The Routledge Handbook of Spanish as a Heritage Language 2017053256, 9781138833883, 9781315735139

Citation preview

THE ROUTLEDGE HANDBOOK OF SPANISH AS A HERITAGE LANGUAGE

The Routledge Handbook of Spanish as a Heritage Language brings together contributions from leading linguists, educators and Latino Studies scholars involved in teaching and working with Spanish heritage language speakers. This state-of-the-art overview covers a range of topics within five broad areas: Spanish in U.S. public life, Spanish heritage language use and systems, educational contexts, Latino studies perspectives and Spanish outside the U.S. The Routledge Handbook of Spanish as a Heritage Language addresses for the first time the linguistic, educational and social aspects of heritage Spanish speakers in one volume making it an indispensable reference for anyone working with Spanish as a heritage language. Kim Potowski is Professor of Spanish Linguistics and Director of the Spanish Heritage Language Program at the University of Illinois at Chicago, USA.

ROUTLEDGE LANGUAGE HANDBOOKS

Routledge Language Handbooks provide comprehensive and state-of-the-art linguistic overviews of languages other than English. Each volume draws on an international team of leading scholars and researchers in the field. As reference works, the handbooks will be of great value to readers in many different fields; linguistic typology at all levels, general linguists, historical linguists, sociolinguists, and students of the individual languages or language families concerned. The Routledge Handbook of Chinese Translation Edited by Chris Shei and Zhao-Ming Gao The Routledge Handbook of Arabic Lingusitics Edited by Elabbas Benmamoun and Reem Bassiouney The Routledge Handbook of African Linguistics Edited by Augustine Agwuele and Adams Bodomo The Routledge Handbook of Chinese Second Language Acquisition Edited by Chuanren Ke The Routledge Handbook of Arabic Second Language Acquisition Edited by Mohammad T. Alwahary For more information about this series please visit: https://www.routledge.com/Routledge-Language-Handbooks/book-series/RLH

THE ROUTLEDGE HANDBOOK OF SPANISH AS A HERITAGE LANGUAGE

Edited by Kim Potowski Spanish List Advisor: Javier Muñoz-Basols

First published 2018 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business  2018 selection and editorial matter, Kim Potowski; individual chapters, the contributors The right of Kim Potowski to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Names: Potowski, Kim, editor. | Munoz-Basols, Javier, editor. Title: The Routledge handbook of Spanish as a heritage language / edited by Kim Potowski. Description: New York, NY : Milton Park, Abingdon, Oxon, [2018] | Includes bibliographical references and index. Identifiers: LCCN 2017053256| ISBN 9781138833883 (hardback : alk. paper) | ISBN 9781315735139 (ebook) Subjects: LCSH: Spanish language—Social aspects. | Spanish language--Influence on English. | English language—Influence on Spanish. | Bilingualism—Handbooks, manuals, etc. | Native language—Study and teaching. | Heritage language speakers—Education. | Language and culture. | Group identity. | Sociolinguistics. Classification: LCC PC4074.75 .R68 2018 | DDC 460—dc23 LC record available at https://lccn.loc.gov/2017053256 ISBN: 978-1-138-83388-3 (hbk) ISBN: 978-1-315-73513-9 (ebk) Typeset in Bembo by Swales & Willis Ltd, Exeter, Devon, UK

CONTENTS

Notes on contributors

ix

  1 Spanish as a heritage/minority language: a multifaceted look at ten nations Kim Potowski

1

PART I

Social issues

15

  2 A historical view of US latinidad and Spanish as a heritage language Andrew Lynch

17

  3 Spanish in U.S. language policy and politics Phillip M. Carter

36

  4 Spanish language use, maintenance, and shift in the United States Devin Jenkins

53

  5 Spanish in linguistic landscapes of the U.S. José M. Franco Rodríguez

66

  6 Linguistics and Latino studies: intersections for the advancement of linguistic and social justice Lourdes Torres   7 Spanish and identity among Latin@s in the U.S. Rachel Showstack

v

78 92

Contents

  8 Spanish as a heritage language and the negotiation of race and intra-Latina/o hierarchies in the U.S. Rosalyn Negrón   9 Queering Spanish as a heritage language Holly Cashman and Juan Antonio Trujillo PART II

107 124

Linguistic studies

143

10 Morphology, syntax, and semantics in Spanish as a heritage language Silvina Montrul

145

11 Heritage Spanish phonetics and phonology Rebecca Ronquest and Rajiv Rao

164

12 The lexicon of Spanish heritage language speakers Marta Fairclough and Anel Garza

178

13 Heritage Spanish pragmatics Derrin Pinto

190

14 Neurolinguistic approaches to Spanish as a heritage language Harriet Wood Bowden and Bernard Issa

203

15 Psycholinguistic perspectives on heritage Spanish Jill Jegerski

221

16 Child heritage speakers’ morphosyntax: rate of acquisition and crosslinguistic influence Naomi Shin

235

17 Sociolinguistic variation in U.S. Spanish Rena Torres Cacoullos and Grant M. Berry

254

18 Spanish dialectal contact in the United States Daniel Erker

269

19 Understanding and leveraging Spanish heritage speakers’ bilingual practices Almeida Jacqueline Toribio and Leah Durán

vi

284

Contents PART III

Educational issues

299

20 Towards the development of an analytical framework for examining goals and pedagogical approaches in teaching language to heritage speakers Guadalupe Valdés and María Luisa Parra

301

21 Outcomes of classroom Spanish heritage language instruction 331 Melissa A. Bowles 22 Critical language awareness and Spanish as a heritage language: challenging the linguistic subordination of US Latinxs Jennifer Leeman

345

23 Differentiated teaching: a primer for heritage and mixed classes Maria Carreira and Claire Hitchins Chik

359

24 Key issues in Spanish heritage language program design and administration 375 Sara Beaudrie 25 Spanish for the professions and community service learning: applications with heritage learners Ann Abbott and Glenn Martínez 26 Spanish heritage speakers studying abroad Rachel Shively

389

403

27 Expanding the multilingual repertoire: teaching cognate languages to heritage Spanish speakers Ana Carvalho and Michael Child

420

28 Developing Spanish in dual language programs: preschool through twelfth grade Kathryn Lindholm-Leary

433

29 What do we know about U.S. Latino bilingual children’s Spanish literacy development? Jeannette Mancilla-Martinez

445

vii

Contents PART IV

Spanish as a minority/heritage language outside of the U.S.

461

30 Spanish in the Antipodes: diversity and hybridity of Latino/a Spanish speakers in Australia and Aotearoa-New Zealand Criss Jones Díaz and Ute Walker

463

31 Spanish as a heritage language in Italy Milin Bonomi and Laura Sanfelici

479

32 Spanish as a heritage language in Germany Carmen Ramos Méndez-Sahlender

492

33 Spanish as a heritage language in Switzerland Verónica Sánchez Abchi

504

34 Chilean Spanish speakers in Sweden: transnationalism, trilingualism, and linguistic systems Maryann Parada

517

35 Spanish as a minority/heritage language in Canada and the UK Martin Guardado

537

36 Language issues for US-raised ‘returnees’ in Mexico Clare Mar-Molinero

555

Index 568

viii

CONTRIBUTORS

Ann Abbott is Associate Professor/Academic Professional at the University of Illinois, UrbanaChampaign, USA and Director of Undergraduate Studies in the Department of Spanish and Portuguese. Her research and publications focus on community service learning, social entrepreneurship, business Spanish, and a growing attention to advocacy and civic engagement. Sara Beaudrie is an associate professor of Spanish Linguistics in SILC at Arizona State University, USA and director of the Spanish Heritage Program. She is the co-editor of two books on heritage language research and pedagogy published by Georgetown University Press, and co-author of Heritage Language Pedagogy: Research and Practice (McGrawHill, 2014). Grant M. Berry is a PhD candidate in Spanish and Language Science at Penn State University, USA. His research interests include language variation and change, sociolinguistics, bilingualism, language processing, and laboratory phonology. Milin Bonomi is a postdoctoral fellow in Hispanic Linguistics at the University of Milan, Italy. Her research focuses on Latino communities in Italy from a sociolinguistic and an educational perspective. Melissa A. Bowles is an associate professor in the Department of Spanish and Portuguese at the University of Illinois at Urbana-Champaign, USA. Her main research interest is in classroom second and heritage language acquisition, and the ways in which instruction differentially affects the two learner groups. Maria Carreira is a professor of Spanish at California State University, Long Beach and the Co-director of the National Heritage Language Resource Center at UCLA, USA. She is a co-author of six Spanish-language textbooks and of Voces: Latino Students on Life in the United States (Praeger, 2014). Phillip M. Carter is an associate professor of Linguistics and English at Florida International University, USA. He has authored numerous articles in sociolinguistics and linguistic ix

Contributors

anthropology, and is co-author of Languages in the World: How History, Culture, and Politics Shape Language (Wiley-Blackwell, 2016). Ana Carvalho is a professor of Spanish and Portuguese linguistics at the University of Arizona, USA, where she also directs the Portuguese Language Program. Her research interests include language variation and change in contact situations, including contexts of bilingualism and language acquisition. Holly Cashman is associate professor of Spanish and chair of Languages, Literatures, and Cultures, core faculty in the Women’s Studies Program, and coordinator of Queer Studies at the University of New Hampshire, USA. Her current research focuses on language practices of LGBTQ+ Latinxs. She is the author of Queer, Latinx, and Bilingual (Routledge, 2018). Michael Child is an assistant professor in the Spanish and Portuguese Department at Brigham Young University, USA. From 2014 to 2016 he was an Assistant Professor (Universitair Docent) of Portuguese and Linguistics at Leiden University in the Netherlands. His interests include second/third language acquisition, bilingualism, Portuguese and Hispanic sociolinguistics, Spanish as a heritage language, and second language pedagogy.​ Leah Durán is an assistant professor of Teaching, Learning, and Sociocultural Studies in the College of Education at the University of Arizona, USA. Her research focuses on bilingualism and biliteracy in young children, and the development of pedagogical approaches which support them. The findings appear in journals such as the Journal of Literacy Research, Journal of Early Childhood Literacy, International Multilingual Research Journal, and the Bilingual Research Journal. Daniel Erker is an assistant professor of Spanish and Linguistics at Boston University, USA, where he is the director of the Spanish in Boston Project. Erker earned his MA in linguistics at CUNY Graduate Center, USA, and his PhD in linguistics at New York University. His research focuses on language variation, contact, and change, especially in Spanish spoken in the USA. Marta Fairclough is Professor of Spanish Linguistics and Director of the Spanish Heritage Language Program at the University of Houston, USA. Her research focuses on Heritage Language Education and US Spanish. She has published a book, two co-edited volumes, and numerous book chapters and articles. José M. Franco Rodríguez is an associate professor in the Department of Communication, Languages and Cultures at Fayetteville State University, USA. His research on Spanish in linguistic landscapes of the USA started in Los Angeles County in 2002 while he was a bilingual teacher in the county. This investigation was followed by a study on Miami-Dade County’s linguistic landscape. Later he applied his methodology to analyse and contrast the linguistic landscape of Almería (Spain). He uses his research findings to teach about Spanish and Hispanics in both Spanish and Education courses. Anel Garza is a faculty member at San Jacinto College, USA. Her areas of teaching include Modern Languages, English to Speakers of Other Languages, and Developmental Studies. Her research interests focus on Heritage and Language Education and US Spanish. x

Contributors

Martin Guardado is an associate professor of Applied Linguistics and the Academic Director of the English Language School at the University of Alberta, Canada. His research interests include heritage language development and English for academic purposes. Claire Hitchins Chik is an associate director of the Title VI National Heritage Language Resource Center at UCLA, USA. Her most recent project is the development of an online certificate, “Teacher Training for the 21st Century: Teaching Heritage Languages.” With her colleagues Olga Kagan and Maria Carreira, she is a co-editor of The Routledge Handbook of Heritage Language Education: From Innovation to Program Building (Routledge, 2017). Earlier in her career, Hitchins Chik worked with a Chinese EFL teacher from Yunnan Province on his autobiography, coauthoring, and getting the book, Mr. China’s Son (1993, 2nd ed. 2002, Westview Press), to publication. Bernard Issa (PhD, University of Illinois at Chicago) is Assistant Professor of Spanish at the University of Tennessee-Knoxville, USA. His research examines how individual difference variables (e.g. motivation, attention, working memory) relate to the processing and development of second language grammar. Jill Jegerski is an assistant professor of Spanish and SLATE (Second Language Acquisition and Teacher Education) in the Department of Spanish and Portuguese at the University of Illinois at Urbana-Champaign, USA. Her primary research interests include bilingual and non-native sentence processing, psycholinguistic research methods, and Spanish as a heritage language. Devin Jenkins (PhD, University of New Mexico) is Associate Professor of Spanish and Chair of the Department of Modern Languages at the University of Colorado Denver, USA. His research focuses on Spanish in the southwest USA, bilingualism, language contact, and population trends. Criss Jones Díaz is a senior lecturer in the School of Education at the Western Sydney University, Australia. Her research and publication interests are primarily in critical and cultural studies with an emphasis on bi/multilingualism, languages and literacies education, and identity negotiation in contexts of diversity and difference. Jennifer Leeman is Professor of Spanish at George Mason University (Fairfax, Virginia, USA) and Research Sociolinguist at the US Census Bureau. Her current research focuses on language and race in the US Census, linguistic anthropological perspectives on Spanish in the USA, language policy, and the sociopolitics of heritage language education. Kathryn Lindholm-Leary is Professor Emerita of Child and Adolescent Development at San Jose State University, USA, where she taught for 28 years. Kathryn has worked with over 75 two-way and developmental bilingual programs from PreK through 12 over the past 30 years and has written books, chapters, and journal articles, and has given presentations to researchers, educators, and parents on the topics of dual language education and child bilingualism. More recently, she worked with the National Academy of Sciences in their report on the development of English/dual language learners. Kathryn has served on advisory boards or as consultant to federal and state departments of education, various professional organizations, and other agencies, school districts, and schools. xi

Contributors

Andrew Lynch, University of Miami, USA, researches language variation, social and structural consequences of Spanish in contact with other languages, ideological and sociolinguistic dimensions of Spanish in the USA, and heritage language acquisition and pedagogy. He is Editor in Chief of Heritage Language Journal. Jeannette Mancilla-Martinez is Associate Professor of Literacy at Vanderbilt University’s Peabody College of Education and Human Development. Her program of research is focused on advancing students’ language and literacy outcomes, with her work spanning the toddlerhood years through adolescence. Clare Mar-Molinero is Professor of Spanish Sociolinguistics and Director of the Centre for Mexico-Southampton Collaboration (MeXsu) at the University of Southampton, UK. She has published widely on Spanish as a global language, language policies, urban multilingualism, and linguistic superdiversity. Glenn Martínez is Professor of Hispanic Linguistics and Adjunct Professor of Nursing at The Ohio State University, USA. His research and publications focus on language and healthcare, Spanish as a heritage language in the USA, and language policy and planning. Carmen Ramos Méndez-Sahlender is Professor of Spanish and Head of the Centre for Innovation in Teaching and Learning at the University of Applied Languages, Munich, Germany. Her research focuses on language teaching at university, learner and teacher beliefs, and heritage languages. Silvina Montrul is currently Professor of the Department of Spanish and Portuguese and Professor of Linguistics at the University of Illinois at Urbana-Champaign, USA, where she is also Director of the University Language Academy for Children and Director of the Second Language Acquisition and Bilingualism Lab. Rosalyn Negrón is an associate professor of Anthropology at the University of Massachusetts, Boston, USA, specializing in urban social anthropology. Rosalyn studies the interpersonal dimensions of race and ethnicity in diverse cities, with a special focus on Latina/os, ethnic flexibility, and social networks. Maryann Parada is Assistant Professor of Hispanic Linguistics at California State University Bakersfield, USA. She recently earned her PhD from the University of Illinois at Chicago, USA, where her dissertation examined the lexical profiles and language maintenance of trilingual Spanish heritage speakers in Stockholm. Her research interests language attitudes and ideologies, names and identity, family language policy, and heritage language pedagogy. María Luisa Parra is Spanish Senior Preceptor at the Department of Romance Languages and Literatures at Harvard University, USA. Her main research interests are pedagogy for Spanish as a heritage and foreign language, and the impact of immigration in Latino children’s bilingual development and process of school adaptation. Derrin Pinto (PhD, University of California at Davis, USA) is currently Professor of Spanish Linguistics and Chair of the Department of Modern and Classical Languages at the University xii

Contributors

of Saint Thomas in Minnesota, USA. He has published studies involving different areas of pragmatics, discourse analysis, and second language acquisition. Kim Potowski is Professor of Spanish Linguistics in the Department of Hispanic & Italian Studies at the University of Illinois at Chicago, USA, where she holds appointments in Latin American and Latino Studies, Curriculum and Instruction, and an affiliation with the Social Justice Initiative. She has directed the Spanish for Heritage Speakers program since 2002 and is the founding director of its summer study abroad program in Oaxaca, Mexico. Her research focuses on Spanish in the USA, including factors that influence intergenerational language transmission and change as well as connections between language and ethnic identity. She has authored, co-authored, and edited over 12 books including Intra-Latino language and identity: MexiRican Spanish (John Benjamins, 2016), El español de los Estados Unidos (Cambridge University Press, 2015), Heritage Language Teaching: Research and Practice (McGraw-Hill, 2014), Language Diversity in the USA (Cambridge University Press, 2011), and Language and Identity in a Dual Immersion School (Multilingual Matters, 2007). Her advocacy for dual language education in promoting bilingualism and academic achievement was the focus of her 2013 TEDx talk “No child left monolingual.” Rajiv Rao is an associate professor of Spanish at the University of Wisconsin-Madison, USA. He employs laboratory approaches to address the phonology of heritage speakers, AfroHispanic intonation, the prosody-pragmatics interface, prosodic phrasing, and intonational phonological theory. Rebecca Ronquest is an assistant professor of Spanish at North Carolina State University, USA. Her research focuses on the phonetic/phonological systems of heritage speakers of Spanish, the Spanish vowel system, laboratory approaches to phonology, and Spanish in the Southeastern USA. Verónica Sánchez Abchi is a researcher at the Institute of Pedagogical Research and Documentation (Neuchâtel, Switzerland) and teaches Spanish didactics at the Higher School of Education in Valais, Switzerland. Her research interests include writing competences in L1, L2, and in heritage languages, and language acquisition in multilingual contexts. Laura Sanfelici is a researcher in Spanish Language and Translation at the University of Genoa, Italy. Her research interests focus on the maintenance of the language of origin by Hispano-American immigrants attending Italian Middle School, and the relationship between languages and immigration. Noami Shin is an associate professor of Linguistics and Spanish at the University of New Mexico, USA. Her primary interests include child language acquisition, bilingualism, language contact, and sociolinguistics. Her research focuses on patterns of morphosyntactic variation, examining how these patterns are acquired during childhood and how they change in situations of language contact. Her articles have appeared in journals such as Journal of Child Language, Language Variation and Change, Language in Society, International Journal of the Sociology of Language, Language Acquisition, and Spanish in Context. Rachel Shively (PhD, University of Minnesota, USA) is an associate professor of Spanish and Applied Linguistics at Illinois State University, USA. Her widely published research focuses xiii

Contributors

on second language pragmatics and study abroad. In 2011 she was awarded the ACTFL-MLJ Pimsleur Award for Research in Foreign Language Education. Rachel Showstack is an assistant professor of Spanish and Linguistics at Wichita State University, USA. She has published articles in Spanish in Context, Language and Intercultural Communication, The Journal of Spanish Language Teaching, and The Journal of Language, Identity, and Education. Almeida Jacqueline Toribio is Professor of Linguistics in the Department of Spanish and Portuguese at The University of Texas, USA. Her dossier reflects scholarship in the areas of language contact and variation, and a trajectory from theoretical to more empirically based approaches. Her research endeavours have been supported by the National Endowment for the Humanities, The Russel Sage Foundation, and the Mellon Foundation, among others, and the findings appear in handbooks, compendia, and journals, including The International Journal of the Sociology of Language, Bilingualism: Language and Cognition, Social Science Quarterly, International Journal of Bilingualism, Lingua, and Linguistic Inquiry. Lourdes Torres is a professor and Chair of Latin American and Latino Studies at DePaul University, USA. She is editor of the journal, Latino Studies. She is the author of Puerto Rican Discourse (Routledge, 1997) and co-editor of Tortilleras: Hispanic & Latina Lesbian Expression (Temple, 2003) and Third World Women and the Politics of Feminism (Indiana University Press, 1991). Her recent essays on language and culture have appeared in Meridians, MELUS, Centro Journal, and International Journal of Bilingualism. Rena Torres Cacoullos is Professor of Spanish and Linguistics at Penn State University, USA. Her work aims to discover grammatical structure through quantitative analysis of natural production data in its social context. Juan Trujillo is coordinator of World Languages and Cultures in the School of Language, Culture, and Society at Oregon State University, USA. His work explores intersections of Mormonism, queer identity, and latinidad from an autoethnographic perspective.​ Guadalupe Valdés is the Bonnie Katz Tenenbaum Professor of Education at Stanford University, USA. Her work has focused on the English-Spanish bilingualism of Latinos in the USA, and on discovering and describing how two languages are developed, used, and maintained in Latino immigrant communities. Ute Walker is a senior lecturer in the School of Humanities at Massey University, New Zealand (Manawatu campus) with research interests in the areas of language and identity, bilingualism, and digitally facilitated language learning. Harriet Wood Bowden (PhD, Georgetown University, USA) is Associate Professor of Spanish at the University of Tennessee-Knoxville, USA. Her research investigates first, second, and heritage language acquisition and neurocognition, and the interaction of multiple learnerinternal and external factors influencing these processes.

xiv

1 SPANISH AS A HERITAGE/ MINORITY LANGUAGE A multifaceted look at ten nations Kim Potowski the university of illinois at chicago, usa

Introduction It is estimated that the combined total number of “native” Spanish speakers1 around the world is between 437 million (Ethnologue 2016, www.ethnologue.com/statistics/size) and 472 million people (Instituto Cervantes 2016), making it the second most commonly spoken language in the world. It is spoken by sizeable populations over 20 countries; the top ten countries with the largest numbers of Spanish-speaking inhabitants are displayed in Table 1.1. This table assumes that all residents in each country (except the U.S.) are Spanish-speaking, although this is not totally accurate. For example, although it is estimated that over 95% of the population in Mexico speaks Spanish, the number in Paraguay is closer to 69% (Moreno Fernández & Otero Roth 2006). The presence of the U.S. on this list surprises some people, particularly its position as home to the third largest Spanish-speaking population among the world’s nations.2 Before continuing, I should make clear the problematic nature of two principal concepts discussed in this chapter and the volume overall. The most basic is the idea of “Spanish.” As noted by Otheguy, García, and Reid (2015, p. 286), named languages such as “Spanish,” “English,” Table 1.1  Estimated Spanish-speaking populations by country Country

Estimated number of Spanish-speaking habitants

Source

  1 Mexico   2 Colombia   3 United States   4 Spain   5 Argentina   6 Peru   7 Venezuela   8 Chile   9 Guatemala 10 Ecuador

120,000,000 49,100,000 46,500,000 46,400,000 43,500,000 31,800,000 31,000,000 18,300,000 16,900,000 16,200,000

2015, www.inegi.org.mx 2017, www.dane.gov.co 2015, www.census.gov 2015,www.ine.es 2016, www.indec.gov.ar 2017, www.inei.gob.pe 2016, www.ine.gov.ve 2017, www.ine.cl 2017, http://fadep.org 2015, www.inec.gob.ec

