, 131–146. Flores-Ferrán, N. (2004). Spanish subject personal pronoun use in New York City Puerto Ricans: Can we rest the case for English contact? Language Variation and Change, 16, 49–73. Flores-Ferrán, N., & Toro, J. (2000). The persistence of dialect features under conditions of contact and leveling. Southwest Journal of Linguistics, 19(2), 31–41. Haegueman, L., & Guéron, J. (1999). English grammar: A generative perspective. Malden, MA: Blackwell Publishers Inc. Liceras, J. (1988). Syntax and stylistics: More on the pro-drop parameter. In J. Pankhurst, M. Sharwood Smith, & P. Van Buren (Eds.), Learnability and second languages: A book of readings (pp. 71–93). Dordrecht: Foris. Liceras, J. (1989). On some properties of the “pro-drop” parameter: Looking for missing subjects in non-native Spanish. In Linguistic perspectives on second language acquisition (pp. 109–133). Cambridge: Cambridge University Press. Lipski, J. (1994). Latin American Spanish. London: Longman. Lipski, J. (1996). Patterns of pronominal evolution in Cuban-American bilinguals. In A. Roca & J.B. Jensen (Eds.), Spanish in contact: Issues in bilingualism (pp. 159–186). Somerville, MA: Cascadilla Press. Lozano, C. (2008). Selective deficits at the syntax-discourse interface: Evidence from the CEDEL2 corpus. In N. Snape, Y.I. Leung, & M. Sharwood Smith (Eds.), Representational deficits in SLA: Studies in honor of Roger Hawkins (language acquisition and language disorders). Amsterdam: John Benjamins. Margaza, P., & Bel, A. (2006). Null subjects at the syntax-pragmatics interface: Evidence from Spanish interlanguage of Greek speakers. In M. Grantham O’Brien, Ch. Shea, & J. Archibald (Eds.) Proceedings of the 8th generative approaches to second language acquisition conference (pp. 88–97). Somerville, MA: Cascadilla Proceedings Project. Martínez Sanz, C. (2011). Null and overt subjects in a variable system: The case of Dominican Spanish. PhD dissertation, University of Ottawa, Canada. Montrul, S. (2004). Subject and object expression in Spanish heritage speakers: A case of morpho-syntactic convergence. Bilingualism: Language and Cognition, 7(2), 1–18. Montrul, S. (2005). Second language acquisition and first language loss in adult early bilinguals: Exploring some differences and similarities. Second Language Research, 21(3), 199–249. Montrul, S., & Perpiñán, S. (2011). Assessing differences and similarities between instructed heritage language learners and L2 learners in their knowledge of Spanish tense-aspect and mood (TAM) morphology. Heritage Language Journal, 8(1), 90–133.

149

The effect of person on subject expression

149

Montrul, S., & Slabakova, R. (2003). Competence similarities between native and nearnative speakers: An investigation of the preterite/imperfect contrast in Spanish. Studies in Second Language Acquisition, 25(3), 351–398. Morales, A. (1997). La hipótesis funcional y la aparición de sujeto no nominal: El español de Puerto Rico. Hispania, 80(1), 153–165. Orozco, R. (2015). Pronominal variation in Costeño Spanish. In A. Carvalho, R. Orozco, & N.L. Shin (Eds.), Subject pronoun expression in Spanish: A cross-dialectal perspective (pp. 17–37). Washington, DC: Georgetown University Press. Otheguy, R., & Zentella, A.C. (2012). Spanish in New York. Language contact, dialectal leveling, and structural continuity. New York: Oxford University Press. Otheguy, R., Zentella, A.C., & Livert, D. (2007). Language and dialect contact in Spanish in New York: Toward the formation of a speech community. Language, 83(4), 770–802. Posio, P. (2015). Subject pronoun usage in formulaic sequences. In A. Carvalho, R. Orozco, & N.L. Shin (Eds.), Subject pronoun expression in Spanish: A cross-dialectal perspective (pp. 59–78). Washington, DC: Georgetown University Press. Potowski, K. (2008). “I was raised talking like my mom”: The influence of mothers in the development of MexiRicans’ phonological and lexical features. In J. Rothman & M. Niño-Murcia (Eds.), Linguistic identity and bilingualism in different Hispanic contexts (pp. 201–220). New York: John Benjamins. Prada Pérez, A. de. (2009). Subject expression in Minorcan Spanish: consequences of contact with Catalan. Ph.D. dissertation, The Pennsylvania State University, University Park, PA. Prada Pérez, A. de. (2015). First person singular subject pronoun expression in Spanish in contact with Catalan. In A.M. Carvalho, R. Orozco, & N.L. Shin (Eds.), Subject pronoun expression in Spanish: A crossdialectal perspective (pp. 121–142). Washington, DC: Georgetown University Press. Prada Pérez, A. de. (Forthcoming). First person singular subject expression in Caribbean heritage speaker Spanish oral production. In L. Ortiz López, R. Guzzardo Tarmargo, & M. González-Rivera (Eds.), Hispanic contact linguistics: Theoretical, methodological and empirical perspectives. Amsterdam, The Netherlands: John Benjamins. Ranson, D.L. (1991). Person marking in the wake of /s/ deletion in Andalusian Spanish. Language Variation and Change, 3(2), 133–152. Richards, M. (2008). Two kinds of variation in a minimalist system. In F. Heck, G. Müller & J. Trommer (Eds.), Varieties of competition (pp. 133–162). Leipzig, Germany: Linguistische Arbeits Berichte 87. Rizzi, L. (1994). Early null subjects and root null subjects. In T. Hoekstra & B.D. Schwartz (Eds.), Language acquisition studies in generative grammar (pp. 151–176). Amsterdam: Benjamins. Rothman, J. (2009). Pragmatic deficits with syntactic consequences?: L2 pronominal subjects and the syntax-pragmatics interface. Journal of Pragmatics, 41, 951–973. Sheehan, M. (2006). The EPP and null subjects in Romance. Ph.D. dissertation, Newcastle University. Shin, N.L. (2012). Variable use of Spanish subject pronouns by monolingual children in Mexico. In K. Geeslin and M. Díaz-Campos (Eds.), Selected proceedings of the 2010 Hispanic linguistics symposium (pp. 130–141). Somerville, MA: Cascadilla Proceedings Project. Shin, N.L., & Otheguy, R. (2013). Social class and gender impacting change in bilingual settings: Spanish subject pronoun use in New York. Language in Society, 42(4), 429–452.

150

Ana de Prada Pérez and Inmaculada Gómez Soler

150

Silva-Corvalán, C. (1982). Subject expression and placement in Mexican-American Spanish. In J. Amastae & L. Elías-Olivares (Eds.), Spanish in the United States: Sociolinguistic aspects (pp. 93–120). Cambridge: Cambridge University Press. Silva-Corvalán, C. (1994). Language contact and change: Spanish in Los Angeles. New York: Oxford University Press. Silva-Corvalán, C. (2014). Bilingual language acquisition: Spanish and English in the first six years. New York: Cambridge University Press. Sorace, A. (2011). Pinning down the concept of “interface” in bilingualism. Linguistic Approaches to Bilingualism, 1(1), 1–33. Toribio, A.J. (2004). Convergence as an optimization strategy in bilingual speech: Evidence from code-switching. Bilingualism: Language and Cognition, 7(2), 165–173. Torres Cacoullos, R., & Travis, C.E. (2010). Variable yo expression in New Mexico: English influence? In S. Rivera-Mills & D. Villa (Eds.), Spanish of the Southwest: A language in transition (pp. 185–206). Frankfurt: Iberoamericana/Vervuert. Torres Cacoullos, R., & Travis, C.E. (2015). Foundations for the study of subject pronoun expression in Spanish in contact with English: Assessing intralinguistic (dis)similarity via intralinguistic variability. In A.M. Carvalho, R. Orozco, & N.L. Shin (Eds.), Subject pronoun expression in Spanish: A cross-dialectal perspective (pp. 81–100). Washington, DC: Georgetown University Press. Travis, C.E. (2005). The yo-yo effect: Priming in subject expression in Colombian Spanish. In R. Gess & E.J. Rubin (Eds.), Theoretical and experimental approaches to romance linguistics: Selected papers from the 34th linguistic symposium on romance languages (pp. 329–349). Amsterdam and Philadelphia, PA: John Benjamins. Travis, C.E. (2007). Genre effects on subject expression in Spanish: Priming in narrative and conversation. Language Variation and Change, 19(2), 101–135.

151

Section 3

Sociolinguistic perspectives

153

8

Evidence of creolized English grammar in the Spanish of Dominicanos on St. Thomas, US Virgin Islands Daniel S. D’Arpa

1. Introduction On the island of St. Thomas, in the US Virgin Islands, where the mainstream language is a creolized Caribbean English (Roy, 1975a; Sprauve, 1974), there is a significant and growing speech community of Spanish-speaking immigrants from Dominican Republic. The variety of Spanish spoken by members of this speech community exhibits linguistic features that are common structures of St. Thomas English Creole (STTEC), including aspects of phonetics, phonology, and syntax. The goal of this article is to document features of this variety of Dominicano Spanish on St. Thomas and to propose further sociolinguistic studies to explore whether these language choices are influenced by social pressures perceived by Dominicanos to negotiate their identity in a diverse but divided society.

2. Geopolitical and sociolinguistic background of St. Thomas St. Thomas is located in the Caribbean Sea, off the east shore of Puerto Rico. St. Thomas, St. John, and St. Croix, each unique in form and history, comprise the three largest inhabited islands in the US territory of the archipelago of the Virgin Islands. Historical accounts of the Virgin Islands (Bastian, 2001; Boyer, 2010; Dookhan & Sheridan, 1994) record that as late as the mid 17th century, after Spain, the Netherlands, England, and France had annihilated the local Indigenous Peoples and challenged each other for control of the region, Denmark claimed St. Thomas for its exploitation, through a nationally sponsored commercial investment program called the Danish West India Company. The company enjoyed Denmark’s military protection and the rights and privileges of a monopoly. Though its European colonizers constructed African slavery in St. Thomas, the Danish West India Company never made a profit. For nearly two centuries, the islands were a financial and social burden to distant Denmark, until in 1917, interested in a strategic military base location, the United States bought the Danish West Indies and renamed the new territory the United States Virgin Islands. Since its earliest European settlement, the population of St. Thomas has always been heterogeneous, with a diversity of nationalities and language varieties (Bastian, 2001; Sabino, 1994; Sprauve, 1975). Danish was never the primary

154

Daniel S. D’Arpa

154

language on St. Thomas; Dutch was the first lingua franca (Bastian, 2001, p. 100) until English overtook it following brief periods of British occupation in the early 19th century and subsequent changes in demographics (Sprauve, 1975, p. 5). According to Bastian, after the arrival of the first shiploads of African slaves, the European “white population [would always be] small in proportion to the black population” (99) and the same diversity of nationalities and languages among Europeans was also true of the enslaved African population on St. Thomas and St. John. Thanks to their diversity, the enslaved people developed a now extinct Dutch Creole known as Negerhollands, a language combining Dutch and African influences. The descendants of those enslaved make up the majority of St. Thomians, whose official language is, today, a Caribbean-styled standard English used for legal, academic, and commercial purposes (Craig, 1975; Roy, 1975b; Sabino, Diamond, & Cockcroft, 2003; Sprauve, 1975); outside of these contexts, local speech is dominated by St. Thomian varieties of creolized Caribbean English, Caribbean varieties of French, and Dominicano Spanish.

3. Dominicano presence on St. Thomas St. Thomas has regularly seen a “historical migration of workers from one island to the other in search of labor” (Sprauve, 1975, p. 6); early in the 20th century, many Virgin Islanders traveled to Santo Domingo, and now Dominicanos are the latest in a long line of immigrant groups to make their place in the community and culture of St. Thomas. Prior to 1960, there had been no Dominicanos on St. Thomas (D’Arpa, 2015); by 1970, the sudden presence of 122 Dominicanos on St. Thomas was a direct consequence of the end to the long-standing Trujillo regime, which for decades had prohibited emigration from Dominican Republic (Sagás & Molina, 2004). By the 1980 census count, Dominicanos on St. Thomas more than doubled, to 258, beginning a trend of exponential growth: by 1990, they numbered just under one thousand, and in 2010, the Dominicano population reached 3,573, of which over three-quarters claimed to have been born in Dominican Republic, indicating this population growth is still fueled more by arriving immigrants than by growing families. Dominicanos on St. Thomas largely report that they have come to find work in manual or domestic labor, and the majority are of prime working age, between 18 and 65 years old. The ratio of Dominicano men to women on St. Thomas is approximately 1 to 1, among whom common scenarios include a single mother who earns the private school tuition for the children she left behind, or a young man who saves to buy property in Dominican Republic so that he might climb the social ladder upon his return. Most Dominicano immigrants to St. Thomas fulfill the transmigrant dream of returning to their home nation after working from several years to various decades on St. Thomas.

4. St. Thomas English Creole The research on STTEC includes A Brief Description and Dictionary of the Language Used in the Virgin Islands edited by John Roy (1975a), which was created to

155

Creolized English grammar on St. Thomas

155

bring awareness to continental teachers in local schools of the legitimacy of Virgin Island English Creole (VIEC). The report contains an extensive taxonomy of linguistic features of STTEC, especially in the section titled “On the syntax, morphology, and lexicon of Virgin Islands English Creole” (1975) by Gilbert A. Sprauve, who had previously compiled a description of STTEC features, especially its sound system, in his doctoral dissertation “Toward a Reconstruction of Virgin Islands English Creole Phonology” (1974). Robin Sabino, in her article, “Plural marking in the Virgin Islands English Creole in the St. Thomas—St. John Community” (1983), added details missing from earlier studies that had oversimplified the plural markings in VIEC as either the basilectal plural marker dem or the acrolectal s and which had ignored several Creole pronominal modifiers that mark plurality, including forms of the possessive, allyou shirt, or “all your (sg./pl.) shirts”; quantifiers, alotta star “a lot of stars”; numbers, de four million, eight million hundred billion pung a ting; and demonstrative, dem blue ticket buay “those blue ticket boys” (Sabino, 1983, p. 5). The sociolinguistic literature on the Virgin Islands includes a study by Robin Sabino et al. (2003), which establishes that STTEC is a “multisystemic repertoire” that affects, but is also affected by, daily use of language in social situations, particularly the effects of the listener on plural markings. They discovered that as discourse aimed at a local audience became more formal, “speakers preserved the Creole language pattern by increasing the plural marking of definite nouns at a substantially higher rate than for indefinite nouns” (92), demonstrating that Virgin Islanders have a great competence for code-switching, which may be more influenced by audience than by speech act. A similar sociolinguistic phenomenon regarding the effect of audience/listener had been described by Robert Di Pietro (1968) in his study of multilingual communities in St. Croix— relevant here to the extent in which the two islands share a social history—in which he noted that when outsiders, and even natives who had been off island too long, do not speak the proper island language in informal speech, an effort is made on the part of the Cruzan to approximate standard English (130). However, a native’s attempts to speak standard English or talk like continentals in the wrong social context can inspire somewhat derogatory expressions like “she was yankin”. Di Pietro predicted that language contact influences would diminish the use of the Virgin Island English–based dialect, ultimately leading to the “demise of diglossia and the spread of bilingualism” in the territory. Alma Simounet (2005), in her study of languages in contact on St. Croix, argues that Spanish is currently in danger of eventually being lost to English. In her own words, her study is a “wake-up call” to protect against the loss of Spanish on the Virgin Islands, and it centers optimistically on “religious institutions as the source of protective resources for the retention of Spanish within a migrant community from Puerto Rico” (263).

5. Methodology My own research on the Spanish spoken by Dominicanos on St. Thomas (D’Arpa, 2015) revealed that although their Spanish is not in danger of being lost, it is changing; the members of this speech community are incorporating elements of

156 Daniel S. D’Arpa

156

STTEC in their spoken Spanish. My experience with St. Thomas speech communities has been by way of membership, participant observation, and one-on-one interviews. For four years as a faculty member of the University of the Virgin Islands, St. Thomas campus, where I taught Spanish courses mostly to native St. Thomians, I made myself known among Dominicano Spanish speakers who invited me to social gatherings, whom I invited to speak to my classes on Hispanic culture and on immigrant experiences on St. Thomas. Eventually, I structured a sociolinguistic questionnaire, appended in my doctoral dissertation (2015, p. 198) and based on a model by Michol Hoffman (2009), to systematically document language and social data from a representative sample of Spanish-speaking Dominicanos on St. Thomas. The survey was divided into two parts: the first contained approximately 40 relatively short questions on topics including demographic data and questions on perceptions of language use and cultural values; the second section consisted of only two or three broad questions intended to elicit unguarded speech by prompting respondents to tell a personal life story of success, failure, near death, or stress. The questions in the first portion of the interview were addressed using the formal usted form, and in the second portion, questions were posed in the informal tú form. The first interviewees were acquaintances who knew me as a professor or fellow community member; these earlier respondents then connected me to their friends and families and so on along social networks. I conducted interviews in barber shops, in restaurants, at church gatherings, at beach barbeques, in supermarket parking lots, in school classrooms, and in taxis. The surveys were conducted entirely in Spanish, and each respondent received all the same questions in the same linear sequence, but they were encouraged to talk as much as they liked on each topic. Interviews typically ran from 30 to 45 minutes. The conversations were recorded with a lapel microphone and a digital recorder. All the recordings were reviewed, and a section of the most unguarded speech detected from each speaker was transcribed. The transcriptions were color coded to more quickly reference linguistic features and then checked against the recordings for accuracy. In a forthcoming publication, I will refine the quantitative approach begun in my dissertation (D’Arpa, 2015), to measure language variation for which answers to social questions on Dominicano identity and assimilation into St. Thomian culture were correlated to the phonetic production of the St. Thomian /ε/. The categories for external social factors included age, sex, place of birth, education, selfascribed identity, social network strength, frequency of visits to Dominican Republic, and perceptions of anti-Hispanic discrimination. The internal linguistic variables accounted for factors in six categories, which included preceding and following phonological segments, syllable type and stress, whether or not the token was borrowed from English, and the relative degree of formality of the interview question. The data showed that linguistic factors alone were not enough to explain the appearance of this allophone in Spanish, making a stronger argument for the influence of social factors. The resulting data further showed that a Spanish speaker on St. Thomas describing their ethnic background by calling

157

Creolized English grammar on St. Thomas

157

themselves St. Thomian was a good predictor that they would pronounce the St. Thomian /ε/ in their Spanish; and second, a Spanish speaker describing their social relationships as more strongly rooted in the Dominicano community with fewer ties to St. Thomian social networks was also a good predictor that they would pronounce the St. Thomian /ε/ in their Spanish.

6. Dominicanos report language change I am not the first to observe that a variety of Spanish spoken by Dominicanos on St. Thomas shows evidence of change; members of this speech community report the change themselves. More than half of the Dominicanos whom I spoke with believed that there will evolve in St. Thomas a dialect of Spanish different from those spoken in Dominican Republic, resulting from contact with the local English (D’Arpa, 2015). According to these interviewees, there is already proof of such a dialect in the speech of young Dominicanos, who, in the words of some respondents, speak a Spanish that is, different, mixed, badly pronounced, distorted, and accented. Several of them noted the effects of language contact, offering insights such as, Porque los hispanos que son [de] aquí copian el acento de aquí y lo hablan en español. Y se notan mucho la diferencia. Because the Hispanics that are from here copy the accent from here and they speak it in Spanish. And the differences are very noticeable. Another respondent agreed: Ello no hablan el español que hablamo nosotro. Lo que son de aquí lo habla como que ni uno ni entiende mucha vece. Aunque le digan epañol, lo hablan muy distinto. They don’t speak the Spanish we speak; those born here speak in a way you often don’t understand. Even if they say it’s Spanish, they speak it very differently. My own collected data show evidence to support these claims.

7. Features of a St. Thomas variety of Dominicano Spanish 7.1 Phonetics and phonology 7.1.1 /s/ > [h], [Ø] Lipski observes that in Dominicano Spanish, “syllable- and word-final /s/ is aspirated or, more frequently, lost [with] an exceptionally high level of hypercorrection” (136). Indeed, Dominicano Spanish on St. Thomas showed the

158

Daniel S. D’Arpa

158

aspiration and loss of /s/ > [h], [Ø]. But in addition to the loss of /s/, whole syllables were observed to be elided, most frequently in the speech of young people, as in this response from a female high school student: [ea na ma cree ke noso tamo, hatamo ablando pañol poke [x] otras hente scuchan.] Ella nada más cree que nosotros estamos, estamos hablando español porque otra gente escucha. She just thinks we’re, we’re talking Spanish because other people are listening. 7.1.2 Metathesis Metathesis is another linguistic feature of Dominicano Spanish on St. Thomas of which speakers themselves are aware. As the following utterance illustrates, speakers can describe this phenomenon in at least lay terms: Dicen “brroadora” para borrador. They say “brroadora” for eraser. The speaker here was an older immigrant describing the changes she perceives in the speech of younger Dominicano Spanish speakers on St. Thomas (i.e., brroadora). I recorded evidence of metathesis from a young girl who selfidentifies biculturally and bilingually as Dominicana and St. Thomian; she produced leo < ello as she explained to me how her native fluency of STTEC leads people to think she is non-Hispanic: Yo le digo que yo hablo epañol; leo (ello) me dicen“dique, yo no sabía”. I tell them I speak Spanish; they tell me, “huh, I didn’t know.” An adult male who had immigrated relatively recently, and who likely exemplifies Dominicano speech with less influence from STTEC, said on the topic of physical differences between people from different islands, Ya aún lejo, po lo meno el sífico. Un lo conoce por el sífico; aquí, la oreja son muy ditinta. Even from far off, at least physically. You know them by their physical features; here their ears are very different. If Gilbert Sprauve’s observation of STTEC (1975, p. 54) still holds, by which “metathesis occurs in only two lexical items: flɪm for ‘film,’ aks for ‘ask,’” then these particular changes in the Spanish spoken by younger Dominicanos

159

Creolized English grammar on St. Thomas

159

on St. Thomas probably do not result from any linguistic influence of STTEC. Rather, this expansion of a variable rule in Dominicano Spanish, which speakers seem to be aware of, might be indexical of social factors such as solidarity with the Dominicano community. Further research may tell. 7.1.3 [ε] The STTEC vowel /ε/ was observed as an allophone in the phonetic system of Dominicano Spanish speakers on St. Thomas. According to Sprauve (1974), who had catalogued the vowel system of STTEC, a significant distinguishing characteristic that sets St. Thomian English apart from other Caribbean varieties of English is the quality of the /E/, which today I would represent with an epsilon [/ε/], can be described as a mid low front vowel or a mid open front vowel that is monophtongal. I emphasize “monophtongal” here since the ungliding, pure quality of the St. Thomas/St. John vowels, especially those in the “mid” range, helps in distinguishing that dialect of Caribbean English–related Creoles from its numerous relatives. Putting aside the qualities of “pure and ungliding,” which may distinguish it as peculiar to STTEC, the vowel [ε] is common in several varieties of Caribbean English; according to Dagmar Deuber (2014, p. 14), a comparison of the vowel systems of Jamaican and Trinidadian standard English as outlined in the Handbook of Varieties of English by Devonish and Harry (2004: 460) and Youssef and James (2004: 515), respectively, indicates a number of parallels: The pronunciation of NEAR and SQUARE words, which are generally merged, is usually monophtongal as well ([e:])/[ε]). Consider the following phonetic transcription of speech from a 45-year-old male electrician, living 18 years on St. Thomas, who produced the STTEC [ε] in his Spanish while speaking on the possible reasons why English-speaking St. Thomians there discriminate against Dominicanos. His understanding is that Hispanic culture at large is perceived as inferior: Polquε, bueno, eh, por ejemplo, el, eh pol eh pol la cultura. La cultura mayormente de lo spano er una cultura siempre baj. Toncε cuando e muy por abajo lo que crεa e problεma. Because, well, it’s, for example, the, it’s because, it’s because of culture. The Hispanic culture has generally been a low culture. So when a culture is devalued, that causes problems. Offering a specific example, he explains how Hispanics who do not speak English are unfairly treated, underserved, and harassed in public offices: E cuando uno no puede hablá inglέmuy bien, pue lo poneŋ a peldé tiempo. Tiende? εsí que no hay un, no hay una combinación pelfeta. Si necesito tre papele, me diceŋ uno. Y jumo no hablo bieŋ, pue me poneŋ a la viaje. Mucho poblema dificil.