1

Kim Potowski

and “Quechua” are not actually independent linguistic objects, but rather constructs that exist only socially according to the definitions and affiliations of its speakers. Similarly, the concept of “countries” is problematic. The emergence of most modern nation-states in the 19th century resulted from wars and annexations, and as a result, many populations of individuals with similar ethnic, religious, and/or linguistic affiliations were arbitrarily split by lines drawn on a map. Thus, when contemplating ideas about “Spanish” and about “countries/nations,” readers should keep present the fact that these constitute a kind of conceptual shorthand that do not always correspond to realities as individuals experience them. Despite these important limitations, I believe that much can be learned about what we generally understand to be the Spanish language via a study of those who claim to speak it in different parts of the world. In most of the 21 nations where Spanish is an official or national language (either de facto or de jure), it exists in contact with at least one other language. For example, it is estimated that 43% of the population of Guatemala, 37% of that of Bolivia, 35% of Peru, and 5.4% of the population of Mexico speaks an indigenous language. Yet in these contexts, Spanish is the dominant language of society – the one typically taught in schools, used in the media, and necessary for economic stability. In other places, Spanish is in contact with another language, but it has equal official status with that language. Such is the case in three autonomous regions of Spain (the Basque Country, Catalonia, and Galicia) where children are frequently educated through varying proportions of Basque, Catalan, and Galician respectively. However, despite its co-official status, in these autonomous regions Spanish is typically seen as having greater social value. For example, there are few if any adults who are monolingual in Basque (Cenoz 2008) or Catalan (Boix-Fuster & Sanz 2008),3 and all Spaniards are obliged to know Spanish but not any other language (del Valle 2000). In any case, in places like these around the world where Spanish is the majority and high status language – both where Spanish exists largely by itself as well as where it is in contact with but shares equal legal status with another language – there has been a great deal of linguistic research documenting micro- and macro-level linguistic features. However, in locations where Spanish is a minority language, and especially where it is a minoritized language, the sociolinguistic reality is very different. The term minority is easy enough to define: it refers to a quantity less than 50%. The Cervantes Institute (Instituto Cervantes 2016) calculated a total of 45.8 million native Spanish speakers residing in non-Spanish-speaking countries (i.e. where Spanish is a minority language) including Australia, Russia, and Switzerland. However, this source likely undercounted the Spanish-speaking population in the U.S., listing it at 42.7 million, which is 3.8 million short of the 46.5 million listed in Table 1.1. If we add these 3.8 million U.S. Spanish speakers to the 45.8 million estimated by the Cervantes Institute, we get a possibly more accurate total of 49.6 million Spanish speakers around the world residing in non-Spanish-speaking countries. It is likely that no language besides English has as many “native” speakers living in places where it is not the common tongue. Clearly the largest number of Spanish speakers in this situation (46.5 million out of 49.6 million, or 94%) live in the U.S., constituting approximately 15% of its national population. Figure 1.1 shows the proportion of U.S. Spanish-speaking residents by county. The areas marked with the darkest shade have a proportion of Spanish speakers greater than 68%, and some areas demarcated with the next darkest shade have over 50%. We notice that in counties in the Southwest, southern Florida, and rural Washington, Idaho, and Kansas, Spanish speakers form the majority of the local population. Despite these local majority concentrations, on the national level Spanish remains a minority and minoritized language in the U.S. Minoritized means that in addition to being in the minority and having no legal status or support, Spanish is marginalized and sometimes outwardly discriminated against. Both local and national discourses frequently frame it as inferior, problematic, and a 2

Spanish as a heritage/minority language

Figure 1.1  P  ercent Spanish-speaking, by county (Source: Modern Language Association language maps, 2010 American Community Survey)

threat to national unity and to children’s advancement in school. These negative forces lead many individuals to abandon Spanish and not pass it down to future generations (see Potowski 2010 for evidence that this is in fact the outcome for almost all non-English languages in the U.S.). Around the world, it is common for minority and minoritized languages, which I will abbreviate as minority/ized, to be replaced by the locally dominant language. Lambert (1977) described this phenomenon as subtractive bilingualism, in which lower status, negative attitudes, and lack of educational opportunities in the language lead to its weakening or total replacement. Thus, unlike studying Spanish in contact with Guaraní in Paraguay or with Galician in Spain, for example, where Spanish is not minority/ized, it is reasonable to hypothesize that there will be different linguistic and social features of Spanish where it is a minority/ized language – not least of which is the fact that under these conditions, the language is very frequently on a path towards loss. In the U.S., non-English languages are commonly referred to as heritage languages. The first part of the 21st century saw the creation of the National Heritage Language Resource Center in addition to numerous Spanish for heritage speakers educational programs (see chapters by Beaudrie and by Valdés & Parra, this volume). However, some scholars reject this term, embracing the position of García (2005) that the word heritage “connotes something that one holds onto vaguely as one’s remembrances but certainly not something that is used in the present or that can be projected into the future” (2005, p. 601) and that: [a]s the languages of the world transcend their traditional territories and English spreads, languages other than English in the U.S. are being controlled through a shift in discourse . . . perhaps best exemplified by the silencing of the word bilingual and replacing it with heritage languages. 2005, p. 605 3

Kim Potowski

Despite these valid criticisms, the title of this volume and of the present chapter utilize the term heritage, although chapter authors use the terms of their choice. It is important to note here that not all minority languages are minoritized. For example, English is spoken natively in Mexico by a relatively small number of people, many of whom refer to themselves as expats (even though they fit most definitions of immigrant). While proficiency in Spanish is necessary for most avenues of success in Mexico, many of these individuals are able to secure employment in white-collar professions and are not shamed for speaking English. Quite the opposite is true, in fact, with the worldwide prestige of English usually granting them high status and earning potential. Even so, as Anderson and Solis (2014) and Mar-Molinero (this volume) show, many Mexican-origin individuals who (in)voluntarily move to Mexico after growing up in the U.S. do not necessarily enjoy all the benefits of native English proficiency that one might expect, which is likely related to their lower socioeconomic position as the children of economically motivated migrants to the U.S. This forces an examination of the role of socioeconomic status in language prestige. Middle class Chilean political exiles in Sweden (Neilson Parada, this volume), for example, likely present a different sociolinguistic profile than poorer Ecuadorian immigrants in Italy (Bonomi and Sanfelici, this volume), the former perhaps more likely to maintain Spanish intergenerationally and the latter to lose it. The purpose of this volume is to explore Spanish as a minority/ized language in different parts of the world, its authors examining macro and/or micro elements of Spanish as well as issues that impact speakers’ use of the language. Most work to date on Spanish as a minority/ized language has focused on adults and has been produced in the U.S. by two groups: linguists, and high school and postsecondary language instructors in heritage Spanish programs. In addition to these two important areas of scholarship, this collection presents the perspectives of researchers working in the fields of primary school education (grades preschool through eight) and Latina/o Studies. The focus thus moves beyond Spanish as a minority/heritage language to include a variety of linguistically relevant considerations about the people who speak it, and includes work focused on adults and on children. I divide this introductory chapter into the same four Parts that structure the volume: (1) social issues, (2) linguistic studies, (3) educational issues, and (4) countries outside of the U.S. These divisions are somewhat arbitrary in that there are obvious intersections between social issues, education, and linguistics both within and outside of the U.S. For example, Jones Díaz and Walker’s chapter on Spanish-speakers in Australia and New Zealand combines social and educational issues, while Bowles’ chapter detailing outcomes of adult classroom heritage language instruction involve some linguistic descriptions of their Spanish systems. Despite these thematic overlaps, the Parts are meant to help readers conceptualize broader themes. Countries outside of the U.S. appear in their own Part because research on Spanish in these locations is scant; except for the lexical availability analysis in Neilson Parada’s chapter about Chile, and Mar-Molinero’s chapter about U.S.-raised Mexican-origin individuals in Mexico, these chapters are limited to offering an introductory overview of how Spanish came to be spoken in each country and considerations regarding its continued vitality.

Part I: Social issues The nine chapters in this Part address a variety of social issues related to Spanish in the U.S. The chapter by Lynch presents an analysis of the forces that brought the construct of latinidad, and the Spanish language as an essential feature of it, into the U.S. cultural landscape during the late 19th and early 20th centuries. He traces the early presence of “Hispanic-positive” novels and Hollywood actors along with the incorporation of New Mexico into the U.S. and the 4

Spanish as a heritage/minority language

scholarship of Aurelio Espinosa, but shows that by the 1930s the country had shifted to more Hispanophobic realities, the likes of which would prompt actress Margarita Carmen Cansino to change her name to Rita Hayworth. Moving to more current times up through April 2017, Carter outlines what he calls the “paradox of Spanish in the United States” via an examination of two phenomena: the use of Spanish in political discourse, and language policies toward Spanish in the U.S. For example, he notes that “the act of speaking Spanish figures differently depending on who is doing it” and that race is always present in questions of language and politics (also a theme in Negrón’s chapter). He concludes with an important consideration of the role of Spanish Language Academies in supporting Spanish in such hostile environments, noting that to date, they have generally shown disdain for the way Spanish is spoken in the U.S. Both of these chapters offer timely insight into the current resurgence of nationalist, anti-immigration discourse in the U.S. under the Trump administration. After this consideration of the effects of the political and discursive landscape in the U.S. on the use of Spanish, Jenkins’ chapter presents a demographic update of current Hispanophone populations around the country. Interesting trends include that Mexicans are the fastest growing Spanish-speaking group in the New York City area (heretofore dominated by Puerto Ricans and Dominicans); Orlando has become a nouveau Puerto Rican enclave; and the Eastern seaboard is the region that has seen the most significant numerical increase in Hispanic residents between the 2000 and 2010 Censuses. He then moves into a deft analysis of several trajectories that lead to predictions about the future of the language. Franco Rodríguez’ chapter follows nicely with a summary of work on the presence of Spanish in the “linguistic landscape” of the U.S., with many studies finding that the visibility of Spanish is disproportionately low in relation to the size of the Hispanic population, in part due to an increasingly English-centered ideological environment. Next, the chapter by Torres explores connections between Spanish-speakers and social justice and the ways in which language functions in the life of Latino communities. Focusing on how the fields of sociolinguistics and Latino studies have responded to prejudicial, status quo notions of Latino ways of speaking, she argues for continued advocacy in advancing language rights along the lines of work done by Shana Poplack, Ana Celia Zentella, Bonnie Urciuoli, Guadalupe Valdés, and Ofelia García. Similarly, Showstack’s chapter explores connections between Latino identity and language, unpacking essentialized assumptions that Spanish has a uniform social meaning for all Latinos and instead demonstrating several ways in which individuals use Spanish to represent identities in interaction within specific social contexts such as family, community, work, and classrooms – this last context a potent site for potentially (re) constructing positive opinions about U.S. Spanish varieties. Negrón further complicates relationships between Latinos and their Spanish varieties by incorporating an analysis of “race” (the purpose of these quotation marks is explained in the chapter) and the social hierarchy it generates both in the U.S. and in Latin America. Through her literature review as well as examples from her own work in New York City, she demonstrates that Latinos frequently negotiate racial categorization through Spanish in both affiliative and differentiating ways, the latter often couched in racial terms. Specifically, Latinos sometimes judge each other’s Spanish as inferior because they see certain people as racially inferior. Finally, focusing on commercially produced textbooks for Spanish heritage speaker courses, Cashman and Trujillo reveal the rampant heteronormativity that serves to erase the experiences of and potentially alienate Latino LGBTQ+ students, simultaneously doing nothing to address racialized violence, economic injustice, and racial privilege that impact queer communities of color. Since these chapters were completed, the Spanish version of the U.S. government website was taken down for several months after Trump’s inauguration; draconian migratory regulations 5

Kim Potowski

were threatened and some enacted; and general moves away from national and linguistic permeability have increased. Many of the chapters in this Part offer recommendations for practice; it is of crucial importance for readers to think about and act on concrete suggestions to combat injustices against Spanish-speaking individuals and communities.

Part II: Linguistic studies In places where Spanish is the only language spoken, it seems logical to posit that all children learn it in more or less the same way, going through similar stages of phonological and morphosyntactic development in the same order. However, linguistic forms and frequencies change over time, indicating that the systems of children differ from those of their parents. There are also obvious differences in the Spanish spoken in different geographic locations and by varying social classes, genders, and ethno/racial groups, to name a few categories. This kind of sociolinguistic variation continues to be amply documented around the Spanish-speaking world, for example in the work compiled in Díaz-Campos (2011) as well as presentations at the biannual Workshops on Spanish Sociolinguistics and other venues. But when another language is present, the two linguistic systems interact and complicate acquisition, use, and social meaning. Do monolingual and bilingual Spanish systems differ, and if so how? And do bilingual Spanish systems differ in cases in which: (a) Spanish is the majority prestige of the two languages, (b) Spanish has equal prestige with the other language, or (c) Spanish is a minority language? The ten chapters in this Part address situation (c), presenting cogent reviews of relevant work. They all focus on Spanish in the U.S. because this is where the majority of work on the topic has been carried out; comparative studies are sorely needed in other countries as well. I begin this summary with an important quote from Pinto’s chapter that all discussions of heritage languages should keep in mind: [c]omparing SHL [Spanish as a heritage language] to monolingual varieties of Spanish should be considered an analytical exercise that does not necessarily imply any of the following: 1) That heritage speakers of Spanish strive, or should strive, to speak like monolingual speakers; 2) That the monolingual variety is superior; 3) That any differences reported for SHL represent deficiencies that need to be corrected; 4) That formal registers are more valid than informal registers, although one or the other may be more appropriate in a particular context. Pinto, this volume The Part begins with Montrul’s chapter on the grammatical aspects of U.S. Spanish heritage language systems, carefully detailing aspects of the highly variable inflectional instances of morphology; interfaces between morphosyntax, semantics, and pragmatics; and complex syntax that these speakers evidence. It walks readers through clear explanations with helpful examples and compares important points in summative tables, such that experts and more novice readers alike can appreciate the complexity and also the frequent systematicity of these bilingual systems. The Part then moves from acquisition/processing into production. Ronquest and Rao summarize current research in heritage Spanish phonetics and phonology, including Voice Onset Times of the consonants /ptk/ as well as very recent work on rhotics (‘r’ sounds). Regarding Spanish vowels, they present interesting findings that even minimal contact with English can affect them, as can whether a task is formal or informal. The penultimate section of the chapter summarizes work in the most understudied area to date, heritage speaker prosody (intonation, stress, and rhythm), while the final section presents a summary of where heritage speaker 6

Spanish as a heritage/minority language

phonology is more like that of monolingual speakers and where it is not, leading to “two seemingly contradictory observations about heritage Spanish pronunciation: ‘sounding’ like [native Spanish speakers] on the one hand, but having a ‘heritage accent’ on the other.” Moving to the lexicon, Fairclough and Garza first dispense with the mistaken notion that vocabulary is “simply memorized lexical units . . . consisting of individual word forms” by explaining the various types of lexical knowledge required to use a word productively. The authors then describe the vast dialectal variation present in U.S. Spanish lexicon, the role of the age of acquisition on the lexicon, and studies on both receptive and productive lexicon with U.S. Spanish speakers. They argue that heritage language instruction should focus on building students’ lexicon because it has been correlated with greater global language proficiency as well as a reduction in students’ linguistic insecurity. The chapter by Pinto explores the relatively scant published work on the pragmatic systems of U.S.-raised Spanish speakers, including the discourse marker use of Chicano children in southern California and of Puerto Rican adults in New York, Miami, and Chicago; the use of non-canonical grammar structures (influenced by English) for requests and complaints; communicative strategies in longer oral narratives; and the use of tú and vos in Houston. He notes that research on pragmatic phenomena – which exhibit unwieldy variety – would benefit greatly from close comparisons with monolingual Spanish and monolingual English data, from a combination of quantitative and qualitative analysis, and the incorporation of interpretation studies instead of just production. Bowden and Issa discuss neurophysiological investigations of heritage language online processing of Spanish (meaning that it takes place in real-time) during recognition and production tasks. These studies use functional magnetic resonance imaging (fMRI), functional near infrared spectroscopy (fNIRS), and event-related potentials (ERP), all of which measure changes in electrical activity or blood flow in the brain. Main findings include that heritage speakers’ two languages may be differentiated on a fine-grained neurological level, yet there is a high degree of overlap in neural substrates that underlie their two languages. They also show increased activation of particular regions of the brain related to executive function when switching between their two languages. Also centered on real-time processing methodology such as timed lexical decision tasks and eye tracking, Jegerski’s chapter elucidates the state of the art in psycholinguistic studies of heritage Spanish systems, including the intriguing finding that heritage speakers look more like second language learners in off-line tasks but more like monolingually raised native speakers in on-line tasks. The chapter by Shin also looks at morphosyntax but focuses on children, finding that in the U.S. context of reduced exposure to Spanish, bilingual children tend to acquire Spanish morphosyntax at a slower rate than monolingual children, and that while English does influence children’s Spanish morphosyntax, as they get older they become increasingly adept at suppressing features that do not correspond to communicative expectations. Such knowledge can assist speech pathologists working with bilingual children, helping them understand what to expect among typically developing bilinguals versus what requires intervention. In addition, the chapter outlines outstanding theoretical issues that are ripe for future research, such as the need to better understand how bilingual children acquire variable grammatical patterns as opposed to categorical ones, as well as when and how bilingual children generalize over lexically restricted patterns. Torres-Cacoullos and Berry examine the understudied area of sociolinguistic variation in U.S. Spanish – that is, correlations between linguistic behavior and speakers’ sociodemographic characteristics. They find patterns that generally replicate those found across the Spanish-speaking 7

Kim Potowski

world, indicating that social factors are just as important in minority language situations as they are in other sociodemographic contexts. They also argue for a data optimization method such as principal component analysis as a way to group speakers on the basis of their linguistic behavior, illustrating with a corpus of New Mexico Spanish. Another type of variation in U.S. Spanish results from the fact that speakers hail from vastly different dialect regions of Latin America. Yet Erker’s chapter on dialectal contact shows that speaker region of origin is only one among many social factors contributing to their linguistic choices. The studies he reviews show both intergenerational continuity and change, as well as the fact that it is typically highly salient linguistic features (such as the pronoun vos and syllable final /s/) that are sensitive to change in the U.S. Finally, Toribio and Durán explore research on macro bilingual practices including codeswitching and the ways in which children are socialized into these practices. They note that many K-12 researchers and classrooms have begun legitimizing code-switching and translanguaging, but the university level has largely not, meaning that when they “[enter] the university Spanish language classroom, heritage speakers are expected to leave their hybrid practices at the door.” In this way, an important resource for developing linguistic and content knowledge among college students is overlooked. The authors advocate for critical language awareness approaches (see Leeman, this volume) that engage Latino students in discussions of bilingual varieties and language mixing practices. At least three areas for future research are suggested by the work in this Part. First, our knowledge would benefit from comparisons of two groups of Spanish-speaking bilingual individuals, both children and adults: those in majority contexts vs. those in contexts where Spanish is a minority language. For example, how does the Spanish spoken by indigenous bilingual speakers in Ecuador compare with that spoken by second generation Colombians in Switzerland or third generation Puerto Ricans in New York? A second area in need of research is the role of contextual factors including societal and familial attitudes. For example, King, Fogle, and Logan-Terry (2008) document that some indigenous Ecuadorian parents believe that early second language exposure confuses children, which leads these parents to promote Spanish only in the home and, concomitantly, to shift away from their indigenous language. Finally, we sorely need more longitudinal studies that follow children over a period of time to see how their Spanish system changes (such as Silva-Corvalán’s 2014 study of her two grandsons).

Part III: Educational issues In the U.S., three-fourths of all Hispanics aged five and older speak Spanish. However, that share is projected to fall to about two-thirds in 2020 (López & González-Barrera 2013). In addition, one out of every four school-age students in the U.S. today is Latino, yet the National Center for Educational Statistics reports that Latino fourth and eighth graders score about two grade levels lower than the national average on tests in math and reading. There is an even larger difference – up to four grade levels – between Hispanic students who are proficient in English and those who are not. We also know that 12% of all U.S. Latino students drop out of high school, compared to 5% of White students. It is painfully obvious that our current educational approaches are not benefitting these students’ Spanish, their English, or their academic achievement. In the fortunate event that these students make it to college, many of them are classified as “heritage Spanish speakers” and take Spanish courses specifically developed based on their strengths and needs. There have been excellent recent collections detailing educational issues related to teaching Spanish to heritage speakers at the college level, including Pascual y Cabo (2016), Fairclough and Beaudrie (2016), and Beaudrie, Ducar, and Potowski (2014). This Part presents a few new explorations of this theme as well as important considerations for promoting 8

Spanish as a heritage/minority language

successful outcomes among Latino children at the K-8 level. Valdés and Parra set the stage by exploring a seven-step process that underlies most attempts at “curricularizing language” and how each step might apply to the teaching of Spanish as a heritage language. Important elements include ideologies of language, race (connecting nicely with the chapter by Negrón), class, and identity (the topic of Showstack’s chapter) as well as theories of second language acquisition and bilingualism. In general, the authors support the goal of raising students’ critical language awareness “as a way to decolonize their thoughts and feelings about their use of the Spanish language, so they can become active users and address their communities’ needs.” Similarly, Leeman’s chapter focuses on critical language awareness, highlighting the unfortunate central role of mainstream educational institutions in the legitimization of linguistic subordination. She challenges educators to examine how they can engage students in questioning dominant ideologies surrounding bilingualism and bidialectalism and alter the status quo. The Part then moves from these larger curricular issues to more micro classroom-level themes. As explained by Bowles, we have several decades of research on instructed second language acquisition – most using a pre-test, treatment, and post-test design – yet few such studies in heritage speaker classrooms. This is despite indications that heritage language development probably differs from that of second language learners. Her chapter seeks to answer two fundamental questions: (1) Is instruction in a classroom setting beneficial for heritage language acquisition? If so, (2) What features make such instruction most effective? She concludes with concrete suggestions for future research in this sorely understudied area, arguing for a new subfield of research called instructed heritage language acquisition. In keeping with the volume’s insistence on bridging areas of inquiry, I would like to also suggest that pre-test/treatment/post-test studies also be carried out on K-8 Spanish speakers in dual language programs (see LindholmLeary’s chapter, this volume) as well as with high school students in heritage speaker programs. Focusing on classroom processes instead of outcomes, Carreira and Hitchins Chik describe how, unlike second language students whose Spanish proficiency is mostly a direct result of prior coursework, heritage speakers’ knowledge and abilities are much more heterogeneous because they derive from life experiences. They then summarize principles in differentiated teaching, offering eight concrete tools for teachers to help their students get the most out of their Spanish classes. This is important both for separate heritage speaker classrooms as well as in Spanish programs where heritage speakers and second language learners come together at some point. Returning again to a more macro perspective, the Part features Beaudrie’s chapter on heritage language program administration. Given the exponential growth of these programs around the nation, as well as the increase in Hispanic Serving Institutions4 (which are logical places for heritage Spanish programs to blossom and which could ideally form a coalition to support these programs), this is a particularly important area of practice and scholarship. Her chapter includes important topics such as course placement procedures, curriculum development,5 and evaluating student success. Abbott and Martínez examine issues in the increasingly popular fields of Spanish for the professions and community service learning and some specific applications for heritage speaker populations. Spanish for the professions grew out of a need to connect second language learners with functional language that was absent from traditional textbooks, including business and healthcare, while community service learning sought to expand opportunities for second language learners to interact with native speakers while developing a sense of social responsibility. Yet these two key areas were absent from heritage language education until recently, and the authors make a compelling case for the transformative educational experiences that these approaches can foster for heritage speakers connected to issues of immigration, health care, and cultural perspectives. Also in the university context, noting that the number of U.S. students studying abroad has more than tripled since the late 1990s, Shively examines issues 9

Kim Potowski

related to heritage speakers studying abroad in Spanish-speaking countries. For many of these students, study abroad may be more of a “return home” than an “immersion in difference” which typically characterizes the second language study abroad experience. But even so, these students often encounter being positioned as “foreign,” which ushers in a certain amount of identity renegotiation. Carvalho and Child explore issues in the acquisition of cognate languages (such as French, Italian, and Portuguese) by heritage Spanish speakers. Specifically, when compared with the acquisition of second language learners, heritage speakers’ acquisition of these languages is characterized by a faster rate, early high competence in receptive skills, and ease of communication. The authors detail efforts of curriculum designers to develop materials that best suit heritage Spanish speakers in these courses – which are technically third language (L3) contexts – using activities that capitalize on their implicit linguistic knowledge, while also keeping in mind that heritage Spanish speakers are a linguistically heterogeneous group. Questions about linguistic transfer and language change that emerge here are also relevant where Spanishspeakers have immigrated to countries where Italian or French are spoken (see the chapters by Bonomi and Sanfelici and by Sánchez Abchi, respectively). The Part concludes with two chapters that focus on the elementary school years, a critical topic given that the seven hours per day that kids spend in school constitute approximately half of their waking hours. Lindholm-Leary focuses on a topic close to my heart, dual language programs, which teach between 50% to 90% of the curriculum in Spanish. Over 30 years of research findings demonstrate that both native Spanish speakers and native English speakers benefit from dual language programs on standardized achievement tests, course grades, school attendance and dropout rates, and student attitudes. I have not seen any other school model with as strong an impact on Latino youth (who at the national level are at risk on many measures) and on Spanish language development, both of which constitute key components of a social justice-driven curriculum. Unfortunately, there are far too few Spanish dual language programs around the U.S., with a recent estimate of under 300 schools for almost 8 million Latino students (Potowski 2016) based on data from the dual language directory of the Center for Applied Linguistics (2007).6 Given the resounding success of dual immersion programs for Latino students’ Spanish, English, and overall academic development, the Spanish heritage language community should be fighting for the creation of more such programs as well as greater availability of high quality teacher preparation programs for the professionals who work there. Finally, the chapter by Mancilla-Martinez explores what we know about how SpanishEnglish bilingual Latino children develop literacy skills in Spanish. Unfortunately, we know relatively little about this process, due in large part to a lack of assessments designed for and normed on Spanish-English bilinguals. The author reviews research on English literacy development and on Spanish literacy development, highlighting important differences between the two and suggesting important areas for how best to support this growing and vulnerable population’s literacy achievement in both languages.

Part IV: Spanish as a minority/heritage language outside of the U.S. The final Part of the book presents profiles of Spanish-speaking communities in nine nations around the globe. As noted in the Introduction, it would be difficult for a single volume to include all contexts where Spanish is a minority language, and this one is no exception. Several interesting contexts that are unfortunately not represented include Brazil, Morocco, Gibraltar, Equatorial Guinea, and the Philippines. However, the countries profiled (shown in Table 1.2) provide an interesting range of contexts in Europe, Australia/New Zealand, and

10

Spanish as a heritage/minority language Table 1.2  Immigrant Spanish speakers in nine countries outside the U.S. featured in Part IV Country

Estimated number of Spanish-speakers

Most common countries of origin

Canada Italy United Kingdom Germany Switzerland Australia Sweden New Zealand Mexico

410,700 400,000 300,000 238,000 160,000 117,000  81,500  27,000 n/a

Mexico, Colombia Peru, Ecuador Spain, Colombia Spain, Mexico Spain, Colombia Chile, El Salvador Chile, Colombia Varied United States

North America. Some of these chapters represent the first publications to my knowledge about Spanish and its speakers in these locations. Jones Díaz and Walker combine quantitative data with qualitative original research with Latin American immigrants in Australia and Aotearoa-New Zealand, with several poignant anecdotes that highlight the role of Spanish in identity construction and cultural maintenance. Due in large part to the similarities between Spanish and Italian, Bonomi and Sanfelici’s chapter about Italy includes the concept of Spanish in-motion to describe the set of hybrid and multiple language practices performed by Latinos there, which occur despite the nation’s largely monolingual ideology that seeks to assimilate migrants through the exclusive use of Italian. Similarly, beliefs about how to best promote German acquisition is a focus of Ramos Méndez-Sahlender’s chapter about Germany, where declining resources are a main factor in the lack of Spanish maintenance programs. The chapter on Switzerland by Sánchez Abchi documents that despite the challenge of extremely diverse backgrounds, students have been shown to write better in Spanish after attending “Language and Culture of Origin” classes. Nielson Parada summarizes linguistic research to date on the Chilean-Swedish community before describing the climate of heritage language support in the country’s educational system. Guardado examines Canada and the UK (a combination I requested in order to represent two Anglophone contexts), summarizing research on morphosyntax, language and identity, and educational experiences among Spanish-speaking immigrants to these two locations. Finally, the inclusion of Mexico in this Part might seem surprising, particularly since we saw in Table 1.1 that this is the Spanish-speaking nation with the largest number of inhabitants in the world. In this chapter, Mar-Molinero focuses on a phenomenon that is increasingly common: that of individuals raised for most or all of their lives in the U.S., who are descendants of Mexican nationals and who have returned to Mexico. Growing up in the U.S. frequently leads to certain linguistic phenomena in their Spanish (explored in Part II) as well as cultural knowledge and experiences that are out of harmony with what is expected of their Mexicoraised peers. The author details ways in which Mexican schools might best meet the needs of these “returnee” students.

Conclusions and future directions Minoritized languages around the world, especially those spoken by immigrants, are very typically on a path towards loss. This volume explores linguistic, social, and educational issues germane to understanding Spanish as a heritage/minoritized language in ten different nations.

11

Kim Potowski

There are, of course, many other important and relevant topics that could have appeared here, including some that I was simply unable to procure (such as Spanish in U.S. mainstream media and in U.S. literature) as well as others that did not occur to me. I conclude by briefly describing four additional areas that may be fruitful for future investigations into the topic of heritage/ minoritized languages.

Social justice In what ways can researchers bring about greater social justice for speakers of Spanish as a minoritized language? Can we work to lessen linguistic prejudice? Help stem language loss? Encourage speakers to develop a sense of pride in their Spanish in the face of negative forces, as does the curricular work of Wolfram (2013) and of Charity Hudley and Mallinson (2014)? As noted by Heleta (2016), a lot of researchers’ work is “largely sitting in academic journals that are read almost exclusively by [our] peers.” Instead we need to share our work with the broader public in order to effect change. For example, it is shocking that the field has not yet produced an educational, mainstream-oriented feature-length documentary about Spanish in the U.S. akin to American Tongues, Do you speak American? or Talking Black in America. Such a video could not only generate pride among U.S. Spanish-speakers but also educate the broader public about the history and value of the language in the present-day U.S.

Mothers as minority/ized language transmitters Mothers may play an especially important role in the development of a minoritized language. For example, Walker (2011) observed that mothers exerted a definitive influence on Spanish language use among Latino children in New Zealand. Potowski (2016) studied the Spanish of ‘MexiRicans’ (individuals who have one Mexican parent and one Puerto Rican parent) raised in Chicago, where Spanish is a minority language and Puerto Rican Spanish is the minority dialect compared to Mexican Spanish. She found that those MexiRicans who exhibited Puerto Rican phonological features always had a Puerto Rican mother, suggesting that in order to develop features of a locally minority dialect within a minority language, having a mother who speaks that minority dialect is necessary. Zentella (1997) found that Puerto Rican girls in a New York City neighborhood developed stronger Spanish than the boys – because they were required to stay near the home with their mothers, while boys were allowed to leave the block – which meant that when they became parents, girls were in a stronger position to pass on Spanish than their U.S.-raised male partners. Work on other languages, too, has made a connection between successful minority language transmission and having a mother who spoke the minority language (Williams 1987 in Wales; Kamada 1997 in Japan; Kondo 1997 in Hawaii; Boyd 1998 in Denmark, Finland, and Sweden; Chiaro & Nocella, 1999 in Britain).7 This suggests that a society might bolster minority language maintenance via mother-friendly policies including postpartum leaves of absence, equal pay, jobsite childcare, and incentives to stay home with children until they go to preschool.

Technology What is the role of technology in the bilingualism of the world’s Spanish-speaking individuals? Trenchs (2013) comments on the value of the Internet, specifically YouTube, for purposes of cultural entertainment tied to ethnolinguistic identity among Chinese immigrant adolescents in Barcelona, while Vincent (2015) explores the role of mobile phones and Internet cafés among 12

Spanish as a heritage/minority language

Latino diaspora members in maintaining cultural and linguistic connections. Yet a digital divide in broadband access continues to exist in the U.S., with only 46% of Latinos (vs. 73% of Whites) having home access in 2015 (Pew Hispanic Center 2016), constituting yet another inequity in Latinos’ socioeconomic realities that might be addressed through dedicated policies.

Notes 1 An airtight definition of the term “native” speaker is elusive, hence my use of quotation marks. Criteria such as “commencing acquisition at birth” or “speaking the language at home” are imperfect because both situations can lead to linguistic systems that could not sensibly be referred to as “native-like.” Monolingualism obviously cannot be a requirement, either.Thus, I use instead the term “Spanish speaker” with the goal of identifying the number of individuals who use Spanish with locally adequate proficiency to accomplish tasks in their lives.This includes, for example, speakers of indigenous languages in Latin America as well as the estimated 3 million non-Latinos in the U.S. and the 400,000 in Italy who identify as home Spanish speakers. Given the near impossibility of ascertaining reliable data for such an imprecise definition, these figures should be taken as estimates only. 2 The U.S. number was calculated by adding 37.6 million home Spanish speakers (American Community Survey, U.S. Census 2011) and an estimated 8.9 million undocumented Spanish speakers from Latin America as of 2014 (Pew Hispanic Center 2016). 3 Rei-Doval (2016) reports that in 2013, 31% of Galicians were monolingual in Galician. 4 A federal designation meaning that 25% or more of the student population is Hispanic. 5 The Hispanidades project is another excellent curricular model that links Latino students from around the U.S. in sharing first person, locally relevant material (www.lrc.columbia.edu/hispanidades/collaborations). 6 Although this is likely an undercounting. Unofficial estimates place the national number of dual language schools in the U.S. between 1,000 (Maxwell 2012) to over 2,000 (Watanabe 2011). But even if there were 2,000 such programs, that would not be nearly enough for almost 8 million Latino students. 7 In her summary of a large body of work, DeHouwer (2009) found no effect for parental gender on language development. It may be only in language minority contexts, such as those explored by Walker, Potowski, Zentella, and others, that the mother might exert greater influence on children’s bilingual language development.

References Anderson, J., & Solis, N. (2014). Los Otros Dreamers. Mexico City: Independent Publisher. Beaudrie, S., Ducar, C., & Potowski, K. (2014). Heritage Language Teaching: Research and Practice. New York: McGraw-Hill. Boix-Fuster, E., & Sanz, C. (2008). Language and identity in Catalonia. In J. Rothman and M. NiñoMurcia (Eds.), Bilingualism and Identity: Spanish at the Crossroads with Other Languages (pp. 87–106). Amsterdam: John Benjamins. Boyd, S. (1998). North Americans in the Nordic region: Elite bilinguals. International Journal of the Sociology of Language, 133, 31–50. Cenoz, J. (2008). Learning through the minority: An introduction to the use of Basque in education in the Basque Country. In J. Cenoz (Ed.), Teaching through Basque: Achievements and Challenges (pp. 1–5). Clevedon, UK: Multilingual Matters. Center for Applied Linguistics. (2007). Directory of two-way bilingual immersion programs in the U.S. Retrieved from www.cal.org/Lwi/directory/. Charity Hudley, A.H., & Mallinson, C. 2014. We Do Language: English Language Variation in the Secondary English Classroom. New York: Teachers College Press. Chiaro, D., & Nocella, G. (1999, April). Anglo-Italian Bilingualism in the UK: A Sociolinguistic Perspective. Paper presented at the 2nd International Symposium on Bilingualism, Newcastle upon Tyne, UK. DeHouwer, A. (2009). Bilingual First Language Acquisition. Clevedon, UK: Multilingual Matters. del Valle, J. (2000). Monoglossic policies for a heteroglossic culture: Misinterpreted multilingualism in Modern Galicia. Language and Communication, 20, 105–132. Díaz Campos, M. (2011). The Handbook of Hispanic Sociolinguistics. Oxford, UK: Wiley-Blackwell. Ethnologue (2016). Retrieved from www.ethnologue.com/statistics/size.

13

Kim Potowski Fairclough, M., & Beaudrie, S. (Eds.). (2016). Innovative Strategies for Heritage Language Teaching: A Practical Guide for the Classroom. Washington, DC: Georgetown University Press. García, O. (2005). Positioning heritage languages in the United States. Modern Language Journal, 89(4), 601–605. Heleta, S. (2016, March 8). Academics can change the world: If they stop talking only to their peers. The Conversation. Retrieved from https://theconversation.com/academics-can-change-the-world-ifthey-stop-talking-only-to-their-peers-55713. Instituto Cervantes. (2016). El español en cifras. Retrieved from www.cervantes.es/imagenes/File/prensa/ EspanolLenguaViva16.pdf. Kamada, L. D. (1997). Bilingual Family Case Studies (Vol. 2). Tokyo: Japan Association for Language Teaching. King, K. A., Fogel, L., & Logan-Terry, A. 2008. Family language policy. Linguistics and Language Compass, 2(5), 907–922. Kondo, K. (1997). Social-psychological factors affecting language maintenance: Interviews with Shin Nisei university students in Hawaii. Linguistics and Education, 9(4), 369–408. Lambert, W. E. (1977). Effects of bilingualism on the individual: Cognitive and socio-cultural consequences. In P. A. Hornby (Ed.), Bilingualism: Psychological, Social and Educational Implications (pp. 15–28). New York: Academic Press. López, M., & González-Barrera, A. (2013). What is the future of Spanish in the United States? In Fact Tank News in the Numbers. Retrieved from www.pewresearch.org/fact-tank/2013/09/05/ what-is-the-future-of-spanish-in-the-united-states/. Maxwell, L. A. (2012, March 28). ‘Dual’ classes see growth in popularity. Education Week. Retrieved from www.edweek.org/media/downloads/files/spotlight-deeper-learning.pdf. Moreno Fernández, F., & Otero Roth. (2006). Demografía de la lengua española. Retrieved from http:// eprints.ucm.es/8936/1/DT03–06.pdf. Otheguy, R., García, O., & Reid, W. (2015). Clarifying translanguaging and deconstructing named languages: A perspective from linguistics. Applied Linguistics Review, 6(3), 281–307. Pascual y Cabo, D. (Ed.). (2016). Advances in Spanish as a Heritage Language. Amsterdam: John Benjamins. Pew Hispanic Center. (2016). Digital Divide Narrows for Latinos as More Spanish Speakers and Immigrants Go Online. Retrieved from www.pewhispanic.org/2016/07/20/digital-divide-narrows-for-latinos-asmore-spanish-speakers-and-immigrants-go-online/. Potowski, K. (2010). Language Diversity in the USA. Cambridge, UK: Cambridge University Press. Potowski, K. (2016). Current issues in Spanish heritage language education. In D. Pascual y Cabo (Ed.), Advances in Spanish as a Heritage Language (pp. 127–141). Amsterdam: John Benjamins. Rei-Doval, G. (2016). Language endangerment and language revitalization in a minority (post) colonial setting: The Galician case. In P. Mohanty & N. Ostler (Eds.), Proceedings of the XX Conference of the Foundation for Endangered Languages (pp. 72–75). Foundation for Endangered Languages & University of Hyderabad, India. Silva-Corvalán, C. (2014). Bilingual Language Acquisition: Spanish and English in the First Six Years. Cambridge, UK: Cambridge University Press. Trenchs, M. (2013). Les vivencies sociolingüistiques i el multilingüisme dels joves d’origen xines a Catalunya. Didactica de la Llengua i de la literatura, 60, 28–39. Vincent, J. (2015). Staying in touch with my mobile phone in my pocket and internet in the cafés. In Rosina Márquez Reiter & Luisa Martín Rojo (Eds.), A Sociolinguistics of Diaspora: Latino Practices, Identities, and Ideologies (pp. 169–180). Oxford, UK and New York: Routledge. Walker, U. (2011). The role of community in preserving Spanish in New Zealand: A Latin American parent perspective. In Potowski, K. & Rothman, J. (Eds.) Bilingual Youth: Spanish in English-Speaking Societies (pp. 331–354). Amsterdam: John Benjamins. Watanabe, T. (2011, May 8). Dual-language immersion programs growing in popularity. Los Angeles Times. Retrieved from http://articles.latimes.com/20 l l/may/08/local/la-me-bilingual-20. Williams, C. H. (1987). Location and context in Welsh language reproduction: A geographic interpretation. International Journal of the Sociology of Language, 66, 61–84. Wolfram, W. (2013). Community commitment and responsibility. In J. K. Chambers & N. Schilling (Eds.), The Handbook of Language Variation and Change (pp. 557–576). Hoboken, NJ: John Wiley & Sons. Zentella, A. C. (1997). Growing up Bilingual. Malden, MA: Blackwell.

14

PART I

Social issues

Q Taylor & Francis Taylor & Francis Group

� http://taylorandfrancis.com

2 A HISTORICAL VIEW OF US LATINIDAD AND SPANISH AS A HERITAGE LANGUAGE Andrew Lynch university of miami, usa

Introduction In this chapter, I review some aspects of early 20th-century latinidad that inform an understanding of contemporary cultural phenomena and Spanish language inquiry in the current era of renewed nationalist sentiment and fervent debate around (im)migration. I first seek to establish the importance of the ‘cultural imaginary’ for studies of Spanish language in the US, and then address three principal phenomena in chronological fashion: (1) the emergence of the concept of ‘Spanish (as) heritage’ in the context of pre-statehood New Mexico during the early 1900s; (2) the construction and (re)production of latinidad in the cultural industries of the 1910s and 1920s—specifically, Hollywood film, New York music and dance, and urban architecture in the Southwest and Florida; and finally, (3) the repercussions of the Great Depression, especially with regards to cultural intervention and (im)migration during the 1930s and 1940s. The purpose of this chapter is not to synthesize or review previously published work on latinidad, but rather to highlight for readers some of the societal trends that brought the cultural construct of latinidad—and Spanish language as an essential feature of it—into the US cultural imaginary during the early 20th century.1 One garners a real sense of history repeating itself when considering the phenomenological parallelism of the 1920s/1990s and 1930s/2000s, periods when prosperity and a Latin cultural ‘boom’ were subsequently fractured by grave economic and political crises in which explicit Hispanophobia grew rampant. During the decade of the 1990s, a so-called ‘Latin fever’ swept the US: Ricky Martin, Enrique Iglesias, and Shakira topped the Billboard charts with songs partially or completely in Spanish; Jennifer López (‘J. Lo’) rose to stardom with her Hollywood portrayal of popular Tex-Mex singer Selena, whose death was publicly mourned by millions in the US and Mexico; salsa outsold ketchup in US supermarkets for the first time in the nation’s history; and Spanish language programs at schools, colleges, and institutions of higher learning across the country expanded significantly. Of course, Jennifer López was not the first Latina superstar, but rather Dolores Del Río, who rose to Hollywood stardom during the early days of silent film in the 1920s. Born in Mexico in 1905, Del Río was spotted by a Hollywood director at a social gala in Mexico City at the age of 20 and went on to become one of the top ten moneymakers in the industry. Like her, there were many other Hispanic-identified actors who 17

Andrew Lynch

accrued great fame in early Hollywood film: Myrtle González, Lupe Vélez, Antonio Moreno, and Ramón Novarro, among others. Things ‘Latin’ or ‘Spanish’ were decidedly stylish during the early 20th century, the same era when Argentine tango, Cuban son, and Dominican merengue emerged on the cultural scene in New York,2 and Spanish-themed architecture appeared in iconic cities such as San Diego, Santa Barbara, Santa Fe, and Miami. Rather predictably, when the ‘foreign’-accented English of so many of Hollywood’s great silent stars could later be heard in talkies, within the context of the Great Depression, their appeal to producers and audiences quickly waned. The career of Ramón Novarro, who was receiving 1300 fan letters per week and earning roughly a million dollars a year in 1923, met a fate similar to that of Antonio Moreno, Lupe Vélez, and numerous others; by 1937, he had lost imminence in Hollywood and, in later years, received only minor and often typecast roles (Rodríguez 2004). The professional outcome of these prominent public personalities underscores the great vitality of the linguistic dimension of cultural imaginaries, and the primordial role that Spanish language and Hispanic-accented English play in the construction and projection of latinidad in the US context (Lippi-Green 2012). Both Novarro and Vélez were what we today consider heritage speakers of Spanish, having arrived in the US from Mexico at the ages of 14 and 13, respectively. But as much as Hispanic-accented English became a liability for their careers by the 1930s, it is equally true that, nearly six decades later, the lack of Spanish-language fluency was in some ways perceived as a liability for both Selena and J. Lo in their early days of fame (Aparicio 2003; Williams 2005). Writing in the context of New York City in the 1990s, Ana Celia Zentella described the ‘chiquitafication’ of US Latinos and the varieties of Spanish spoken by them. As she recalled the great impact that the Carmen Miranda-inspired dancing Chiquita Banana Girl had on her as a child during the 1950s, she remarked that: I had no idea how the Latin bombshell cliché was shaping the expectations that the world outside of my barrio had of my sister and me, daughters of a Mexican father and Puerto Rican mother, but we certainly thought that the number one box-office attraction in the country was having a wonderful life. 1996, p. 1. Zentella’s widely cited call for an ‘anthropolitical’ approach to the study of Spanish in the US reflects the continuous participation of the cultural imaginary in the forces that shape language and the ideas that we have about it.