160

Daniel S. D’Arpa

160

That’s when if you don’t speak English well, then they waste your time. You see? So there is no perfect solution. If I need three papers, they tell me one. And since I don’t speak well, they send me on a goose chase. [There are] lots of difficult problems. Another respondent who pronounced the STTEC [ε], and who similarly reported on racism against Dominicanos, was a 22-year-old female student and hotel staff worker. She adds that the prejudice against Dominicanos is based not only on language differences but also on non-black phenotype features like lighter skin and on xenophobic constructions of rights and privileges that are based on birthplace (i.e., born on St. Thomas or not). She says, Aquí son muy racista aquí todo el tempo. Si tú no eres dε εquí, es difícil conseguir las cosas. Te dan más; spetsialmente si no sabes hablar el idioma. Pero aquí son muy racista. Spetsialmente con lo blanco, lo hispano. They’re very racist here all the time. If you’re not born here, it’s hard to get on. They give [it] to you more, especially if you don’t speak the language. But they’re very racist here—especially against whites, Hispanics. 7.2 Morphophonology 7.2.1 Plural /-se/ lexicalized in singular nouns In some varieties of Dominicano Spanish, “plural nouns are formed by adding se rather than s: casa-cásase (house(s)), mujer-mujérese (woman/women)” (Lipski, 2008, p. 137). In Dominicano Spanish on St. Thomas, evidence of this feature was recorded for certain singular nouns: En mi paíse, la mayoría de la gente somo negro. In my country, most of us are black people. This speaker, who clearly elides the final -s in plural verbs and nouns—somo(s) negro(s) (we are black people)—has combined the grammar rules of elided final -s and plural -ses to produce paíse (country) as the default singular form of países (countries). There is no ambiguity regarding the singular or plural meaning, because while *en mis países (*in my countries) may be grammatically acceptable, it is not a logical utterance regarding a speaker’s heritage; so it should be clear that there is a singular form here, in paíse. More evidence was demonstrated by another speaker whose variation of this morphophonology in the same sentence included the singular masculine indefinite article un: Por lo meno, de un paí pobre que salí a un paíse como ete que, si uté trabaja, echa para alante. At least, coming from a poor country to a country like this where, if you work, you get ahead.

161

Creolized English grammar on St. Thomas

161

7.2.2 Word-final /-m/ as plural There may be evidence of a new plural noun marker in at least one idiolect of Dominicano Spanish spoken on St. Thomas. However, since this was observed in the case of only one speaker, I cautiously submit it here as an important curiosity, and I propose that future investigations consider whether this feature is descriptive of the larger speech community. According to Sprauve (1975, p. 17), the opposition of features /count/ versus /mass/ is present in English Creole on a fairly parallel basis with standardized English, as shown in the following examples: sing. indefinite definite

pl. a goat de goat

(some) goat(s) de goat (dem/-s)

For some plural nouns, a 15-year-old female student respondent clearly pronounced a word-final /-m/. If this is a result of dialect contact and leveling with STTEC, in which dem is a plural marker with great salience and markedness, then this word-final /-m/ on Dominicano Spanish nouns could be a simplification of dem functioning to mark plurality in nouns in Dominicano Spanish. Consider the use of final -m in the following sentences: Pero lo hipano sí tienen problema sí. Tienen hipa-, eh, problemam. Como hablan epañol. But Hispanics do have problems. They have Hispa-, uh, problems. Since they speak Spanish. Te tratan como que tú ere baseura, [llamándonos] dem santeus, dominicanom. It’s like they treat you like you’re garbage, [calling us] dem Santo, Dominicano dem. The first of these two examples rules out the possibility that this is just a mistake of careless speech because, on the contrary, it is an immediate repetition of the phrase “tienen problema” (they have problems) and is therefore more likely to be emphatic speech with the intention to clarify or drive the point home. There were numerous examples by many speakers in my interviews of /n/ > /m/ for assimilation into a proceeding bilabial consonant (e.g., que vengam porque), but the final -m in question here appears before a stop in a noun phrase marked with plurality and may therefore demonstrate a new morphophonological variant of plurality among nouns in Dominicano Spanish, resulting from contact with the STTEC plural marker dem. 7.3 Verb phrase syntax 7.3.1 /taŋ/ Evidence may suggest that a tendency in STTEC to use unbound morphemes may be making its way into a variety of Dominicano Spanish spoken on St.

162

Daniel S. D’Arpa

162

Thomas, particularly to express the habitual mood and with the morpheme /taŋ/. STTEC is analytical rather than inflectional (Craig, 1975, p. 42), employing aspectual particles (e.g., a, di, da, doz) where standardized English employs inflections. In St. Thomas English Creole, there exists a habitual mode (Sprauve, 1975, p. 15), which is indicated by the use of the preverbal function word does/du. In the sentence, “he doe(s) study real hard,” the phrase “doe(s)/ du study” is the equivalent of standardized English “studies (habitually).” The present progressive, which is signaled by the participle—in, as in “you wearin de hat an de wrong side” (you’re wearing the hat on the wrong side)—can be combined with the habitual mood to produce a sentence such as, “he doe(s) be studying real hard” (he is [always] studying real hard). Consider the following sentence in Dominicano Spanish: A vece cuan yo ando coŋ con mi tíos, o y ello taŋ tienε: el radio lae (Engl. “like”) duro en carro, gente taŋ diciendo “¡o: eh quita esa [m]ú:sica!”. Sometimes when I go with my uncles, oh and they doe(s) have the radio like loud in [the] car, people doe(s) be saying “Hey, cut that music!” Here, the morpheme /taŋ/ (from Spanish están) is easily recognizable in the construction [taŋ disiendo] (están diciendo); however, in this utterance, it is not an example of the *present progressive, which in standardized Spanish indicates an event happening concurrently with the utterance. Rather, this speaker has employed it to indicate a type of habitual action. Indeed, it looks very much like a grammatical mode of STTEC that Sprauve (1975, pp. 15–16) described as a recreative ongoing mode, used to report—with drama and often prejudice—an incident of the recent past—dey come callin(g) me name; dey go (around) tellin lies; dey keep on hittin—and it differs meaningfully from you come walkin? where the inflection—in—functions to create an adverb of manner. The Spanish sentence just transcribed seems to meet Sprauve’s criteria in that it describes a recreative ongoing event as evidenced by the phrase “a veces cuando yo ando” (sometimes when I go), and it reports, with drama and prejudice, an account of cultural differences and conflict in recent history. Therefore, the Spanish morpheme /taŋ/ in taŋ diciendo may have taken on the meaning of dey come saying, dey go (around) saying, or dey keep (on) saying. However, the investigation of this feature should not end here, because the verb ending -iendo is a compelling reason to consider this phrase, “taŋ diciendo”, as a translation of another, related but separate verb morphology found in STTEC, one that is formed by combining the habitual and present progressive. Sprauve offered the following example of this mode, which he did not name: “He doe(s) be studying real hard, or ‘He is (always) studying real hard’” (15). If this meaning is the preferred interpretation of the utterance, then the morpheme /taŋ/ could be translated as the doe(s) of STTEC. In fact, I chose this second option in my earlier translation in order to create a simplified rule for /taŋ/ that also appears in an earlier phrase in the same sentence, ello taŋ tienε, but with a different

163

Creolized English grammar on St. Thomas

163

meaning. In the earlier sentence, the most logical interpretation of ello taŋ tienε (they doe[s] have) would express a habitual action. In STTEC, habitual action is indicated by the structure doe[s] + verb = habitual: he doe(s) study real hard. Does/du study is the equivalent of standardized English “studies (habitually)”: she doe(s) drive fas, or “she drives fast” (15, 20). It might be that in a variety of Dominicano Spanish spoken on St. Thomas, the morpheme /taŋ/ has been assigned new meaning in some contexts, which includes expressing the aspect of habitual actions in verb phrases.

8. Conclusion This article has described linguistic features of a variety of Dominicano Spanish spoken on St. Thomas, US Virgin Islands. Many features of this variety are consistent with those of Dominicano Spanish in other environments (Lipski, 2008), whereas other features recall linguistic structures of STTEC, namely the phonetic variation of [ε], the word-final /-m/ to mark plurality, and the morpheme /taŋ/ to express the aspect of habitual action. The language samples presented in this chapter were selected from segments of the interviews in which the respondents discussed discrimination from St. Thomians, to support claims that the legacy of colonialism persists in St. Thomas (de Albuquerque & McElroy, 1999), where immigrants of Dominican Republic are the latest group to contribute to its diversity and one with the least social capital. Recurring themes in respondents’ answers suggest that there is a pervasive social anxiety among Dominicanos on St. Thomas to prove they belong and have rights to scarce cultural and economic resources. Dominicano Spanish speakers who understand that the “use of Creole English [is] an affirmation of national and ethnic pride and . . . the language of true Caribbean identity” (Nero, 2001, p. 7) may seek to negotiate their identity and make known their social aspirations by way of sociolinguistic performances that incorporate elements of STTEC into their Spanish, consistent with past sociolinguistic research (Labov, 1972; Meyerhoff, 2015; Suarez, 2010; Turell, 2001). Further research is needed on Dominicano Spanish in contact with STTEC, including variationist analyses of speech from a sample of the community, for which a survey should be deliberately designed to elicit speech that contains the observed features, such as plural constructions, storytelling in habitual actions, and sets of tokens in Spanish and English for pronunciation. Members of the speech community, ideally bilingual speakers of STTEC and Dominicano Spanish, should be recruited and trained to help construct and administer the survey.

References Bastian, J.A. (2001). A question of custody: The colonial archives of the United States Virgin Islands. The American Archivist, 64(1), 96–114.

164

Daniel S. D’Arpa

164

Boyer, W. (2010). America’s Virgin Islands: A history of human rights and wrongs (2nd ed.). Durham, NC: Carolina Academic Press. Craig, D.R. (1975). Towards a Standard English language programme for Creoleinfluenced speakers. In J. Roy (Ed.), A brief description and dictionary of the language used in the Virgin Islands. St. Thomas, VI: V.I. Dept. of Education. D’Arpa, D.S. (2015, January 1). Dominican Spanish in contact with St. Thomas English Creole: A sociolinguistic study of speech variation on St. Thomas, U.S. Virgin Islands. ProQuest Dissertations Publishing. de Albuquerque, K., & McElroy, J.L. (1999). Correlates of race, ethnicity and national origin in The United States Virgin Islands. Social and Economic Studies, 48(3), 1–42. Deuber, D. (2014). English in the Caribbean: Variation, style and standards in Jamaica and Trinidad. Cambridge: Cambridge University Press. Devonish, H., & Harry, O.G. (2004). Jamaican Creole and Jamaican English: Phonology. In B. Kortmann & E. W. Schneider (Eds.), A handbook of varieties of English (pp. 441–471). Berlin, Germany: De Gruyter. Di Pietro, R.J. (1968). Multilingualism in St. Croix. American Speech: A Quarterly of Linguistic Usage, 43(2), 127. Dookhan, I., & Sheridan, R.B. (1994). A history of the Virgin Islands of the United States. Kingston, Jamaica: Canoe Press. Hoffman, M. (2009). Sociolinguistic interviews. In J. Holmes & K. Hanzen (Eds.), Research methods in sociolinguistics (pp. 25–41). Malden, MA: Wiley Blackwell. Labov, W. (1972). Sociolinguistic patterns. Philadelphia, PA: University of Pennsylvania Press. Lipski, J.M. (2008). Varieties of Spanish in the United States. Washington, DC: Georgetown. Meyerhoff, M. (2015). Introducing sociolinguistics. New York, NY: Taylor & Francis. Nero, S.J. (2001). Englishes in contact: Anglophone Caribbean students in an urban college. New York, NY: Hampton Press. Roy, J. (Ed.). (1975a). A brief description and dictionary of the language used in the Virgin Islands. St. Thomas, VI: V.I. Dept. of Education. Roy, J. (1975b). A note on some of the consonant variations of Virgin Islands English Creole. In A brief description and dictionary of the language used in the Virgin Islands. St. Thomas, VI: V.I. Dept. of Education. Sabino, R. (1983). Plural marking in the Virgin Islands English Creole in the St. Thomas— St. John community. Penn Review of Linguistics, 7, 3–11. Sabino, R. (1994). They just fade away: Language death and the loss of phonological variation. Language in Society, 23(4), 495–526. Sabino, R., Diamond, M., & Cockcroft, L. (2003). Language variety in the Virgin Islands: Plural marking. In M. Aceto & J.P. Williams (Eds.), Contact Englishes of the Easter Caribbean. Philadelphia, PA: Benjamin Publishing. Sagás, E., & Molina, S. (2004). Dominican migration: Transnational perspectives. Gainesville, FL: University Press of Florida. Simounet, A. (2005). La religión y la retención lingüística: el caso de una iglesia pentecostal en Santa Cruz, Islas Vírgenes estadounidenses. In Contactos y contextos lingüísticos : el español en los Estados Unidos y en contacto con otras lenguas.—(Lingüística iberoamericana; 27). Madrid: Iberoamericana Vervuert. Retrieved from http://digital. casalini.it/an/2519795 Sprauve, G.A. (1974). Towards a reconstruction of Virgin Islands English Creole phonology. Ph.D. Dissertation, Princeton University.

165

Creolized English grammar on St. Thomas

165

Sprauve, G.A. (1975). On the syntax, morphology, and lexicon of Virgin Islands English Creole. In J. Roy (Ed.), A brief description and dictionary of the language used in the Virgin Islands (pp. 4–22). St. Thomas, VI: V.I. Dept. of Education. Suarez, D. (2010). The paradox of Linguistic hegemony and the maintenance of Spanish as a heritage language in the United States. Journal of Multilingual and Multicultural Development, 23(6). Retrieved from www.tandfonline.com/doi/abs/10.1080/ 01434630208666483 Turell, M.T. (2001). Multilingualism in Spain: Sociolinguistic and psycholinguistic aspects of linguistic minority groups. Bristol, UK: Multilingual Matters. Youssef, V., & James, W. (2004). The creoles of Trinidad and Tobago: Phonology. In B. Kortmann & E. W. Schneider (Eds.), A handbook of varieties of English (pp. 508–524). Berlin, Germany: De Gruyter.

9

Puerto Rican evaluations of varieties of Spanish Eva-María Suárez Büdenbender

1. Introduction The main objective of this paper is to examine Puerto Rican islanders’ evaluations of their own variety in comparison to their evaluations of other varieties of Spanish. Recent studies on self-evaluations of Puerto Ricans on the island point to the possibility that linguistic insecurity, evident among immigrant communities on the US mainland (e.g., Zentella, 1997, 2002, 2007; Otheguy, Zentella, & Livert, 2007; Otheguy & Zentella, 2012), is not found among Puerto Ricans on the island to the same degree and therefore appears to be a byproduct of immigration (Mojica de León, 2014). In spite of apparent high value placed on the native variety, other varieties, such as those spoken in Spain and Argentina, continue to garner the highest prestige among Caribbean Spanish speakers (e.g., Alfaraz, 2002, 2014; Mojica de León, 2014). Other factors to consider are (a) the conflicted relationship between Puerto Ricans and the influence of English on their native variety and an increase in bilingualism among the younger and more educated urban population (e.g., DomínguezRosado, 2015; Loureiro-Rodríguez, Guzzardo, & Vélez Avilés, 2016) and (b) non-linguistic factors, such as perceived socioeconomic status of a country, which are thought to drive the high prestige of other varieties of Spanish (Alfaraz, 2002, 2014; Mojica de León, 2014). The current study will seek to confirm these tendencies and gain a deeper understanding of motivations. A brief summary of linguistic features will offer some details on the particularities of this Caribbean Spanish variety and will be followed by a discussion of linguistic and non-linguistic factors in language (self-)evaluations.

2. Linguistic features of Puerto Rican Spanish PRS belongs to the Caribbean/Antillean zone and therefore shares a number of features at all linguistic levels (phonetics/phonology, morpho-syntax, and lexicon) with other varieties spoken in and along the Caribbean basin. Among these are the aspiration or loss of syllable-final /s/ (la[h] casa[ø], or “the houses”),1 the velarization of /n/ (e[ŋ] la escuela, or “in (the) school”)2 (e.g., Holmquist, 2003; Lipski, 1994; López Morales, 1992; Navarro Tomás, 1948), and the lateralization

167

Puerto Rican evaluations of Spanish

167

of /r/ in coda position (i.e., the realization of coda /r/ as [l] as in pa[l]te instead of parte “side”), a very common feature across all parts of Puerto Rico that can be found among speakers from various socioeconomic groups (e.g., López Morales, 1992; Navarro Tomás, 1948). Widespread is the uvular articulation of multiple trill /r/ in syllable-initial position (e.g., López Morales, 1992; Navarro Tomás, 1948), which can be observed across Puerto Rico but is especially common among the lower social classes, inhabitants of rural areas, and among men (Lipski, 1994, p. 334). However, it can also be found even in the most formal discourses (Navarro Tomás, 1948, p. 93). Other features include the weakening of the intervocalic /d/, particularly in the suffix -ado (e.g., Lo he entrega[d]o. “I have handed it in.”) (Lipski, 2008), the affricate pronunciation of /j/ word/phrase initially but not intervocalically (Saciuk, 1977; Lipski, 2008), the aspiration [h] of the posterior fricative /x/, and the fricative [ʃ] pronunciation of the affricate /tʃ/ much like English “show” (Lipski, 2008). Examples of the syntactic features that deviate from the perceived norm include the use and distribution of overt pronouns (Tú me avisa cuando tú esté lista. “Let me know when you are ready.”) (e.g., Lipski, 1994; Orozco, 2015), word order and NP characteristics (e.g., Dauphinais & Ortiz López, 2013), and the use of infinites in optative clauses (Eso te pasa por tú ir demasiado rápido. “This happens to you because of you going too fast.”) (e.g., Aponte Arlequín, 2014). Other features are non-inverted subject pronouns in questions (¿Qué tú quieres? “What do you want?” instead of the more standardized ¿Qué quieres (tú)?) (e.g., Lipski, 2008). Moreover, personal pronouns can be found as preposed lexical subjects especially in constructions with para (para yo llamarte, or “for me to call you”). However, these structures also appear in other Caribbean dialects (Lipski, 1994, 2008) and in other varieties of Spanish, such as the Canary Islands (e.g., Almeida, 1990; Lorenzo Ramos, 1976). Another aspect of PRS frequently observed is the influence of English on PRS. This appears in the form of calques such as ¿Cómo te gustó la playa? or “How did you like the beach?” compared to the more standardized sentences ¿Te gustó la playa? or ¿Cómo lo pasaste? (e.g., Lipski, 1994, p. 335; Lloréns, 1971). Several of the linguistic features of PRS, such as the lateralization of coda /r/, the uvular articulation of multiple trill /r/ in syllable-initial position, and calques from English, are a deviation from those varieties considered more standard, such as those spoken in Spain or Colombia and therefore have been shown to be highly stigmatized by speakers of other varieties of Spanish (e.g., Alfaraz, 2002, 2014; Domínguez-Rosado, 2015; Lipski, 1994; Mojica de León, 2014).

3. Issues influencing linguistic evaluations of Puerto Rican Spanish Recently, PRS speakers living in Puerto Rico have evaluated their variety positively (Mojica de León, 2014). However, studies on other Caribbean varieties did not show positive evaluations in spite of similarities between the varieties

168 Eva-María Suárez Büdenbender

168

(Alfaraz, 2002, 2014; Coello Millán, 2014; Jara Murillo, 2008; Orozco, 2010; Severino Cerda, 2014; Sobrino Triana, Montero Bernal, & Menéndez Pryce, 2014). This outcome is quite surprising given the linguistic similarity between the different Caribbean Spanish varieties. Alfaraz (2002) attributes this behavior to the fact that speakers of less prestigious varieties frequently disassociate themselves from others by downgrading speakers of similar forms of speech (see also Preston, 1989, 2002). This process of disassociation was also found among Puerto Rican immigrant communities in Orlando, Florida (Lamboy, 2011), who sought to distance themselves from their compatriots in New York City linguistically and socially, due to the negative images connected to Puerto Rican immigrants in that city. This exemplifies that existing belief systems are not solely driven by linguistic factors but are closely intertwined with social or sociocultural beliefs (e.g., Crowley, 2003; Lippi-Green, 1997). What is more, these beliefs, once engrained in a community, appear to be stable over time (Alfaraz, 2014). Cross-dialectally, Puerto Rico’s low linguistic prestige has been connected with comparatively high poverty levels among all Hispanics in the United States but also linguistic features that deviate strongly from the perceived standard and thus are particularly vulnerable to negative evaluations (Alfaraz, 2002, 2014; d’Anglejan & Tucker, 1983; Lambert, Hodgson, Gardener, & Fillenbaum,1960; Long & Preston, 2002; Niedzielski, 1999; Preston, 1989, 1999). Associations between a particular linguistic feature and an assumed socioeconomic status, level of education, sexual orientation, or ethnicity exist in the minds of speakers can be triggered by very fine phonetic detail (e.g., Babel, 2014; Docherty, Langstrof, & Foulkes, 2013; Drager, 2006; Foulkes & Docherty, 2006; Hay & Drager, 2007). In the Caribbean, recent perception studies within Puerto Rico revealed a sensitivity of the listeners toward socially indexed phonetic features (e.g., Delgado-Díaz & Galarza, 2014; Holmquist, 2003; Mack, 2010, 2011). Alfaraz (2002, 2014) found perceptions of race to be a factor in linguistic evaluations in Miami Cubans. The Spanish spoken in countries that are perceived to have phenotypically darker speakers (e.g., Dominican Republic, Cuba, and Puerto Rico) received lower ratings than speakers of those countries who are perceived to be phenotypically lighter or more European. According to Duany (2002), the Puerto Rican national narrative highlights the Spanish heritage, while stereotypes continue to stigmatize the black population of the island. This same discourse has attempted to marginalize and exclude immigrant communities, such as Cubans and Dominicans, from this national imagery. Although, current perceptions of Cubans and Cuban Spanish are fairly positive (Mojica de León, 2014), the relationship with Cuban immigrants remains distant. Cubans have been perceived as dishonest and selfish and to this day might struggle to integrate into Puerto Rican society (Duany, 2002, p. 26). As for Dominicans, racial and linguistic stereotypes depict this immigrant group as dumb, dirty, and violent, and many ridicule Dominican Spanish as uneducated. The term “negro” (“black”) has become synonymous with “dominicano” (“Dominican”) (Duany, 2002, pp. 25–27). Therefore race, or the perception thereof, may also play a part in Puerto Ricans’ linguistic evaluations in the current study.