Language and the cultural imaginary In basic theoretical terms, cultural imaginaries are important to scholars of language because they give presence—and in a psychological sense, life—to language in society, and to ‘imagined’ communities at a national level. In his seminal work on the emergence of nation-states following the invention of the printing press in Western Europe, Anderson (1983) affirmed that: “all communities larger than primordial villages of face-to-face contact (and perhaps even these) are imagined. Communities are to be distinguished not only by their falsity/genuineness, but by the style in which they are imagined” (p. 6). He affirmed that a nation is an imagined political community, imagined because its members “will never know most of their fellow members, meet them, or even hear of them, yet in the minds of each lives the image of their communion” (p. 6). Dawson (1994) defined the ‘cultural imaginary’ as “those vast networks of interlinking discursive themes, images, motifs and narrative forms that are publicly available within a culture 18

US latinidad, Spanish as heritage language

at any one time, and articulate its psychic and social dimensions” (p. 48). This is certainly the case when imaginaries are made concrete through institutionalized policies and practices, political and commercial discourse, architecture, place names, signage, advertising, literature, theater, music, film and television shows. All of these cultural phenomena serve, semiotically, to create places and to situate language in context at both micro and macro levels. Images and thoughts of places and of characterological figures evoke in our social minds the use of particular languages, dialects or sociolects; likewise, language use may also serve to evoke or establish places and communities psychologically, socially, economically, and politically. Cultural imaginaries of US latinidad are important and relevant to readers of the present volume in two principal ways. First, patterns of language choice and cross-generational change and shift are, to some extent and in some regards, the product of cultural imaginaries. As Marcontoni (2015) remarked, for many US Latinos, “to promote one’s latinidad is to promote the culture itself, and culture is intimately tied to the language in which it is communicated.” He also noted that (cf. Schwartz 2011): What has been prevalent recently have been videos which showcase the growing community of Latinos who choose not to learn Spanish for one reason or another (and it is a choice) but are frustrated by the expectation of Latinos and whites alike for them to know the language. A plausible hypothesis would be that if Spanish is imagined as an integral aspect of US culture and society, the population in general might be more likely to accept its use in public life, and the bilingual population in particular may be more likely to speak the language more frequently and across a wider array of settings (cf. Callahan 2009 on language choice in US service encounters). This would constitute a fundamental approach—in addition to the empirical study of language attitudes and perceptions—to accounting for the actual reasons behind language choice and, concomitantly, language shift. Second, cultural imaginaries of latinidad play a vital pedagogical role in the academic success of US Latinos. Explaining the source of his own motivations to study and acquire fluency in Spanish, Marcantoni (2015) recalled how the “shame tactics” of family members and other native speakers discouraged him from speaking Spanish, yet his desire to read the works of Gabriel García Márquez in the original language motivated him to study Spanish, thus setting him on the path of choosing to speak it more often and, in turn, acquire fluency. He wrote that (cf. Lynch & Potowski 2014): I encourage native speakers, when you come across a Latino who isn’t a fluent Spanish speaker, that you ask them about their interests and maybe recommend a musician, a filmmaker or a writer . . . Maybe they still won’t learn the language, but they will create a positive relationship with the culture and, in turn, with Spanish as well. Even when unspoken, Spanish language is necessarily part and parcel of the Latino dimensions of the US cultural imaginary. Cañas’ argument regarding the concept of Latino New York is illustrative: [i]f it is legitimate to speak of a Hispanic communal identity in New York City, it is because people have continued to use the Spanish language in an active and dynamic form, not only as a vehicle of communication and creation, but also as a principal point of reference for a Latino sensibility, however latent, in those Hispanics who speak only English. 2010, p. 252 19

Andrew Lynch

Similarly, award-winning US Latina novelist Cristina García, who writes in English, has asserted that an “underlying pulse” of Spanish language shapes and informs her discourse, and that limited, emblematic uses of Spanish in her work reflect a “memory-packed punch” delivered by the “mother tongue” (Lynch 2011). In this respect, Spanish as heritage language may be aptly conceptualized through the lens of a collective consciousness and cultural imaginary, recognizing that ideas about ‘heritage’ may serve to empower and, at the same time and in some respects, coerce (Nieto-Phillips 2004). Such were the circumstances of political debate around language in pre-statehood New Mexico, which is the first of the three major topics that I highlight in this chapter.

Issues and topics The emergence of Spanish as ‘heritage’ language in the Southwest The Spanish legacy As the US entered the throes of modernity during the Gilded Age of the late 19th century, in polluted manufacturing centers such as Chicago and Pittsburgh and along expansive railroad networks, romanticist currents prompted an inward look at the spoils of industrialization and a longing for a return to simpler, more rustic ways of life amidst the backdrop of an untouched natural landscape. For Easterners, this meant looking westward, to New Mexico in particular. At the time, the Spanish settlers’ inability to attain modernity came to be viewed by some as a virtue rather than a fault, and a romanticized cultural imaginary of days-gone-by emerged (Weber 2000). One of the most highly influential and widely popularized works of the time was journalist Charles Lummis’ (1911) book Some Strange Corners of Our Country, first published in 1892, in which the natural marvels of the region and the social customs of its indigenous peoples are described—and at times exoticized—in depth of detail. Drawing parallels between the “savage peoples” of the Southwest and the Congo, and comparing the region’s deserts to the Sahara, Lummis explicitly urged proud Americans to learn something of the wonders of the “most remarkable” and “most neglected” area in the US. He opened the book stating that “when these and so many other wonders are a part of America, we, who are Americans, should be ashamed to know absolutely nothing of them” (1911, p. 7), and concluded by affirming that “the trip abroad may at least be postponed until we are ready to tell those we shall meet in foreign lands something of the wonders of our own” (p. 270). With the same propagandistic sort of awe with which he recalled having photographed the flagellation and crucifixion practices of the Penitentes during Lent in 1888 (pp. 90–93), he shares with the reader photographs of the “stone autograph-album” at El Morro, a wall-like sandstone promontory where numerous Spanish explorers, among them Juan de Oñate, carved their names as they passed through the area in the 16th and early 17th centuries. There, in Lummis’ words, “the heroes wrote their autographs upon a vast perpendicular page of stone, with their swords which had won the New World for pens!” (p. 167). He observed that: All the old inscriptions are in Spanish—and many in very quaint old Spanish, of the days when spelling was a very elastic thing . . . All around these brave old names . . . are Saxon names of the last few decades. Alas! Some of these late-comers have been vandals, and have even erased the names of ancient heroes to make a smooth place for their “John Jones” and “George Smith.” pp. 168–169 20

US latinidad, Spanish as heritage language

In these lines, Lummis captured the early-20th-century Zeitgeist within which the contemporary US cultural phenomenon of ‘Spanish heritage’ first became manifest. Within that cultural imaginary, Anglo modernity—at times cast in a negative light—was ideologically confronted with a glorified Spanish colonial past. Walt Whitman seemingly foresaw the coming cultural trend. In an open letter written to the people of Santa Fe in 1883, he wrote that: Character, literature, a society worthy the name, are yet to be establish’d, through a nationality of noblest spiritual, heroic and democratic attributes . . . To that composite American identity of the future, Spanish character will supply some of the most needed parts. Whitman 2010, pp. 179–180 Anglo author Helen Hunt Jackson’s critically acclaimed 1884 novel Ramona had pervasive cultural influence in the nascent imaginary of the Spanish heritage in relation to indigenous populations of the Southwest. Of course, the confrontation of ‘Indian,’ ‘Anglo,’ and ‘Hispanic’ had been ongoing for three centuries, ever since the physical skirmishes between early colonists in the present-day south of Georgia and north of Florida. Moreover, the climate within which Lummis wrote was rife with the conflict of the Spanish-American War of 1898. In the days following the final battle of the war, some four months before the Treaty of Paris would be signed, the New York Times published an editorial accusing Spanish-speaking New Mexicans of having sympathized with Spain. The author of the editorial clearly laid the blame on language, and urged that future policy in Puerto Rico must not follow suit: The trouble with these disaffected, semi-traitorous citizens is that they have been allowed to attend schools in which only the Spanish language was spoken . . . As long as this is permitted, of course a considerable majority of the inhabitants will remain “Mexican” and will retain a pseudo-allegiance to the land which, if the matter were one of reason instead of instinct, they would detest much more vehemently than the real American does. All through the territory there are to be found today thousands of children, as well as thousands of adults, to whom English is almost unknown, and therefore a more or less detested, tongue. Such a state of affairs is disgraceful as well as dangerous . . . In the present attitude of these “Mexicans”, who do not know that they are Americans, there is to be found a strong hint as to the course which should be pursued in Puerto Rico. Absolutely no official recognition should be given hereafter to the Spanish language in that island. New York Times, 24 August 1898 Some would suggest that such opinions reflected the centuries-long historical influence of the leyenda negra.3 What perhaps lent credibility to the fabricated accusations that Nuevomexicanos sympathized with Spain during the conflict was the real extent of Spanish language use throughout the recently annexed Southwest and the growing local sense of pride in the region’s Spanish heritage.4

Language and New Mexican identity ‘Spanish-American’ identity emerged in New Mexico during this time, partly in relation to a growing ‘Hispanophile’ movement in the US cultural industries, of which the burgeoning New Mexican tourism industry was one aspect, and partly as an articulation of resistance to political 21

Andrew Lynch

and social marginalization. As Nieto-Phillips explains, “the Spanish heritage was both the object of Anglos’ fascination and a source of ethnic agency as Nuevomexicanos (of various echelons) struggled to reclaim some degree of control over their political destiny and cultural assets” (2004, p. 8). The cultural ‘heritage’ assets of New Mexico, of which Spanish language constituted a vital part, were fundamental to the future state’s claim to the title ‘land of enchantment.’ Thus came the need to defend the language in the face of Anglo political efforts to abolish it. In the midst of the political and cultural debates leading up to New Mexico’s 1912 statehood, a young college student named Aurora Lucero delivered a stirring speech in favor of bilingual education and the preservation of the Spanish language as part of an oratorical competition at New Mexico Normal University in December of 1910. Her speech came just six months after President William Howard Taft signed the 1910 Enabling Act for New Mexico and Arizona’s admission into the Union, which stipulated that the constitutions of both states must include provisions to carry out public school education always in English and to establish a requirement for service as a state official or legislator the ability to read, write, speak, and understand English (Crawford n.d.).5 Lucero asserted that: Spanish is the language of our parents. Today, it is our own, and it will be the language of our children and our children’s children. It is the language bestowed upon us by those who discovered the New World. We are American citizens, for certain, and . . . must learn the language of our country . . . Yet we need not negate in the process our roots, our race, our language, our traditions, our history, or our ancestry, because we shall never be ashamed of these. On the contrary, they shall make us proud. cited in Nieto-Phillips 2004, p. 173 Lucero’s intentions in some ways echoed those of Lummis, in that both called upon the US to take stock of its Spanish patrimony and take steps to preserve it.6 ‘Hispanophilia’ was on the cultural horizon, not just in the work of Lummis but in other important contemporary public figures as well, among them politician Lebaron Bradford Prince (Governor of New Mexico Territory from 1889 to 1893) and philologist Aurelio Espinosa who, like Lucero, urged that Spanish language must be taught in schools. Ultimately, the English language-related provisions of the 1910 Enabling Act were not upheld, and New Mexico became the nation’s first officially bilingual state.7 However, New Mexico’s legislation proved essentially symbolic. The state constitution contained only two articles pertaining to the role of Spanish: 1) all laws must be published in Spanish—a requirement that has not been enforced since the 1940s; and 2) teachers must be trained in Spanish—a directive that was consistently interpreted as applying only to teachers of monolingual Spanish-speaking children (Bills 1997, p. 170). The present-day website of the American Association of Teachers of Spanish and Portuguese (AATSP), founded in 1917, describes the rise of Spanish language study at the time in rather more skeptical terms: American isolationism gave Spanish a boost when German was dropped from many schools during the First World War. Spanish became the language of choice, not through any love of the language, but for simple expediency . . . So Spanish developed a constituency and a foothold in American education, but for unattractive and unsatisfactory reasons. AATSP n.d. One of the founding members of the AATSP and one of its most fervent advocates of Spanish language education was Aurelio Espinosa, a pioneer in the study of Spanish in the US. In 1910, 22

US latinidad, Spanish as heritage language

the same year that he joined the faculty of Stanford University, he affirmed in the Journal of American Folklore that: “The abundant material which has already been found in New Mexico and Colorado would seem to furnish ample proof that vast treasures of folk-lore are to be found in Texas, California, and Arizona” (Espinosa 1910, p. 395). Espinosa’s interest in documenting the Spanish heritage in the US Southwest was somewhat personal: he was born in 1880 in southern Colorado to a Nuevomexicano family who traced their lineage to the original Spanish settlers accompanying Juan de Oñate at the end of the 16th century (Nieto-Phillips 2004, pp. 178–179). In the inaugural issue of the journal Hispania, of which he was the first editor-inchief, Espinosa declared in 1917 that: [a]side from giving to problems of pure pedagogical interest the great attention which they deserve, [Hispania] will also attempt to interpret sympathetically to our pupils and teachers of Spanish the history and culture of the great Spain of the past and present. p. 19 While addressing the Mexican and Nahuatl elements in New Mexican Spanish in his now classic three-tome Studies, Espinosa was, according to some scholars, more generally concerned with the continuity of Peninsular Spanish in the Southwestern US. As Nieto-Phillips pointed out, Espinosa identified as ‘Spanish’ about 1,000 of the 1,400 local dialect features that he documented in New Mexican Spanish; he attributed another 300 to English, 75 to Nahuatl, and only 10 to autochthonous Indian origin (pp. 180–181). Nieto-Phillips affirmed that “at the heart of Espinosa’s argument was his faith in Nuevomexicano’s linguistic purity” (p. 181), similar to Limón’s (2014) interpretation: Espinosa’s studies of the Southwest at times seem to mount a cultural defense, what some would term a “resistance,” to Anglo modernity and occupation. Paradoxically, such a stance is taken in the name of an idealized Spanish language and folklore in New Mexico, one sustained by an ideology based on cultural purity. p. 4588 In time, of course, and as Espinosa had perhaps anticipated, English language use would exert an ever greater influence on New Mexican Spanish (Sanz-Sánchez 2014), and Spanish as spoken in Mexico would ultimately prevail through continuous patterns of migration. Bills and Vigil (1999) pointed out that traditional New Mexican Spanish throughout the 20th century was profoundly influenced at all levels by contemporary varieties of Mexican Spanish; they predicted that by the mid-21st century, traditional New Mexican varieties would be near extinction (p. 58).

Hispanophilia Architecture Espinosa’s scholarship was characteristic of the Hispanophile cultural trend that took root during the 1910s. Architecture was a key element of the Spanish heritage imaginary. During the same years in which Espinosa published Studies in New Mexican Spanish and large numbers of Mexicans entered the Southwest fleeing the conflicts of the Revolution, plans were made to host the Panama-California Exposition in San Diego in commemoration of the opening of the Panama Canal. The venue where the 1915–1917 exposition would be held was given the name Balboa Park, in honor of the first Spanish explorer to reach the Pacific Ocean from the 23

Andrew Lynch

New World. For construction of the site, Spanish Colonial Revival architecture was chosen, recalling Mission and Pueblo Revival styles of the Southwest and taking principal inspiration from the Spanish or Mediterranean Revival styles of the late 1800s evident in Florida.9 The exposition brought national and international acclaim to architect Bertram Goodhue’s work in Balboa Park, and established the vogue of Spanish Colonial Revival style across the US during the 1920s, the best examples of which are found in the cities of San Diego and Santa Barbara, California, and Coral Gables, Florida. Today, one of Miami’s most affluent neighborhoods, Coral Gables, was conceived by George Merrick in 1921 and incorporated in 1925. Hundreds of grandiose Mediterranean Revival-style residences line some 65 miles of streets bearing Spanish names, all of which would assume new significance with the mass arrival of Cubans in Miami following Fidel Castro’s 1959 Revolution: Ponce De León, Segovia, Sevilla, Aragón, Valencia, Granada, Alhambra, etc. Pérez (1999) affirmed that, during the 1920s, “the vogue of Havana insinuated itself into the vision of Miami: foreign, tropical, exotic, as ambience and circumstance through Spanishlanguage usage and Royal Palms landscaping” (p. 432). Indeed, in the mid-1920s, about 75% of tourists in South Florida also visited Havana (New York Times 1926), embodying a symbiotic relationship between the two places in the US imaginary. Coral Gables’ legendary Biltmore Hotel is among the more notable examples of Spanish Colonial Revival architecture in the US, along with Miami’s Freedom Tower, Pima County Courthouse in Tucson, St. Mary’s Basilica in Phoenix, Corpus Christi Cathedral, El Paso’s Plaza Theatre, Mission Dolores Basilica of San Francisco, St. Vincent Catholic Church in South Los Angeles, and the City Halls of Pasadena and Beverly Hills. Gebhard (1967) noted that Spanish Colonial Revival style—which “had become the architecture of Southern California” by the end of the 1920s—“had little, if any, real roots in the historic past of the area” and was “almost totally a myth created by newcomers” during the late 19th century (p. 131).10 The lead architect of the flimsily constructed San Diego exposition that gave impetus to the building fad affirmed that it was meant to be temporary, like stage scenery or “the fabric of a dream” (Goodhue 1916, p. 7). Goodhue was explicit regarding the relationship of the Spanish cultural imaginary to Southern California’s ‘rightful’ heritage: [i]t is perhaps strange to say quite flatly that so many buildings that have given pleasure to so many should be destroyed; but, after all, this was the paramount idea in the minds of the Fair’s designers, and only by thus razing all of the Temporary Buildings will San Diego enter upon the heritage that is rightfully hers. 1916, p. 9 Ultimately, several of the buildings were either retained or reconstructed, and remain to the present day, constituting the ‘Spanish’ image of Balboa Park.

The Hollywood film industry Nowhere was the contemporary concept of Spanish heritage brought more to life than in Hollywood. In keeping with the popular cultural imaginary of the time, Latino film stars were sometimes marketed—or chose to market themselves—as ‘Spanish’ rather than Latin American. The official studio biography of Ramón Novarro, born José Ramón Samaniegos in Durango, Mexico in 1899, stated that he was a Spaniard when, in fact, his family background was assuredly Mexican.11 Although his mother’s family was partly of Spanish descent (her ancestors were rumored to have sailed from Spain with Cortés), they also claimed Aztec ancestry. His father is 24

US latinidad, Spanish as heritage language

known to have graduated high school in Las Cruces, New Mexico and attended the University of Pennsylvania; his paternal grandfather was one-time Governor of Chihuahua and the first Mexican to be elected to the city council of El Paso, Texas (Rodríguez 2004, pp. 48–49). Novarro’s cousin Dolores Del Río was also described as a “Spanish actress” when she first rose to stardom.12 Pedro de Córdoba, who began his 116-film career starring in Cecil B. DeMille’s Carmen in 1915, was born in New York City to a Parisian French mother and a Cuban father. Because his father’s grandparents had immigrated to Camagüey from Córdoba, Spain, the press and his own management highlighted his Spanish ancestry: [i]nstinctively when one sees him, in one’s mind rises visions of brave toreadors . . . graciously fascinating women in waving mantillas and eyes dancing . . . cathedrals of potent meaning with incense and the black robed figures of priests; a pen sketch of the atmosphere in which Pedro would seem at home. Chamberlain Brown’s Office, cited in Rodríguez 2004, p. 41 Actress Anita Page, star of the first sound film to win the Academy Award for Best Picture, was born Anita Pomares, the granddaughter of Salvadoran immigrants in New York City. In 1929, when The Broadway Melody won the Oscar, she was described in Photoplay magazine as “a blond, blue-eyed Latin” with “a dash of Spanish ancestry” (cited in Rodríguez 2004, p. 14). Numerous other Hollywood actors of the 1910s and 1920s changed their names to capitalize on the ‘Latin’ vogue of the time: New Mexican Joe Page rose to stardom as Don Alvarado; Viennese immigrant Jacob Krantz became Ricardo Cortez; and Mexican-born Paula Marie Osterman allegedly tried out three different names before Hollywood executives decided that Raquel Torres “sounded Spanish enough” (Rodríguez 2004, p. 16). In a 1928 feature in Photoplay, Joan Crawford was pictured in a lace mantilla and Spanish shawl with a caption that described her as “more Spanish than the Spaniards themselves” (cited in Rodríguez 2004, pp. 4–5).

The Tango craze Hollywood’s first famed “Latin Lover,” Italian-American actor Rudolph Valentino (born Rodolfo Guglielmi), rose to stardom dancing tango in the nation’s top-grossing film of 1921, The Four Horsemen of the Apocalypse (based on the novel written by Spanish author Vicente Blasco Ibáñez). Most attribute the ‘tango craze’ that spread throughout Paris, London, and ultimately New York in the 1910s to the success of the musical comedy The Sunshine Girl, which premiered at London’s Gaiety Theatre in February of 1912. Groppa (2004) observed that, in this production, the British performers “appeared on stage dressed like Spaniards,” including Andalusian hat and Manila shawl, thus giving “a poor example to a nascent Hollywood film industry that never shook off this stereotype” (p. 11). One year later, the show opened in New York’s Knickerbocker Theatre, marking the entry of this characteristically Buenos Aires bordello-born style of music and dance into the US cultural imaginary.13 By January 1914, The New York Times ran a seven-column headline declaring “All New York Now Madly Whirling in the Tango”; there were some 700 dance halls across the city open from noon until dawn of the next day (Groppa 2004, p. 21). The fact that tango became entirely in vogue among the city’s aristocrats and the local social scene of the upper echelon secured its success. Perhaps because the principal protagonists of the tango craze in New York were not Argentinian (or even Argentine-American), the linguistic dimension of the music was never fully conveyed. Groppa speculated that the reason so many prominent Argentinian tango professionals did not prosper in New York at the time was the language barrier. This was still 25

Andrew Lynch

true in the 1930s for the world’s best-known tango artist, Carlos Gardel, whose acting career with Paramount Pictures was curtailed by his reported inability to master English; none of his highly successful Spanish-language films were subtitled or dubbed into English. Tucci (1969) recounted that at the New York premier of his 1934 film Cuesta abajo at the Teatro Campoamor in East Harlem (i.e. Spanish Harlem), crowds filled the street to catch a glimpse of the star, who was received with a 15-minute ovation at the theater. Undoubtedly, the throngs of fans who awaited him were mostly Caribbean-origin Spanish speakers. New York’s Hispanic population increased from 22,000 in 1916 to 134,000 in 1940; according to the 1930 Census, 41% of that population was Puerto Rican and another 18% were Cuban or Dominican (Haslip-Viera 2010, p. 36). Puerto Rican (im)migration was driven by the economic prosperity of the 1920s and the Jones Act of 1917, which granted Puerto Ricans US citizenship. By the early 1930s, formerly Jewish-predominant East Harlem had become a majority Puerto Rican neighborhood.14 In sum, during the 1910s and 1920s—a time of economic prosperity across the US— things ‘Spanish,’ and to lesser extent ‘Latin’ or ‘Spanish-speaking,’ entered the nation’s cultural imaginary for the first time, through seemingly localized phenomena: economic development and tourism in New Mexico; architectural styles in specific urban settings of Florida and the Southwest; tango halls in New York City; and the spaces of Hollywood movies and the identities of actors who appeared in them. Because the film of this era was silent, language was a nonissue. With the advent of talkies in the late 1920s and early 1930s—coinciding with the nation’s most catastrophic economic downturn and the resonance of political unrest in Europe—Latino actors, and particularly those who spoke Hispanic-accented English, began to be typecast. It was with good reason that US Spanish heritage speaker Margarita Carmen Cansino changed her name to Rita Hayworth shortly after entering the Hollywood film industry in the mid-1930s. During this time, Hispanophilia quickly waned as the real-life consequences of the Wall Street Crash were felt, and political tensions between populist, nationalist, socialist, and communist ideologies grew across Europe and Latin America.

Repercussions of the Great Depression The Good Neighbor Policy During the 1930s, in keeping with his longstanding Good Neighbor Policy, President Franklin D. Roosevelt pledged that the US would no longer intervene or interfere in the domestic affairs of the nations of Latin America. His administration undertook various efforts of cultural diplomacy aimed at establishing more positive relationships with Latin American governments. Special emphasis was placed on Pan American Day in public schools, and exchange programs for art exhibits and for university professors and students were established. Latin American leaders were received in Washington with pomp and circumstance, and the 1939 World’s Fair in New York City, with the motif “The World of Tomorrow,” showcased international relations. Brazil’s Ambassador of Good Will for the Fair was Carmen Miranda, who went on to become the muse of Roosevelt’s Good Neighbor campaign, appearing in 14 Hollywood films. Like Lupe Vélez before her in the 1930s, Miranda embodied and performed the stereotype of the vivacious and ‘hot-blooded’ Latina who spoke highly accented English.15 In conjunction with the Office of Coordinator of Inter-American Affairs (CIAA), Hollywood’s Production Code Administration was charged with presenting Latin Americans in a favorable light, introducing Latin American themes and settings in film, and ensuring linguistic ‘authenticity’ (Rodríguez 2004, p. 82). Adams (2007) pointed out that this policy was “not entirely altruistic or indicative

26

US latinidad, Spanish as heritage language

of a new and higher level of ethnic consciousness but in large measure economic” as Hollywood looked increasingly toward Latin America during the 1930s, when European markets were falling under the control of Nazi Germany (p. 292). The elites of Latin America who controlled the markets targeted by Hollywood executives were concerned with promoting a particular image of their own, to wit: “the modernity and architectural wonders of large Latin American cities with skyscrapers, automobiles, and surprisingly clean and bustling city streets with fashionably western-dressed [light-skinned] men and women” (Adams 2007, p. 294).

Immigration and deportation While Hollywood exported preconceived images of Latin America to Latin America itself, Washington deported Mexicans. From 1929 to 1935, according to the Immigration & Naturalization Service (INS), about 82,000 Mexicans were formally removed from the US, some through voluntary proceedings (U.S. Citizenship and Immigration Services 2014). However, the number reported in 1936 by the U.S. Consulate General in Mexico City for the years 1930–1935 was much higher: 345,839. Other historical estimates place the figure higher still (Koch 2006). The official website of U.S. Citizenship and Immigration Services (2014) states that: In 1930, as the extent of the Depression became more clear some Americans accused Mexicans, as well as other aliens, of holding jobs needed by U.S. citizens. At the same time, local relief agencies began to feel the strain of using decreasing resources to serve an increasingly needy populace. Many agencies felt pressure to exclude foreign-born applicants from receiving aid. Some agencies and local governments began requiring applicants to show proof of legal residence. Others used the threat of federal immigration law, which held that immigrants who became “public charges” could be deported, to discourage them from requesting aid. These conditions likely caused many Mexicans to consider returning to their native country. However, personal accounts such as that related by Ignacio Piña, a Mexican-American child living in Montana in 1931, offer other insight into the experiences of many of those deported: “They came in with guns and told us to get out . . . They didn’t let us take anything,” not even a trunk that held birth certificates proving that he and his five siblings were U.S.-born citizens. The family was thrown into a jail for 10 days before being sent by train to Mexico. Piña says he spent 16 years of “pure hell” there before acquiring papers of his Utah birth and returning to the USA. Koch 2006 By 1942, the Roosevelt administration would initiate the Bracero program through diplomatic agreement with the Mexican government, bringing nearly a quarter of a million Mexican workers into the US over the next five years and appropriating nearly $120 million for the program, “every penny of which should be regarded as a direct subsidy to the large-scale employers of farm labor in a period of unprecedented prosperity” (McWilliams 1990, p. 238). As a result of the strenuous push-and-pull dynamics of cross-border movements and heightened patriotism in the midst of World War II, the Zoot Suit Riots erupted in Los Angeles in 1943, linked to ‘Pachuco’ urban forms of dress and creative Spanish-language use among young Mexican-Americans.

27

Andrew Lynch

The emergence of ‘Spanglish’ as political construct Simultaneously, Operation Bootstrap got underway in Puerto Rico as a late outgrowth of the Good Neighbor policy, on the heels of US federal attempts to linguistically Anglicize the population. In defense of Spanish language use, Puerto Rican intellectuals emphatically recalled the Island’s Spanish heritage. In Problemas de la Cultura Puertorriqueña (1934), Emilio Belaval affirmed that: [s]omos españoles hasta la médula, y en nuestro contenido nacional hay una cantidad respetable de españolismo vital . . . A la afirmación que más miedo le tenemos es a admitir que somos españoles, y por eso, es que hemos resistido sin que hasta ahora se nos rompa el espinazo. cited in Vélez 2000, pp. 12–13 Manuel Rivera Matos (1940) wrote that: “sabemos que el espíritu y la expresión de un pueblo son consustanciales con el idioma en el cual está contenido toda nuestra herencia cultural”; and Antonio Sáez (1940) declared that: “el destino de nuestra cultura está fatalmente determinado por el destino de nuestra lengua. La lengua es algo más que un instrumento de comunicación. Es una manera de ver el mundo” (both cited in Vélez 2000, p. 13).16 But it was perhaps Salvador Tió’s argument that would have the most longstanding and far-reaching ideological implications for future discussions of Spanish-English bilingualism in the US: “Teoría del espanglish.” Tió’s playful tongue-in-cheek commentary, published in Diario de Puerto Rico in 1948, is credited with giving rise to use of the term ‘Spanglish’ in popular and academic spheres. In the opening lines of the article, Tió declared that bilingualism is “dos lenguas muertas”; “una disgregación del pensamiento”; “máquina de fabricar gagos.” Ironically affirming that he is opposed to bilingualism because of the damage that Puerto Ricans might do to the English language (“No debemos cargar con la responsabilidad histórica de acabar con un gran idioma”), he proposes—in jocose tone—that ‘Spanglish’ must be regarded as the unitary language of Puerto Rico: “El espanglismo pretende que usemos una lengua como si fuesen dos.” More than 60 years later, Otheguy and Stern (2010) maintained that the now pervasive use of the term ‘Spanglish’ has proven pernicious, principally because the extraordinary degree of fusion with English language structures that is suggested by this etymologically hybrid term ideologically alienates Spanish speakers in the US from those living elsewhere. Moreover, they argued, it perpetuates the widely held misconception that the language spoken by US Latinos is not ‘really Spanish’ or is an ‘incomplete’ version of what should be ‘Spanish’ (see Otheguy 2016 on this latter point). They offered a historical explanation for this false belief: A strategy of scorn and contempt of Spanish speakers was established in the USA in the 1940s and 1950s in the wake of the early waves of Latin American immigration. Many academics and commentators of the time demeaned the Spanish of these immigrants because it was not Castilian Spanish. That which you speak, the immigrants were admonished, is not Spanish, because it does not reflect the norms of north-central Spain . . . Many of them accepted this criticism and decided that the language that they had brought from Mexico, Puerto Rico, Cuba, or elsewhere, was of little value . . . Yesterday’s strategy of depriving immigrants of their Spanish language because it was not Castilian has been transmuted, today, into the attempt to take it from them by labeling it as Spanglish. Otheguy & Stern 2010, p. 97 28

US latinidad, Spanish as heritage language

Implications and conclusions The economic and political circumstances of everyday life that developed during the days of the Great Depression and the conflict of World War II would provide the ideological and theoretical basis of the contemporary sociolinguistic study of Spanish in the US, principally in locales along the US-Mexico border and in New York City vis-à-vis the Island. Among the scholars whose now classic works reflected those circumstances well were Rosaura Sánchez (1972, 1983) and Joshua Fishman and colleagues (1971) (see Klee & Lynch 2009 for further references). Through a Marxist analysis of Spanish language use in the Southwest during the 1970s, Sánchez exposed the pervasiveness of US consumerist ideology along both sides of the border, emphasizing the impact of upward economic mobility on Spanish language use and the ‘push-and-pull’ effects of migration on linguistic identities. In his seminal work North from Mexico first published in 1948, McWilliams affirmed that: Migration from Mexico is deeply rooted in the past. It follows trails which are among the most ancient on the North American continent. Psychologically and culturally, Mexicans have never emigrated to the Southwest: they have returned . . . As finally fixed, the border was a border of the borderlands rather than a national boundary. 1990, pp. 62–63 Referring to the Pachuco speech phenomenon that spread from El Paso to the cities of California during the 1920s and 1930s, McWilliams characterized the Mexican borderlands as a region of “cultural fusion” (pp. 258–261). His idea perhaps conveyed a sense similar to that of the term ‘transculturation’ coined by Cuban anthropologist Fernando Ortiz in 1947, and later applied to the sociolinguistic situation of Cuban-Americans in Miami by the end of the millennium (Otheguy, García, & Roca 2000). Many of the older Chicanos who formed part of Sánchez’s research surely lived the traumatic effects of the Bracero program, the stark anti-immigrant attitudes of the 1930s and 1940s, as well as the deportation practices that stemmed from those attitudes. The social stigmatization, stereotyping, and conflict of identities experienced by that generation likely led Mexican-origin actors such as Lupe Vélez to be typecast in Hollywood—in much the same way that Sofía Vergara has been in the present-day popular television series Modern Family— and others such as Rita Hayworth to attempt to erase their Latino heritage altogether. In urban areas, segregation and marginalization created the sort of living conditions that conditioned the cultural and institutional complexities of language choice at the crux of Fishman’s seminal work Bilingualism in the Barrio, the same conditions reflected in the iconic Broadway production West Side Story (Negrón-Muntaner 2000), and in Zentella’s widely cited ethnography Growing Up Bilingual (1997). The issues of perceived/constructed ‘Spanish’ heritage, ‘Latin’ identity, cultural production, and (im)migration prior to World War II that I have described in this chapter, in relation to local social and economic currents, continue to give shape to the US cultural imaginary in the present day, informing the theoretical and ideological bases of our work as teachers and researchers of Spanish language in the current US political climate. What notions of Latina/o and/or Spanishspeaking identity during the early 20th century might suggest, in retrospect, is that the construction of latinidad is driven largely by economic and political motivations. Greater prosperity is more likely to condition broader acceptance, visibility, and general interest in ‘things Latin’ in the US context; economic downturns and political crisis or instability give way to a heightened sense of Hispanophobia and concomitant English-only attitudes. It is surely no coincidence that 29

Andrew Lynch

the mainstream vogue of things Latin came to a rather abrupt halt with the Wall Street Crash in 1929 and the onset of the Great Depression during the early 1930s, just as the burst of the Dot-com bubble in 2000 and the September 11 attacks of the following year soon quelled the ‘Latin fever’ of the 1990s. During the ‘Roaring Twenties,’ markets soared, real estate boomed, and consumerism flourished as a consequence of low unemployment rates. Similarly, the 1990s were characterized by great economic prosperity, with the US federal government reaching not only a zero deficit by the final years of the Clinton administration but actually entering into a budget surplus (FactCheck.org 2008). Although the economy per se was obviously not what produced the Latino cultural trend, prosperity surely facilitated it. While observing the relative prominence of latinidad in the national cultural imaginaries of both time periods, one must always bear in mind that anti-immigrant sentiment was still very much present. On the one hand, the political context was rife with anti-immigrant sentiment and English-only imperatives in the early 20th century. We recall Teddy Roosevelt’s 1907 Proclamation that: “We have room for but one language here, and that is the English language.” In 1906, Roosevelt signed The Naturalization Act into law, which required immigrants to demonstrate English language ability to be granted citizenship, despite the absence of any official language policy at the federal level in the US. Some scholars attribute these restrictive language policies to the government’s effort to subdue revolutionary immigrant workers. Yet, on the other hand, there was a growing appreciation of things Spanish, and an effort to recall and reconstruct the nation’s Spanish legacy through architecture, tourism, and advocacy for Spanish language education. This apparent contradiction of ideological terms was not a fluke of the time; it can still be observed in our present moment. The findings of a study published in 2011 by Branton, Cassese, Jones and Westerland highlighted the difference between attitudes toward acculturation, attitudes toward the Latino presence in the US, and attitudes toward Latino visibility in the mass media versus anti-immigrant attitudes per se. Analyzing data taken from the National Election Studies in 2000 and 2004, these authors demonstrated that acculturation fear, anti-Latino sentiment, and media exposure were determining factors that conditioned attitudes toward immigration after the 9/11 terrorist attacks; yet in the 2000 data, before the attacks, they observed that none of those ‘Latino’ factors bore a correlation with attitudes regarding immigration. This study suggests that anti-immigrant and English-only sentiment may in some respects be independent of attitudes toward Latino visibility in the media and in cultural production, or the Latino presence in the US more generally. This would perhaps explain some of the apparent contradiction of the 1990s, when a so-called ‘Latin fever’ swept US popular culture and Spanish emerged nationally as a legitimate voice of commerce, advertising, and retail in relation to transnational markets and neoliberalist economic strategy; and concomitantly, Spanish language programs at schools, colleges, and institutions of higher learning expanded significantly. Yet it was also in 1998 that the state of California banned bilingual education programs in public schools, immediately followed by Arizona and Massachusetts. In addition to noting that acculturation fear, anti-Latino sentiment, and media exposure bore little or no relationship to anti-immigrant attitudes in the year leading up to the 9/11 attacks, Branton et al. (2011) also made an interesting observation regarding the volume of immigrationrelated stories published in selected newspapers from across the US: there was no significant increase of immigration-related stories during 2001–2005 (in comparison to 1996–2000); but in major newspapers published beyond the Border region, there indeed was a significant increase in that same time period. This was the case of the Charlotte Observer, the Minneapolis Star-Tribune, the New York Times, and the Washington Post. This finding is perhaps suggestive of the variations of local versus national imaginaries, and localized factors or forces that condition language ideologies. Although anti-immigrant and English-only ideologies pervaded the US during the 30

US latinidad, Spanish as heritage language

early 20th century, attitudes toward things ‘Latino’ in the cultural imaginary could oft be quite positive, especially at local levels. Irrespective of the economic circumstances—prosperity or austerity—racism remains a constant variable. The European-origin (i.e. Spanish or ostensibly Spanish-background), ‘whitewashed’ images of latinidad so prevalent in 1920s Hollywood remain to the present day, as Dávila (2000) observed: Representations of latinidad in the Spanish TV networks, when not revolving around generic representations that prioritize white Hispanic actors and talent, have generally reduced different Latino/a subgroups to particular cultural indexes . . . The dominance of white Hispanic types in Spanish TV negates and leaves little room for acknowledging Latinos/as’ racial and ethnic diversity. 2000, pp. 40–4117 Linked to such generic representations of Latinos in the mass media is an equally generic, i.e. normatively ‘neutral,’ Spanish-speaking voice (Valencia & Lynch 2016). Renewed and at times quite virulent debates regarding institutionalized racial discrimination, (im)migration control along the US-Mexico border, the use of Spanish in US public life, the role of Spanish language in mass media and marketing, as well as bilingual education policy and practice in recent years, serve to remind us that the construction of latinidad is ongoing and, in some respects, has seemingly—and in some regards lamentably—evolved very little over the past century. My hope is that, through this chapter, readers may confront the present volume with a historically more nuanced understanding of the necessary relationship between ‘Latina/o’ cultural imaginaries and Spanish language use and acceptance in the US context, and the potential contribution of these imaginaries to language research agendas.

Notes 1 Aparicio (2003) notes that the definition of latinidad depends upon the academic discipline from which the concept is approached. In the realm of cultural studies, she states that: “Latinidad has been partly defined as the ways in which the entertainment industry, mainstream journalism, and Hollywood have homogenized all Latinos into one undifferentiated group, thus erasing our historical, national, racial, class, and gender subjectivities” (p. 91). Despite the stereotypes and essentializing or particularistic images that the construct often serves to reiterate and perpetuate, Aparicio argues that latinidad as a site of potential homogenization—or, on the other hand, cultural wars—can be fruitfully exploited to reveal convergences and divergences along common lines: “By rethinking Latinidad through media and popular culture as a site through which we produce knowledge about a Latino other and we explore our (post)colonial analogies, we are, in fact, proposing a ‘decolonial imaginary’” (p. 94). 2 Roberts (1979) suggested that: “The 1930s saw the introduction of Latin music as a substyle within US popular music as a whole; the late 1970s saw its clear emergence as a major ingredient in the sound of almost all American popular idioms” (p. 186). 3 As explained by Powell (1971), “The basic premise of the Black Legend is that Spaniards have shown themselves, historically, to be uniquely cruel, bigoted, tyrannical, obscurantist, lazy, fanatical, greedy, and treacherous” (p. 11). Many attribute this ‘Hispanophobic’ bias in historical documentation to propaganda of the British Crown and the Protestant Reformation beginning in the early 16th century; others argue that it predates the 16th century. According to Weber (2000), the English prejudice toward the Spanish was transplanted to North American soil, tainting contemporary US-Latin American relations and fomenting anti-Hispanic and racist sentiment within the context of the US. Along those lines, González Pino and Pino argued in 2007 that that the leyenda negra still influenced the attitudes and perceptions of Spanish heritage language (SHL) students in the US Southwest.Their survey study revealed that more than half of SHL students considered that the public perception of Spain’s heritage in both Latin America and the United States was negative; “half appeared not proud of the Spanish heritage,

31

Andrew Lynch

4 5

6

7

8

9

10

11

12

and 63% picked negative characterizations of it” (p. 235). As one reviewer pointed out, such negative stereotypes regarding non-Anglos and immigrants in general have traditionally existed in the US. Sanz-Sánchez (2014) observed that “New Mexico, where both bilingualism and shift to English seem to have been kept largely at check until the early 20th century, was a clear exception to [the] general process of language shift” in other areas of the Southwest (p. 225). The Naturalization Act of 1906, signed into law by President Theodore Roosevelt, required that immigrants demonstrate English language ability in order to be granted citizenship. As Ovando explained, the restrictive language policies of the turn of the century likely came about “as an ideological support for the new imperialism, as a practical instrument of colonial rule, and as a form of social control at home, directed at revolutionary immigrant workers” (2003, p. 18). The New Mexico state legislature created the Spanish-American Normal School at El Rito in 1909 with the express purpose of preparing New Mexican Spanish-speaking natives to become public school teachers in areas where the Spanish language was prevalent; the school is now Northern New Mexico University. In early debates around statehood in New Mexico in 1890, both Anglo Republicans and Nuevomexicanos objected to English-only provisions. The Republicans’ reasoning was that, once New Mexico joined the Union, the influx of European and Anglo-American immigrants would naturally produce language shift from Spanish to English in all spheres of public life (Nieto-Phillips 2004, pp. 76–77). Indeed, this is what ultimately happened. Limón affirmed that: “In examining Aurelio Espinosa’s writings on the Spanish Civil War, we see the way that his understanding of Spanish identity extended from the realm of the imaginary past and into the realm of present-day politics” (2014, p. 464). Espinosa’s Catholic conservatism and right-wing political orientation shaped his relationship with Spanish philologists of his time, such as Ramón Menéndez Pidal, whom Espinosa condemned for supporting the Republican cause in Spain (Limón 2014, p. 463). Explicit in his support of Franco’s Nationalist party during the 1930s, Espinosa was, according to Limón, probably apprehensive about Franklin D. Roosevelt’s New Deal project: “Against these menacing versions of modernity being enacted in the United States and across the Atlantic, Espinosa’s folklore studies attempted to promote and exalt traditional Catholicism and what he viewed as premodern forms of social organization in the Southwest” (p. 464). Early examples in Florida are the Ponce de León Hotel (now Flagler College) and the Alcázar Hotel (now Lightner Museum) in St. Augustine, both completed in 1887. Castillo San Marcos, the original fortress constructed by the Spanish to defend St. Augustine, is among the extremely few preserved or reconstructed architectural remains of the original Spanish rule of Florida. Founded by Pedro Menéndez de Avilés in 1565, the city is today considered America’s oldest. Gebhard identifies two chronological phases of Spanish Colonial Revival style architecture: 1) Mission Revival (1880s–1900s), which was loosely inspired by early California mission constructions, as well as Pueblo or Santa Fe Revival (1880s–1920s), inspired by the provincial Spanish colonial buildings in the Río Grande River Valley of New Mexico; and 2) Mediterranean Revival (1910s–early 1930s), combining elements principally from Spain and Italy, but sometimes Mexico and Islamic North Africa as well (1967, pp. 131–132). Fluent in various languages, Novarro took pride in his Spanish-speaking background and was actively involved in outreach to the Latino community. He built Teatro Íntimo in his home, where Los Angeles youth performed plays and songs entirely in Spanish. He also starred in highly successful Spanishlanguage films in Mexico. In 1929, at the height of her celebrity, Del Río declared that: Someday I would like to play a Mexican woman and show what life in Mexico really is . . . I am eager to play in stories concerning my native people, the Mexican race. It is my dearest wish to make fans realize their real beauty, their wonder, their greatness as a people. Carr 1979, p. 42

13 Groppa affirms that tango and jazz emerged simultaneously, and in parallel fashion: Having both risen at the end of the nineteenth century as a result of rhythms, races, and a dissatisfaction of the mixture, jazz, like the tango, originated in dance halls and slum brothels. Both were rejected socially but were eventually accepted and considered authentic art forms of their respective countries. Groppa 2004, p. 15

32

US latinidad, Spanish as heritage language 14 Though Puerto Ricans were significantly present in New York City by the 1930s as well as other urban areas of the US by the 1940s, it was not until the 1957 Broadway musical (and 1961 blockbuster film) West Side Story that they would be interpellated in the US mainstream cultural imaginary. As NegrónMuntaner explains: The film opened a discursive space from which to speak for the “real” Puerto Rican community . . . West Side Story locates Puerto Rican identity at the crossroads of colonialism, racialization, and shame by addressing not just one Puerto Rican but a whole community as abject. Yet, in hailing Puerto Ricans and immediately constituting their subjectivity as criminal, this group acquires several previously denied possibilities, including social and visual representability. Negrón-Muntaner 2000, pp. 84–86 15 In her recurring lead role as Carmelita Lindsay in the seven-film Mexican Spitfire series, which was associated with Hollywood’s Good Neighbor initiatives, Vélez exaggerated accent, grammatical imprecision, hand gestures, and facial expression to comic effect. This typecast character secured her place in Hollywood until her suicide in 1944, after so many other Latino actors had fallen out of work in the 1930s. 16 It was not until 1949, when Puerto Rico’s first democratically elected governor Luis Muñoz Marín took office, that Spanish was declared the language of instruction in public schools. 17 Dávila reminds the reader that: Debates about the representativity of any media are not unique to Latino-oriented and Spanish-language media. Such controversy is, in fact, common to any media whose mass scope inherently excludes entire segments of its putative audience . . . In contrast to the socalled general market media, however, the Hispanic media have historically functioned and promoted themselves as the “spokespersons” for the totality of U.S. Latinos, a pretense that makes even more problematic the real biases of their representations. Dávila 2000, p. 38

Further reading Aparicio, F. and Chávez-Silverman, S. (Eds). (1997). Tropicalizations: Transcultural Representations of Latinidad. Hanover, NH: University Press of New England. Dávila, A. (2001). Latinos Inc.: Marketing and the Making of a People. Berkeley, CA: University of California Press.

References Adams, D. (2007). Saludos Amigos: Hollywood and FDR’s Good Neighbor Policy. Quarterly Review of Film and Video, 24, 289–295. American Association of the Teachers of Spanish and Portuguese. (n.d.). History of the AATSP. Retrieved from www.aatsp.org/?page=History. Anderson, B. (1983). Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso. Aparicio, F. (2003). Jennifer as Selena: Rethinking latinidad in media and popular culture. Latino Studies, 1, 90–105. Bills, G. (1997). New Mexican Spanish: Demise of the earliest European variety in the United States. American Speech, 72, 154–171. Bills, G. and Vigil, N. (1999). Ashes to ashes: The historical basis for dialect variation in New Mexican Spanish. Romance Philology, 53, 43–67. Branton, R., Cassese, E., Jones, B. and Westerland, C. (2011). All along the watchtower: Acculturation fear, anti-Latino affect, and immigration. The Journal of Politics, 73, 664–679. Callahan, L. (2009). Spanish and English in U.S. Service Encounters. New York: Palgrave Macmillan. Cañas, D. (2010). New York City. In C. Remeseira (Ed.), Hispanic New York. A Sourcebook (pp. 245–300). New York: Columbia University Press. Carr, L. (1979). More Fabulous Faces. New York: Doubleday.