169

Puerto Rican evaluations of Spanish

169

Frequently, a side effect of a variety’s low prestige is the internalization of this negative perception by the speakers themselves, which in turn can give rise to linguistic insecurity (e.g., Gardener-Chloros, 1991; Paltridge & Giles, 1984). Among Puerto Ricans living on the US mainland, linguistic insecurity seems widespread (Zentella, 1997, 2007). However, to date, there is little evidence of linguistic insecurity on the island. Mojica de León (2014) in her study on linguistic attitudes in San Juan did not find evidence of linguistic insecurity. Speakers ranked their native variety as the one they most prefer (Mojica de León, 2014). However, similar to previous studies (e.g., Alfaraz, 2002, 2014; Jara Murillo, 2008; Orozco, 2010), the newer data also underline a continued high estimation of the varieties spoken in Spain, which are largely considered the most “correct” and are highly valued. The discourse of Puerto Rican national identity still highlights its cultural and ethnic ties to Spain, and Spanish ancestry is to this day highly valued, along with an independent Puerto Rican identity that, at least in recent years, has started to incorporate Taíno and African elements, albeit cautiously (Duany, 2002, p. 19). Sensitive in this context is the relationship with English and its perceived influence on PRS and the effects of increasing degrees of bilingualism on the island. A recent matched guise study by Loureiro-Rodríguez et al. (2016) shows that, on the one hand, PRS is highly and positively valued and strongly linked to a Puerto Rican identity but that, on the other hand, the ability to speak English is indicative of a higher level of education and urban origin, affecting in particular younger and urban Puerto Ricans and their identity (see also Domínguez-Rosado, 2015). In sum, although linguistic insecurity among Puerto Ricans has been studied among émigrés, few studies to date have examined the perceptions of Puerto Ricans living on the island, specifically knowledge about Puerto Rican linguistic (in)security at home. The present study seeks to contribute to the discussion on the interrelatedness of linguistic features, their evaluations, and the perceived socioeconomic, educational, and possibly racial features of its speakers. Specifically, we want to examine inferences made by non-linguists about language and the attitudes that emerge from these beliefs. For this reason, the methodology used here pertains to research conducted in the area of perceptual dialectology. This area of research examines the beliefs and attitudes of non-linguists, or folk beliefs, first initiated by Hoenigswald (1966) and later on established by Preston (e.g., Preston, 1989, 1999; Long & Preston, 2002). A growing number of studies in this research tradition examine the attitudes of everyday Spanish speakers, particularly Caribbean Spanish speakers (e.g., Alfaraz, 2002, 2014; Jara Murillo, 2008; Orozco, 2010; Severino Cerda, 2014), allowing us to understand the role of historical, social, political, and educational values attached to language attitudes, even among varieties as linguistically similar as those spoken in the Spanish-speaking Caribbean. More concretely, this study is modeled after the methodology used in Alfaraz (2002, 2014), where Miami-Cuban participants were asked to rate all 20 Spanish-speaking countries along the dimensions of “correctness” and “pleasantness.” Examining the perceived beliefs of the

170

Eva-María Suárez Büdenbender

170

participants about their own variety as well as that of others is the main focus of this study.

4. The present study The objective is to investigate Puerto Rican islanders’ evaluations of their own variety and in comparison to their evaluations of other varieties of Spanish and to examine the question of how much the economic strength of a country would influence its ratings. To this end, the study was guided by the following questions: Research Question 1: How will Puerto Ricans evaluate their own variety compared to other varieties of Spanish? And to get more detail, we ask the following two questions in addition: 1 2

Where do they feel the most “correct” variety of Spanish is spoken? Where do they believe the most “pleasant” variety of Spanish is spoken?

Research Question 2: Is there any evidence that the ratings are driven by nonlinguistic factors such as the perceived levels of education or socioeconomic status of the speakers? Based on the results of previous studies discussed in the preceding section, it is hypothesized here that ratings on “correctness” and “pleasantness” will reflect the influence of historical beliefs (e.g., higher ratings for Spanish from Spain). Moreover, ratings on “correctness” and “pleasantness” will reflect the influence of perceived socioeconomic strength and higher levels of education among speakers. 4.1 Materials A two-part questionnaire was used: the first part inquired about participant information (gender, age, level of education, region of origin, and profession), which was followed by four open-ended questions with regard to linguistic observations (items 6–9 in the questionnaire). These were included to gain participants’ insights into their experiences and reasoning in their evaluations. The second part of the questionnaire consisted of perception scales of “correctness of accent” and “pleasantness of accent” taken from previous studies in the area of perceptual dialectology (e.g., Alfaraz, 2002; Hoffman & Walker, 2010; Orozco, 2010; Preston, 1999). The format of this part of the questionnaire was adapted from Alfaraz (2002), where all 20 Spanish-speaking countries were presented individually with 7-point Likert scales for each dimension (“correctness” and “pleasantness”). The 7-point Likert scale was chosen for this study to make the results comparable to other studies of perceptual dialectology. A possible weakness with this scale (as opposed to a 4-point scale) is that participants not sure of their answer may default to a 4. This will have to be taken into account when interpreting the results.

171

Puerto Rican evaluations of Spanish

171

In half of the questionnaires, all 20 Spanish-speaking countries were presented in alphabetical, order and in the other half of the questionnaires, the countries were presented in reverse alphabetical order. Again, this method was taken from Alfaraz (2002, 2014), and the presentation of reverse alphabetical order included to counterbalance any possible conditioning effects that an alphabetical order might have entailed. A sample questionnaire (alphabetical order) can be found in the Appendix. 4.2 Participants After taking out incomplete questionnaires, a total of 77 surveys were analyzed for this study.3 Of these, 51.9% were from men and 48.1% from women, and 22.1% of participants belonged to the age group 18–29, 15.5% to the age group 30–39, 23.4% to the age group 40–49, 20.8% to the age group 50–59, and 18.2% to the age group 60 and older. The participants were overall highly educated: 70.1% of the participants had attended university, 19.5 % had reached 10th to 12th grade in high school, 9.1% had studied until the 7th to 9th grade, and 1.3% had left school between the 4th and 6th grades. No participant had a level of education below 4th grade. Although the level of education in the participant pool is rather high, it is in accordance with recent trends in San Juan—in particular, among the younger population (U.S. Census Bureau, estimates for, 2010). Similar high numbers of post high school education can be found all over the island. Therefore, this sampling, albeit small, is representative of the population of San Juan as a whole. 4.3 Procedure Data collection was conducted via stratified random sampling, allowing for stratification for gender, age group, and level of education to ensure equal distribution for each of these sociolinguistic factors and to avoid a preponderance of, for instance, younger speakers vs. older speakers. Face-to-face interviews were conducted in various areas of urban San Juan, PR (Barrio Obrero, Santurce, Miramar, Loiza, Tras Talleres, Viejo San Juan), where participants were recruited. Data collection took place throughout a larger geographical area of the capital with the aim of gaining access to a wide spectrum of possible participants and to stratify for gender, age, and level of education. For the open-ended questions, the participants were encouraged to answer in as much detail as possible. The interviewer took notes of the answers, or in some cases, the participants wrote their answers on the questionnaire themselves. This occasional change in procedure was permitted since many of the written responses allowed the participants, many of whom have high literacy, to express their thoughts more clearly. After completing the initial portion of the interview, the participants were either guided orally through the questions and 7-point Likert scales wherein the interviewer marked the answers on a questionnaire, or the participants

172

Eva-María Suárez Büdenbender

172

7 6 5 4 3 2 1 0

Correctness

Figure 9.1 Representation of results for ratings on “correctness of accent.”

chose to fill out the scales on their own. In either case, all participants were asked to answer all questions completely so as not to leave any questions blank. 4.4 Data analysis The quantitative data on all 77 questionnaires were entered into Excel and then subsequently entered into SPSS for analysis. Descriptive procedures were used to summarize information on percentages on gender groups, age groups, and level of education. Moreover, means were calculated for the ratings for each country in each one of the two dimensions: “correctness of accent” and “pleasantness of accent.” The results of the means calculations are represented in Figures 9.1 and 9.2 and in the discussion of results in the ensuing section.

5. Results and discussion Figure 9.1 represents the results for the ratings for “correctness of accent.” It is plotted by average ratings given to each variety/country. The variety rated highest in this dimension was Spain (5.90), closely followed by Colombia (5.36), Argentina, and Venezuela. Puerto Rico still received average ratings hovering around 5 (5.12, 5.09, and 5.09, respectively). Other Caribbean varieties, such as Panama (4.82) and Cuba (4.66), received less enthusiastic ratings. The Caribbean neighbor to Puerto Rico, Dominican Republic, received the lowest ratings of all in this dimension (3.73). This outcome falls in line with previous studies on perceptual attitudes by speakers of Caribbean Spanish (e.g., Alfaraz, 2002, 2014; Jara Murillo, 2008; Orozco, 2010). The low estimation of Dominican Spanish is surely related to the low levels of education and socioeconomic status

173

Puerto Rican evaluations of Spanish

173

7 6 5 4 3 2 1 0

Pleasantness

Figure 9.2 Representation of results for ratings on “pleasantness of accent.”

of many Dominican migrants living in Puerto Rico (e.g., Duany, 2005; Suárez Büdenbender, 2013). The results for the dimension of “pleasantness of accent” are presented in Figure 9.2. Similar to Figure 9.1, Spain (6.09) and Colombia (5.73) hold the first two places. In contrast to the results for “correctness of accent,” Puerto Rico (5.67) can be found in third place, followed by Argentina (5.57) and Venezuela (5.34). Another Caribbean island variety, Cuban Spanish, received average ratings (4.89), and again, in last position is Dominican Republic (4.07) with the least pleasant accent according to the participants. Based on the ratings in both dimensions, Puerto Ricans hold their native variety in high esteem in the dimension related to solidarity (“pleasantness”), although they do concede perceived weaknesses when comparing their native tongue to those varieties thought to reflect more “standard” forms as can be seen in the ratings for “correctness.” The evaluations for both dimensions fall in line with previous work (Mojica de León, 2014), revealing little evidence of linguistic insecurity among Puerto Ricans, but an awareness of the distance of their native variety to a perceived standard in at least some linguistic areas. The responses to the open-ended questions offer more insights into the rationale behind these ratings. The participants’ responses on their thoughts on the “correctness” of PRS were mostly positive (item 2), however, they also commented on its perceived incorrect features in pronunciation or lack of education of the speaker (1). 1

Regular. Todavía existen muchas palabras mal escritas o habladas. Normal. There still exist many words that are written or pronounced incorrectly. (male, San Juan, manager, 30–39 years old)

174

Eva-María Suárez Büdenbender

174

The participants’ view on the Spanish spoken in the Caribbean (item 3) underlines the importance of education in speaking (2), whereas others mentioned the phonetic/phonological differences to more standard varieties (3). Other participants sought to highlight the practical aspects of language (4). 2

Entre menos educación más pobre es el español. Among those with less education also the Spanish [language] is poor. (male, San Juan, manager, 30–39 year old)

3

Es bastante correcto. Nos cortemos las “R”. It’s fairly correct. We cut the “R.” (female, San Juan, sales person, 50–59 years old)

4

Sí, es entendible. It’s comprehensible. (female, Fajardo, waitress, 40–49 years old)

The ratings of the native variety reveal an ambiguity whereby, on the one hand, there is an acute awareness of the perceived flaws of PRS vis-à-vis a more “correct” and abstract standard but, on the other hand, there is an emotional connection to the native variety, as observed in higher ratings on the “pleasantness” scale (see also Mojica de León, 2014). As van Dijk (1998) and Domínguez-Rosado (2015) point out, the roots for these perceptions can be found in social and sociocultural beliefs that influence speakers’ evaluations at a subconscious level. The Puerto Rican cultural identity narrative has always closely tied itself to its European/Spanish roots, in spite of fostering its own postcolonial identity (Duany, 2002). An important aspect to this identity is the language, and for many, the Puerto Rican vernacular is an emblem of their identity. Simultaneously, their perceptions of English and bilingualism are highly negative, at least in part due to a failed attempt by the US government to impose English as an official language in 1948 (Duany, 2002, p. 19). Therefore, it is not surprising that the current belief system also includes the perceived negative influence of English on their native variety, a trend mostly associated with younger speakers’ use of code-switching and anglicisms, called “Spanglish”: 5

No, últimamente la juventud no ha dado importancia a las leyes de la lengua española. También tienen muchas mezclas con el anglicismo y/o Spanglish. No, lately, the youth have not paid much attention to the importance of the rules of the Spanish language. They also have [use] many mixtures with anglicisms and/or Spanglish. (female, Puerto Rico, sales person, 20–29 years old)

There is evidence that across generations, attitudes appear to be changing toward a growing acceptance of English in Puerto Rico. Especially among more highly

175

Puerto Rican evaluations of Spanish

175

educated, young urban Puerto Ricans, the Puerto Rican identity appears to “unlink” from the exclusive use of Spanish and inclusion of knowledge and use of English in conjunction with Spanish (Domínguez-Rosado, 2015, p. 76; Loureiro-Rodríguez et al., 2016). Of course, the discussion of language in Puerto Rico often connects the realm of linguistic identity with the local political situation, its relationship with the United States, and political self-determination (for a more detailed discussion, see, e.g., Domínguez-Rosado, 2015; Valdez, 2016). On the island, national discourse toward the diaspora has been negative and critical of émigrés’ definition of the Puerto Rican identity. A key factor here is the immigrants’ use of language, specifically the perceived negative effects of bilingualism with English among second- and third-generation Puerto Ricans living on the US mainland. There is the worry that English may also contaminate Spanish on the island and therefore threaten the linguistic identity there (e.g., Duany, 2002; Zentella, 2007). However, recently, there is evidence of second- and third-generation Puerto Ricans living on the mainland claiming the Puerto Rican identity while exhibiting varying degrees of fluency in Spanish, which has forced a reevaluation of their attitudes toward bilingualism and the Puerto Rican identity on the island (Domínguez-Rosado, 2015). Important in the context of Puerto Rican self-perceptions is also their evaluations of other varieties in comparison to their own. It was hypothesized that sociohistorical beliefs would influence the participants’ ratings. This is particularly true for Spain, which remains for many the “mother country,” the place of origin of the language and as such must speak the most correct form of the language: 6

Asumo que España por ser la madre patria y mucho de nuestro lenguaje es heredero de ellos. I assume Spain, as it is the mother country and much of our language comes from them. (male, San Juan, cafeteria manager, 30–39 years old)

Spain retains a historical importance in the minds of many speakers, not just the current speaker pool (e.g., Alfaraz, 2002; Orozco, 2010). Zentella (2007, p. 28) in her examination of in-group versus out-group dynamics among Latinos in the US mainland highlights the perceived superiority of Spanish in Spain and of the highland South American dialects and the negative influence of English on Spanish as main factors in Latin American national identity. Specifically, the author argues that the belief in a linguistic norma culta in the minds of speakers often correlates with perceived economic, racial, and ethnic divides in the speakers’ homelands. This, she states, could contribute to Puerto Ricans’ negative evaluations of their own variety (Zentella, 2007, p. 28). The current data support the conclusion that these evaluations are not a by-product of immigration but instead exist already in the narrative of the country of origin. Furthermore, a contributing factor in the evaluations of Spanish from Spain, and also some of the other higher rated varieties, such as Mexican Spanish,

176 Eva-María Suárez Büdenbender

176

Argentinian Spanish, and Colombian Spanish, could be Puerto Ricans’ exposure to these varieties through popular media and television (e.g., telenovelas, movies, etc.). Much like in the case of Cuban and Dominican Spanish (e.g., Duany, 2002, 2005), exposure to these varieties fosters linguistic and nonlinguistic associations and access, albeit a virtual one, to these varieties could allow Puerto Ricans to gain a more nuanced understanding between the linguistic features of these varieties and their speakers. A limited knowledge of or experience with varieties spoken in countries such as Honduras, Paraguay, and El Salvador could explain the lack of comments (positive or negative) on these varieties or their speakers and the neutral ratings given to them in both dimensions (around 4, the most neutral response in the 7-point Likert scale). Additionally, perceptions of economic strength and higher levels of education have been shown to correlate with linguistic attitudes toward varieties spoken in these countries (Crowley, 2003; Lippi-Green, 1997). In fact, the highest ranked countries on “correctness” and “pleasantness” of language have some of the highest GDPs among Spanish-speaking countries, a correlation previously pointed out by Alfaraz (2002, 2014). However, these evaluations appear to not take into account shifts in economic power, as Colombia and highly ranked Argentina have declined considerably in comparison to other countries, such as Mexico, Chile, and Panama, all of which have outranked Argentina and Colombia (e.g., World Economics, 2012). However, the prestige of Argentinian and Colombian Spanish has not suffered, pointing to the fact that linguistic prestige can remain stable for some time, possibly reflecting past economic realities. The results do highlight work from the area of sociophonetics, where social information about a speaker or group of speakers can influence their evaluations on said variety and can emerge simply by context (conversational context in this study) (e.g., Babel, 2014; Hay, Warren & Drager, 2006). These connections and associations can also help to make sense of the ratings given to Cuban and Dominican Spanish. Puerto Ricans are well acquainted with both varieties, as Cubans and, more recently, Dominicans form large immigrant groups on the island. Whereas Cubans are seen as industrious but selfish and not trustworthy, Dominicans and Dominican Spanish have been depicted as dumb and lazy (Duany, 2002, 2005). In particular, the evaluation of Dominican Spanish is not unique to Puerto Rico. Dominican Spanish is cross-dialectally disparaged (e.g., Alfaraz, 2002, 2014; Duany, 2002, 2005; Mejía Pardo, 1993; Mojica de León, 2014; Suárez Büdenbender, 2013), a fact largely attributed to the economic weakness of the Dominican island, the immigrants’ relatively low levels of education and resulting largely blue-collar employment (Duany, 2002, 2005). The low evaluations specifically of Cuban and Dominican Spanish also evidence a form of disassociation from more disparaged, albeit linguistically similar, varieties. Much like the positions of previous researchers (Alfaraz, 2002, 2014; Lamboy, 2011), it is posited here that the disassociation with Cuban and Dominican Spanish in the present study is motivated by an awareness of stigma attached to both varieties, though especially to Dominican Spanish, which is based on perceived

177

Puerto Rican evaluations of Spanish

177

socioeconomic and educational differences. It is also possible that the disassociation is further motivated by another non-linguistic element, an attempt to distance ones’ self from speakers perceived to be phenotypically darker. Previous work has pointed to perceptions of racial difference as another underlying factor that could influence perceptions (e.g., Duany, 2002, 2005; Mejía Pardo, 1993; Mojica de León, 2014; Suárez Büdenbender, 2013). Specifically, Mejía Pardo (1993) points out that criticism of Dominican Spanish abounds in her study on Puerto Rican attitudes toward the immigrant group. But, whereas outspoken criticism of the groups’ education, social status, or even race or phenotype is not socially acceptable, criticism and at times ridicule of Dominican Spanish is. This evaluation is reinforced by Duany (2002, 2005), who emphasizes the stigmatization of the black population in Puerto Rico and the strong anti-Dominican discourse on the island. Race and perception thereof remain ambiguous, where the dominant narrative regards the Puerto Rican nation as a racial melting pot and racial prejudice as nonexistent. Nevertheless, at the heart of the Puerto Rican ideal is still the white male peasant from the inner island, el jíbaro, and the importance of African or Taíno heritage remain an afterthought (Duany, 2002). It is therefore not surprising to find these beliefs mirrored in the evaluations on this study, where countries with perceived stronger European features like Spain and Argentina receive high ratings and countries known to have stronger African heritage can be found toward the lower end of the ratings. Nor are these results unique. Previous work has found similar evaluations in both dimensions (“correctness” and “pleasantness”) among speakers in Veracruz, Mexico (Orozco, 2010), in Venezuela (Coello Millán, 2014), and in Costa Rica (Jara Murillo, 2006). Alfaraz completed the most extensive analysis in her studies (2002, 2014) and found ethnic composition to be a salient factor in Miami Cubans’ perceptions of varieties of Spanish, who favored European populations, whereas regions with stronger Indigenous or African ancestry, like the Andean regions or the Caribbean, were rated low. Specifically, speech associated with speakers of African origin (Afro Cubans, Puerto Ricans, and Dominicans) received low ratings in this study. Undoubtedly, the tendency in Puerto Rico and other Caribbean nations (e.g., Duany, 2002, 2005) to “whiten” the national identity in an effort to bind the nation and its peoples to the former European colonial nation comes at the price to suppress the value of African and Indigenous contributions. This devaluation is mirrored in the speakers’ preference for particular linguistic varieties over others.

6. Conclusion In sum, there exists a robust linguistic security among PRS speakers in San Juan. Puerto Ricans appreciate their native tongue and hold it in high regard. An important factor in Puerto Rican self-evaluations is the ambiguous relationship to bilingualism with English, undoubtedly due to the complex relationship with the United States and ongoing debates about the political status of the island. Negative evaluations of “Spanglish” appeared frequently in the data, and although little

178

Eva-María Suárez Büdenbender

178

work has examined Puerto Ricans’ thoughts on and evaluations of bilingualism, such as attitudes toward borrowing, code-switching, and other contact phenomena, there is evidence of a slow acceptance of bilingualism with English on the island. Also, further study merits the possible distancing of the Puerto Rican participants from other Caribbean speaker groups, a finding previously observed among Cuban and Dominican Spanish-speaking communities in the United States. Likewise, the study has been able to add to our understanding of the correlations between socioeconomic status, level of education, ethnic composition, and linguistic prestige. It appears that the latter remains stable in spite of possible economic upheavals in the countries where they are spoken, but these findings require further examination. Moreover, the present results also contribute to the discussion on the correlations between perceptions of ethnic composition and linguistic perceptions, helping to strengthen the argument that linguistic evaluations are the more socially acceptable avenue to express social and racial prejudice.

Notes 1 The weakening of /s/ in coda position is not a process limited to Caribbean Spanish but that has been observed in countless countries and regions (e.g., Spain, the Pacific Coast from Colombia to Chile, Argentina, etc. (Lipski, 1985b, 1994)). 2 Likewise, the velarization of /n/ is not limited to the Caribbean and has been found to occur in a wide variety of other regional variants of Spanish, including varieties spoken in Latin America (e.g., Klee & Lynch, 2009; Lipski, 1994). 3 Overall, 87 participants took part in this study. However, due to incompleteness of the questionnaires, the data of ten participants had to be excluded from the study.

References Alfaraz, G. (2002). Miami Cuban perceptions of varieties of Spanish. In D. Long & D.R. Preston (Eds.), Handbook of perceptual dialectology: Volume 2 (pp. 1–13). Amsterdam: Benjamins. Alfaraz, G. (2014). Dialect perceptions in real time: A restudy of Miami-Cuban perceptions. Journal of Linguistic Geography, 2, 74–86. Almeida, M. (1990). Diferencias sociales en el habla de Santa Cruz de Tenerife. La Laguna: Instituto de Estudios Canarios. Aponte Arlequín, H.A. (2014). Desafíos del español caribeño: el debate sobre el modo y la microvariación modal. Unpublished PhD dissertation, University of Puerto Rico (Río Piedras). Babel, A. (2014). Stereotypes versus experience: Indexing regional identity in Bolivian Valley Spanish. Journal of Sociolinguistics, 18(5), 604–633. Coello Millán, H.Y. (2014). Actitudes lingüísticas en Venezuela. Exploración de creencias hacia la variante nacional, la lengua española y el español dialectal. Bergen Language and Linguistics Studies, 5, 1407–1532. Crowley, T. (2003). Standard American English and the politics of language. Basingstoke: Palgrave, Macmillan. d’Anglejan, A., & Tucker, R.G. (1983). Sociolinguistic correlates of speech style in Quebec. In R.W. Shuy & R.W. Fasold (Eds.), Language attitudes: Current trends and prospects (pp. 1–27). Washington, DC: Georgetown University Press.