33

Andrew Lynch Crawford, J. (n.d.). Language rights and New Mexico statehood. Language Policy Web Site & Emporium. Retrieved from www.languagepolicy.net/archives/nm-con.htm#5. Dávila, A. (2000). Talking back: Hispanic media and U.S. latinidad. CENTRO Journal, 12, 37–47. Dawson, G. (1994). Soldier Heroes: British Adventure, Empire, and the Imagining of Masculinities. London: Routledge. Espinosa, A. (1910). New-Mexican Spanish folk-lore. Journal of American Folklore, 23, 395–418. Espinosa, A. (1917). Hispania. Hispania, 1, 19–23. FactCheck.org. (2008). The budget and deficit under Clinton. Retrieved from www.factcheck. org/2008/02/the-budget-and-deficit-under-clinton/. Fishman, J., Cooper, R. and Ma, R. (1971). Bilingualism in the Barrio. Bloomington, IN: Indiana University, Research Center for the Language Sciences. Gebhard, D. (1967). The Spanish colonial revival in southern California (1895–1930). Journal of the Society of Architectural Historians, 26, 131–147. Goodhue, B. (1916). The Architecture and the Gardens of the San Diego Exposition. San Francisco, CA: Paul Elder & Co. Groppa, C. (2004). The Tango in the United States. A History. Jefferson, NC: McFarland & Co. Haslip-Viera, G. (2010). The evolution of the Latino community in New York. In C. Remeseira (Ed.), Hispanic New York. A Sourcebook (pp. 33–56). New York: Columbia University Press. Klee, C. and Lynch, A. (2009). El español en contacto con otras lenguas. Washington, DC: Georgetown University Press. Koch, W. (2006, 4 April). U.S. urged to apologize for 1930s deportations. USA Today. Retrieved from http://usatoday30.usatoday.com/news/nation/2006-04-04-1930s-deportees-cover_x.htm. Limón, R. (2014). The science of folklore: Aurelio Espinosa on Spain and the American Southwest. Journal of American Folklore, 127, 448–466. Lippi-Green, R. (2012). English with an Accent. Language, Ideology, and Discrimination in the United States, 2nd ed. New York: Routledge. Lummis, C. (1911 [1892]). Some Strange Corners of Our Country. New York: Century Co. Lynch, A. (2011). Novelist Cristina García reflects on language and bilingualism (Andrew Lynch interviews). Retrieved from www.youtube.com/watch?v=kvi43MB_QuU. Lynch, A. and Potowski, K. (2014). La valoración del habla bilingüe en Estados Unidos: fundamentos sociolingüísticos y pedagógicos en Hablando bien se entiende la gente. Hispania, 97, 32–46. Marcantoni, J. (2015). How native speakers ruined Spanish for everyone else. Latino Rebels, 27 September 2015. Retrieved from www.latinorebels.com/2015/09/27/how-native-speakers-ruinedspanish-for-everyone-else/. McWilliams, C. (1990 [1948]). North from Mexico. The Spanish-Speaking People of the United States. New York: Greenwood. Negrón-Muntaner, F. (2000). Feeling pretty: West Side Story and Puerto Rican identity discourses. Social Text, 63, 83–106. New York Times. (1898, 24 August). Topics of the times. p. E3. [No author named.] New York Times. (1926, 11 July). Would make Cuba a tourists’ mecca. p. E3. [No author named.] Nieto-Phillips, J. (2004). The Language of Blood. The Making of Spanish-American Identity in New Mexico, 1880s–1930s. Albuquerque, NM: University of New Mexico Press. Otheguy, R. (2016). The linguistic competence of second-generation bilinguals: A critique of ‘incomplete acquisition’. In Tortora, C., den Dikken, M., Montoya, L. and O’Neill, T. (Eds.), Romance Linguistics 2013. Selected Papers from the 43rd Linguistic Symposium on Romance Languages (pp. 301–319). Amsterdam: John Benjamins. Otheguy, R., García, O. and Roca, A. (2000). Speaking in Cuban: The language of Cuban Americans. In McKay, S. and Wong, S. (Eds.), New Immigrants in the United States (pp. 165–188). Cambridge, UK: Cambridge University Press. Otheguy, R. and Stern, N. (2010). On so-called Spanglish. International Journal of Bilingualism, 15, 85–100. Ovando, C. (2003). Bilingual education in the United States: Historical development and current issues. Bilingual Research Journal, 27, 1–24. Pérez, L. (1999). On Becoming Cuban: Identity, Nationality and Culture. Chapel Hill, NC: University of North Carolina Press. Pino, B. G. and Pino, F. (2007). An exploration of cross-cultural antagonism in heritage students of Spanish: La leyenda negra in a new guise. NAAAS Conference Proceedings (pp. 226–238). Scarborough, ME: National Association of African-American Studies.

34

US latinidad, Spanish as heritage language Powell, P. W. (1971). Tree of Hate: Propaganda and Prejudices Affecting United States Relations with the Hispanic World. New York: Basic Books. Roberts, J. S. (1979). The Latin Tinge. The Impact of Latin American Music on the United States. New York: Oxford University Press. Rodríguez, C. (2004). Heroes, Lovers, and Others. The Story of Latinos in Hollywood. Washington, DC: Smithsonian. Sánchez, R. (1972). Nuestra circunstancia lingüística. El Grito, 6, 45–74. Sánchez, R. (1994 [1983]). Chicano Discourse: Socio-historic perspectives. Houston, TX: Arte Público Press. Sanz-Sánchez, I. (2014). ‘Como dicen los americanos’: Spanish in contact with English in territorial and early statehood New Mexico. Spanish in Context, 11, 221–242. Schwartz, A. (2011). Mockery and appropriation of Spanish in White spaces: Perceptions of Latinos in the United States. In Díaz-Campos, M. (Ed.), The Handbook of Hispanic Sociolinguistics (pp. 646–663). Malden, MA: Wiley Blackwell. Tucci, T. (1969). Gardel en Nueva York. New York: Webb. U.S. Citizenship and Immigration Services. (2014). INS Records for 1930s Mexican repatriations. Retrieved from www.uscis.gov/history-and-genealogy/our-history/historians-mailbox/ins-records1930s-mexican-repatriations. Valencia, M. and Lynch, A. (2016). Migraciones mediáticas: la translocación del español en televisoras hispanas de Estados Unidos. Cuadernos AISPI. Revista de la Associazione Ispanisti Italiani, 8, 171–196. Vélez, J. (2000). Understanding Spanish-language maintenance in Puerto Rico: Political will meets the demographic imperative. International Journal of the Sociology of Language, 142, 5–24. Weber, D. (2000). La frontera española en América del Norte (trans. J. Ferreiro). Mexico City: Fondo de Cultura Económica. Whitman, W. (2010 [1883]). The Spanish element in our nationality. In C. Remeseira (Ed.), Hispanic New York. A Sourcebook (pp. 179–182). New York: Columbia University Press. Williams, B. (2005). The bridges of Los Angeles County: Marketing language in the Chicano cinema of Gregory Nava. Canadian Journal of Film Studies, 14, 54–70. Zentella, A. C. (1996). The ‘chiquitafication’ of U.S. Latinos and their languages, OR why we need an anthropolitical linguistics. SALSA III: Proceedings of the Third Annual Symposium about Language and Society at Austin. Texas Linguistic Forum, 36, 1–18. Zentella, A. C. (1997). Growing Up Bilingual. Malden, MA: Blackwell.

35

3 SPANISH IN U.S. LANGUAGE POLICY AND POLITICS Phillip M. Carter florida international university, usa

Introduction: the paradox of Spanish in the United States At the height of the 2012 U.S. Presidential election cycle, the nonprofit group known as the Commission of Presidential Debates announced its panel of moderators for the season’s events. All of them happened to be English-monolingual Anglo Whites. The U.S.-based Spanish language television network Univisión, recognizing that they had been effectively shut out by this decision, filed a request that an additional debate be held to focus on Hispanic-related issues. Under their proposal, the debate would be held in Spanish. The Commission denied the request, a move that prompted Univisión to host their own events – one night for each of the candidates, President Barack Obama and Governor Mitt Romney. The Spanish language events ended up being a ratings success for Univisión: 1.6 million viewers tuned in to watch Romney and 2.7 million to watch Obama. The entire spectacle – from the Commission’s decision to block the participation of a Spanish-language medium, to Univisión’s decision to host its own events in Spanish – was awash in language politics. Univisión’s further decision to hold its events in Miami, the U.S. city where Spanish is most closely linked with economic, sociocultural, and political success (Carter & Lynch 2015), is also hardly accidental. I begin with this anecdote from the world of language politics to illustrate what I call “the paradox of Spanish in the United States.” In the case of the 2012 Presidential debates, both candidates agreed to a Spanish-speaking forum – presumably to gain access to and favor with Spanish speakers in the United States, whom Escobar and Potowski (2015, pp. 9–10) estimate to number 48.6 million1 – but only after that forum was first denied by the official debate commission. As I will demonstrate later in this chapter, political candidates in the United States often make use of Spanish in campaigns and political advertisements, even when their party platforms articulate policies that can be understood as anti-Latinx, anti-Spanish language, or both. The paradox of Spanish is further evident in the explosion of official English and anti-bilingual education laws at the state level in the places with the largest Spanish-speaking populations (e.g. Florida and California) and during the periods with the greatest growth in Latinx populations. In this Chapter, I explore this paradox by examining the use of Spanish in political discourse, talk about Spanish in politics, and discuss language policies toward Spanish in the United States and pay attention to the broader political and cultural contexts in which these policies find traction. The first section of the chapter introduces the paradox of Spanish in the United States in 36

Spanish in U.S. language policy, politics

both political and sociocultural terms. The second section explores Spanish in recent political discourse, including in the 2016 Presidential election cycle. In the third section I give an overview of the ways in which language policies have shifted with the political winds in the history of the United States. I pay particular attention to the progressive policies toward multilingualism passed during the 1960s and 1970s, as well as the restrictive policies that followed in the 1980s and 1990s. I then consider the implication of these policies for language rights in the workplace. Then I consider the way language attitudes toward Spanish in the United States are shaped by language academies, focusing on two recent debates involving North American linguists and the Real Academia Española (RAE): the response to the official definition of espanglish in the RAE dictionary, and the response to the publication of two “style guides” by the Academia Norteamericana de la Lengua Española entitled Hablando bien se entiende la gente. Finally, I make conclusionary remarks and suggestions for linguists interested in influencing language policy.

Discourse about Spanish in politics and beyond In the 2016 Presidential election season, Spanish was at issue during both the Democratic and Republican primaries and later during the general election between Hillary Clinton and Donald Trump, where the various racial meanings that attend Spanish in the United States were made explicit. On the one hand, the election cycle saw the rise of Anglo Spanish speakers – Jeb Bush as a Republican candidate for President during the primary season, and Tim Kaine as Hillary Clinton’s Vice-Presidential running mate during the general election. Very little attention was given to the use of Spanish by the two Cuban American presidential candidates, Marco Rubio and Ted Cruz, who competed for the Republican ticket. On the other hand, the election cycle ultimately saw the complete and total repudiation of Spanish, which coincided with strong antiimmigrant and anti-Latinx policy positions from the political right and the eventual winner of the general election, Donald Trump. Soon after Kaine was announced as Clinton’s running mate, he delivered an address to an audience of Spanish-speakers in Miami, in which he spoke passionately – in Spanish – about living in Honduras and learning Spanish. The audience seemed mostly persuaded, but soon after, Trump surrogates attacked. Scottie Nell Hughs, a Trump supporter and political commentator, remarked on television that “What Mr. Trump did, he spoke in a language all Americans can understand – that is English. I didn’t have to get a translator for anything going on,” in reference to Kaine’s Spanish. During the primary season, Trump famously excoriated Jeb Bush for speaking Spanish, suggesting that “he should really set an example by speaking English in the United States.” Trump’s remarks about Bush, and his surrogate’s remarks about Kaine, suggest a form of nationalism rooted in nativism, and undergirded by the ‘one nation, one language’ ideology described by Anderson (1982). Trump’s nostalgia for an English-only past that never really existed did not keep his campaign from printing signs in Spanish – Hispanics [sic] para [sic] Trump – with the English spelling of ‘Hispanic’ and the grammatical error para for por or con. Returning to the Univisión Presidential forum from the 2012 election, the paradox of Spanish in the United States introduced in the prior section is further evident in the discursive content about Spanish during the two debates. For example, during the visit from Governor Mitt Romney, event co-moderator María Elena Salinas, pointing out that the official Republican party platform calls for English to be the official language of the United States, asks, “¿cuál debería ser el papel del español en los Estados Unidos?” [What role should Spanish play in the United States.] Receiving the question via simultaneous translation, Governor Romney responds: 37

Phillip M. Carter

Well, you know, English is the language of government in this country and that’s the way it’s been for some time, as you know. But I take some inspiration from the comments of Governor Luis Fortuño who is an extraordinary governor of Puerto Rico. He said that Spanish is the language of our heritage, English is the language of opportunity and so if I were, I mean I have a son who speaks pretty fluent Spanish and I’m delighted – my son Craig over there. Look, I’d like our kids to speak foreign languages; if you’re lucky enough to grow up in a home where English and Spanish is [sic] spoken or English and another language, great, learn the other language, be able to communicate with other people in the world, but also teach our kids English so they can achieve the kind of opportunity that is associated with the language which is, really as Governor Fortuño says the language of opportunity here. Romney’s response is noteworthy for the discussion about the politics of Spanish in the United States for at least five reasons. First, declaring English to be the language of government in the United States sets up the assumption of official English monolingualism that interlocks with a broader ideology of English monolingualism (Macías 1985; Wiley 2000), while ignoring the reality of the United States’ internal multilingualism2 and the fact that there is no official language of government (Wiley 2004). Second, by narrativizing Spanish as the “language of our heritage” and English as “the language of opportunity,” language is given a temporality, with Spanish discursively situated in the past and English in the future. Third, by praising his son for learning Spanish, Romney’s response contributes to what linguistic anthropologists and sociolinguists have described as “the double figuration of Spanish,” (Carter 2014; Hill 2008; Urciuoli 1996; Zentella 1997) in which Spanish is constructed as a valuable resource for Anglos while it is constructed as a “barrier” or cultural baggage for U.S. Latina/os. Fourth, Spanish is incorrectly labeled as “foreign language,” thus normalizing the ideological relationship between the U.S. nation state and English along the lines of the une langue, une nation ideology described by Anderson (1982) and Tetel Andresen and Carter (2016), among others, as well as the long history of Spanish in the United States, including Spanish colonial history. At the same time, it also contributes to the ideology described by Santa Ana (2002, p. 211) in which English is understood to be separate, distinct, and “above” all other languages, where the term “language” becomes a cover term for all forms of speech other than English. Finally, the implicature of the clause “but also teach our kids English” is that there is a place in the United States where children – presumably children from Spanish-speaking families – do not learn English, an idea that distorts the reality in which ethnolinguistic minority languages, including Spanish, have been found to undergo cross-generational language shift by or during the third generation (Potowski 2010; Veltman 1983, 1988). Romney’s remarks – presented to a Latinx audience in Miami and broadcast on Univisión – illustrate a final point about Spanish in U.S. politics that must be made explicit: wherever Spanish is mentioned, questions of race are already at stake. This is true not only because language in the United States is always already racialized (Charity Hudley 2017), but also because Spanish in the United States is constructed as an icon of latinidad (Irvine & Gal 2000). By invoking Fortuño, Romney is able to voice hegemonic positions on English and Spanish without appearing racially insensitive. Taken together, these comments, and the spectacle surrounding the 2012 Presidential debates in which they found traction, are further reflective of the broader sociopolitical context in the United States in which Spanish is paradoxically visible (e.g. Presidential candidates did in fact agree to participate in the Univisión debate held in Spanish) and vilified (comments about Spanish construct it as inferior to English). In terms of politics qua politics, this paradox is most immediately obvious when we compare the use of Spanish by politicians in political campaigns, 38

Spanish in U.S. language policy, politics

which Callahan (2004, p. 117) notes is still “cause for comment,” with platform policies related to language. There is no shortage of examples of politicians who campaign in Spanish while espousing English-only or even anti-Spanish positions. Perhaps the most extreme example involves Tom Tancredo, a former congressman from Colorado who later ran for President and Governor of that state in 2014. During his stint in Congress, Tancredo was known for his hardline policies on immigration as well as for his cause célèbre: making English the official language of the United States. During his 2008 Presidential campaign, Tancredo – in keeping with his politics – promised to not advertise in Spanish and did not do so. By the time of his gubernatorial campaign, however, his campaign had released a Spanish language website called Viva Tancredo as well as several Spanish language advertisements. The promise of reaching voters via a Spanish language medium was apparently greater than the English-only politics Tancredo so vehemently articulated. The paradox of Spanish in U.S. politics was once again at play. The discrepancy between practice and politics as it pertains to Spanish is also evident in the Nuestro Himno controversy of 2006, which has been described in detail by Chavez (2008). In the mid-2000s, as debate about immigration reform became prominent in national conversations, immigration activists took to the streets to protest H.R. 4437, proposed legislation that would have classified “illegal” immigrants as felons. In the spring of 2006, millions of people participated in protests across the country – the largest being held in cities such as Chicago, Dallas, and Los Angeles.3 The Spanish-language media – including the Univisión and Telemundo networks – were key in mobilizing participation in the events, where protestors were seen carrying both Mexican and U.S. flags. Protestors also sang Nuestro Himno,4 a Spanish-language version of the English-language Star Spangled Banner. Although the Spanish-language version is not a literal translation, its lyrics evoke the sentiment of the English-language version. In addition to being played during protests, the song also gained popularity on Spanish language radio stations. The English-language media eventually picked up on the song, and their reporting fueled a backlash and national language controversy at a level unseen since the so-called “Ebonics controversy” of 1996 (Rickford 1999; Wolfram 1998). The controversy around Nuestro Himno commanded so much media attention that then-President George W. Bush weighed in publicly on the song. He told reporters that “people who want to be citizens of the United States should learn English and ought to learn to sing the national anthem in English” (Vandehei 2006). As with Romney’s comment that we “should also teach our kids English,” Bush’s remarks on Nuestro Himno also rest on the assumption that those who choose to sing the song in Spanish could not also do so in English, which of course ignores the reality of code choice in U.S. Latinx communities. The comments also assume that the only people who would sing the national anthem in Spanish are those who “want to be citizens,” rather than those who already are. Citizenship is thus discursively tied to English-speaking, as Spanish is situated as being both outside of the purview of citizenship and outside the borders of the nation. Reflecting again this paradox, George W. Bush himself had previously used Spanish language slogans such as Juntos Podemos and Viva Bush in his Presidential campaigns. Other politicians have been more straightforward in articulating anti-Spanish positions. In the 2008 Presidential election cycle, Republican candidate Newt Gingrich stated that “we should replace bilingual education with emergence, [sic] with immersion in English, so people learn the common language of the country, so they learn the language of prosperity, not the language of living in a ghetto.” Though it was not stated explicitly, “the language of living in a ghetto” was presumably Spanish. Like other comments that situate Spanish in the past and English in the future, these remarks work with another false dichotomy: prosperous places (where English is spoken) and ghettoized spaces (where Spanish is spoken). Further, the comments about bilingual education distort the reality that students receive instruction in both languages, Spanish and 39

Phillip M. Carter

English, though the comments imply that bilingual education is in effect education in Spanish only. In this case, Gingrich issued a public apology, but as Hill (2008) points out, apologies rarely, if ever, achieve the same breadth of circulation as the original comments. Thus, the linkage of “Spanish” to “ghetto” was further entrenched in the discursive reality. Discourse by politicians about Spanish as it pertains to Puerto Rico is also noteworthy given (1) the historical presence of Spanish on the island, (2) the official status of both languages, English and Spanish, and (3) ongoing debate about the political relationship between the island and the United States. Puerto Rico is currently an official territory of the United States, a status that could change should Puerto Ricans vote for independence or U.S. statehood.5 We have already seen in this chapter that Presidential candidate Romney invoked comments made by Puerto Rican governor Luis Fortuño, by situating Spanish as “the language of our heritage” and English as “the language of prosperity,” which implied that the role of English on the island should be expanded. A somewhat more hardline position was articulated by Republican primary candidate Rick Santorum when he visited the island as part of a primary campaign in 2012. Although Santorum said he supported ongoing sovereignty for Puerto Ricans as well as the ongoing right to self-determination, his comments implied that there were in fact linguistic limits to this self-determination, namely, that he would only support statehood if English were to be adopted as the primary language of the island. He told a reporter: Like any other state, there has to be compliance with this and any other federal law . . . And that is that English has to be the principal language. There are other states with more than one language such as Hawaii but to be a state of the United States, English has to be the principal language. The point that I wish to make is not that Santorum’s comments are factually incorrect – the U.S. Constitution does not stipulate an official language as such, neither does it impose in any way a language requirement for statehood. Instead, I would like to emphasize that broad, culturally ratified discourses about the nation, reflected in talk about language by politicians, again assumes the naturalness of the citizens of the state. As Santa Ana (2002) points out, while “American English signals allegiance and full citizenship” (p. 236), Spanish speaking and bilingual education are understood to be “a symbolic threat to the social order” (p. 237). For Latinx politicians, who remain relatively few at all levels of government, the relationship to Spanish varies by politician. Florida Representative to the U.S. Congress Ileana RosLehtinen, the first Cuban and the first Hispanic woman elected to U.S. Congress, has opposed English-only legislation within her party and at the same time can be seen speaking Spanish to her constituents and granting interviews to Spanish language television stations across South Florida. In contrast, her colleague Marco Rubio, a fellow Republican who also addresses the media in Spanish, has been at the fore of restrictive language policies. In 2013, he proposed an amendment to the immigration reform bill being debated by the U.S. Senate that would require immigrants to learn English before getting a green card. In 2012, Mexican American mayor of San Antonio Julio Castro delivered what was widely perceived to be a career-boosting keynote speech at the Democratic National Convention. In his speech, he used a few lines of Spanish, including Que dios los bendiga. After the speech, the major controversy was not that Castro had used Spanish words to start with, but whether or not he was qualified to do so since he was not “fluent” in the language. Here again, the odd paradox of Spanish in U.S. politics is clear – Latinx politicians can use Spanish and go unnoticed in doing so, unless they are considered “not really fluent.”6 During the 2016 election cycle, two Cuban American men – Marco Rubio (Florida) and Ted Cruz (Texas) – ran for U.S. President on the Republican ticket. An exchange between 40

Spanish in U.S. language policy, politics

the two men during a Republican primary debate illustrates both the paradox of Spanish in the United States in general and the especially peculiar way Spanish is constructed for Latinx politicians. During a portion of the debate focusing on immigration reform, Cruz noted that “Marco [Rubio] went on Univision [English phonology] – in Spanish – and said he would not rescind President Obama’s illegal executive amnesty.” The line, which implied that Rubio was communicating a different message in Spanish than in English, fed perfectly into the narrative that Spanish allows Latinx people to be secretive and dangerous (Barrett 2006; Carter 2014), and played to jeers and boos from the Republican studio audience. It appeared, for a moment, that Cruz had successfully deployed an anti-Spanish trope against Rubio. As Cruz continued to talk, the camera panned to Rubio, who appeared to chuckle to himself, before replying, “First of all, I don’t know how he knows what I said on Univision [English phonology], because he doesn’t speak Spanish.” As he began “second of all,” over a background of laughter and applause, Cruz interjected: “Marco, si quieres díselo ahora, ahora mismo díselo, en español si quieres.” Rubio continued in English. In the exchange, reference to Spanish was first deployed to disqualify Rubio, and then Spanish was used in a verbal repartee in which the two men tried to one-up each other for what appeared to be the entertainment of the mostly Anglo studio audience. The political meanings of Spanish are thus multiple and contradictory, even within the same speech event.