179

Puerto Rican evaluations of Spanish

179

Dauphinais, A., & Ortiz López, L. (2013). Word order and NP characteristics in Cuban Spanish: Pragmatic and sociolinguistic variation. Paper presented at the Hispanic Linguistics Symposium 2013, University of Ottawa, Ottawa, Canada. Delgado-Díaz, G., & Galarza, I. (2014). Implicaciones sociolingüísticas en la percepción: El caso de la /r/ posterior en el español de Puerto Rico. Paper presented at the Hispanic Linguistics Symposium 2014, University of Indiana, Bloomington, IN. Docherty, G., Langstrof, C., & Foulkes, P. (2013). Listener evaluation of sociophonetic variability: Probing constraints and capabilities. Linguistics, 51(2), 355–380. Domínguez-Rosado, B. (2015). The unlinking of language and Puerto Rican identity: New trends in sight. Newcastle Upon Tyne: Cambridge Scholars Publishing. Drager, K. (2006). Social categories, grammatical categories, and the likelihood of “like” monophthongization. In Proceedings of Australian international conference on speech science and technology (pp. 384–387). Auckland: Auckland University Press. Duany, J. (2002). The Puerto Rican nation on the move: Identities on the island and in the United States. Chapel Hill, NC: The University of North Carolina UP. Duany, J. (2005). Dominican migration to Puerto Rico. A transnational perspective. Centro Journal, 17(1), 475–511. City University of New York, Centro de Estudios Puertorriqueños. Foulkes, P., & Docherty, G.J. (2006). The social life of phonetics and phonology. Journal of Phonetics, 34, 409–438. Gardener-Chloros, P. (1991). Language selection and switching in Strasbourg. Oxford: Oxford University Press. Hay, J., & Drager, K. (2007). Sociophonetics. Annual Review of Anthropology, 36, 89–103. Hay, J., Warren, P., & Drager, K. (2006). Factors influencing speech perception in the context of a merger-in-progress. Journal of Phonetics, 34, 458–484. Hoenigswald, H. (1966). A proposal for the study of folklinguistics. In W. Bright (Ed.), Sociolinguistics (pp. 16–26). The Hague: Mouton. Hoffman, M., & Walker, J.A. (2010). Ethnolects and the city: Ethnic orientation and linguistic variation in Toronto English. Language Variation and Change, 22(1), 37–67. Holmquist, J.C. (2003). Coffee farmers, social integration and five phonological features: Regional socio-dialectology in west-central Puerto Rico. In L. Sayahi (Ed.), Selected proceedings of the first workshop on Spanish sociolinguistics (pp. 70–76). Somerville, MA: Cascadilla Proceedings Project. Jara Murillo, C.V. (2006). El español de Costa Rica según los ticos: un estudio de lingüística popular. San José, CR: Editorial UCR. Jara Murillo, C.V. (2008). Lingüística popular: el español de Costa Rica según los ticos y algunos centroamericanos residentes en el país. Revista internacional de lingüística iberoamericana, 11, 55–99. Klee, C.A., & A. Lynch. (2009). El español en contacto con otras lenguas. Washington, DC: Georgetown University Press. Lambert, W., Hodgson, R., Gardener, R., & Fillenbaum, S. (1960). Evaluational reactions to spoken language. Journal of Abnormal and Social Psychology, 60, 44–51. Lamboy, E.M. (2011). Language and identity construction: Can we talk about a New Puerto Rican in the United States? In L.A. Ortiz-López (Ed.), Selected proceedings of the 13th Hispanic linguistics symposium (pp. 70–80). Somerville, MA: Cascadilla Proceedings Project. Lippi-Green, R. (1997). English with an accent: Language, ideology, and discrimination in the United States. London and New York: Routledge.

180

Eva-María Suárez Büdenbender

180

Lipski, J.M. (1985b). Creole Spanish and vestigial Spanish: Evolutionary parallels. Linguistics, 23, 963–984. Lipski, J.M. (1994). Latin American Spanish. London and New York: Longman. Lipski, J.M. (2008). Varieties of Spanish in the United States. Washington, DC: Georgetown University Press. Lloréns, W. (1971). El habla popular de Puerto Rico. Río Piedras: Editorial Edil, 2nd edition. Long, D., & Preston, D.R. (Eds.). (2002). Handbook of perceptual dialectology (Vol. 2). Amsterdam: Benjamins. López Morales, H. (1992). El español del caribe. Editorial Mapfre: Madrid. Lorenzo Ramos, A. (1976). El habla de los Silos. Santa Cruz de Tenerife: Caja General de Ahorros. Loureiro-Rodríguez, V., Guzzardo Tamargo, R., & Vélez Avilés, J. (2016). Actitudes de estudiantes universitarios en torno a las lenguas en Puerto Rico. Presentation at the 8th International Workshop of Spanish Sociolinguistics, San Juan, Puerto Rico, April 2016. Mack, S. (2010). Perception and identity: Stereotypes of speech and sexual orientation in Puerto Rican Spanish. In C. Borgonovo, M. Español-Echevarría, & P. Prévoste (Eds.), Selected proceedings of the 12th Hispanic linguistics symposium (pp. 136–147). Somerville, MA: Cascadilla Proceedings Project. Mack, S. (2011). A sociophonetic analysis of /s/ variation in Puerto Rican Spanish. In L.A. Ortiz-López (Ed.), Selected proceedings of the 13th Hispanic linguistics symposium (pp. 81–93). Somerville, MA: Cascadilla Proceedings Project. Mejía Pardo, D. (1993). Macroestructuras, superestructuras y proposiciones de opiniones en 17 relatos de puertorriqueños acerca de dominicanos. Unpublished M.A. thesis, Graduate Program of Linguistics, University of Puerto Rico, Río Piedras. Mojica De León, C.M. (2014). Una mirada hacia las actitudes lingüística en Puerto Rico. Bergen Language and Linguistics Studies, 5, 1249–315. Navarro Tomás, T. (1948). El español en Puerto Rico. Río Piedras: Editorial Universitaria. Niedzielski, N. (1999). The effect of social information on the perception of sociolinguistic variables. Journal of Language and Social Psychology, 18, 62–85. Niedzielski, N., & Preston, D.R. (2003). Folk linguistics. New York: Mouton de Gruyter. Orozco, R. (2010). Veracruz speech perceptions: A preliminary study. Paper presented at NWAV 39, November 2010, San Antonio, TX. Orozco, R. (2015). Pronominal variation in Colombian Costeño Spanish. In A.M. Carvalho, R. Orozco, & N. Lapidus Shin (Eds.), Subject pronoun expression in Spanish: A crossdialectal perspective (pp. 17–38). Washington, DC: Georgetown University Press. Otheguy, R., & Zentella, A.C. (2012). Spanish in New York: Language contact, dialectal leveling, and structural continuity. Oxford and New York: Oxford University Press. Otheguy, R., Zentella, A.C., & Livert, D. (2007). Language and dialect contact in Spanish in New York: Toward the formation of a speech community. Language, 83, 770–802. Paltridge, J., & Giles, H. (1984). Attitudes toward speakers of regional accents of French: Effects of regionality, age, and sex of listeners. Linguistische Berichte, 90, 71–85. Preston, D.R. (1989). Perceptual dialectology. Dordrecht, The Netherlands: Foris. Preston, D.R. (Ed.). (1999). Handbook of perceptual dialectology (Vol. 1). Amsterdam: Benjamins. Preston, D.R. (2002). Language with an attitude. In J.K. Chambers, P. Trudgill, & N. Schilling-Estes (Eds.), The handbook of language variation and change (pp. 40–66). Malden, MA: Blackwell.

181

Puerto Rican evaluations of Spanish

181

Saciuk, B. (1977). Las realizaciones múltiples o polimorfismo del fonema /y/ en el español puertorriqueño. Boletín de la Academia Puertorriqueña de la Lengua Española, 5, 133–153. Serverino Cerda, G.M. (2014). Actitudes lingüísticas en República Dominicana. Conciencia e identidad lingüísticas en la ciudad de Santo Domingo. Bergen Language and Linguistics Studies, 5, 1316–1345. Sobrino Triana, R., Montero Bernal, L.E., & Menéndez Pryce, A.J. (2014). Actitudes lingüísticas en Cuba. Cambios positivos hacia la variante nacional de lengua. Bergen Language and Linguistics Studies, 5, 290–408. Suárez Büdenbender, E.M. (2013). ‘Te conozco bacalao’—Investigating the influence of social stereotypes on linguistic attitudes. Hispania, 96(1), 110–134. U.S. Census Bureau. American Community Survey, estimates for 2010. Retrieved from https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=CF Valdez, J.R. (2016). The battleground of metaphors: Language debates and symbolic violence in Puerto Rico (1930–1960). Journal of Historical Sociolinguistics, 2(1), 1–25. Van Dijk, T.A. (1998). Ideology: A multidisciplinary approach. London: Sage. World Economics. (2012). Measuring Latin American GDP. Retrieved from www.worldeconomics.com. Zentella, A.C. (1997). Growing up bilingual: Puerto Rican children in New York. Malden, MA: Blackwell. Zentella, A.C. (2002). Latina/o languages and identities. In M. Suárez-Orozco & M. Páez (Eds.), Latinos: Remaking America (pp. 321–338). Berkeley, CA: University of California Press. Zentella, A.C. (2007). ‘Dime con quién hablas, y te diré quién eres’: Linguistic (in)security and Latina/o unity. In J. Flores & R. Rosaldo (Eds.), Companion to Latina/o studies (pp. 25–37). Malden, MA: Blackwell Publishing Ltd.

Appendix Questionnaire A (alphabetical order of countries)*

*Please be aware that in half of the questionnaires the presentation of countries was given in counter alphabetical order. FAVOR DE CONTESTAR LAS SIGUIENTES PREGUNTAS (A) No _____ 1

¿De qué ciudad, provincia es Ud.? _____________________________________________________________

2

Sexo

F&

3

Edad

20–29 &

4

¿Cuál es el grado escolar más alto que completó? 1º–3º &

5

M&

4º–6º &

30–39 &

7º–9º &

40–49 &

10º–12º &

50–59 &

60+ &

Universidad &

Si usted trabaja, ¿en qué trabaja? __________________________________

6

¿Le parece que el español que se habla en Puerto Rico es muy correcto? ¿Por qué? ________________________________________________________________ 7

¿Le parece que el español que se habla en el Caribe es muy correcto? ¿Por qué? ________________________________________________________________ 8 ¿En qué país se habla el español más correcto? ¿Por qué? ________________________________________________________________ 9

¿Cómo es diferente el español de Puerto Rico del español que se habla en otros países? ________________________________________________________________ (This portion adopted from Alfaraz, 2002)

183

Puerto Rican evaluations of Spanish

183

¿Cuán correcto le parece el español que se habla en los siguientes países? más correcto 7

6

menos correcto 5

4

3

2

1

Países Argentina ______

Honduras ______

Bolivia ______

México ______

Chile ______

Nicaragua ______

Colombia ______

Panamá ______

Costa Rica ______

Paraguay ______

Cuba ______

Perú ______

Ecuador ______

Puerto Rico ______

El Salvador ______

Rep. Dominicana ______

España ______

Uruguay ______

Guatemala ______

Venezuela ______

¿Cuán agradable le parece el español que se habla en los siguientes países? más agradable 7

6

menos agradable 5

4

3

2

1

Países Argentina ______

Honduras ______

Bolivia ______

México ______

Chile ______

Nicaragua ______

Colombia ______

Panamá ______

Costa Rica ______

Paraguay ______

Cuba ______

Perú ______

Ecuador ______

Puerto Rico ______

El Salvador ______

Rep. Dominicana ______

España ______

Uruguay ______

Guatemala ______

Venezuela ______

10 Aquí no se cogen las guaguas Language and Puerto Rican identity in San Diego Ana Celia Zentella

1. Introduction When my Mexican father and Puerto Rican mother met and married in NYC in the 1920s, they belonged to a small and diverse group of Latin American immigrants; neither could count on many compatriots. Only one thousand Puerto Ricans lived on the United States mainland in 1910, most in NYC (Zentella, 2000), and the number of Mexicans in the Big Apple was negligible. Today, LatinUs1 constitute 29.4% of all New Yorkers (8.5 million); Dominicans, at 29% of the LatinU population, have recently eclipsed Puerto Ricans, who account for 28%, while Mexicans account for 15% (Bergad, 2016, p. 7). In cities across the United States, diverse groups of Spanish-speaking immigrants are in close contact and raising children of mixed parentage, like my Puerto-Mex/Mexirican sister and me. These communities offer intriguing possibilities for research, because while studies of languages in contact around the world abound, Bonnici and Bayley (2010) found that contact between the varieties of Spanish spoken in the United States and Canada has been understudied. Escobar and Potowski (2015) and Erker (2018) summarize and compare the research to date. We still lack complete answers to questions raised by Bayley et al. (2012) concerning the possible formation of a Spanish koine— that is, “a stabilized contact variety which results from the mixing and subsequent leveling of features of varieties which are similar enough to be mutually intelligible” (Kerswill, 2002, p. 671). Among the most salient questions is the following: to what extent, for example, are the various dialects of Spanish found in major US cities converging? To what extent do speakers of a Spanish dialect that diverges from the variety spoken by a majority of Latinos in a particular community shift in the direction of the predominant variety in the area? Are cases of dialect convergence transitory phenomena that are likely to be eclipsed as subsequent generations shift to English or does the evidence suggest that more stable varieties are likely to develop? (Bayley et al., 2012, p. 48) Seeking answers, Otheguy and I conducted the largest-scale study of Spanish in the United States to date, investigating the rate of—and constraints on—the presence

185

Puerto Rican identity in San Diego

185

or absence of subject personal pronouns (SPPs) in six groups of LatinUs in NYC (Otheguy & Zentella, 2012). Unlike English, Spanish allows for including or omitting I/you/he she/it/we/they, with some exceptions. In the Caribbean and other regions settled by Andalusians, where the deletion of syllable-final s is common, rates of overt subject pronouns outpace the rates of inland areas where the pronoun is not needed to disambiguate the subject of verbs that lose their syllable-final s, such as to distinguish tú comes (you familiar) from yo/Ud, él, ella come (I/you formal eat, he, she eats). An analysis of 65,000 tokens in the speech of 140 Puerto Ricans, Dominicans, Cubans, Colombians, Ecuadorans, and Mexicans revealed that SPP rates increase with time spent in NYC in all groups, and some SPP constraints that distinguish Caribbean and Latin American Mainland Spanish varieties diminish across the generations. Long-term immigrants and their children employed SPPs in ways that moved closer to that of their noncompatriotas and reflected the influence of English, suggesting the formation of an NYC variety of Spanish, although Erker (2018) points out that “unraveling the respective effects” of language and dialect contact “is very likely impossible,” especially regarding assimilation of non-stigmatized features like SPPs.

2. Puerto Rican and Mexican contact in the Midwest and Southwest of the United States Our NYC speakers were from different neighborhoods and had varied educational, class, and racial backgrounds. The likelihood of koineization increases when groups share similar backgrounds, neighborhoods, workplaces, churches, and schools for decades. But the most extensive analyses of the Spanish of working-class Puerto Ricans and Mexicans who have long been in close contact, as has been true for 60 years in Lorain, Ohio, and Chicago, Illinois, discovered that neither group adopts the other’s ways of speaking. In Lorain, where the Puerto Rican community eclipsed the older Mexican community in the 1980s, Ramos-Pellicia (2014) found that old tensions maintain boundaries; lexical borrowing and phonological convergence are avoided due to powerful ideologies regarding the inferiority of the other group’s speech and its English-influenced “corruption of the Spanish language” (p. 43). Similarly, Gosh Johnson’s study of Puerto Ricans and Mexicans in a Chicago high school revealed that despite contact, essentialized ethnic identities—that is, the construction of Mexicans as non-assimilated immigrants vs. Puerto Ricans as “modern citizens”—proved too salient a boundary to interaction and the formation of integrated social networks for dialect leveling to occur. Analysis of syllable-final /r/ and syllable-final /s/ indicated “no linguistic convergence occurring” (Johnson, 2005, p. 126). In contrast, the Mexican and Puerto Rican high schoolers that Rosa (2019) studied in Chicago shared integrated and deeply intimate social networks, characterized by joking about features of each other’s Spanish to mark distinctions. Although they sometimes adopted the others’ stereotypical vocabulary and other linguistic features knowingly or unknowingly, they were more united by a “strong investment in what was ideologically constructed as unaccented

186 Ana Celia Zentella

186

English” (Rosa, 2015, p. 41). For these students, identities were shared in relation to English rather than Spanish. Another investigation of contact between Puerto Rican Spanish (hereafter PRS) and Mexican Spanish (hereafter MS) in Chicago compared three generational levels: G1 = immigrant, arrived 9 years ago or older; G2 = Chicago born or arrived by age six with G1 parents; G3 = Chicago born, with one or both G2 parents. Potowski and Torres (forthcoming) asked 39 Puerto Ricans and 37 Mexicans to identify ten pictures to determine whether they produced the MS or PRS variant for orange-fruit (naranja/china), pacifier (chupete/bobo), bus (camion/autobús/guagua), eyeglasses (lentes/espejuelos, gafas, anteojos), earrings (aretes/pantallas), swimming pool (alberca, piscina), drinking straw (popote/ sorbeto), banana (plátano, guineo), kidney beans (frijoles/habichuelas), and cake (pastel/bizcocho). Interviewees were also asked whether they knew another word for each item, and who used that word. Puerto Ricans knew more Mexican vocabulary (8.89) than Mexicans knew Puerto Rican vocabulary (3.08). The authors attribute this disparity to the numerical superiority of Mexicans; in 2010, Mexicans accounted for 79.7% of Chicago’s 779,000 LatinUs, whereas Puerto Ricans constituted only 10.2% (Potowski, 2016, pp. 31–33). They also acknowledged the impact of television, music, and movies in MS. The widespread influence of MS was corroborated when Puerto Ricans on the island knew more Mexican vocabulary (4.67) than Mexicans in Mexico knew Puerto Rican vocabulary (0.80). On the other hand, when the three generations of Puerto Ricans in Chicago were compared, the generation with the most exposure to the other community (G3) did not know significantly more “outsider” items than their immigrant (G1) counterparts, because in their generation, English was replacing Spanish. Even if a Chicago Mexirican dialect were being forged, it would not be shared by the third generation. In other predominantly Mexican areas of the United States, such as Texas and California, Puerto Ricans comprise an even smaller segment of the LatinU community and have yet to create identifiable barrios. They live amid overwhelming numbers of Mexicans who don’t understand when they ask, “¿Dónde puedo coger la guagua?” (“Where can I catch the bus?”). What are the effects of linguistic and cultural swamping on the ways of speaking and identities of Puerto Ricans far from their primary island and mainland bases? In San Antonio, Texas, where most (63%) residents are Mexican, only 1% are Puerto Ricans. In such a setting, it seems logical that PRS would change in the direction of MS, but Bayley and colleagues (2012) discovered that social and professional networks conditioned speakers’ choices regarding null and overt pronouns. Seven of ten Puerto Ricans expressed pronouns at the same rate as compatriots in San Juan and New York (45%), but they belonged to predominantly PRS-speaking networks. In contrast, the three Puerto Ricans with predominantly Mexican or Mexican American networks used overt SPPs “at a rate (23%) that is similar to the rate of the Mexican immigrant and Mexican American speakers reported in San Antonio and to the rate reported for Mexican American speakers in other studies” (Bayley et al., 2012, p. 58).

187

Puerto Rican identity in San Diego

187

The San Antonio study and those conducted in Chicago prove that when Puerto Ricans and Mexicans live in the same neighborhood or attend the same school, social contact does not necessarily result in linguistic accommodation or even extensive knowledge of the other community’s ways of speaking, unless interaction is intense, as with in-group members. Potowski’s 2016 study of MexiRicans (the children of one Mexican parent and one Puerto Rican parent) in Chicago offers further corroboration. Exposure to both dialects provided MexiRicans with “more familiarity with lexical items from both dialects than . . . generationally matched Mexicans and Puerto Ricans in Chicago” (p. 255). Nevertheless, there was no evidence of koineization: their Spanish “showed a great deal of variability, pointing away from the existence of a set of features as . . . a ‘MexiRican dialect’” (p. 255). Similarly, my sister and I, who know the words to both La Borinqueña and the Himno Nacional de México and who danced in the Ballet Folklórico Mexicano de NY on the same weekends we danced salsa at Club Broadway, can talk at ease with Mexicans or Puerto Ricans, sometimes sounding like them, depending on the extent to which we identify with them. In sum, the variables that determine linguistic leveling are more numerous and complicated than calculating numerical superiority or extent of interaction. As I noted in my 1990 study of 25 lexical variations in the four leading varieties of Spanish in NYC at the time (Puerto Rican, Dominican, Cuban, Colombian), “a number of social and economic realities . . . impinge upon communication and linguistic change, leveling, and/or diffusion” (Zentella, 1990, p. 1097). The Cubans and Colombians did not adopt or know many words shared by Puerto Ricans and Dominicans, although they were the most frequently heard, because the lower educational and income levels of Puerto Ricans and Dominicans—and racial discrimination against them—resulted in negative attitudes toward their Spanish. Studies of Spanish in contact must take into account the racism and classism that shape attitudes toward distinct dialects, as well as the political views of speakers that may also influence borrowing and accommodation. These variables were relevant in our study of PRS in San Diego.

3. Puerto Ricans in San Diego The 20,468 Puerto Ricans that constitute under 1% of San Diego County’s population (3,095,313) live 30 miles away from the Mexican border and among 600,000 Mexicans; in the city of San Diego, 0.6% of residents are Puerto Rican (8,200) (Rodriguez Ayuso, Geerman Santana, & Marazzi Santiago, 2013). In 94 individual interviews and questionnaires, I probed Puerto Rican attitudes toward their linguistic context and the extent to which they were aware of the Mexican lexicon.2 Of particular interest are the familial, generational, economic, and political variables that make most Puerto Ricans continue to speak in traditional ways. Puerto Ricans in the west differ markedly from those in Chicago and traditional northeast enclaves: “Those who left the city for regions westward . . . ranked fairly highly [in] education, English language fluency, New York City birth, white-collar occupations, and household income” (Marzán, Torres, & Luecke,

188 Ana Celia Zentella

188

2008, p. 10). In San Diego, Puerto Ricans have established vibrant cultural organizations, including La Casa de Puerto Rico, founded in 1972, although there were only 578 Puerto Rican residents in 1970 (López, 2010, p. 1). The number of Puerto Ricans has increased by 38% since 2000 (Rodriguez Ayuso et al., 2013), and many are thriving: 30% have a BA/BS and 75% have, at minimum, a high school diploma; the median family and household income is US$42,000, and slightly less than half (48%) speak Spanish at home (López, 2017). Another distinguishing feature, and one which brought many Puerto Ricans to San Diego, is their U.S. Navy background: 14% are veterans. The strength of the reincorporated House of Puerto Rico San Diego (HPRSD) in 1988 is apparent in the construction of its cottage/museum in the city’s prominent Balboa Park in 2006, and its numerous activities (House of Puerto Rico). In addition, the bomba and plena musicians, singers, and dancers of another group, originally Areito Borincano (AB) and now Bomba Liberté, have helped Puerto Rican culture thrive in San Diego since 2001. Our 94 interviewees participated in either HPRSD or AB. Here is a summary of their backgrounds:3 1 sex [n = 94]: 59% male, 41% female 2 age [n = 90]: ranging from 15 to 72 years; majority (64%) 39–59 years old 3 place of birth [n = 94]: 52% b. PR; 34% b. NYC; 14% b. CA 4 age of arrival to the United States if born elsewhere [n = 58]: 83% arrived at/ before 25 years old 5 number of years in CA [n = 88]: majority (53%) have lived in CA < 20 years; 16% for 31–44 years 6 frequency of travel to PR [n = 90] 62% visit every 1–3 years 7 education [n = 92]: 77% at least some college; 19% some HS; 4% elementary 8 country where educated [n = 90]: 75% at least some education in the United States 9 occupation [n = 80] a b c d 10 11

42% = skilled, blue collar, clerical 40% = professional, business owner, white collar 10% = unskilled 8% = graduate students or military

household income [n = 87]: majority (85%) = $30–70,000>; 5% = < $15,000. language proficiency [n = 94]: 91% understand, speak, read, and write English “very well” or “as if it were their first language” a

12

proficiency in understanding, speaking, reading and writing Spanish is 5–17% less than their proficiency in English

language use [n = 94]: most (62%) spoke Spanish to their parents, and 72% speak Spanish frequently in San Diego

189

Puerto Rican identity in San Diego a

13

189

speak 11% more Spanish with Mexicans than with Puerto Ricans, and bilingualism dominates their interactions at work (53%), with children (46%), and with spouse (46%)

lived with a Mexican [n = 89]: most (52%) have lived with a Mexican for 1–35 years; 40% for ten years or more

A slight majority of our informants (52%) were born in PR, and the great majority (83%) had migrated by age 25. Most (64%) were mature adults in their 40s and 50s, half (53%) had lived in California for fewer than 20 years, and they were well educated: three-quarters had attended college and had some education in the United States. The majority worked as professionals or business owners, or they were graduate students or in the military, and they earned $30–70,000 per year. Most visited PR every few years, but in San Diego, they spoke more Spanish to Mexicans than to Puerto Ricans. Bilingualism characterized their interactions at work and home, and they rated their English proficiency higher than their Spanish proficiency. Of major interest was the fact that the majority (52%) had lived with a Mexican (40% for ten years or more), but this did not diminish—and may even contribute to—the strength of their commitment to PRS.