Spanish in municipal, state, and legal contexts Doral, Florida is a middle-class boomtown located in northwest Miami-Dade County. Not unlike other Miami-Dade municipalities, Latinos make up a solid majority of Doral’s overall population, 79.5% according to the 2010 Census. What is remarkable about Doral is that it is home to the largest Venezuelan community in the United States, for whom it is often referred to as “Doralzuela.” The Venezuelan population there is not only large, but also affluent. Venezuelan-owned businesses predominate. The Spanish language television network Univisión and its English language sister network Fusion are both headquartered there. And Spanish abounds. In 2010, Spanish speakers comprised 82% of the Doral population (U.S. Census 2010). Thus, a proposal brought about by Venezuelan-born mayor Luigi Boria in 2013 to designate Doral to be a “multilingual and multicultural city”7 seemed to make good intuitive sense, especially considering that the effort was framed in terms of promoting economic development. The resolution passed, but with language that seemed to formally designate English to be the official language of the city, while leaving the status of Spanish, which the resolution was ostensibly written to address in the first place, somewhat less clear. The resolution reads, “the City of Doral’s official and main language is English, while also acknowledging Spanish as the second official most used and spoken language in the City of Doral.” Section 2 of the resolution states that “The City of Doral will continue conducting its Council and Advisory Board meetings along with all other official government business in the English language.” Again, the paradox of Spanish in the United States is evident: a document that seeks to celebrate the bilingualism of a Latinx-majority city ends up designating English as the official language of the city, declaring English to be the only language of government business, and with ambiguous language declares Spanish not to be the “co-official” language but “the second official most used and spoken language,” which apparently cannot now be used to conduct government business. The language of the Doral resolution was of course not written in a sociopolitical vacuum. Language policy with respect to Spanish in South Florida has ebbed and flowed with the political winds. In 1973, Miami-Dade County passed an ordinance that declared the county to be bilingual and bicultural. The ordinance was in place for exactly seven years until it was swiftly 41

Phillip M. Carter

overturned by voters in a referendum in 1980, at which time the county became officially “English only.” As the Latinx population of the county expanded during the course of the 1980s due to political crises in Cuba and elsewhere in Latin America, the Anglo White population contracted. These demographic changes were eventually reflected in the Miami Dade County Commission, who in 1993 voted to overturn the English-only ordinance passed in 1980. To further complicate matters, English was made the only official language of the State of Florida, when 84% of voters voted in favor of that legislation. That law has not been overturned and remains current today (as well as in 31 other states). The ebbs and flows of multilingual policy in Florida reflect a broader reality about the role of policy in U.S. language history; namely, as Fishman (2004) points out, language policies move with the changing winds of the broader sociopolitical climate. Crawford (1989) points out that during the colonial era, multilingualism was protected since language was understood to be an integral part of cultural heritage, which the colonists were eager to protect. The pre-Revolutionary Continental Congress even published some of its materials in French and German, in addition to English (Crawford 1989). Hakuta (1986, p. 165) notes that during the post-Revolutionary debates about whether or not to declare a national language, it was finally decided that a single language was “incompatible with the spirit of freedom.” Fitzgerald (1993) writes that support for bilingual education – even monolingual education in the ethnic language – held widespread support from the colonial period to the mid-1800s. By the end of the nineteenth century, attitudes toward bilingualism began to shift, and for the first time, anti-bilingual policies began to emerge in the United States. English was imposed as the language of instruction in Puerto Rico following the Spanish-American War (Fitzgerald 1993). In 1906 the Nationality Act was passed, requiring immigrants to the United States to speak English for naturalization. And in the era of World Wars I and II, anti-German sentiment resulted in a sharp decline in German language and German/English bilingual schools, despite the decision in Meyer v Nebraska (1923) that held that Nebraska’s 1919 law restricting foreign language education was in violation of the 14th amendment of the U.S. Constitution (Fishman 2004). In the 1960s and 1970s, the pendulum swung back slightly in the other direction, with the passage of the Civil Rights Act of 1964, which prohibited discrimination on the grounds of national origin. The Bilingual Education Act was passed in 1968 and provided that low-income children in school districts with a high percentage of language-minority students have access to bilingual education. In 1974 the Act was renewed and even expanded as the income restrictions were removed from law (Fitzgerald 1993). The pendulum shifted once again during the late 1970s and early 1980s. During this period, funding for renewals of the Bilingual Education Act was overwhelmingly tied to English-only instruction (Crawford 1989). In 1981, Senator S.I. Hayakawa proposed an amendment to the U.S. Constitution that would declare English to be the official language of the United States (Wiley 2004). Hayakawa later launched an organization called “U.S. English,” which in addition to promoting a Constitutional amendment, also pushed a broad English-only platform within the United States. As Crawford (1999) points out, 21 states passed laws that restricted the use of minority languages during the 1980s and 1990s. Although the ‘English as official language’ amendment to the Constitution never passed at the federal level, despite being put forth several times, many states have passed amendments to state constitutions or statutes that designate English as the official language. Florida and California – two states with significant Spanish speaking populations – have amended their constitutions to make English the sole official language. Today, 31 states have some kind of Englishonly designation. This is remarkable considering that prior to 1981 – the year that Hayakawa proposed amending the U.S. Constitution – only four states had official language provisions. 42

Spanish in U.S. language policy, politics

The expansion of official English laws at the level of the state has moved in parallel with the broader sociopolitical climate as well as with the growth of non-English speaking communities in the United States. This fact is unlikely to be a coincidence. No state currently designates Spanish to be an official or co-official language. This includes the state of New Mexico, whose Constitution of 1912 set forth plans for a bilingual government in which English and Spanish could be used in the state legislature and laws could be published in both languages.8 The Constitution stopped short of declaring English or Spanish to be the official language of the state, and today Spanish is no longer used in the state legislature. In 1995, New Mexico became the first state to adopt a “State Bilingual Song,” entitled “New Mexico: Mi Lindo Nuevo México.” State and municipal laws, such as those mentioned earlier, are, of course, not merely symbolic. Though the impacts vary state-by-state, official English laws can restrict or curtail voting in languages other than English as well as access to a full range of governmental services. They can also shape the outcomes of legal proceedings where language questions are at stake and more indirectly can shape opinions about language diversity in institutional contexts and in the workplace. Zentella (2014) provides an overview of legal cases involving Spanish speaking on the job. She notes that the Equal Employment Opportunity Commission (EEOC) invokes protections against national origin discrimination articulated in the Civil Rights Act of 1964 as a means of banning on-the-job linguistic discrimination. Two guidelines address language specifically: 1) employers may impose English-only rules only if there is a business necessity; and 2) when employers impose English-only rules, the workers must be informed of the policy and they must be informed of the consequences of disobedience. (Zentella 2014, p. 624) Despite these guidelines, Zentella notes that employers are nevertheless able to dismiss workers for speaking languages other than English simply by claiming that it is bad for business. She describes a number of cases in which workers were fired for speaking Spanish among themselves on the job, as was the case of workers at a Chicago manufacturing plant, who were fired for greeting each other in Spanish. The paradox of Spanish in the United States discussed in this chapter is again evident in those cases in which U.S. Hispanics who were hired for being able to speak Spanish were later fired for speaking Spanish. She cites the Premier Operator Services case in which 13 telephone operators based in Dallas, who were hired to speak Spanish to customers, were prohibited from otherwise speaking Spanish on the job, including during all breaks. When the employer received notification from the EEOC that a complaint had been filed, all 13 employees were immediately fired. Occasionally, cases involving on-the-job language discrimination find their way to U.S. courtrooms. Zentella (2014, p. 625) describes Gutiérrez v. Municipal Court (1988), which considered the legality of English-only rules established for courthouse employees in Huntington Park, California, which effectively banned Spanish speaking among Spanish-speaking clerks who were expressly hired to communicate with the Spanish-speaking public. The Ninth Circuit Court ruled that the English-only policy was in violation of the EEOC guidelines having to do with national-origin discrimination. Judges and municipal police departments have also played a role in criminalizing Spanish. In 2009, Ernestina Mondragón was stopped by Dallas police for making an illegal U-turn, for which she was fined. She was also fined for not speaking English. It turns out that between 2006–2009, some 39 people were fined during traffic stops in Dallas for not speaking English. There is no law in Dallas requiring that drivers speak English (Linthicum 2009). And in 1995, a Texas judge ruled that a mother who spoke Spanish to her daughter was committing a type of child abuse. The judge told the mother that “you’re abusing that child and you’re relegating her to the position of housemaid” (Verhovek 1995). 43

Phillip M. Carter

Anti-Spanish regulations also find their way into U.S. schools and communities, and the formal and informal pressures against speaking Spanish at school are well documented by linguistic anthropologists (e.g., Hill 2008; Santa Ana 2002; Valdés 2001). For example, in 2014, the principal of a middle school in Hempstead, Texas made an announcement over the school intercom that speaking Spanish would not be permitted. The principal was eventually fired (Gray 2014). In 2005, 16-year-old Zach Rubio was suspended from a public school in Kansas for speaking Spanish in the hallway. According to the student, the conversation consisted of two lines, the first a question from a friend who asked, “¿Me prestas un dólar?” to which Rubio replied “no problema” (Reid 2005). In 2004, a teacher at a school in Arizona was accused of hitting or slapping eight students for speaking Spanish in class (Ryman & Madrid 2004). In these cases, the paradox of Spanish in the United States is especially clear. U.S. schools routinely offer Spanish as a “foreign” language – presumably because there is some perceived value in knowing it – at the same time that native speakers are discouraged for speaking Spanish or even penalized for doing so. This points to the ‘double figuration’ of Spanish introduced in the first section of this chapter, in which Spanish figures as a valuable resource for one group (i.e., non-Latinos) while being problematic cultural baggage for U.S. Latin@s. Zentella (1997, p. 283) puts it this way, “Why is the bilingualism of the well-to-do a source of linguistic security, while the bilingualism of the poor is a source of insecurity and disadvantage?”

Attitudes and perceptions: Spanish in the U.S. and language academies Spanish in the United States is not only in a dialectical relationship with English and subject to ideologies about English monolingualism, it is at the same time in a dialectical relationship with ideologies about Spanish qua Spanish. Ideologies about Spanish have been explored in the sociolinguistic and linguistic anthropology literatures on Spanish in the United States. For example, in her study of Puerto Ricans living in el bloque in New York City, Zentella (1997, pp. 47, 8) discusses the stigma attached to what she calls their “spectrum of linguistic codes,” which includes “Spanglish” and several styles of English and Spanish. She further notes that community members had a great deal of awareness about the stigma associated with Puerto Rican Spanish in New York City. Alfaraz (2002, 2014) has approached the study of attitudes toward Spanish in Miami with methods of perceptual dialectology. She studied explicit language attitudes among Miami Spanish speakers toward a number of national origin varieties, including Pre-Revolutionary Cuban Spanish, the variety spoken by the first wave of Cuban immigrants to arrive in Miami in the 1960s, and Post-Revolutionary Cuban Spanish, those immigrants who have arrived since the Mariel boatlift in the 1980s. In her work, participants ranked these varieties according to two criteria: pleasantness and correctness. For both criteria, participants ranked Peninsular Spanish most favorably, followed by Pre-Revolutionary Cuban Spanish. Post-Revolutionary Cuban Spanish ranks toward the bottom, along with Mexican, Puerto Rican, and Dominican varieties for both traits examined, correctness and pleasantness. For a summary of attitude studies on Spanish in the United States, I direct readers to Escobar and Potowski (2015), and for a history of the development of these attitudes I direct readers to del Valle (2007, 2013). Taken together, the studies show that heritage language speakers of Spanish in the United States often express positive attitudes toward Spanish in responses to direct methods, even when they report using it very little themselves. At the same time, the literature in this area shows that U.S. Latinos are largely aware of which varieties of Spanish are considered prestigious and which are not, including varieties that may be considered U.S. Spanish. While the origins of perceptions and attitudes are complex (see Preston 2010), in this final section I would like to consider the role of language authorities, especially language academies, 44

Spanish in U.S. language policy, politics

in perpetuating beliefs about the “quality” of Spanish dialects both in the United States and elsewhere. Of course, no discourse is completely imposed top-down, but the role of language academies in circulating popular narratives about Spanish in the Spanish-speaking world – including the United States – cannot be denied. I begin the discussion outside of the United States. On February 16, 2015, the director of La Real Academia Española, Dario Villanueva, participated in a live radio interview with Uniminuto Radio, a station based in Bogota. When asked about the Academy’s assessment of Spanish in Colombia, Villanueva opined “los colombianos hablan tan bien el español que sorprende positivamente la habilidad con que se expresan y eso nos estimula a mejorar y a darnos cuenta de los defectos y errores que nosotros mismos podamos cometer.” This response was not met with any push-back on the part of the journalists, who took Villanueva’s comments to be unproblematic. I begin with this anecdote in order to put forth two points about the role of RAE in the Spanish-speaking world. First, the positions of the RAE – and those of the Asociación de Academias de la Lengua Española in which the RAE is now technically included – are broadly accepted throughout Spain and Latin America (Del Valle 2009, 2011; Senz & Alberte 2011; Woolard 2012). Second, the perceptions they help to construct find their way into the United States. To return to Alfaraz’s (2002) study of Miami Cubans’ perceptions of 21 Spanish dialects, Colombian Spanish was found to be the fourth most “correct,” behind Peninsular Spanish, PreRevolution Cuban Spanish, and Argentine Spanish. In terms of pleasantness, it was also ranked as fourth, in the same ranking order. Again, attitudes and perceptions are not produced unilaterally by language authorities, though their influence – both in terms of the breadth of their reach and the nature of the ideas they put forth – is difficult to deny. Two controversies involving Spanish in the United States and the Real Academia Esapañola illustrate the problematic way the RAE understands U.S. Spanish and underscore the tense relationship between linguists who work on Spanish in the United States and the RAE. The first controversy involves the publishing of the 23rd edition of the Diccionario de la Real Academia de la Lengua Española by RAE in 2014. As with all updated editions, the 23rd edition contained updated definitions and lexical entries for new words approved by the Academy. This edition was noteworthy, however, for its inclusion of the word espanglish (Spanglish), which was defined as “a form of speech used by some Hispanic groups in the United States, in which they mix deformed elements of vocabulary and grammar from both Spanish and English.”9 In Spain and Latin America, RAE’s decision to include espanglish was met with purist ideologies. In the United States, linguists took issue with the way in which the term was defined, especially the notion that Spanglish is based on “deformed” elements of English and Spanish. The definition – which seems to be uninformed by perspectives of linguists who study Spanish in the United States – was widely considered to be a direct slight at the varieties of Spanish spoken by U.S. Latina/os. A group of prominent linguists whose work centers on Spanish in the United States – led by José del Valle and Ana Celia Zentella – wrote a scathing critique of the definition, which I have reproduced in incomplete form below. Readers can find the entire open letter on the website elcastellano.org.10 Por medio de la presente, los abajo firmantes le queremos expresar a la Real Academia Española y a la Asociación de Academias de la Lengua Española (sin la más mínima esperanza de que les importe un bledo) nuestro más alegre y entusiasmado rechazo del artículo dedicado a la palabra espanglish en la vigésima tercera edición del DRAE. Por un lado, la brevedad y simpleza de la definición revelan una deplorable técnica lexicográfica y una patética incapacidad para reproducir en beneficio del usuario los 45

Phillip M. Carter

matices semánticos y resonancias sociolingüísticas que el término exhibe en los distintos usos que de él se hacen. Por otro, la referencia a los fenómenos asociados con el contacto español-inglés en EE.UU. como «deformaciones» muestra que, a pesar de los notables esfuerzos de las academias por limpiarse los mocos del purismo y mostrar una cara limpia y abierta, es la propia institución académica (y no los hablantes de espanglish) la que, al excluir las conclusiones de los trabajos de investigación que muestran la sistematicidad de estas prácticas lingüísticas, perpetra escandalosas tergiversaciones de la naturaleza y funcionamiento del lenguaje Claro está que los abajo firmantes no se sorprenden en absoluto ni de aquella incompetencia ni de esta irresponsabilidad, actitudes de las cuales estas instituciones suelen hacer gala. Acostumbrados estamos ya a contemplar cómo esta lamentable gestión de la normatividad convive con la competentísima organización de fastos lingüísticos y de los consiguientes viajes y cuchipandas financiadas por quienes (imaginamos) se sienten beneficiarios de la acción geopolítica de las academias e indiferentes ante sus dislates lingüísticos. The authors obviously pull no punches, taking issue with the scant length of the definition, the lack of nuanced meaning, the absence of critical perspectives from those who have studied the language, and the insistence that Spanglish be understood as involving deformations of Spanish. The espanglish definition is not the only occasion during which North American linguists and the Real Academia Española have disagreed. The Academia Norteamericana de la Lengua Española (ANLE) is one of 22 members of the Asociación de las Academias de la Lengua Española. In 2014, the director of the ANLE and several of his associates published a language guide entitled Hablando bien se entiende la gente (Piña-Rosales, Covarrubias, Segura, & Fernández 2010), which the authors described as “un esfuerzo de servicio comunitario realizado en aras del buen decir,” the purpose of which was to “despejar dudas ortográficas y gramaticales.” Of course, the very notion that grammatical points need to be cleared up by language authorities is itself a prescriptivist notion; the work is from the outset thus mired in what Lippi-Green (1997) has termed “standard language ideology,” the belief in what language is or should be. But beyond this general orientation, the authors articulate a vision of language consistent with what we expect from language academies, but which is otherwise totally out of synch with the positions of contemporary Linguistics. They note, for example, that “[e]l maltrato laboral es sin duda un asunto muy serio, pero lo es también . . . el maltrato de nuestro idioma” (Piña-Rosales et al. 2010, p. 139). The implicature, of course, is that those responsible for this maltrato are U.S. Latinos, who mix their Spanish with their English, who use English-based calques, who misuse false friends. and who rely on unnecessary neologisms. While the book is not intended to be a scholarly work, the complete lack of reference to any of the literature published by linguists who have studied the phenomena the book covers is of course also noteworthy. The omission of scholarly perspectives on “Spanglish” as it pertains to calques, loanwords, and other lexical phenomena is especially of note given the immensity of the published literature on the topic (e.g. Otheguy 1993; Otheguy & Stern 2010; Poplack 1982; Zentella 1997, inter alia). Their view of language mixing is especially clear, as they write in the opening pages of the book: “¿no sería preferible hablar bien los dos idiomas, sin mezclarlos, que hablar mal los dos?” (Piña-Rosales et al. 2010, pp. 1–2). A critical response to Hablando bien se entiende la gente was published by Andrew Lynch and Kim Potowski in the journal Hispania in 2014 (Lynch & Potowski 2014). Their critique reframes the objects presented in ANLE’s guide in terms of the literature on Spanish in the United States and corrects some of the misinformation presented in the ANLE guide. 46

Spanish in U.S. language policy, politics

For example, HBSELG presents a number of terms as “extranjerismos” from English, which Lynch and Potowski show are in fact well attested in print in monolingual varieties of Spanish. For example, the ANLE claims that supuesto, as in supuesta víctima, is the result of English influence. However, Lynch and Potowski (2014) show that in Spain’s very own newspaper, El País, there were 280 searchable instances of supesta víctima. In the final section of their paper, the authors set forth their vision of a more productive way to promote the Spanish language in the United States.11 Rather than a style guide designed to “improve” the speech of U.S. Hispanics, they suggest a practical book that encourages bilinguals to use Spanish more frequently, not only with parents but also with friends in a variety of social contexts. They suggest that the ANLE’s resources may be better spent providing information for Spanish-speaking parents about how to make speaking Spanish in the home effective and promoting bilingual education in U.S. schools. Several months after the publication of Lynch and Potowski’s (2014) critique of ANLE’s book, Hispania published a response letter written by Gerardo Piña-Rosales (2014), the Director of the Academia Norteamericana de la Lengua Española. In it, Piña-Rosales concedes nothing, arguing in the first page that “sus juicios sobre Hablando bien se entiende la gente obedecen a un principio de laissez-faire irreconciliable con cualquier intento normativo de la lengua que hablamos y escribimos.” He adds that if Lynch and Potowski were to write a book themselves, it would be titled Lo que importa es que hables, aunque hables mal. After the publication of Piña-Rosales’ response, the debate spilled over to the Internet, where a repartee ensued between sociolinguists, who gave negative reviews to Hablando Bien on Amazon.com, and representatives of ANLE who defended their work. Sociolinguist Kim Potowski wrote reviews to two editions of the book Hablando Bien Se Entiende La Gente 1 and 2. In both reviews, she notes that the book is “sociolinguistically uninformed” and is intended to “shame U.S. bilinguals into conforming to someone else’s way of speaking.” She makes the point that language shaming promotes crossgenerational language shift and accuses the ANLE of mocking Spanish speakers in the United States. Reminiscent of the recommendations made at the end of Lynch and Potowski’s (2014) article, she also adds the following suggestions for the ANLE: If they really wanted to promote Spanish use, they’d do something useful. Like promote dual language K-8 schools. Contribute to greater TV programming offerings in Spanish relevant to U.S. Latino youth. Lobby the government for more favorable language policies and a federal cabinet dedicated to multilingualism. And sponsor trips abroad for heritage speaker high school and college students. A number of other critical reviews of the books were published on the Amazon site, most of which focused on the same issues raised by Potowski. Many comments defend the text, including one written by Porfirio Rodríguez, a member of the ANLE, which he titled “ANLE is a great asset to the improvement of the Spanish language in the United States.” In the review he accuses linguists of mistaking their book “with a sociolinguistic investigation” and misunderstanding the intentions of the ANLE. Addressing Potowski and the other linguists who review the book, Rodríguez writes: They want to impose their views that lack objectivity and scientific evidence. They do not present any valid theories in their arguments; they just present their opinions that seem to be more the result of emotions, than the product of good reasoning. They seem to be using personal attacks, rather than acknowledge the good work that ANLE has been doing. 47

Phillip M. Carter

These debates illustrate the ongoing, fundamental differences between linguists and Academy grammarians about the nature of language, particularly as it pertains to the Spanish language in the United States.