4. Puerto Rican knowledge of the Mexican lexicon Because most considered vocabulary the primary difference between PRS and MS, I requested translations of 22 popular Mexican expressions/phrases and the Mexican way of saying 13 Puerto Rican items. In both tasks, the majority were unable to answer correctly. Table 10.1 indicates that between 4% and 70% of the Puerto Ricans whom we interviewed knew how to translate one or more of 22 Mexican expressions/phrases, but the majority knew only 32% (7/22) of this list. Most (59%) of our interviewees were males, which explains why the majority of the group knew #16 vato); more than twice as many men (n = 33) as women (n = 15) could translate it (p = 0.04061) because gender-related items are more likely learned by members of that gender, even outsiders. For similar reasons, significantly more women (n = 19) than men (n = 11) knew what #9 se alivió means (p = 0.0092570). Other female-related words related to jewelry (#2 arracadas), clothing (#3 huanga), and little children (#7 morros) were unknown to the majority, but it was surprising to learn that popular male-linked words (#5 mandilón, #14 huey), and racist and sexual labels (#6 bizcocho, #13 mayate) were unfamiliar to most, including men, as were everyday border-related terms (#8 pocho, #10 pollo, #12 chilango). As for discourse markers, popular ones (#18 órale, #19 padrísimo, #22 ándale) left their impact more than the less frequent expressions (#4 me dejaste pelón, #11 me vale gorro). The mexicanismos that most knew well had been learned the hard way; confusion and misunderstandings are effective teachers. Many reported miscommunications involving the use of #21 ahorita to mean “right now” instead of “soon,”

190

Ana Celia Zentella

190

Table 10.1 Mexican words translated into Puerto Rican words Mexican words

Puerto Rican translation

English translation

N = number # who of answered respondents correctly

% who answered correctly

1. cayó una mosca en la leche

como perro en misa

92

4

4%

2. arracadas

pantallas

unwelcome person, like a fly in the soup/ dog at mass hoop earrings

92

7

8%

3. huanga

ancho

loose fitting

92

12

13%

4. me dejaste pela’o pelón

without money/ 91 broke

18

20%

5. mandilón

sentado en el baúl

hen-pecked

88

22

25%

6. bizcocho

chocha

vagina

51

13

25%

7. morros

nenes

little children

92

26

28%

8. pocho

agringado/ americanizado

Americanize/ US-born Mexican

92

27

29%

mojado/cruzó la frontera

“wet-back”

89

27

30%

10. se alivió

dió a luz

she gave birth

92

30

33%

11. me vale gorro

me importa tres pepinos

I don’t care about that

92

37

40%

12. chilango

del Distrito Federal

from Mexico City

91

36

40%

13. mayate

un negro

black person

92

37

40%

14. huey

compai/pana

friend/“bro”

92

44

48%

15. chavala

nena/chamaca

girl

92

46

50%

16. vato

tipo

dude

92

49

53%

17. popote

sorbeto

straw

92

49

53%

18. órale

chévere

9. pollo

okay/fine

92

52

57%

19. padrísimo ‘tá cabrón brutal/chévere

very good

92

52

57%

20. coger

chichar/meter

to have sex

51

31

61%

21. ándale

dale, ok

right/go ahead

92

63

69%

22. ahorita

ahora

now

92

64

70%

191

Puerto Rican identity in San Diego

191

its customary meaning in PR and Dominican Republic (RAE, 2005, p. 33). One person was surprised when a visitor showed up unexpectedly: Ella decía que los voy a visitar ahorita y ahorita era que ya estaba en la puerta. She said I’m going to visit you SOON and SOON turned out she was already at the door. (E: Male, 41 years old, b. NY, ages 11–23 in PR, 8 years in San Diego, upper middle class) More conflicts were caused by words considered as obscenities. An undocumented woman from our church was horrified by my husband’s refusal when she offered to drive us to the airport: “O no,” Papi said, “porque los agentes de la migra te cogen” (“because the border patrol agents will catch you”); she understood she would be raped (see #20). Sometimes a misunderstanding can lead to blows, as occurred when one interviewee’s husband mistook a friendly title for an insult: Entonces viene el manager y le dice, “Mira R,” no sé qué- yo no sé qué fue lo que le dijo, y R le contestó y él le dijo, “Mira cabrón,” y vino mi esposo, sacó el puño y lo tiró al piso. “¿Por qué tú me has-? ¡Pero R!” “¡A un puertorriqueño no se le dice esa palabra, porque yo estoy seguro que mi esposa está en mi casa y ella me es fiel!”. . . . Lo corrieron, como dicen los mexicanos, lo corrieron. Then the manager comes and tells him, “Look R,” I don’t know—I don’t know what he told him, and R answered, and he said, “Look cabrón,” and my husband went and punched him and knocked him to the floor. “Why did you—? But R!” “You don’t say that word to a Puerto Rican, because I’m sure that my wife is at home and she is faithful to me!” . . . They fired him, like Mexicans say, “lo corrieron.” (G: Female, 49 years old, b. PR, 19 years old in San Diego, white collar) It cost R his job to learn that cabrón is a comradely expression among Mexican males, not meant to call out a cuckold as it often does in PR Spanish (although cabrón may also be used by Puerto Ricans males with no pejorative meaning). This anecdote also reveals that Puerto Ricans do incorporate mexicanismos like lo corrieron in their conversation, although G calls attention to its Mexican origin, perhaps for the Puerto Rican interviewer’s benefit. Table 10.2 reveals that most interviewees could only produce the Mexican version for four of 13 Puerto Rican expressions/phrases (31%). Once again, items linked to traditionally female domains, like #1 “make the bed,” were foreign to the majority, but it was surprising that two-thirds could

192

Ana Celia Zentella

192

Table 10.2 Puerto Rican words translated into Mexican words PR words

Mexican translation

English translation

N = number of respondents

# who answered correctly

% who answered correctly

1. hacer la cama

tender la cama

to make the bed

92

13

14%

2. te vemos mañana

te miramos mañana

we’ll see you 92 tomorrow

14

15%

3. a la fuerza

a lo macho

forced to

91

14

15%

4. cavetes

agujetes

shoelaces

92

15

16%

5. se ve (linda, se mira inteligente, gorda, etc.)

she looks pretty, etc.

90

23

26%

6. cerveza

chela

a beer

88

28

32%

7. chiringa

papalote

kite

92

32

35%

8. OK

ándale/ órale

alright/okay

92

40

43%

9. dame pon

raite/ aventón

(give me) a ride

92

42

47%

10. nene

chamaco/ chavo

little children

91

46

51%

11. cartera

bolsa

pocketbook

92

48

52%

12. COGER la agarrar guagua

to TAKE the 89 bus

47

53%

13. coger LA GUAGUA

to take THE BUS

54

71%

. . . el camión

76

not come up with the Mexican slang for a beer (#6 chela), shoelaces (#4 cavetes), kite (#7 papalote], a ride (#9 aventón), and the very widespread OK (#8 órale, ándale). Also, most did not know/remember how Mexicans say something was done by force (#3 a huevo/lo macho) or the Mexican use of #2 mirar(se) instead of #5 ver(se) in expressions like “s/he/it LOOKS pretty”; or “X will SEE you tomorrow/next week,” although these were surely understood in context. Cross-tabs uncovered the relevant statistically significant variables, including knowledge/use of Spanish (.0003): those who spoke Spanish with their children knew 65% of the 35 words in Tables 10.1 and 10.2, those who spoke Spanish and English with them knew 44%, and those who spoke only English knew 27%. Whether they were educated in Spanish also proved significant (0.0021): those who had studied in PR answered 53% of the questions correctly, those educated in both PR and the United States answered 36% correctly, and those who were educated only in the United States answered 29% correctly. As expected, having lived with a Mexican or speaking a lot of Spanish with Mexicans proved the most significant of the variables (0.0000): the former

193

Puerto Rican identity in San Diego

193

knew 49% of the words and the latter knew 46%. Interviewees who did not live with a Mexican only produced correct versions for 27%, those who spoke little Spanish with Mexicans knew less (22%), and those who never spoke to Mexicans knew the least (17%). Having lived with a Mexican was a powerful predictor of successful answers, but not always, as the following four interviewees proved: 1 2 3

4

J: male, 28 years old, b. NYC, six years in San Diego, working class, never lived with Mexican—did not know 91% of terms, including “ahorita” N: female, 38 years old, b. NYC, ages 18–20 in PR, 37 years in San Diego, lower working class, lived eight years with a Mexican—knew 68% of terms R: male, 55 years old, b. PR, 22 years in CA, white collar, upper middle class, Mexican wife of 13 years—did not know 67% of terms, including “ahorita” C: female, 59 years old, b. PR, ages four to 36 in NYC, 18 years in CA, lower working class, never lived with a Mexican, but church and neighborhood are Mexican—knew 74% of terms

J and N reflected the expected effect of the “lived with a Mexican” variable: J had not lived with a Mexican and did not know 91% of the words, including ahorita, whereas M, who was also born and raised in NYC but had lived with a Mexican for eight years knew 68% of them. In contrast, R, an upper middle-class male who had spent 22 years in California, 13 of them married to a Mexican, still did not know what ahorita meant. The lower-working-class woman (C), who had spent more years in NYC than PR or San Diego and had not lived with a Mexican, was familiar with three-quarters of the words because of her involvement in her church and neighborhood. However, as we shall see, C’s knowledge of MS did not reflect positive attitudes toward accommodation. Many in PR link party affiliation to language attitudes, according to Pérez Casas (2016, p. 55): “If someone says you are not a Puerto Rican because you are speaking English, then that person is probably someone who supports independence. . . . Statehood supporters think English is superior to Spanish, and commonwealth supporters are always in the middle.” Given this link, we wondered whether political views were significantly related to knowledge of mexicanismos, and they were (0.0127). Independence supporters knew half (51%) of the words, the pro-commonwealth supporters knew 43% of them, and statehooders knew only approximately one-third (35%). Those with no status preference knew the fewest (18%). These results indicate that a stronger commitment to Spanish and independence or associated US status was linked to greater knowledge of Mexican words. But support for a Puerto Rican identity distinct from that of both the United States and Mexico was evident in the nearly unanimous preference for PRS. This is an important finding that invites more investigation concerning the link between language loyalties, national identity, and political views.

194

Ana Celia Zentella

194

5. Spanish, Puerto Rican identity, and Mexican contact in San Diego PRS was almost unanimously preferred to MS (93%), primarily for cultural and/ or nationalist reasons (69%)—such as soy puertorriqueño (“I’m Puerto Rican”)— or because it is what they know/are comfortable with, such as “it is part of me,” “it is unique,” “it is mine.” Some (11%) deemed it more aesthetically pleasing, while a few (2%) viewed MS as más formal (“more formal”) citing its standard pronunciation of final s and velar r, and menos mezcla (“less mixing”); 5% preferred both varieties equally. When asked about Mexican attitudes toward PRS, the differences were more pronounced. The majority (56%) think Mexicans do not like PRS; they have heard it being criticized for “dropping sounds” and being “too fast,” “hard to understand,” “slangy,” and “vulgar.” In contrast, the remaining 44% believe Mexicans like the PRS accent, that they find it “different,” “animated,” and alegre (“happy”). Given the negative attitudes most have heard, it was unexpected to learn that the majority (70%) would not consider it either a compliment or an insult if they were told they spoke like a Mexican, although 28% did view it as an insult, and only 2% considered it a compliment. Also unexpected was that while the majority (61%) insist almost no Puerto Ricans in San Diego sound Mexican, slightly more than one-quarter (27%) have been told by other Puerto Ricans that they speak Spanish like Mexicans. In fact, many admit they sometimes accommodate Mexicans, primarily by incorporating Mexican vocabulary (53%); one-third (34%) have adopted some pronunciations. For some, this is a conscious effort to seek unity, one well received by Mexicans: Cuando uno dice una frase como ellos, ellos se alegran y para mí es positivo. . . . Aquellos puertorriqueños que hablan como mexicanos están tratando de unir . . . lo hacemos como un gesto de buena voluntad. When someone says a sentence like they do, they are glad and for me that’s positive. . . . Some PRs that talk like them are trying to unite . . . we do it as a gesture of good will. (A: Male, 45 years old, b. NYC, ages 10–18 in PR, 12 years in San Diego, upper middle class) In fact, almost half (44%) acknowledge that some mexicanismos have become part of their Spanish; they mentioned lo corrieron (“they fired him”), panza (“belly”), gűero”(“blonde”), tirar (“to throw away”), pinche (“darned/damned”), and no mames (“don’t mess around”). Pronunciation is another matter. Few (9%) have adopted Mexican phonology, and for C, that is a line that should not be crossed; her reaction against those Puerto Ricans who sound like Mexicans was violent: Son comemierda . . . Me cae como un peo. Yo se lo digo en la cara. Yo le digo “Yo, motherfucker, tú eres puertorriqueño, por qué tú me estás

195

Puerto Rican identity in San Diego

195

hablando de vmmmi, vmii, tú estás cantando.” . . . Nosotros no cantamos un cojón. They’re shiteaters . . . Hits me like a bomb [fart]. I tell them to their face. I tell him/her, ‘Yo, motherfucker, you’re Puerto Rican, why are you talking to me like vmmmi, vmii? You’re singing.’ . . . We don’t sing shit. (C: Female, 59 years old, b. PR, ages four to 36 in NYC, 18 years in CA, lower working class) The Mexican intonation pattern was often criticized as “singing,” but the defense of PRS was based more on national identity than on linguistic likes/dislikes. Interviewees maintained that a firm PR identity in San Diego precluded dialectal shift or accommodation. Almost three-quarters (72%) believe it is not inevitable that PRS will change under MS influence, and an even greater percentage (85%) would not “consider it OK” if PRS did change. But even if it did, most insist that that would not prove damaging: 81% believe you can speak like a Mexican and still be Puerto Rican; only 19% did not agree. This dominant view reflected the reality at home, where only 17% spoke only Spanish to their children. More than double that percentage raised their children in English (37%), but almost half (46%) spoke both. As a result, 37% of their children did not speak Spanish, although the majority (63%) did. But the majority of their children did not sound like Puerto Ricans when they spoke Spanish (44% sounded like Anglos, 33% like Puerto Ricans, 23% like Mexicans), so it was natural for parents to decouple language from identity in San Diego. Their children claimed Puerto Rican identity even if they didn’t speak like Puerto Ricans. G lamented the fact that her sister’s US- born children did not speak Spanish yet they insisted on claiming their Puerto Ricanness: Los hijos de mi hermana, que nacieron aquí, no hablan español. Ellos dicen que son puertorriqueños. . . . su papá es americano y todo. . . . Es una batalla entre como ellos se identifican y como los otros los identifican. Ellos siempre están orgullosos y les gusta usar pues la ropa que diga PR. . . . Ellos están constantemente relacionando pues todo de PR con ellos, pero, uno los mira y uno los ve que son bien americanos. My sister’s children, who were born here, don’t speak Spanish. They say they are Puerto Rican. . . . Their Dad is American and all. . . . It’s a battle between how they identify and how others identify them. They are always proud, and they like to wear clothes that say PR. . . . They are constantly relating everything from PR to themselves, but you look at them and you see that they’re real Americans. (G: Female, 49 years old, b. PR, 19 years in San Diego, white collar) Perhaps if her own children had been raised in San Diego instead of PR, G would be more understanding of her sobrinUs’ situation, or maybe their looks were the

196 Ana Celia Zentella

196

determining factor. In any case, strong identification with a PR identity was not incompatible with solidarity with Mexicans.

6. Solidarity with Mexicans Both female (G) and male (A) interviewees expressed cultural, linguistic, and political reasons for striving toward solidarity with Mexicans, including shared personality traits and language: G: Los mexicanos son más como nosotros que los americanos. . . . tendemos a ser mas amigables, más conversadores. El americano es más recto, más serio. Mexicans are more like us than the Americans. . . . We tend to be friendlier, more conversationalists. The American is stiffer, more serious. A: [el español] es una manera de identificarte you know, decir ‘Mira soy hispano, tengo las mismas costumbres que tú, y estoy seguro que nos gustan un montón de cosas como tú.’” Spanish is a way of identifying yourself . . . to say, “Look, I’m Hispanic, I have your same customs, and I’m sure we like a lot of things you like.” On the other hand, A stipulated that relations were best with Mexicans who were bilingual: Eso sí, el mexicano que no sabe hablar inglés, o el mexicano que nada más habla inglés, tenemos problemas comunicándonos. That said, the Mexican who doesn’t know English, or the Mexican who only speaks English, will have problems communicating. As A’s views and the insertion of English in his and others’ comments make clear, bilingual code-switching is part of PRS in San Diego. Besides linguistic empathy, many recognized that Mexicans worked hard, and they lamented the anti-Mexican attitudes of Anglos, and even some Puerto Ricans: E: Yo hablo con ellos [maintenance workers], “Buenos días, cómo están.” Sí, porque el americano promedio no interactúa con ellos. Para ellos esa gente son invisibles. I talk with them, “Good morning, how are you.” Yes, because the average American doesn’t interact with them. For them those people are invisible. C: They work. En California yo he visto tú sabes que la gente—americana— hasta los puertorriqueños . . . dicen que les están quitando el trabajo a ellos . . . pero el trabajo está ahí, y eres tú que no lo estás cogiendo, pues

197

Puerto Rican identity in San Diego

197

ellos lo están cogiendo para soportar [sic] su familia, hasta limpiar un toilet . . . ellos están trabajando pa’ su familia . . . aunque tengan papeles chuecos.” They work. In California I’ve seen you know that people—Americans, even Puerto Ricans . . . say they are stealing their jobs, but the work is there, and you’re the one who isn’t taking it, because they are taking it to support their family, even cleaning a toilet. . . . They are working for their family . . . even if they have fake papers. However, E cited political differences among Puerto Ricans, claiming that estadistas (“statehooders”) wanted nothing to do with Mexicans, perhaps because many were undocumented and/or lower working class: Yo he visto personas que son PNP de clavos pasaos que no quieren saber de mexicanos. No quieren que sus hijas salgan con mexicanos. I’ve seen people who are hard core PNP [statehood party] members who want nothing to do with Mexicans. They don’t want their daughters to go out with Mexicans. None of our informants expressed this attitude, and many lamented the legal problems Mexicans face and how they are treated, even by compatriots: J: Y no me gusta como tratan a los mexicanos aquí. Son los mismos mexicanos que los tratan—los mismos, la inmigración, y los tratan a ellos como perros. I don’t like how they treat Mexicans here. It’s Mexicans themselves who treat them—immigration, and they treat them like dogs. G: Me da mucha pena, verdad, porque ellos a lo que vienen—la mayoría, noventa y nueve por ciento lo que vienen es a buscar mejorar, pero las leyes aquí pues no les permite, entonces cuando uno ve todas las cosas que están sucediendo—. O sea por ley, no debiera ser pero por la cuestión de humanidad, pues me da pena, qué se yo. It makes me very sad, really, because they come—the majority, 99% what they come for is to better themselves, but the laws here don’t allow them, so when you see what’s going on—. So legally it shouldn’t be but for humanitarian reasons, it makes me sad, I dunno. Whether or not they adopted Mexican terms and pronunciations or criticized those who did, solidarity with Mexicans was further cemented by the fact that both groups suffered at the hands of compatriots in Mexico and Puerto Rico, as A explained: Tenemos algo en común. . . . Ah, ¿que los tuyos a ti te discriminan? A mí también los míos me discriminan (“We have something in common. . . .

198

Ana Celia Zentella

198

Oh so your own people discriminate against you? Mine discriminate against me too”).

7. Conclusion: language and the construction of Puerto Rican identity in San Diego Despite the omnipresence of Mexicans in San Diego, the distance from Puerto Rico and the proximity to Mexico, interviews with 94 Puerto Ricans reveal little lasting impact of MS on their Puerto Rican ways of speaking. Only those who have been married to Mexicans for decades have adopted some MS features, but the majority of all our interviewees’ primary networks did not include Mexicans, and they did not use and/or understand common Mexican lexical items. Lack of adaptation to the primary Spanish dialect of the region is also an “act of identity” rooted in nationalism; most are middle class and belonged to organizations that promote Puerto Rican culture in San Diego. As Le Page and Tabouret-Keller (1985) point out regarding “acts of identity”: The individual creates for himself the patterns of his linguistic behavior to resemble those of the group or groups with which from time to time he wishes to be identified, or so as to be unlike those from which he wishes to be distinguished. (1985, p. 181)

Linguistic, cultural, and political solidarity with Mexicans leads to some instances of accommodation, but these are fluid and in continuous construction, inhibiting dialect leveling and koineization. Are the results of the Boricua experience in San Diego applicable to other settings? Erker (2018, fn2) reminds us that Spanish-speaking communities across the United States vary widely in their size, settlement history, and socio-demographic profile. As such, it is crucial that any investigation be sensitively attuned to particular aspects of the community that may bear on local linguistic practices and attitudes. The anthro-political linguistic perspective I have advocated since 1995 incorporates the impact of the racialization of a people and their language(s) resulting from anti-immigrant hysteria and violence, linguistic profiling, English-only laws, the disparagement of bilingual education, and the rejection of the poor. In San Diego, the presence of educated Puerto Ricans—many former military—with solid jobs and citizenship that allow them to visit Puerto Rico to renew linguistic ties obscures the distinct reality of lower working-class Puerto Ricans in the county. Their language practices and attitudes remain to be studied, but in all cases, the impact of English cannot be ignored—for example, the newsletter of the House of Puerto Rico, originally written in Spanish in 1975, became bilingual in 1981, and was in English only by 1987 (López, 2010). More to the point, in the

199

Puerto Rican identity in San Diego

199

second and third generations of every family, Spanish is being lost. Consequently, as Rosa (2015) found, ethnic identity is less a barrier to dialect leveling than is English hegemony. Conflicts between coger and agarrar, or guagua and camion, will not be a problem; future generations will be trying “to catch the bus.”

Notes 1 Instead of the traditional Spanish binary gender markers (o/a), or the x, which does not distinguish singular from plural and is difficult to pronounce, I advocate the universal U/Us. 2 I am indebted to student research assistants who helped with interviews, transcriptions, statistical analyses, and/or the construction of tables: Gabriel Ramírez, Diego García Montufar, José Fuste, Lupita Ruiz-Tolentino, and Bridget Vaughan. 3 The varying Ns reflect the number of respondents who answered a particular question.

References Bayley, R., Cárdenas, N.L., Treviño Schouten, B., & Vélez Salas, C.M. (2012). Spanish dialect contact in San Antonio, Texas: An exploratory study. In M. Díaz-Campos & K. Geeslin (Eds.), Selected proceedings of the 14th Hispanic linguistics symposium (pp. 48–60). Cascadilla Proceedings Project. Retrieved from www.lingref.com/cpp/ hls/14/paper2655.pdf Bergad, L.W. (2016, December). The Latino population of New York City 1990–2015— Latino data project—Report 65. Center for Latin American, Caribbean, and Latino Studies, CUNY Graduate Center. Retrieved from http://clacls.gc.cuny.edu/files/2017/ 03/Latino-Data-Project-Report-65.-The-Latino-Population-of-New-York-City-1990– 2015.-December-2016.pdf Bonnici, L.M., & Bayley, R. (2010). Recent research on Latinos in the USA and Canada, Part 2: Spanish Varieties 1. Language and Linguistics Compass, 4(2), 121–134. Erker, D. (2018). Spanish dialectal contact in the United States. In K. Potowski (Ed.), Handbook of Spanish as a heritage/minority language (pp. 269–283). New York: Routledge. Escobar, A.M., & Potowski, K. (2015). El español de los Estados Unidos. Cambridge: Cambridge University Press. House of Puerto Rico. Welcome to Borinquen in San Diego. Retrieved from www.houseofpuertorico.com/ Johnson, E.G. (2005). Mexiqueño?: A case study of dialect contact. Doctoral dissertation, University of Pennsylvania. Retrieved from http://repository.upenn.edu/pwpl/ vol11/ iss2/8 Kerswill, P.E. (2002). Koineization and accommodation. In J.K. Chambers, P. Trudgill, & N. Schilling-Estes (Eds.), The handbook of language variation and change (pp. 669– 702). Malden, MA and Oxford: Blackwell. Le Page, R., & Tabouret-Keller, A. (1985). Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press. López, D. (2010, December 9). A history of La Casa de Puerto Rico (San Diego): 1972– 1987: The first 15 years. El Boricua: House of Puerto Rico San Diego Newsletter.

200

Ana Celia Zentella

200

Retrieved from www.houseofpuertorico.com/new-articles/elboricua-SpecialEdition_120910.pdf López, D. (2017, July/August). California Puerto Ricans generally doing “better” economically. El Boricua: House of Puerto Rico San Diego Newsletter, p. 2. Marzán, G., Torres, A., & Luecke, A. (2008). Puerto Rican Outmigration from NYC: 1995–2000. Centro de Estudios Puertorriqueños Policy Report, 2(2). Retrieved from https://centropr.hunter.cuny.edu/sites/default/files/working_papers/Outmigration091108.pdf Otheguy, R., & Zentella, A.C. (2012). Spanish in New York: Language contact, dialectal leveling, and structural continuity. New York: Oxford University Press. Pellicia, M.R. (2014). Language ideologies in action: When different Latino linguistic identities collide. In R. Orozco (Ed.), New directions in Hispanic linguistics, (pp. 22–49). Newcastle: Cambridge Scholars Publishing. Retrieved from www.cam bridgescholars.com/download/sample/59753 Pérez Casas, M. (2016). Codeswitching and identity among Island Puerto Rican bilinguals. In R. Guzzardo Tamargo, C.M. Mazak, & M.C. Parafita Couto (Eds.), Spanish-English codeswitching in the Caribbean and the US (pp. 37–60). Amsterdam: John Benjamins. Potowski, K. (2016). Intralatino language and identity: Mexirican Spanish. Philadelphia, PA: John Benjamins Publishing Company. Potowski, K., & Torres, L. (forthcoming). Spanish in Chicago. New York: Oxford University Press. Real Academia Española. (2005). Diccionario Pan-hispánico de dudas (Spanish ed.). Bogotá: Santillana. Rodriguez Ayuso, I.R., Geerman Santana, K., & Marazzi Santiago, M. (Eds.). (2013). Puerto Rican diaspora atlas 2010. Instituto de Estadísticas de Puerto Rico: Estado Libre Asociado de Puerto Rico. Retrieved from www.estadisticas.gobierno.pr/iepr Rosa, J. (2015). Nuevo Chicago?: Language, diaspora, and Latina/o panethnic formations. In R. Marquez & L.M. Rojo (Eds.), A sociolinguistics of diaspora: Latino practices, identities, and ideologies (pp. 31–47). New York: Routledge. Rosa, J. (2019). Looking like a language, sounding like a race: Raciolinguistic ideologies and the learning of Latinidad. New York: Oxford University Press. Zentella, A.C. (1990, December). Lexical leveling in four New York City Spanish dialects: Linguistic and social factors. Hispania, 73(4), 1094–1105. Zentella, A.C. (2000). Puerto Ricans in the United States: Confronting the linguistic repercussions of colonialism. In S.L. McKay, & S.C. Wong (Eds.), New immigrants in the United States: Readings for second language educators (pp. 137–164). New York: Cambridge University Press.

201

11 Caribbean Spanish influenced by African American English US Afro-Spanish language and the new US Caribeño identity Teresa Satterfield and José R. Benkí Jr. 1. Introduction We present an emergent US Spanish contact variety that systematically combines iconic characteristics of US African American English (AAE) with Caribbean Spanish. We coin the term “US Afro-Spanish” (USAS) for these urban US Spanish speech patterns. While contact between Caribbeans and African Americans in the US East Coast has taken place since at least the 19th century (Hoffnung-Garskof, 2010), modern USAS likely arose in the Bronx (metropolitan New York) in the 1960s and 1970s from sustained interactions between Spanish-dominant immigrants from the Caribbean and African Americans, the latter being monolingual English speakers using in-group AAE (Satterfield, 2012). USAS is thus a by-product of varieties viewed prescriptively as degenerate forms of their respective source languages, standard (Latin American) Spanish and standard (American) English. Given that 70% of US Latinos and African Americans continue to reside in low-income urban areas or inner-ring suburbs (American Community Survey, 2015), examining multilingual urban language contact should be common practice in US-based studies; however, to date, few researchers have addressed the effect of “non-standard” or “ethnically marked” English on US Spanish. See the examples after this paragraph. Example (1a) depicts spontaneous USAS speech produced by an anonymous Caribbean Latino, age 21 in Miami (Satterfield & Alexander, 2006). The USAS utterance (1a) precedes its AAE correlate in (1b). For readers finding it difficult to decipher the USAS or AAE forms, a standard American English (SAE) gloss is provided in (1c): (1) a b c

USAS: “Estamos en la casa; tú sabes cómo lo hacemos, somos calle. To’ el mundo pa’arriba y tú, mami, retrásalo.” AAE: “We in da house; ya know how we do, we street. Everybody up and you, baby, back that thang up.” SAE: “We are here; we are the best at what we do, we are smart. Everyone dance and you, attractive woman, come over here.”

Our extensive USAS corpus demonstrates that rule-governed USAS features are present across linguistic modules. AAE-indexed lexical variants are in frequent

202

Teresa Satterfield and José R. Benkí Jr.

202

usage as calqued USAS lexical tokens in (1a) (Satterfield & Alexander, 2006). High-frequency USAS morphosyntactic constructions (e.g., ella loca; no espero nada de nadie),1 correspond directly to McWhorter’s (1998) continuum of “light” AA(V)E features (e.g., copula deletion (“she crazy”) and negative concord (“I don’t ‘spect’ nothing from nobody”). “Deep” AA(V)E syntax that is isolated from dominant SAE speech and regarded as basilectal has also surfaced in USAS (e.g., gerund adjective: “you style copyin’, clown,” or estilo copiando guasón) (Satterfield & Schuen, submitted). The current inquiry focuses on phonetic and phonological properties of USAS, examining AAE effects on Caribbean Spanish in glottalization, final obstruent deletion, and an emergent innovation that we term “hyper-aspiration.” In hyper-aspiration of USAS, which occurs in /esCV/ sequences such as la esquina, the aspirate variant [h] for the (s) is produced after the voiceless plosive, resulting in a sequence that is phonetically and phonologically similar to a strongly aspirated voiceless plosive in most varieties of English. Thus, the full obstruent [s] in la esquina [laeskina] in standard Spanish contrasts with the aspirate [h] in [laehkina] in Caribbean Spanish, and with the hyper-aspiration [kʰ] (and elided /e/) in [lakʰina] in USAS. Within varieties of American English, glottalization and final obstruent deletion have long been considered distinctive AAE markers (Fasold & Wolfram, 1970; Wolfram, 1994; Bailey & Thomas, 1998; Rickford, 1999; Craig, et al., 2003). We show that similar patterns are manifested in US Spanish. Evidence for USAS’s phonetic and phonological variations comes from acoustic analyses of spoken interviews and conversations involving US adult bilingual Caribbean Latinos. To accurately delineate the contributing factors for USAS, data were also analyzed along socio- and psycholinguistic variables. The chapter proceeds as follows: Section 2 provides a brief phonological overview of USAS, Caribbean Spanish and American English varieties. Section 3 outlines the experiment methodology used. Section 4 provides results from the spectrograms. Section 5 contains phonetic analyses and advances a sociolinguistic account for emergent USAS sound patterns and innovations. Finally, Section 6 offers concluding remarks.

2. Background: Spanish and English source varieties 2.1 Spanish input language Standard Caribbean Spanish conserves basic Spanish structures while exhibiting key phonological features not present in the educated speech of other (nonCaribbean) Spanish varieties. Properties include the deletion of word/syllablefinal obstruents, the aspiration of syllable-final /s/, the velarization of word-final /n/, the lateralization of /r/, the rhotacism of /l/, and the elision of intervocalic /d/ in specific contexts (e.g., Cedergren, 1973; Terrell, 1975; Guitart, 1997). Research indicates that the Spanish spoken in the Caribbean (as in other Hispanophone regions) has been significantly affected by its African populations,

203

Spanish influenced by African Americans

203

specifically by their experiences in second language (L2) acquisition (Granda, 1988; Lipski, 1986, 2000, 2005, 2014; Schwegler, 1996, 2011, 2016). Previously mentioned variations such as the syllable-final aspiration of /s/, intervocalic /d/ elision, /r/-/l/ shifts, and morphological leveling all link to historical Afro-Hispanic contexts (Lipski, 1993, 2005, 2014). 2.2 English input varieties Given the English varieties relevant to USAS emergence, AAE is often defined as an English dialect spoken primarily by people of African heritage in the United States. However, in areas with close socialization between African Americans and other ethnic groups, such as certain boroughs of New York City, a greater number of “non-black” AAE speakers typically exist. Speakers of AAE are typically bi-dialectal, since they combine SAE with “light” and “deep” AAE, to varying degrees (McWhorter, 1998). Features that distinguish AAE from SAE include distinct rule-governed phonological properties, many found in Creoles and dialects of other populations of West African descent (Rickford, 1997). As noted, AAE word-final obstruents undergo phonetic/phonological devoicing, glottalization, and deletion. For example, the final alveolar voiced plosive (d) in “Kool-Aid,” produced in SAE as [kʰuleɪd], can be devoiced as [kʰuleɪt], glottalized as [kʰuleɪʔ] or [kʰuleɪʔd], or deleted as [kʰuleɪ]. Glottalization involves either a complete glottal closure and glottal stop, during which the vocal cords are completely adducted and consequently cease vibration, transcribed as [ʔ], or a partial glottal adduction transcribed as [ʔd] and marked by the vocal cords vibrating irregularly (see discussion and citations in 3.3). Reducing vocally homogeneous final consonant clusters occurs such that best (two final voiceless consonants) is produced as [bεs], and stand (two final voiced consonants) is produced as [stæ̃ n]. AAE word-final devoicing is often accompanied by glottalization, unlike Southern white English and other American English varieties, which also devoice obstruents but do so without glottalization (Fasold, 1981). “Nuyorican English” is an abbreviated term for “New York Puerto Rican English,” though it is a variety customarily used by Latinos of many nationalities, and throughout the US East Coast, not merely in New York (Wolfram, 1974; Labov, Ash, & Boberg, 2006). It is a “non-standard” English variety that may play a secondary role in the formation of USAS. “Nuyorican” is differentiated from SAE and AAE chiefly on phonetic grounds. Specifically, Nuyorican speakers exhibit Spanish vowel and prosody qualities in their English production even if they are not native Spanish speakers. 2.3 Research hypotheses We hypothesize that glottalization will be an identifiable phonological feature differentiating USAS from other varieties of (Caribbean) Spanish. Glottalization is a salient feature of AAE, particularly when it cooccurs with the absence of oral closure as a variant for voiceless obstruents (e.g., [ʔ] vs. more standard variants

204

Teresa Satterfield and José R. Benkí Jr.

204

[t] and [ʔt]) or as a variant of voiced obstruents (e.g., [ʔ] and [ʔd] vs. the standard variant [d]). Spectrographic evidence for glottalization has previously been reported in Puerto Rican adolescent Spanish speakers on the island, particularly in wordfinal position and likely based on US contact (Valentín-Márquez, 2006). We predict that if urban bilingual US Latino speakers use the indexical marker of glottalization when speaking English, it will also emerge in their spoken Spanish. Final obstruent devoicing, another distinguishing feature of AAE, is not predicted to be a productive process in USAS. It is extremely difficult to ascertain whether this feature arises in USAS. First, given the highly productive process of word-final obstruent deletion in Caribbean Spanish, instances of unreduced word-final obstruents are rarely observed. Second, /d/ is the only phonologically underlying voiced obstruent in Spanish that appears in word-final position with any reasonable frequency. Variationist analyses of word-final (d) are additionally challenged, since most occurrences are based on the pronoun “usted” (formal “you”), which is restricted in usage and obtains in formal registers, thus not hypothesized to be productive in the youth-oriented speech of USAS. Data in Spanish will serve to document features of USAS, while instances in English will demonstrate that association with AAE is the origin of these features. Furthermore, we claim that incorporating these distinctive AAE traits into the Spanish structure indexes the politicized identities that serve to position individual Hispanic speakers in relation to their identity as US non-whites while allowing them to identify as non-African American by maintaining Spanish.

3. Experimental methodology The corpus of USAS data for the larger ongoing project encompass 200 speakers and over 10,000 texts largely from (written) social media but also (spoken) videotaped interviews. The experiment seeks to highlight two domains of language behavior. First, the speech patterns of bilingual US-Caribbean Latinos are examined at the phonological level. Second, this speech production is interpreted based on socio- and psycholinguistic variables. 3.1 Selecting speakers/participants The majority of speakers in the USAS corpus are Latino males under the age of 50, with a median age of 25.7 years. They exhibit varying degrees of bilingual competence, ranging from strong AAE or Nuyorican English dominance to strong Caribbean Spanish dominance. To isolate phonological patterns across bilingual speakers, we delimited our sample to only those bilingual individuals who exhibited balanced competence in Spanish and English (AAE) and for whom we had obtained multiple high-quality spoken speech samples in both English and Spanish. Consistent with the bilingual acquisition literature (Bialystok, 2001), we controlled for speaker’s level of bilingual proficiency based on chronological age, onset age of exposure to both languages, and amount of time spent in the United States. We included an additional variable of residence in the

205

Spanish influenced by African Americans

205

Eastern United States during childhood and/or adolescence as a control for Caribbean heritage. The above criteria resulted in a strong match with three speakers. The three participants share common backgrounds related to age, race, socioeconomic status, profession (music industry), and childhood experiences. Irrespective of birthplace, all three report that their families relocated both in and out of the United States. Each speaker has resided in the US East Coast for extended periods during childhood/adolescence and adulthood. All have native or native-like command of Spanish and English and effortlessly grasp in-group nuances of US and Caribbean cultures. All self-identify as non-white Latinos, with one speaker explicitly categorizing himself as black or Afro-Latino. 3.2 Materials Recorded videos of speakers posted on media websites were downloaded to computer disk as MPEG-4 files. Audio soundtracks in MPEG-1 format (116 kbps or 128 kbps bit rate, 48 kHz sampling rate) were loaded for analysis by using Praat 5.1 (Boersma & Weenink, 2009). For the current experiment, we present acoustic data as acoustic waveform and spectrogram displays to precisely illustrate glottalization phenomena. While space limitations preclude detailed statistical analysis of impressionistic transcriptions, the examples provided are representative of speakers’ production. 3.3 Spectrographic analyses Glottalization is audible and can be transcribed in an impressionistic transcription, but it is a gradient acoustic-phonetic phenomenon. The most reliable determinations of glottalization are made using a combination of acoustic waveform and spectrogram displays, along with an audio recording. Glottalization consists of irregular laryngeal vibration or a complete glottal stop. When realized as irregular vibration, glottalization differs from normal modal phonation by irregular laryngeal vibration (Redi & Shattuck-Hufnagel, 2001) and is measured and identified by jitter and sometimes shimmer (Deliyski, 1993). Jitter consists of abrupt changes in the period of the glottal cycle, while shimmer consists of a Table 11.1 List of study participants Participant

Gender

Race*

Birthplace

Age

Childhood

Adulthood

A

Female

Non-white

Añasco, PR

45

New York City

Puerto Rico Miami, FL

B

Male

Non-white

Miami, FL

40

Miami, FL Cuba

Miami, FL

C

Male

Black

Santurce, PR

45

Miami, FL

Puerto Rico

*Race = self-identified classification.

206

Teresa Satterfield and José R. Benkí Jr.

206

series of glottal cycles with alternating high-peak and low-peak amplitude pulses.2

4. Experiment results

Amplitude

We present acoustic waveform and broad-band spectrogram displays (5 ms analysis window for Figures 11.1, 11.2, 11.4, and 11.5; 3 ms window for Figures 11.3, 11.6, and 11.7) of speech containing glottalization. Since the speakers produced all instances of glottalization with the absence of oral closure, the glottal pulses are visible in the spectrogram at the times indicated on the time axis as vertical striations spanning a broad range of frequencies but most prominently in vocal tract resonances of vowel (formants) immediately preceding glottalization. Unless considerable background noise occurs in the recording, pulses are also visible in the waveform. Where glottalization is realized by one or more irregular glottal pulses (i.e., jitter), we see an abrupt lengthening of the space between pulses as the duration of the period sharply increases. For utterances ending in glottal stops, the utterance-ending glottal pulse is visible in both the waveform and spectrogram, with a substantially longer interval between the glottal stop and the preceding (penultimate) glottal pulse. Figures 11.1–11.3 show word-

Figure 11.1 Waveform and spectrogram of “see it” [siɪʔ] by speaker C (VH1, 2007). A glottal stop occurs at the irregular glottal pulse immediately after the tick at t = 43.43 seconds.

Spanish influenced by African Americans

207

Amplitude

207

Figure 11.2 Waveform and spectrogram of “no Kool-Aid you” [nokʰuleɪʔju] by speaker B (mun2, 2008a). Glottalization, realized by irregular glottal pulses, occurs around t = 41.93 seconds.

final consonant deletion and glottalization in English, consistent with AAE. In all three cases, there is an absence of oral closure (i.e., deletion) that is simultaneous with glottalization. The first example (Figure 11.1) is an interview in English, with speaker C (VH1, 2007) producing a full glottal stop [ʔ] for the voiceless alveolar plosive (t) in “see it” in a prepausal context. The glottal stop occurs at the irregular glottal pulse immediately following the tick at t = 43.43 seconds. Speaker B produces strong glottalization in English without oral closure lasting several glottal pulses for the voiced alveolar plosive (d) “Kool-Aid,” in Figure 11.2 (mun2, 2008a). Here, the glottalization occurs at the word-final position but not prepausally. Speaker A produces glottalization in English as well (mun2, 2008b). In Figure 11.3, speaker A produces strong glottalization without oral closure for (d) at the end of “you did” in a prepausal context similar to speaker C. Given the presence of glottalization in the speakers’ English production, we examined their Spanish speech. As hypothesized, the speakers all produce glottalization in word-final position in Spanish (Figures 11.4–11.6); however, the patterns are not uniform across speakers. Speaker A produces glottalization as

Teresa Satterfield and José R. Benkí Jr.

208

Amplitude

208

Figure 11.3 Waveform and broad-band spectrogram “you did” [judɪʔ] by speaker A (mun2, 2008b). Glottalization, realized by irregular glottal pulses, occurs around t = 191.42 seconds. In the background, the interviewer’s overspeech is apparent around 191.20 seconds, and the impulse noise near the end of this clip is not a glottal stop but is instead the sound of speaker A’s fingernails clicking.

variants of word-final obstruents, typically dental (d) and alveolar voiceless fricative (s). The word-final obstruent in her speech is elided or produced with an absence of oral closure (consistent with both AAE and Caribbean Spanish). The weakened or absent consonant is replaced with a glottal stop or glottalization— which is consistent with AAE, and as we claim, USAS. In the video excerpt (WOW.cl, 2008) in Figure 11.4, speaker A produces a full glottal stop instead of the word-final standard Latin American Spanish variant [s] or the “standard” Caribbean Spanish variants [h] or [;] in voz (“voice”). We note that like most Caribbean Spanish speakers, the three speakers produce the full range of syllable/word-final (s) variants: [s], [h], and [;]. Speaker C’s word-final glottal stops in Spanish do not consistently appear in place of elided or weakened obstruents but are often present after a word-final vowel. The excerpt in Figure 11.5, taken from a video interview (mun2, 2007), shows a typical example in which speaker C produces pareja as [paɾ ehaʔ] (couple), with no elided or reduced obstruent.

Spanish influenced by African Americans

209

Amplitude

209

Figure 11.4 Waveform and spectrogram of “con una voz” [konunaboʔ] by speaker A (WOW.cl, 2008). A glottal stop occurs at t = 19.446 seconds.

Speaker B displays both patterns of glottalization in Spanish, shown in Figure 11.6 for the utterance un skit ahí (“a skit there”). He inserts the loanword “skit” (with no epenthetic [e] before [sC], but it does produce a Spanish vowel [i]). He ends with a glottal stop—but no oral closure, similar to the pattern displayed by speaker A. However, at the end of the utterance immediately following ahí (vowel-final), speaker B produces a glottal stop, similar to patterns displayed by speaker C. Our results support the USAS hypothesis in that the bilingual speakers’ English and Spanish utterances contain the AAE phonological feature of glottalization. The experiment also confirms that phonological markers such as the aspiration of syllable-final /s/, as described in the standard reference literature on Caribbean Spanish, continue to be robust in USAS. An unpredicted outcome in the experiment is that salient /s/-aspiration came to be reanalyzed by USAS speakers, resulting in an independent feature emerging in voiceless obstruents following /s/. We use the term “hyper-aspiration” because the phenomenon involves aspiration in a context that has not previously been identified in Caribbean Spanish. “Hyper-aspiration” appears to be restricted to /esCV/

Teresa Satterfield and José R. Benkí Jr.

210

Amplitude

210

Figure 11.5 Waveform and spectrogram of “una pareja” [unapaɾ ehaʔ] by speaker C (mun2, 2007). A glottal stop occurs at t = 344.75 seconds.

sequences, as in the word “esquina” (“corner”). In standard Latin American Spanish, esquina is produced as [eskina]. In standard Caribbean Spanish, (s) is aspirated [ehkina] or deleted [ekina]. The same word with “hyper-aspiration” in USAS appears as [ekʰina], with the aspirate occurring after the voiceless obstruent. Interestingly, homomorphemic [sCʰV] sequences are generally not produced in AAE either, since voiceless plosives following /s/ in English are not aspirated, even when the following vowel is stressed. In Figure 11.7, speaker B produces “hyper-aspiration” (and elision of /e/) in the phrase “la esquina” [lakʰina] (“the corner”). There is considerable noise in the recording, but the key acoustic-phonetic events (beginning of [k] closure, release burst for [k], and onset of voicing following [k]) are identifiable as labeled in the figure. The key result is the long interval between the [k] release with its characteristic burst at 2 kHz and the onset of voicing for voice onset time (VOT) = 58 ms later. This relatively long VOT is characteristic of voiceless aspirated velar plosives. In comparison, the 27 ms VOT following [k] in “un skit ahí” in Figure 11.6 (uttered by speaker B less than a minute later) is typical of Spanish velar voiceless unaspirated plosives.

Spanish influenced by African Americans 211

Amplitude

211

Figure 11.6 Waveform and spectrogram of “un skit ahí” [ũskiʔaiʔ] by speaker B (mun2, 2008a). Glottal stops occur at t = 104.40 and t = 104.79 seconds.

5. Discussion 5.1 Detailing the data points The spectrographic data confirm the presence of a phonological feature heretofore considered unique to AAE. All three bilingual speakers demonstrate glottalization in word-final position in their Spanish utterances, either as variants of final obstruents or after a final vowel. The speakers also produce glottalization in English in word-final position, consistent with glottalization patterns documented in AAE. In both languages, the glottalization cooccurs with the absence of oral closure, highlighting the prevalence of final consonant deletion in the source varieties of AAE and Caribbean Spanish. While the current study describes sound patterns most unique to AAE, there are numerous less-salient phonological properties of AAE that these speakers regularly employ. In English-speaking contexts, all three individuals function exclusively in AAE rather than in SAE. Collectively, the speakers show a consistent pattern of glottalization in AAE; however, they produce subtle individual variations in glottalization in Spanish. Speaker A’s Spanish glottalizations occur as variants of (deleted) final obstruents, similar to the instances attested by Valentín-Márquez (2006) in the

212

Teresa Satterfield and José R. Benkí Jr.

212

Amplitude

/ /

Figure 11.7 Waveform and spectrogram of “la esquina” [lakʰina] with hyper-aspiration by speaker B (mun2, 2008a). The release of the /k/ occurs at t = 72.868 seconds, and the onset of voicing for /i/ occurs at t = 72.926 seconds, with a resulting voice onset time (VOT) = 58 ms.

contact-induced speech of female Puerto Rican adolescents on the island. Speaker C’s glottalizations occur primarily after final vowels, although he elides final consonants like other speakers of Caribbean Spanish. Speaker B’s speech displays both types of glottalizations. The internal variability exhibited in the Spanish pattern in conjunction with the absence of variability in the parallel English pattern is consistent with our claim that glottalization is a nascent USAS innovation, stemming from contact with AAE. Speaker B produces another novel feature in /esCV/ sequences (in words such as esquina (“corner”) and estamos (“we are”)), which we term “hyper-aspiration” because of its marked quality. While non-USAS Caribbean Spanish speakers elide or aspirate a word-internal /s/ in situ, speaker B produces aspiration after the release of the consonant following /s/. This aspiration occurs in an unusual position for Spanish (speaker B’s voice onset times (VOTs) for voiceless plosives are otherwise very short, consistent with monolingual Spanish speakers), and thus, it is quite salient. Anecdotally, hyper-aspiration was not detected in speech samples of Puerto Rican speakers A and C, although we have observed hyper-aspiration in

213

Spanish influenced by African Americans

213

the Spanish of younger US-born Miami Latinos not profiled in the current experiment. Interestingly, the hyper-aspiration phenomenon exhibited by speaker B appears on the surface to be phonetically similar to a more generalized aspiration of /esCV/ sequences found in synchronic Andalusian Peninsular Spanish data and described as “postaspiration” (Torreira, 2012). In Andalusian dialects, all voiceless stops /p, t, k/ are realized with postaspiration, including contexts in which the plosive is preceded by an elided or aspirated /s/, giving rise to both preaspiration and postaspiration (Torreira, 2012; Ruch & Peters, 2016).3 Significantly, reports of other Spanish varieties exhibiting patterns of postaspiration are nonexistent in the literature (Torreira, 2012). Investigations on aspiration in Puerto Rican and Buenos Aires Spanish did not yield any evidence of postaspiration (Torreira, 2007). Thus, it does not seem probable that hyper-aspiration in USAS arose from Andalusian Spanish; moreover, no other source varieties of Spanish for USAS display this feature. In the broader context, the results validate our premise that USAS is unique— not only in that certain traits are shared and reinforced between the two source languages (Caribbean Spanish and AAE) but also in that these new speech forms are distinguished by the systematic use of existing Spanish features that USAS extends beyond standard and popular monolingual Spanish grammars. Crucially, USAS is not Anglicized Caribbean Spanish, and it should not be tagged as “Nuyorican” Spanish, which is a restructured form of Spanish spoken by latergeneration East Coast-born bilingual Latinos. Nuyorican Spanish is highly stigmatized by native Spanish speakers in the countries of origin (Zentella, 1997); although it closely adheres to Spanish phonetic features, it typically exhibits the English-dominant syntax, morphology, and lexicon that violate grammatical norms of Spanish. In contrast, USAS maintains acceptable Hispanophone patterns, selectively incorporating distinguishing features within a grammatical Spanish base. Corpus videos of USAS speakers in Spanish show no indication of stigmatization from their interlocutors, who are native Spanish speakers from Spanish-dominant countries (e.g., Chilean interlocutor with speaker A and Dominican and Mexican interlocutors with speaker C). 5.2 Factors motivating USAS USAS cannot be viewed as a product of language interference, because only specific features of AAE are targeted and because they do not result in a Spanish variety empirically distinct from standard (Caribbean) Spanish. Psycholinguistic and historical research on language contact and bilingualism underscores crosslinguistic transfer of common features, or linguistic convergence (e.g., Sánchez, 2003; Bullock & Gerfen, 2004), as a principal outcome of language contact, particularly in communities undergoing language attrition or language death. While properties such as final obstruent deletion occur in similar contexts in both Caribbean Spanish and AAE, we challenge the claim that such convergence inevitably leads to language loss in the non-dominant language. Despite the long-standing

214

Teresa Satterfield and José R. Benkí Jr.

214

tendency of minority language shift to monolingual English in the United States, Spanish language use—by both immigrants and US-born Hispanics—remains relatively stable according to US demographic reports (Lipski, 2008). Studies (e.g., García & Cuevas, 1995; Otheguy & Zentella, 2012) indicate that in urban settings such as New York City, the use of Spanish in the family and friendship networks, as well as its transgenerational transmission, is still significant. Problematically, proponents of convergence often treat internally motivated innovations, such as the hyper-aspiration feature attested in USAS, as anomalies or symptoms of language death (Guion, 1993) rather than as successful efforts to expand the language within a bilingual context. We pair sociolinguistic methods with acoustic results to motivate principled explanations for the existence of USAS and to show how factors within particular language-contact situations trigger adaptive bilingual strategies (Satterfield, 2003; Satterfield & Barrett, 2004) as a way to preserve language and to resist language loss. Following Woolard (2008), we examine how language ideology provides a broader sociolinguistic framework for understanding USAS and its new phonological patterns. Studies (e.g., Duany, 2000, 2003; Rivera, Marshall, & Pacini Hernández, 2009; Flores, 2011) document (Afro-)Caribbeans’ assimilation pressures in the United States. While immigrants may self-identify their nationality as Puerto Rican or Cuban, they find that upon entering the United States, they become “black”—or, at best, “non-white”—due to their phenotypic traits. Given a priori racial classifications, Hispanics perceived to be “of color” must often integrate themselves into US society differently from those immigrants perceived as white. Contact between Africans and Hispanics in Latin America has existed since the 15th century (Lipski, 2005; Schwegler, 2006). In the United States, however, the dominant society characterizes native US-born people of dark skin color and brown/black immigrants as one and the same. As a result, communities of color become relatively “united” and come to share insights into the realities of being non-white in the United States. As we claim, incorporating the distinctive AAE traits into the Spanish structure likely indexes the politicized identities that serve to position individual Hispanic speakers in relation to their identity as US non-whites while allowing them to identify as non-African American via the maintenance of Spanish (Toribio, 2000). Within the context of a society in which a dominant language and ideology exist, specific grammatical forms come to be iconic markers of ethnic identity (Gal & Irvine, 2000). Younger US Latinos in particular struggle to distinguish themselves from both their native-born American black and white counterparts and to establish an identity that is separate and unique. Nevertheless, as the iconic relationship between linguistic and ethnic difference strengthens, linguistic forms have moved in line with shifts in political consciousness. A sense of (black) racial solidarity is increasingly apparent in the use of AAE, as illustrated in our data. Similarly, the appearance of glottalization further suggests that Latinos of younger generations are strategically introducing structures that index points of

215

Spanish influenced by African Americans

215

distinction between US Spanish and standard Caribbean Spanish grammars, selecting precisely those features that most index phonological and social difference. In sum, the data presented document a changing pattern of US Spanish grammar that, we posit, stems from an iconoclastic shift in the language ideology and demographics of the US Hispanic population. Younger generations of US Latinos now reject prevailing ideologies that devalue (or ignore) minority languages and instead choose to embrace less prestigious varieties. Satterfield and Barrett (2004) find that young Maya-Spanish bilinguals involved in Indigenous language preservation efforts in Guatemala intentionally accentuate differences between the two languages through an emergent word order and innovative code-switching. Similarly, we hypothesize that hyper-aspiration in the current study constitutes acousticphonetic exaggeration, which is best understood as a strategy of differentiation occurring in intense language-contact situations in which bilingual speakers attempt to maintain separation between their two languages due precisely to complex language ideologies.

6. Conclusion The current study contributes to ongoing investigations of a new US Spanish language variety, US Afro-Spanish (USAS). Spectrographic analyses confirm the occurrence of an innovative set of phonological variations that continue to evolve and diffuse into US urban Spanish-speaking communities. Implementing sociolinguistic methodologies provides a more nuanced analysis, leading to our conclusions that linguistic behaviors contributing to this new variety are produced by highly proficient Spanish-English bilinguals, in turn revealing representative language ideologies of younger Latinos who overtly avoid linguistic indexicalities of “whiteness” across their bilingual repertoire. To round out this account, we address USAS’s status as a linguistic code. Insofar as USAS’s emergence is comparable to other contact scenarios, we note that O’Shannessy (2005, p. 31) describes Light Warlpiri, an outback Australian contact variety combining Indigenous Warlpiri, Kriol, and English, as a “new language.” Cheshire and colleagues (2011) avoid labels such as “dialect” or “variety” to report on the emergence of an English-language form in inner-city London, multicultural London English (MLE). They refer to MLE as a “new multiethnolect.” Unlike USAS, MLE represents contact between multiple varieties of first-language (L1) and L2 English rather than crosslinguistic contact. However, similar to USAS’s stigmatized source languages, low prestige English varieties produced MLE. Cheshire and colleagues (2011) note that MLE is driven by young nonAnglo, second-generation English speakers, particularly those of Anglophone Afro-Caribbean culture. African American language and culture are also linked to MLE, reminding us of the African diaspora’s profound impact. For the present, the larger question of USAS’s classification as a linguistic code remains open. Although the existence of a distinctly identifiable Afrolect within Caribbean Spanish is often rejected (Lipski, 1985, 2005), we argue that USAS, as an evolving sociolect or ethnolect, may well be on a black Spanish trajectory. What is certain is

216

Teresa Satterfield and José R. Benkí Jr.

216

that young US-Caribbean Latinos continue to navigate intricate urban social networks and to construct new identities as 21st-century American youth of color.

Notes 1 Negative concord as double negation is completely grammatical in standard Spanish; the distinction here is that the relative frequency of negative concord in USAS is much higher than the frequency found in standard (Caribbean) Spanish corpora (Satterfield & Schuen, submitted). See Ortiz-López (2007, 2010) for related statistics on double negation in monolingual Dominican Spanish. 2 We transcribe all instances of glottalization by using [ʔ], regardless of whether the glottalization consists of irregular laryngeal vibration or a full glottal stop. Our descriptions accompanying each example, however, distinguish between irregular laryngeal vibration and full glottal stops. Other transcription conventions include the use of square brackets to indicate a fine phonetic transcription or event, parentheses around a phonetic symbol to indicate a sociolinguistic variable for which other variants might be produced, and slashes to indicate an abstract phonological representation. 3 We note that the origins of postaspiration phenomena in Andalusian Spanish are not documented in previous studies. One might speculate that postaspiration in Andalusian Spanish is also the result of historical contact with speakers from the African diaspora. However, at present, the basis of the more restricted hyper-aspiration exhibited in USAS does not appear to be genetically connected to Andalusian Spanish.

References American Community Survey. (2015). U.S. Census Bureau. Retrieved March 31, 2017, from www.census.gov/acs/www/data/data-tables-and-tools/data-profiles/2015/. Bailey, G., & Thomas, E. (1998). Some aspects of African-American English phonology. In S. Mufwene et al. (Eds.), African-American English (pp. 85–109). London: Routledge. Bialystok, E. (2001). Bilingualism in development: Language, literacy, and cognition. Cambridge: Cambridge University Press. Boersma, P., & Weenink, D. (2009). Praat: Doing phonetics by computer (Version 5.1.05). [Computer program]. Retrieved August 1, 2009, from www.praat.org/. Bullock, B., & Gerfen, C. (2004). Phonological convergence in a contracting language variety. Bilingualism: Language and Cognition, 7(2), 95–104. Cedergren, H. (1973). The interplay of social and linguistic factors in Panamá. PhD Dissertation, Cornell University. Cheshire, J., Kerswill, P., Fox, S., & Torgersen, E. (2011). Contact, the feature pool and the speech community: The emergence of multicultural London English. Journal of Sociolinguistics, 15(2), 151–196. Craig, H., Thompson, C.A., Washington, J.A., & Potter, S.L. (2003). Phonological features of child African American English, Journal of Speech Language and Hearing Research, 46, 623–635. Deliyski, D. (1993). Acoustic model and evaluation of pathological voice production. Proceedings of Eurospeech, 3, 1969–1972. Duany, J. (2000). Nation on the move: The construction of cultural identities in Puerto Rico and the diaspora. American Ethnologist, 27(1), 5–30. Duany, J. (2003). Nation, migration, identity: The case of Puerto Ricans. Latino Studies, 1(3), 424–444.

217

Spanish influenced by African Americans

217

Fasold, R. (1981). The relation between Black and White speech in the South. American Speech, 56(3), 163–189. Fasold, R., & Wolfram, W. (1970). Some linguistic features of Negro dialect. In R. Fasold & R. Shuy (Eds.), Teaching standard English in the inner city (pp. 41–86). Washington, DC: Center for Applied Linguistics. Flores, J. (Ed). (2011). The Afro-Latino reader. Durham, NC: Duke University Press. Gal, S., & Irvine, J. (2000). Language ideology and linguistic differentiation. In P. Kroskrity (Ed.), Regimes of language (pp. 35–84). Santa Fe, NM: School of American Research Press. García, O., & Cuevas, M. (1995). Spanish ability and use among second-generation Nuyoricans. In C. Silva-Corvalán (Ed.), Spanish in four continents: Studies in language contact and bilingualism (pp. 184–195). Washington, DC: Georgetown University Press. Granda, de G. (1988). Los esclavos del Chocó: Su procedencia africana (Siglo XVIII) y su posible incidencia lingüística en el español del área. Thesaurus, XLIII(1), 64–80. Guion, S. (1993). Structure exaggeration in language death. In SALSA I: Proceedings of the first annual symposium about language and society (pp. 159–166). Austin, Texas: Linguistic Forum 33. Guitart, J. (1997). Variability, multilectalism, and the organization of phonology in Caribbean Spanish dialects. In F. Martínez-Gil (Ed.), Issues in the phonology and morphology of the major Iberian languages (pp. 515–536). Washington, DC: Georgetown University Press. Hoffnung-Garskof, J. (2010). A tale of two cities: Santo Domingo and New York after 1950. Princeton, NJ: Princeton University Press. Labov, W., Ash, S., & Boberg, C. (2006). The atlas of North American English (p. 24). Berlin: Mouton-de Gruyter. Lipski, J. (1985). Black Spanish: The last frontier of Afro-America. Crítica (San Diego), 1, 53–75. Lipski, J. (1986). Convergence and divergence in bozal Spanish: A comparative study. Journal of Pidgin and Creole Languages, 1(2), 171 ff. Lipski, J. (1993). On the non-Creole basis for Afro-Caribbean Spanish. Ms. Albuquerque, NM: Latin American Institute, University of New Mexico. Lipski, J. (2000). Spanish-based creoles in the Caribbean. Ms. State College, PA: Penn State University. Lipski, J. (2005). A history of Afro-Hispanic language: Five centuries, five continents. Cambridge: Cambridge University Press. Lipski, J. (2008). Varieties of Spanish in the United States. Washington, DC: Georgetown University Press. Lipski, J. (2014). The many facet of Spanish dialect diversification in Latin America. In S. Mufwene (Ed.), Iberian imperialism and language evolution in Latin America (pp. 38–75). Chicago, IL: University of Chicago Press. McWhorter, J. (1998). Word on the street: Debunking the myth of a “pure” standard English. New York: Basic Books. mun2. (2007). Interview with Tego Calderón from an episode of the show Vivo posted on September 7, 2007. Retrieved August 15, 2009 from www.holamun2.com/shows/vivo/ vivo-tego-calderon mun2. (2008a). Episode of La Esquina posted on November 19, 2008. Retrieved August 15, 2009 from www.holamun2.com/shows/pitbulls-la-esquina/pitbulls-la-esquinaprice-is-tight-ghetto-fabulous.

218

Teresa Satterfield and José R. Benkí Jr.

218

mun2. (2008b). Interview of Ivy Queen on The Shuffle, posted August 27, 2008. August 15, 2009 from www.holamun2.com/shows/the-mun2-shuffle/the-mun2-shuffle-2124ivy-queen. Ortiz-López, L.A. (2007). La negación en la frontera domínco-haitiana: variantes y usos (socio)lingüísticos. In K. Potowski & R. Cameron (Eds.), Spanish in contact: Educational, social and linguistic inquiries (pp. 211–236). Amsterdam: Johns Benjamin Publishers. Ortiz-López, L.A. (2010). El español y el criollo haitiano: Contacto lingüístico y adquisición de segunda lengua. Madrid/Frankfurt: Iberoamericana/Vervuert. O’Shannessy, C. (2005). Light Walpiri: A new language. Australian Journal of Linguistics, 25(1), 31–57. Otheguy, R., & Zentella, A. (2012). Spanish in New York: Language contact, dialectal leveling, and structural continuity. Oxford: Oxford University Press. Redi, L., & Shattuck-Hufnagel, S. (2001). Variation in the realization of glottalization in normal speakers. Journal of Phonetics, 29, 407–429. Rickford, J. (1997). Prior creolization of African American Vernacular English? Sociohistorical and textual evidence from the 17th and 18th centuries. Journal of Sociolinguistics, 1(3), 315–336. Rickford, J. (1999). African American vernacular English. Maiden, MA: Blackwell. Rivera, R., Marshall, W., & Pacini Hernández, D. (Eds.). (2009). Reggaeton. Durham, NC: Duke University Press. Ruch, H., & Peters, S. (2016). On the origin of post-aspirated stops: Production and perception of /s/ + voiceless stop sequences in Andalusian Spanish. Laboratory Phonology: Journal of the Association for Laboratory Phonology, 7(1), 1–36. Sánchez, L. (2003). Quecha-Spanish bilingualism: Interference and convergence in functional categories. Amsterdam: John Benjamins. Satterfield, T. (2003). Economy of interpretation: Patterns of pronoun selection in transitional bilinguals. In V. Cook (Ed.), Effects of the second language on the first (pp. 214– 233). Clevedon: Multilingual Matters. Satterfield, T. (2012). East Coast 2061: Agent-based models tracking the evolution of U.S. Afro-Spanish. Ms. University of North Carolina NEH Institute. Satterfield, T., & Alexander, R. (2006). Espanglish con más flow: Evidence of nonlanguage shift and AAVE-influenced Spanish. Unpublished Ms. University of Michigan. Satterfield, T., & Barrett, R. (2004). Generation gap: Explaining new and emerging wordorder phenomena in Mayan-Spanish bilinguals. In Proceedings from the first international symposium on bilingualism in Latin America (pp. 285–300). Buenos Aires: ESSARP. Satterfield, T., & Schuen, S. (submitted). Slanguage to sociolect? The development of U.S. Afro-Spanish as a result of social media popularization. Spanish in Context. Schwegler, A. (1996). Lenguas criollas en Hispanoamérica y la contribución Africana al español de América. Signo y Seña, Special Issue. Universidad de Buenos Aires: Instituto Lingüística. Schwegler, A. (2006). Bozal Spanish: Captivating new evidence from a contemporary source (Afro-Cuban “Palo Monte”). In L. Thornburg & J. Fuller (Eds.), Studies in contact linguistics: Essays in honor of Glen G. Gilbert (pp. 71–101). New York: Peter Lang. Schwegler, A. (2011). Palenque (Colombia): Multilingualism in an extraordinary social and historical context. In M. Díaz-Campos (Ed.), The handbook of Hispanic sociolinguistics (pp. 446–472). Oxford: Blackwell Publishers.

219

Spanish influenced by African Americans

219

Schwegler, A. (2016). Combining population genetics (DNA) with historical linguistics: On the African origins of Latin America’s Black and Mulatto populations. In S. Sessarego & F. Tejedo (Eds.), Spanish language and sociolinguistic analysis (pp. 33–88). Amsterdam: Johns Benjamins. Terrell, T. (1975). Functional constraints on deletion of word final /s/ in Cuban Spanish. In C. Cogen (Ed.), Proceedings of the first annual meeting of the linguistic society of Berkeley (pp. 431–437). Berkeley, CA: University of California Press. Toribio, J. (2000). Language variation and the linguistic enactment of identity among Dominicans. Linguistics, 38, 1133–1159. Torreira, F. (2007). Coarticulation between aspirated-s and voiceless stops in Spanish: An interdialectal comparison. In The 9th Hispanic linguistics symposium (pp. 113–120). Somerville, MA: Cascadilla Press. Torreira, F. (2012). Investigating the nature of aspirated stops in Western Andalusian Spanish. International Phonetic Association Journal of the International Phonetic Association, 42(1), 49–63. Valentín-Márquez, W. (2006). La Oclusión Glotal y la Construcción Lingüística de Identidades Sociales en Puerto Rico. In T. Sagarra & J. Toribio (Eds.), Selected proceedings of the 9th Hispanic linguistics symposium (pp. 326–341). Somerville, MA: Cascadilla Press. VH1. (2007). Tego’s plans. Interview posted on February 27, 2007. Retrieved August 15, 2009 from www.vh1.com/video/play.jhtml?id=1552984&vid=135630 Wolfram, W. (1974). Sociolinguistic aspects of assimilation: Puerto Rican English in New York City. Washington, DC: Center for Applied Linguistics. Wolfram, W. (1994). The phonology of a sociocultural variety: The case of African American vernacular English. In J. Bernthal & N. Bankson (Eds.), Child phonology: Characteristics, assessment, and intervention with special populations (pp. 227–244). New York: Thieme. Woolard, K. (2008). Why dat now? Linguistic-anthropological contributions to the explanation of sociolinguistic icons and change. Journal of Sociolinguistics, 12(4), 432–452. WOW.cl. (2008). Ivy Queen conversó con WOW.cl. Interview posted on September 10, 2007 Retrieved August 15, 2009 from www.wow.cl/prontus_wow/site/artic/20070910/pags/ 20070910154109.html. Also posted on www.youtube.com/watch?v=YBRL6ZyDVbY. Zentella, A. (1997). Growing up bilingual. Oxford: Blackwell Publishers.

Index

Note: Page numbers in italic indicate a figure. Page numbers in bold indicate a table. Page numbers followed by an “n” denote a note on the corresponding page. accent: correctness of 172, 172; pleasantness of 173, 173 acceptability judgment tasks 131, 134–136, 138–142, 146 African American English (AAE): Caribbean Spanish influenced by 201–216; final obstruent deletion 202, 204, 213; glottalization 202–209, 206–211, 211, 214; hyper-aspiration 202, 209–210, 212, 212–215, 213n3; lateral deletion and vocalization of /l/ in 56 age: lateralization of /ɾ/ and 43; presence of backed /r/ and 19, 20, 22, 24, 27–29; presence of normative /r/ and 18, 19, 21–22, 22, 28; /r/ variants and 90–91, 102–104, 103–104; velarization of /r/ and 107 Alba, O. 75–76 Alers-Valentín, H. 89, 91 Alfaraz, G. G. 58, 169, 177 Alonso, A. 41, 57 Álvarez-Nazario, M. 42 Alvord, Scott M. 2, 54–67 Andalusian Peninsular Spanish 213, 213n3 apparent-time hypothesis, Labov’s 115 approximated trills 13, 14, 15 Argentinian Spanish 176–177 Arias, Alba 2 Artstein, R. 143 aspiration: hyper-aspiration 202, 209–210, 212, 212–215, 213n3; of /s/ 157–158, 202–203, 209, 212–213

sound and 25, 26, 28–29; individual speaker differences in Puerto Rico 29, 30; individual speaker differences in western Massachusetts 29, 29; intervocalic position of 9; origin and 19, 20, 27–29; place and 27; in Puerto Rico 24–25, 26; regional distribution in U. S. 28; socio-indexical factors 9, 31; spectrograph 13, 15; as stigmatized 28, 31; task and 19, 21, 25, 27–29; in Western Massachusetts 22, 24; in Western Massachusetts and Puerto Rico 19–21, 20–21 Barrera-Tobón, Carolina 3, 113–125 Bastian, J. A. 154 Bayley, R. 184 Belizean Spanish 42 Benkí, José R, Jr. 4 bilingualism: among US Latinos 113; Catalan-Spanish 57; changes in speaker’s Spanish and 114; in Miami Cubans 67; in Puerto Ricans in Grand Rapids 93; in Puerto Rico 3, 28, 166, 169, 174–175, 177–178; in San Diego 189, 196, 198; subject expression and 128–129, 132, 146; US Afro-Spanish language and 202, 204, 209, 211, 213–215; in Virgin Islands 155 Boruga, Felipe Polanco 72 Brief Description and Dictionary of the Language Used in the Virgin Islands, A (Ray) 154 Buenos Aires Spanish 213

backed /r/ 9–11, 15–17, 19, 21–22, 24–29, 31–32; age and 19, 20, 27–29; following

Cabo Rojo, Puerto Rico 10, 27, 106 Canary Islands 167

221 Cardi B 80–81, 80n34, 81n35 Caribbean Spanish 1; influenced by African American English 201–216; Puerto Rican evaluations of 166–183, 172–173; see also specific types Carter, P. M. 55, 58, 65 Cassano, P. V. 41 Catalan 56–57 Chesire, J. 215 Chicago, Illinois 26, 185–187 Chicano English 56–57, 65–66 Clemons, Aris Moreno 2 coda sibilant elision, retention, and insertion in Dominican Spanish 71–85 coda-s 9, 71–85; see also /s/ code-switching 3, 103, 155, 174, 178, 196, 215 Colombian Spanish 176 community size, role of 36–27 contact influence, with Miami-Cuban Spanish and English 54–55, 57–58, 64–65, 67 co-referentiality 129, 131 correctness of accent, evaluation of 172, 172 Costa Rican Spanish 40, 42 Creole: creolized English grammar in the Spanish of Dominicanos on St. Thomas 3, 153–163; Dutch 154; Virgin Island English Creole (VIEC) 155 Cuba/Cuban: /l/ and deletion as /ɾ/ variants 42; Miami-Cuban Spanish and English /l/ 54–67; perceptions of Cubans 168 Cuban Spanish: correctness of accent, evaluation of 172, 172; evaluation by Puerto Ricans of 168, 172–173, 172–173, 176; pleasantness of accent, evaluation of 173, 173 /d/: devoicing of 203–204; elision of intervocalic 202–203; word-final position 204 Danish West India Company 153 dark /l/ laterals 56–57, 59–61, 60, 63–66 darkening 56, 62, 65; see also velarization D’Arpa, Daniel S. 3 Decker, B. D. 44–45 deletion: cupula 202; final obstruent 202–204, 207, 211, 213; /l/ 56–57; /r/ 10, 26, 35, 37, 40–42, 45–46, 49; /s/ 71–74, 107, 185 devoicing: of /d/ 203–204; final obstruent 203–204

Index

221

Di Petro, Robert 155 does/du 162–163 Domínguez-Rosado, B. 174 Dominican Spanish: coda sibilant elision, retention, and insertion in 71–85; correctness of accent, evaluation of 172, 172; creolized English grammar in the Spanish of Dominicanos on St. Thomas 153–163; /e/ 157, 159–160, 163; evaluation by Puerto Ricans of 168, 172–173, 172–173, 176–177; /l/ and deletion as /ɾ/ variants 42; metathesis 157; pleasantness of accent, evaluation of 173, 173; /ɾ/ vocalization 41; /-se/ added in plural nouns 160; /taŋ/ 161–163; verb phase syntax 161–163; word-final /-m/ as plural 161, 163 Dominicans, stereotypes of 168 Duany, J. 177 Dueber, Dagmar 159 /e/ 157, 159–160, 163 /e/ elision 202, 210 education: coda-/s/ in Dominican Spanish 76–77; evaluations of Spanish varieties by Puerto Ricans 173–175; lateralization of /ɾ/ and 43, 45, 50; preverbal subject placement and 120–121, 120–121; /r/ variants 90–91, 100–101, 101 “El Sobrín” 59 elision: of /d/ 202–203; of /e/ 202, 210; /s/ 71–72, 74, 77–78, 80–82, 160; of syllables 157 English: Chicano 56–57, 65–66; creolized English grammar in the Spanish of Dominicanos on St. Thomas 153–163; Miami-Cuban Spanish and English /l/ 54–67; see also specific varieties of English English proficiency: among US Latinos 113; preverbal subject placement and 118, 118–122, 121–122; Puerto Ricans in San Diego 189 Erker, D. 124, 184–185 Espinosa, A. 41 exposure to English, pronominal subject placement and 114–115, 118, 118, 120, 120–121 Fernández, D. 92 final obstruent deletion 202–204, 204, 207, 211, 213 final obstruent devoicing 203–204

222

Index

first-person singular (1sg) 128–133, 135–140, 136, 138–139, 143–146 flap, description of 89 flap /ɾ/ 41, 89 following segment: English lateral /l/ and 62–63, 62–64, 66; Spanish lateral /l/ and 61, 61, 66 Frog, Where Are You? (Mayer) 11 gender: coda-/s/ in Dominican Spanish 77; English lateral /l/ and 62–63, 62–63, 66; lateralization of /ɾ/ and 40; light/dark difference for /l/ 66; /ɾ/ use and 43, 45; /r/ variants and 90–91, 101–103, 102, 103, 105–106, 106; Spanish lateral /l/ and 60, 61, 66; /s/-realization and 75n13; trill variation and 27 generational differences: Miami Cubans and /l/ laterals 54–55, 58, 60, 61–63, 62–67; pronominal subject placement 114–115, 118, 118, 120, 120–121; /r/ variants 91, 102–104, 103–104; rhotic realizations of Puerto Rican community in Western Massachusetts and Puerto Rico 10, 12, 12–13, 17, 21, 29, 31–32; Spanish lateral /l/ and 60, 61; in the use of /l/ and /ɾ/ in communities of PRS speakers in the U. S. 35, 37–38, 40, 45–50, 46, 48 glottalization 202–209, 206–211, 211, 214 GOLDVARB X for Windows 40 Gómez Soler, Inmaculada 3 Graml, C. 91 Grand Rapids, Michigan 2, 10, 27, 88–108 [h] (/s/ variant) 106–107, 202 hablar fisno 72, 75–80, 82 Hammond, R. 90 Handbook of Varieties of English (Devonish and Harry) 159 Hazlet, New Jersey 37–38, 42, 46, 46–50 heritage speakers of Spanish, subject expression and 128–147 Hoenigswald, H. 169 Hoffman, Michol 156 Holmquist, J. 43, 91 Holyoke, Massachusetts 2, 7–8, 10, 12 [hɾ] 13 [hr] 13, 15, 16 hyper-aspiration 202, 209–210, 212, 212–215, 213n3 hypercorrection: qualitative 75, 78, 78n26, 80, 82; quantitative 77–78, 82

222 identity: coda-/s/ and negotiation of 80–82; Puerto Rican in San Diego 184–199; US Afro-Spanish language and the new US Caribeño 201–216 immigrant groups: English lateral /l/ and 62–63, 62–63; light/dark difference for /l/ 65; Miami Cubans and /l/ laterals 54–55, 58, 60, 61–63, 62–67; Spanish lateral /l/ and 60, 61 immigration, of Latino populations 7–8 infinitives, lateralization of /ɾ/ and 47, 47, 49 insecurity, linguistic 166, 169, 173 intervocalic position of /r/ 8–9, 11, 98 Jersey City, New Jersey 91, 99 Johnson, Gosh 185 /l/: assimilation 46, 57–58; dark /l/ laterals 56–57, 59–61, 60, 63–66; differences in the use of /l/ and /ɾ/ in communities of PRS speakers in the U. S. 35–50; light /l/ laterals 56–57, 59, 59–60, 64–66; Miami-Cuban Spanish and English 54–67 Labov, W. 39, 115 lambdacism 45, 57, 59 Lamboy, E. M. 11, 45, 91, 99 language contact: Caribbean Spanish influenced by African American English 201–216; language and Puerto Rican identity in San Diego 3–4, 184–199; pronominal subject placement and 113–125; subject expression and 128, 131–133, 143–144, 146 Lastra de Suárez, Y. 41 lateralization 2, 10, 26, 31, 35, 37, 40, 42–47, 47, 49–50, 92, 107, 107, 166–167, 202 laterals: dark /l/ 56–57, 59–61, 60, 63–66; English 56–57; light /l/ 56–57, 59, 59–60, 64–66; Miami-Cuban and English /l/ 54–67; Spanish 56–57; variation in Cuban Spanish 57–58 laterals in Miami-Cuban Spanish and English /l/ study 54–67 Latino/Latinos: English proficiency among US 113; percentage of US population 113 lexical subjects 129, 138–142, 141–142, 167 life stage: /r/ variants 102–104, 103–104, 105, 106; see also generational differences

223 light /l/ laterals 56–57, 59, 59–60, 64–66 Lipski, J. M. 42, 57, 157 liquid neutralization 9, 57 London, multicultural London English (MLE) in 215 López Morales, H. 43, 90 Lorain, Ohio 10, 26, 31, 37–38, 42, 44–45, 47, 48, 48–49, 185 /-m/, word-final in plural nouns 161, 163 Massachusetts, Puerto Rican community 7–32 Matta de Fiol, E. 90 Mayer, Mercer 11 Medina, Danilo 73 Medina-Rivera, A. 28, 43, 90 Mejía Pardo, D. 177 metathesis 147 Mexican American Spanish 41–42 Mexican Spanish 4, 26, 36, 42, 175, 186 Mexicans, contact with Puerto Ricans in San Diego 3–4 Miami English 55, 65–66 Miami-Cuban Spanish and English /l/ 54–67; contact influence 54–55, 57–58, 64–65, 67 migration between Puerto Rico and Massachusetts 7–8, 11, 27 Milroy, L. 93–94 Mojica de León, C. M. 169 Montrul, S. 132 Moore, Carol 59 Morris, J. 56 Mullen, K. 65 multicultural London English (MLE) 215 /n/, velarization of 166, 166n2, 202 Navarro Tomás, T. 9, 42, 50, 90 negative concord 202n1 Negerhollands 154 New York City 3, 113–125, 168, 185, 187, 201, 203, 214 New York Puerto Rican English (Nuyorican) 203, 213 normative /r/ 8, 15–18, 21–22, 24, 27–29, 31; age and 18, 19, 21–22, 22, 27; individual speaker differences in Puerto Rico 31, 31; individual speaker differences in western Massachusetts 30, 31; in Lorain, Ohio 10; origin and 17, 17–18, 22, 23, 27; place and 18, 19, 27; in Puerto Rico 24, 25; spectrograph 13, 14; stress and 24, 25; task and 18, 18,

Index

223

22, 23, 27; in Western Massachusetts 21–22, 22–23; in Western Massachusetts and Puerto Rico 17–18, 17–19 null subject 128–131, 133–134, 137, 140, 142–146, 146n1, 146n2 null-subject language, Spanish as 114, 128 Nuyorican 203, 213 origin: presence of backed /r/ and 19, 20, 27–29; presence of normative /r/ and 17, 17–18, 22, 23, 27 O’Shannessy, C. 215 Otheguy, R. 184 Otheguy-Zentella corpus 115–116, 123–124 overt lexical subjects 129, 138–142, 141–142 overt pronominal subjects 128–135, 137–139, 138–142, 141–146 overt subject pronouns 114, 116, 118, 124, 130, 167, 185 Pérez Casas, M. 193 person and animacy, hierarchies of 143 person effect on subject expression of Spanish heritage speakers study 128–147; first-person singular (1sg) 128–133, 135–140, 136, 138–139, 143–146; generativist and variationist approaches 128–133, 136, 146; overt lexical subject 129, 138–142, 141–142; overt pronominal subjects 128–135, 137–139, 138–142, 141–146; previous research on 128–133; third-person singular (3sg) 128, 130–133, 135–137, 140–142, 140–146 phonological borrowings 36 place: presence of backed /r/ and 27; presence of normative /r/ and 18, 19, 27; trill variation and 27 pleasantness of accent, evaluation of 173, 173 plosive 202–203, 207, 210, 212–213 plural nouns: /-se/ added in 160; word-final /-m/ 161, 163 Poplack, S. 45 position: English lateral /l/ and 62, 62, 64, 64, 66; lateral variation of /l/ 56–60, 61–62, 62, 64, 64, 66; /r/ in PRS 88, 89, 107, 107; of /ɾ/ 35, 37, 39–43, 46, 46, 46–49, 47; /r/ variants in PRS 98, 98–99, 105; Spanish lateral /l/ and 60, 61, 66; /s/-realization and 75n13

224

Index

post aspiration 213, 213n3 post-tap /r/ 15, 16, 26, 31 post-verbal subjects 114 Potowski, K. 187 Prada Pérez, Ana de 3 pre-aspirated /r/ 13, 15, 16 prestige, linguistic 2–3, 9, 31, 49, 92, 102, 104, 106, 166, 168–169, 176, 178 Preston, D. R. 169 preverbal rate 116–124, 118–122; education and 120–121, 120–121; English proficiency and 118, 118–122, 121–122; exposure to English and 118, 118, 120, 120–121; socioeconomic status and 119–120, 119–122, 122 preverbal subjects 114–124 pro-drop 114, 123, 125 pronominal subject placement variation 113–125 pronominal subjects see overt pronominal subjects pronouns: non-inverted subject pronouns in questions 167; overt subject 114, 116, 118, 124, 130, 167, 185; subject personal 185–186; subject pronoun expression 128–147 Prosper-Sánchez, G. D. 43 Puerto Rican Spanish (PRS): differences in the use of /l/ and /ɾ/ in communities of PRS speakers in the U. S. 35–50; language and Puerto Rican identity in San Diego 3–4, 184–199; linguistic features of 166–167; rhotic realizations of Puerto Rican community in Western Massachusetts and Puerto Rico 7–32; sociolinguistic distribution /r/ in Grand Rapids, Michigan 88–108; varieties, evaluations of 166–183 Puerto Rican Spanish (PRS), evaluation study of 166–183; data analysis 172, 172–173; issues influencing 167–170; linguistic features of 166–167; materials 170–171; participants 171; procedure 171–172; questionnaire 182–183; research questions 170; results and discussion 172–177 Puerto Rican Spanish (PRS) in western Massachusetts and Puerto Rico study 10–32; acoustic analysis and coding 13; conclusion 31–32; discussion 25–31; equipment 12; frequency distribution 15, 16; materials 11; methodology 10–11; participants 12, 12; phonetic realizations

224 13–15, 14–16; qualitative results 13–16; quantitative results 16–25; research questions and hypothesis 10–11; results for Puerto Rico 24–25, 26; results for Western Massachusetts 21–24, 22–24; results for Western Massachusetts and Puerto Rico 17–21, 17–21 Puerto Ricans: community in Massachusetts 7–32; diaspora 7–8, 28, 175; evaluations of varieties of Spanish 166–183; knowledge of Mexican lexicon 189–193, 190, 192; linguistic security in San Juan 177; Mexican contact in the Midwest and Southwest US 185–187; population in Massachusetts 8; in San Diego 187–189 qualitative hypercorrection 75, 78, 78n26, 80, 82 quantitative hypercorrection 77–78, 82 Quilis, A. 57, 95–97 /r/: assimilation 58; frequency distribution of variants of 15, 16; lateralization 2, 10, 26, 31, 35, 37, 40, 42–47, 47, 49–50, 92, 107, 107, 166–167, 202; phonological aspects of 88–90, 89; position of 88, 89, 107, 107; post-tap 15, 16, 26, 31; rhotic realizations of Puerto Rican community in Western Massachusetts and Puerto Rico 7–32; sociolinguistic distribution of Puerto Rican Spanish /r/ in Grand Rapids, Michigan 88–108; “tap” /ɾ/ 8, 13, 15, 16, 26, 31, 41, 45, 89, 102, 105; velarization of 89–92, 98–102, 99, 104–105, 107; vibrante múltiple “trill” /r/ 8; see also backed /r/; normative /r/; standard variant /r/; trill /r/ /ɾ/: differences in the use of /l/ and /ɾ/ in communities of PRS speakers in the U. S. 35–50; gender and 102, 102, 103, 105–107, 106; integration into the Puerto Rican community and 104–105; island Puerto Rican Spanish 42–44; lateralization 35, 37, 40, 42–47, 47, 49–50; life stage and 103–104, 105, 106; position of 98, 98–99, 105; in PRS community in Grand Rapids, Michigan 88, 94, 95–96, 96, 97; retroflex 10, 35–37, 40–42, 45, 47–50; spectrogram of 96, 96; US Puerto Rican Spanish 44–46 [ʀ] (/r/ variant) 90, 94, 94, 98

225 [ʁ] (/r/ variant) 90, 94, 94 R (software) 16 /r/ variants 88–107; age and 102–104, 103–104; education and 90–91, 100–101, 101; gender and 101–103, 102, 103, 105–106, 106; integration into the Puerto Rican community and 104–105, 104–105; life stage and 102–104, 103–104; linguistic factors 98, 98–99; position 98, 98–99, 105; social factors 99, 99; socioeconomic level and 99–101, 99–101; spectrographic analyses of 95–97, 95–97; see also specific variants racial differences, evaluation of Spanish and 177 Ramos-Pellicia, Michelle F. 26, 31–32 40, 43, 45, 108, 185, 210 Raña Risso, Rocío 3, 113–125 Ray, John 154 recreative ongoing mode 162 retroflex /ɾ/ 10, 35–37, 40–42, 45, 47–50 retroflex [ɻ] 41–42, 48–49 retroflex trill 31 retroflexion 10, 26, 35, 41–43, 46 rhotic realizations of Puerto Rican community in Western Massachusetts and Puerto Rico 7–32 /r/-l/ shifts 203 Rodríguez, Ernesto 71 Rogers, Brandon M. A. 2, 54–67 /s/: aspiration 157–158, 202–203, 209, 212–213; coda-/s/ and negotiation of identity 80–82; deletion 71–74, 72–75, 73, 74, 76, 107, 185; Dominican Spanish on St. Thomas, aspiration and loss in 157–158; elision 71–72, 74, 77–78, 80–82, 160; fricative [h] variant 106–107; [h] variant 106–107, 202; insertion 75n13, 78–79, 79, 82; intrusion 75, 78, 78n25, 80, 82; realization 72–79, 75n13, 82; retention 76–78, 80, 82; syllable-final position of 72–73, 73, 106, 185, 202; variants 106–107 Sabino, Robin 155 San Antonio, Texas 186–187 San Diego, California 3–4, 184–199 San Juan, Puerto Rico 9, 12, 42–43, 90, 169, 171, 173–175, 177, 186 Sánchez, R. 41 Satterfield, Teresa 4 /-se/, in plural nouns 160

Index

225

Shouse de Vivas, D. 8, 45 sibilant see /s/ Silva-Corvalán, C. 38, 58, 130 Simounet, Alma 155 Sir Nube Negra 71 socioeconomic level: preverbal subject placement and 119–120, 119–122, 122; /r/ variants and 99–101, 99–101 sociolinguistic distribution of Puerto Rican Spanish /r/ in Grand Rapids, Michigan 88–108 Southern American English: lateral deletion and vocalization of /l/ in 56; r-colored vowels in 44 Spain, evaluations of Spanish from 169, 175–177 Spanglish 174, 177 Spanish: creolized English grammar in the Spanish of Dominicanos on St. Thomas 153–163; Miami-Cuban Spanish and English /l/ 54–67; non-pro-drop variety of 123–125; as null-subject language 114, 128; preverbal subjects in 114–115; see also specific varieties of Spanish speaker egocentrism 130 speech connectivity 130, 133, 136–138, 142 Sprague, Gilbert A. 155, 157, 161–162 Springfield, Massachusetts 2, 7–8, 10, 12 St. Thomas, US Virgin Islands 153–163; Dominicano presence on 154; geopolitical and sociolinguistic background 153–154 St. Thomas English Creole (STEC) 3, 153–163 standard American English (SAE) 201–203, 211 standard variant of /r/ 88–89, 91, 94, 95, 95–96, 96, 98, 101–102, 105–106; gender and 102, 102, 105–106, 106; integration into the Puerto Rican community and 104–105; life stage and 105, 106; position of 98, 98–99, 105; spectrogram of 95, 95 Statistical Package for the Social Sciences (SPSS) 60, 116, 137, 172 stress, presence of normative /r/ and 24, 25, 28 Suárez Büdenbender, Eva-María 3 subject personal pronouns 3, 185–186 subject pronoun expression: effect of person on 128–147; language contact and 128, 131–133, 143–144, 146; null subject 128–131, 133–134, 137, 140,

226

Index

142–146, 146n1, 146n2; overt lexical subjects 129, 138–142, 141–142; overt pronominal subjects 128–135, 137–139, 138–142, 141–146 subject pronouns: expression 128–147; non-inverted in questions 167; overt 114, 116, 118, 124, 130, 167, 185 switch reference 129–132, 139, 144, 146 syllable-final position: of /s/ 72–73, 73, 106, 185, 202; see also word-final position syllable-initial position: light /l/ in 56; /r/ restricted to in Spanish 88; uvular articulation of multiple trill /r/ in 167 /taŋ/161–163 TAM (tense/aspect/mood) 130–131, 135, 136–137, 138–142 tap, description of 89 tap /ɾ/ 8, 13, 15, 16, 26, 31, 41, 45, 89, 102, 105 task: acceptability judgment tasks 131, 134–136, 138–142, 146; presence of backed /r/ and 19, 21, 25, 27–29; presence of normative /r/ and 18, 18, 22, 23, 27; /s/-insertion 75n13 Terrell, T. D. 43, 90 third-person singular (3sg) 128, 130–133, 135–137, 140–142, 140–146 Thomas, E. R. 56 threshold hypothesis, Raña Risso’s 115 Tió, S. 92 Toribio, Almeida Jacqueline 2 trill frication 15, 16, 26, 31 trill /r/: description of 7–8; distribution of 8–9; rhotic realizations of Puerto Rican community in Western Massachusetts and Puerto Rico 7–32; uvular articulation of multiple in syllable-initial position 167; see also /r/ trill variation 9–11, 21, 25, 27, 29, 32 US Afro-Spanish (USAS) 201–216 Valentín-Márquez, W. 2, 9–11, 26, 28, 107, 211 Vaquero, M. 95–97 Van Coetsem, F. 36

226 van Dijk, T. A. 174 Van Hofwegen, J. 56, 66 Varela, B. 54 velarization: of /l/ 56, 60–61, 64; of /n/ 166, 166n2, 202; of /r/ 89–92, 92, 98–102, 99, 104–105, 107 velarization of /r/ 89–92, 107; gender and 101–102, 102; negative attitudes toward 92, 92; percentage in Puerto Rican communities in US 99, 99; socioeconomic level and 99, 99–101 verb phrase 129, 161–163; syntax in Dominican Spanish on St. Thomas 161–163 verbs, lateralization of /ɾ/ and 47, 47, 49 vibrante múltiple “trill” /r/ 8; see also trill /r/ vibrante “tap” /ɾ/ 8; see also tap /ɾ/ Virgin Island English Creole (VIEC) 155 voiceless plosive 202, 210, 212 vowels: Miami-Cuban English 54–55, 60, 63, 65; Miami-Cuban Spanish 54–55, 60–61, 61, 63, 65 Warlpiri 215 “Who Did Patrick’s Homework” (Moore) 59 Woolard, K. 214 word final obstruent 203–204, 208 word order 3, 114, 122, 167, 215 word-final devoicing 203 word-final position: of /ɾ/ 35, 40–43, 46, 46, 46–49, 47; of /d/ 204; of /l/ 66; /-m/ as plural 161, 163; see also position [χ] (/r/ variant) 89 [x] (/r/ variant) 89–91, 94, 95, 97, 97, 97; evaluations of 91–92; gender and 101–102, 102, 103, 105, 106; integration into the Puerto Rican community and 104, 104–105; life stage and 103–104, 105, 106; position of 98, 105; socioeconomic level and 99–100, 99–100; spectrogram of 97, 97; as stigmatized 106–107 [xr] (/r/ variant) 89–90 Zentella, Ana Celia 3–4