Conclusions: undoing the paradox In 2015, at the start of the 2016 Presidential election campaign season, Hillary Clinton announced her candidacy via a web-based video that featured vignettes of ostensibly ordinary Americans doing ostensibly ordinary things. One of the vignettes depicted two brothers preparing to open their first business. This vignette was different from the others featured in the three-minute clip in that the men spoke to each other and to the camera in Spanish, as English subtitles appeared on the bottom of the screen. Not long after, Jeb Bush announced his own candidacy during a speech given in Miami in which he spoke English and Spanish. The use of Spanish – either by a candidate or surrogates of the candidate – will now likely be a prerequisite in future presidential campaign announcements. But Jeb Bush took things a step further by granting a Spanish language interview to ABC news anchor David Muir, who like Bush, speaks Spanish as a second language. While the interview was itself considered newsworthy, it was not met at the time with the kind of controversy one might expect given the political climate toward Spanish described in this chapter, or the kind of hostility with which it would likely be met given the climate toward Spanish after the 2016 election. The optics of two powerful non-Latinx men speaking to one another in Spanish in an English-language medium seemed to suggest, on the one hand, a shift toward greater acceptance of the Spanish language in the mainstream. At the same time, it is difficult to imagine the same interview on the same network taking place among two Latinx participants. The media’s fascination with Tim Kaine makes the same point. Kaine, an Anglo from Virginia, learned Spanish in Honduras, and was praised effusively for his bilingualism by the mainstream English-language media. To my knowledge, no Latinx politician running for national office has ever been praised for his/her bilingualism. The paradox of Spanish in the United States discussed throughout this chapter is itself always racialized. Santa Ana (2002) notes that the idea that “Latinos do not speak English” is a foundational part of the “Anglo-American narrative” about Latinx people. This explains a good deal of the political rhetoric about Spanish explored in this chapter. As linguists, we must continue to heed Zentella’s (1997) call for anthropolitical linguistics, as well as Wolfram’s (1993, p. 227) principle of linguistic gratuity, “investigators who have obtained linguistic data from members of a speech community should actively pursue ways in which they can return linguistic favors to the community.” The biggest favor we can return to U.S. Latinx speech communities is educating the broader public about the nature of bilingualism, the history of Spanish in the United States, and the realities of multilingual education. We should also seek to influence language policy by providing linguistically informed perspectives to policy stakeholders (community members, educators, politicians) and by actively resisting popular narratives about Spanish that influence public opinion and lead to restrictive language policies. Political rhetoric implying that U.S. Latina/ os do not know English, that Spanish is a threat to the state, or that bilingual education is code for Spanish-only education can only find traction when widespread misinformation about basic language matters prevails. Sociolinguists, linguistic anthropologists, and other scholars working on issues related to Spanish in the United States should make public engagement a foundational, rather than ancillary part of our professional commitments. This may include speaking with the press about news stories related to Spanish, partnering with university Media Relations offices to promote socially responsible research on language in U.S. Latinx communities in the media,

48

Spanish in U.S. language policy, politics

authoring Op/Eds and opinion articles in local and national newspapers, and engaging in other forms of extension. Linguists interested in models for community engagement and social change should consult Charity (2008), and linguists interested in engaging future educators on issues of language awareness should see Mallinson and Charity Hudley (2010). Finally, although academy grammarians are unlikely to abandon the prescriptivism that currently defines their professional practice, we should not give up on our relationship with RAE and ANLE altogether, either. If they could take up even some of the suggestions made by Lynch and Potowski (2014) we, as sociolinguists, could have a potentially powerful ally. In an era in which Spanish language content has been removed from Whitehouse.gov, the primary online platform for communicating information from the executive branch of the government to the public, our perspectives and our activism are more urgent than ever.

Notes 1 Escobar and Potowski’s (2015) estimate includes 34.8 million speakers of Spanish aged 5 years and older as reported in the 2010 U.S. Census, 11 million undocumented immigrants from Latin America (Pew Hispanic Center 2013), and the estimated 2.8 million non-Latinos who use Spanish in the home (González-Barrera & López 2013). 2 Although the United States has always been a multilingual nation in practice due to the presence of indigenous and immigrant languages, “the linguistic culture” of the United States is at the same time staunchly monolingual (Schiffman 1996). 3 See Pallares and Flores-González (2010) for interdisciplinary scholarship on the immigrant rights movement in Chicago. 4    Amanece, lo veis?, a la luz de la aurora? lo que tanto aclamamos la noche al caer? Sus estrellas, sus franjas flotaban ayer, En el fiero combate en señal de victoria. Fulgor de lucha, al paso de la libertad, Por la noche decían: «¡Se va defendiendo!» ¡Oh, decid! ¿Despliega aún su hermosura estrellada Sobre tierra de libres la bandera sagrada? 5 The most recent plebiscite vote regarding the political status of the island took place in 2012. The referendum asked two questions: (1) do you agree that Puerto Rico should continue to have its present form of territorial status? and (2) regardless of your selection in the first question, please mark which of the following non-territorial options you would prefer (Sovereign Free Associated State, Independence, Statehood).To the first question, a slim majority (54%) answered ‘no.’To the second question, a majority (61%) chose ‘statehood.’The problem with the results is that in the second question, there was no option for ‘territorial status,’ which means that the 46% of voters who answered ‘yes’ to the first question were forced to choose an answer to the second question that did not match their preference. Regardless of this or any future voting results, the U.S. Congress would have to approve any change in the territorial status of the island. 6 Koike and Graham (2006) show how the issue of fluency in Spanish, and its relation to ‘authentic’ Hispanic identity, came into play during the 2002 campaigns of Dan Morales and Tony Sánchez, candidates running for the Texas Democratic Party’s nomination for governor in 2002. 7 Should the proposal have passed, Doral would not have been the first municipality to make such a designation. On March 14, 2007, Chicago Mayor Richard Daley signed a resolution adopted by the Chicago City Council to recognize and celebrate the multilingualism of the city. Readers can find the full text of the resolution here: http://multilingualchicago.org/resolution. 8 In addition to Spanish in New Mexico, French is also given special status in Louisiana, and the states of Alaska and Hawaii designate non-English co-official languages. 9 The original, Spanish definition of espanglish found in the dictionary of the RAE is as follows: “Modalidad del habla de algunos grupos hispanos de los Estados Unidos, en la que se mezclan, deformándolos, elementos léxicos y gramaticales del español y del inglés.”

49

Phillip M. Carter 10 The letter to RAE was signed by dozens of linguists working in Spain, Latin America, and the United States. It was also signed by popular Mexican screen actor Gael García Bernal. 11 Dada esta realidad, pensamos que con los objetivos que articula la ANLE se correspondería mejor un libro práctico que animara a los hispanos bilingües a hablar la lengua con más frecuencia, a hablarla en la calle y con las amistades y los padres, estudiarla en la escuela y en la universidad, viajar a países de habla hispana para experimentar el uso de la lengua en un contexto mono- lingüe. O bien una guía que ofreciera a los padres y los abuelos consejos de cómo fomentar el uso del español en casa, que recalcara la importancia de la educación bilingüe y explicara por qué es tan valiosa y necesaria para los niños. Nos parece que podría ser mucho más adecuada y efectiva una campaña lingüística alentadora, que animara a los jóvenes bilingües a hablar más la lengua en vez de censurar su uso.

References Alfaraz, Gabriela. 2002. Miami Cuban perceptions of varieties of Spanish. In Daniel Long and Dennis Preston (eds.) Handbook of Perceptual Dialectology, Volume 2, ed. by, 1–12. Amsterdam: John Benjamins. Alfaraz, Gabriela. 2014. Dialect perceptions in real time: A restudy of Miami Cuban perceptions. Journal of Linguistic Geography 2: 74–86. Anderson, Benedict. 1982. Imagined Communities. New York: Knopf Doubleday Publishing Group. Barrett, R. (2006). Language ideology and racial inequality: Competing functions of Spanish in an Anglo-owned Mexican restaurant. Language in Society 35: 163–204. Callahan, Laura. 2004. Spanish/English Codeswitching in a Written Corpus. Amsterdam: John Benjamins. Carter, Phillip M. 2014. National narratives, institutional ideologies, and local talk: The discursive production of Spanish in a ‘new’ U.S. Latino community. Language in Society 43: 209–240. Carter, Phillip M. and Andrew Lynch. 2015. Multilingual Miami: Current trends in sociolinguistic research. Language and Linguistics Compass 9: 369–385. Charity, Anne H. 2008. Linguists as agents for social change. Language and Linguistics Compass 2: 923–939. Charity Hudley, Anne H. 2017. Language and racialization. In García, Ofelia, Nelso Flores, and Massimiliano Spotti (eds.) The Oxford Handbook of Language and Society. Oxford, UK: Oxford Handbooks. Chavez, Leo. 2008. The Latino Threat: Constructing Immigrants, Citizens, and the Nation. Palo Alto, CA: Stanford University Press. Crawford, J. 1989. Bilingual Education: History, Politics, Theory, and Practice. Trenton, NJ: Crane Publishing. Crawford, James. 1999. (4th edition). Bilingual Education: History, Politics, Theory and Practice. Trenton, NJ: Crane Publishing. Del Valle, Jose (ed.). 2007. La lengua, patria común? Ideas e ideologías del español. Madrid / Frankfurt: Iberoamericana/Vervuert. Del Valle, José. 2009. Total Spanish: the politics of a Pan-Hispanic grammar. PMLA 124(3): 880–886. Del Valle, José. 2011. Política del lenguaje y geopolítica: España, la RAE y la población latina de Estados Unidos. In Senz, S. and M. Alberte (eds.) El dardo en la academia. Esencia y vigencia de las academias de la lengua española. Barcelona: Melusina. Del Valle, José. 2013. A Political History of Spanish: The Making of a Language. Cambridge, UK: Cambridge University Press. Escobar, Anna María and Kim Potowski. 2015. El español de los Estados Unidos. Cambridge, UK: Cambridge University Press. Fishman, Joshua. 2004. Multilingualism and non-English mother tongues. In Figegan, Edward and John Rickford (eds.) Language in the USA: Themes for the 21st Century. Cambridge, UK: Cambridge University Press, 115–132. Fitzgerald, Jill. 1993. Views on bilingualism in the United States: A selective historical review. Bilingual Research Journal, 17, 35–56. González-Barrera, A. and M. H. López. (2013). A Demographic Profile of Mexican-Origin Hispanics in the United States. Pew Research Center Hispanic Trends. May 1, 2013. www.pewhispanic. org/2013/05/01/a-demographic-portrait-of-mexican-origin-hispanics-in-the-united-states/. Gray, Lisa. 2014. Principal who told kids not to speak Spanish will lose job. Houston Chronicle. March 19, 2014. Hakuta, K. 1986. Mirror of Language: The Debate on Bilingualism. New York: Basic Books.

50

Spanish in U.S. language policy, politics Hill, Jane. 2008. The Everyday Language of White Racism. Malden, MA: Blackwell. Irvine, Judith T. and Susan Gal. 2000. Language ideology and linguistic differentiation. In Kroskrity, Paul (ed.) Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, NM: School of American Research, 35–84. Koike, Dale and Clayton P. Graham. 2006. Who is more Hispanic: The co-construction of identities in a U.S. Hispanic political debate. Spanish in Context 3: 181–213. Linthicum, Kate. 2009. Dallas police draw ire for citing ‘non-English-speaking drivers’. Los Angeles Times. October 27, 2009. Lippi-Green, Rosina. 1997. English with an Accent: Language, Ideology, and Discrimination in the United States. New York: Routledge. Lynch, Andrew and Kim Potowski. 2014. La valoración del habla bilingüe en los Estados Unidos: Fundamentos sociolingüísticos y pedagógicos en Hablando bien se entiende la gente. Hispania 97: 32–46. Macías, Reynaldo. 1985. Language and ideology in the United States. Social Education 49: 97–100. Mallinson, Christine and Anne Charity Hudley. 2010. Communicating about communication: Multidisciplinary approaches to educating educators about language variation. Language and Linguistics Compass 4: 245–257. Otheguy, Ricardo. 1993. A reconsideration of the notion of loan translation in the analysis of US Spanish. In Ana Roca and John Lipski (eds.) Spanish in the US: Linguistic Contact and Diversity. Berlin: Mouton de Gruyter, 21–45. Otheguy, Ricardo and Nancy Stern. 2010. On so-called Spanglish. International Journal of Bilingualism 15: 85–100. Pallares, Amalia and Nilda Flores-González (eds.) 2010. ¡Marcha! Latino Chicago and the Immigrant Rights Movement. Chicago, IL: University of Illinois Press. Pew Hispanic Center. 2013. A Nation of Immigrants: A Portrait of the 40 Million, Including 11 Million Unauthorized. Washington, DC: Pew Hispanic Center, January 29. Piña-Rosales, Gerardo, Jorge I. Covarrubias, Joaquín Segura and Daniel Fernández, eds. (2010). Hablando bien se entiende la gente. New York: Academia Norteamericana de la Lengua Española / Doral. Impreso. Piña-Rosales, Gerardo. 2014. Letter to the editor: En respuesta a un artículo publicado en Hispania. Hispania 97(3): 355–356. Poplack, Shana. 1982. Sometimes I’ll start a sentence in Spanish y Termino en Español: Toward a typology of code-switching. In Jon Amastae and Lucía Elías-Olivares (eds.) Spanish in the United States: Sociolinguistic Aspects. Cambridge, UK: Cambridge University Press, 230–263. Potowski, Kim. (Ed.) 2010. Language Diversity in the USA. Cambridge, UK: Cambridge University Press. Preston, Dennis. 2010. Language, people, salience place: Perceptual dialectology and language regards. Dialectologia 5: 87–131. Reid, T. R. 2005. Spanish at school translates to suspension. Washington Post. December 9, 2005. Rickford, John. 1999. The Ebonics controversy in my backyard: A sociolinguist’s experiences and reflections. Journal of Sociolinguistics 3: 267–276. Ryman, Anne and Ofelia Madrid. 2004. State investigating teacher accused of hitting students. The Arizona Republic. January 16, 2004. Santa Ana, Otto. 2002. Brown Tide Rising: Metaphors of Latinos in Contemporary American Public Discourse. Austin, TX: University of Texas Press. Schiffman, Harold. 1996. Linguistic Culture and Language Policy. London: Routledge. Senz, Silvia and Monserrat Alberte (eds.) 2011. El dardo en la academia. Esencia y vigencia de las academias de la lengua española. Barcelona: Melusina. Tetel Andresen, Julie and Phillip M. Carter. 2016. Languages of the World: How History, Culture, and Politics Shape Language. Malden, MA: Wiley-Blackwell. Urciuoli, Bonnie. 1996. Exposing Prejudice: Puerto Rican Experiences of Language, Race and Class. Boulder, CO: Westview. U.S. Census Bureau. 2010. 2010 Census. U.S. Census Bureau. www.census.gov/2010census/data/ (accessed January 1, 2016). Valdés, Guadalupe. 2001. Learning and Not Learning English: Latino Students in American Schools. New York: Teacher’s College Press. Vandehei, Jim. 2006. President wants anthem sung in English. The Washington Post. Washington, DC.

51

Phillip M. Carter Veltman, Calvin. 1983. Language Shift in the United States. Berlin: Mouton. Veltman, Calvin. 1988. Modeling the language shift process of Hispanic immigrants. International Migration Review 22: 545–562. Verhovek, Same Howe. 1995. Mother scolded by judge for speaking Spanish. The New York Times. August 30, 1995. Wiley, Terrance. 2000. Continuity and change in the function of language ideologies in the United States. In Thomas Ricento (ed.), Ideology, Politics, and Language Policies: Focus on English. Amsterdam: John Benjamins, 67–85. Wiley, Terrance. 2004. Language planning, language policy, and the English-only movement. In Ed Finegan and John Rickford (eds.) Language in the USA: Themes for the 21st Century. Cambridge, UK: Cambridge University Press, 319–338 Wolfram, Walt. 1993. Ethical issues in language awareness programs. Issues in Applied Linguistics 4: 225–255. Wolfram, Walt. 1998. Language ideology and dialect: Understanding the Oakland Ebonics controversy. Journal of English Linguistics 26: 108–121. Woolard, Kathryn. 2012. Introducción. Las ideologías lingüísticas como campo de investigación. In Schieffelin, B., K. Woolard and P. Kroskrity (eds.) Ideologías lingüísticas: teoría y práctica. Ediciones La Catarata / Unesco, 19–69. Zentella, Ana Celia. 1997. Growing up Bilingual: Puerto Rican Children in New York. Malden, MA: Blackwell. Zentella, Ana Celia. 2014. TWB (Talking while bilingual): Linguistic profiling of Latina/os and other linguistic torquemadas. Latino Studies 12: 620–635.

52

4 SPANISH LANGUAGE USE, MAINTENANCE, AND SHIFT IN THE UNITED STATES Devin Jenkins university of colorado denver, usa

Introduction The United States has a formidable presence in the Spanish-speaking world. Of the 50.5 million1 Hispanics that live in the United States, 35.4 million of them speak Spanish. That figure ranks it as the fifth-largest Spanish-speaking country in the world; if estimates of up to 10 million undocumented and uncounted immigrants are factored in, then that ranking jumps up to second in the world, behind only Mexico. In either case, the United States has more Spanish speakers than Venezuela, Peru, Chile, Bolivia, Ecuador, Paraguay, Uruguay, and any Caribbean or Central American country, and it counts more Puerto Ricans (4.6 million) than Puerto Rico itself (3.7 million). As another point of comparison with a neighboring country, the total population of Canada is, at 33 million, less than the Spanish-speaking population of the United States. The United States, especially the Southwest, has undergone dramatic growth in Hispanic2 and Spanish-speaking communities during the last generation. This growth has been both numerical and geographical. We have not only seen more Spanish speakers and increased densities of these speakers in those areas with established Spanish-speaking communities, but as this chapter will detail, we have also seen a geographical expansion from these regions into areas that do not have the same historical connection to the Spanish language. As we consider this growth and these communities in the context of Spanish for Heritage Speakers, it is worth considering where the language is being maintained and where higher incidences of shift occur. As new language communities are introduced, it is worthwhile to examine how language is, or is not, being maintained and used in longer-standing language communities and determine whether parallels can be drawn.

Where is Spanish spoken in the United States? The undisputable majority of Latinos in the United States live in the Southwestern region of the country. Of the 50.5 million total, 24.5 million—almost half—live in the two states of California and Texas. If we include in the total count the Southwestern states of Arizona, Colorado, Nevada, and New Mexico, that figure increases to nearly 28 million, or 56% of all Latinos in the United States, even though these six states only account for 25% of the total 53

Devin Jenkins

population of the country. The states that make up the Eastern Time Zone, by comparison, represent 46% of the total U.S. population, but less than 30% of Hispanics live in this area. Mexicans comprise, by a large margin, the largest group of Latinos in the United States, representing 31.8 million (63%) of that population. In the six aforementioned Southwestern states, Mexicans represent an even larger portion, accounting for 83% of Hispanics in that region. These numbers have, in part, justified a disproportionate number of studies on Spanish in the Southwest and its relative historical synonymy with Mexicans in the United States (cf. Amastae and Elías-Olivares 1982; Beltramo et al. 1975; Silva-Corvalán 1994; Villa and Rivera-Mills 2010). Puerto Ricans make up the next largest Hispanic group, with a total U.S. mainland population of 4.6 million. The majority live in the Northeastern United States: 1.5 million Puerto Ricans live in New York or New Jersey, with another 366,000 in Pennsylvania (concentrated around Philadelphia), and 266,000 in Massachusetts. Outside of the Northeast, the largest Puerto Rican state is Florida, with 847,550 (with nearly half of that number residing in Orlando, and more recent estimates placing that figure north of 1 million), with half of the state’s Puerto Rican population arriving, and continuing to arrive in significant numbers, since 2000. New York, on the other hand, while still the largest, has remained virtually flat in its total Puerto Rican population since 1980. Illinois (with the vast majority in Chicago) counted 183,000 Puerto Ricans. The top three states (New York, Florida, New Jersey) account for over 50% of the Puerto Rican population. Beyond Mexicans and Puerto Ricans, other Latino groups also stand out. The Cuban population (1.7 million) is the most concentrated of Latino groups. Fully 68% (1.2 million) of Cubans in the United States live in Florida. After that, the largest groups correlate with the most populated states, namely California (88,607), New Jersey (83,362), New York (70,803), and Texas (46,541). The Dominican population, the fastest growing in the United States, numbers 1.4 million and is concentrated heavily in the Northeast. The top five states are the same as the top five states for Puerto Ricans, with New York and New Jersey accounting for 62% of all Dominicans in the United States. Central American groups (collectively totaling 4 million) can be found in large population centers throughout the United States. While California has the largest Central American populations in the country, significant Central American populations can be found in Florida, New York, and Texas as well. Salvadorans (1.65 million in in the United States) make up the largest Hispanic group in Maryland and Washington, DC. Spanish-speaking populations in the Northeast and in Florida are heavily concentrated in urban areas and are fairly well-defined in geographic terms. The Spanish-speaking Southwest, on the other hand, is much larger and less concisely defined. To define the Spanish-speaking Southwest, a minimal definition would be to categorize it as those four states that share a border with Mexico, namely Arizona, California, New Mexico, and Texas. These four states have the highest Hispanic densities and the longest running relationships with the Spanish language, including both recent immigrants and families that have lived in the area for centuries. The push-pull migration in the area promotes constant contact with Mexico. Of the population of Texas, 29.2% claims Spanish as a home language. In California and New Mexico, the percentage is 28.5%; in Arizona the figure surpasses 20% (Figure 4.1). In discussions on the Spanish-speaking Southwest, however, it is important to include the state of Colorado, which shares a linguistic history with the state of New Mexico. The southern third of the state was founded by New Mexicans in the middle of the 19th century, and the regional toponyms reflect this history. The Latino density in Colorado also justifies its inclusion in the Southwest, as more than one in five inhabitants of the state are Latinos, and 11.9% of

54

Spanish use, maintenance, shift in US

Figure 4.1  Hispanic or Latino population as a percent of total population by county, 1010 Source: United States Census Bureau [Public domain], via Wikimedia Commons

the population speaks Spanish. The inclusion of Colorado in studies of Southwestern Spanish reflects its significance as a member of this linguistic community (Bills et al. 1995; Hudson et al. 1995, among several). A state that has recently merited attention as part of the Spanish-speaking West is Nevada. In 1980, the Hispanic population (dominant in any language) represented only 6.7% of the total population of the state. In 1990, that number had increased to 10.4% and by 2000 had soared to 19.7% of the state. The Hispanic population continues to grow and, as of 2010, 26.5% of the population is Latino, with 19.6% claiming Spanish as a home language (Figure 4.2). Jenkins (2009a, 2009b, 2010) was among the first to include Nevada in studies on Spanish in the region. Beyond these six states, other states warrant discussion as part of this linguistic region as well. The remaining states in the extreme West, namely Idaho, Oregon, Utah, and Washington, deserve consideration for a couple of reasons. First is the growth of the Hispanic populations in the region since the late 1990s, reaching a notable density. In the 2010 Census, the Hispanic population of Oregon totaled 11.7%, while the other three states each measured 11.2%. Hispanophone concentrations for each of these states are nearing 10% and, if the current trajectories hold, will each pass that benchmark before the next Census. In addition to the growth, there is no fundamental demographic difference between Latinos in the Northwest and those in more Southern states other than the higher densities and proximity to the border; the origins of Hispanics in the Northwest, especially those that account for recent growth, are essentially the

55

Devin Jenkins

Figure 4.2  Largest detailed Hispanic origin group by state, 2010 Source: United States Census Bureau [Public domain], via Wikimedia Commons

same as those in the Southwest.3 Mexicans make up the majority of the Latino population in every state west of the Mississippi, other than Hawaii. Even to the east of the Mississippi, many states share this same characteristic. It bears mentioning that Illinois, with the majority of the Latino population concentrated in the Chicago Metropolitan Area, counts more Mexicans than any other state outside of California, Texas, and Arizona. As the Mexican population grows in virtually every corner of the country, sociolinguistic phenomena in these areas are ripe for exploration, well beyond the Southwest, or even the West. Hispanics of Mexican descent form the majority group in 40 of the 50 United States (Table 4.1). New Mexico, an indisputably Southwestern state, claims a Mexican population of 590,890.4 By comparison, Florida counts more Mexicans with 629,718. States like Georgia (519,502) and North Carolina (485,960) also merit attention. In fact, of the ten states with the largest Mexican populations, three are east of the Mississippi, and a fourth, Washington, is decidedly not Southwestern. Growth has not been exclusive to any one group or any one region of the country. In fact, the Latino population grew 43% between 2000 and 2010, from 35.3 to 50.5 million. The Mexican population rose by 54.1%, from 20.6 to 31.8 million residents, representing 78% of the Latino population growth. Mexicans have grown at rates much faster than other Latino groups in areas traditionally dominated by other Latino groups like New York, New Jersey, and Florida. In New York, for example, Mexicans grew at a rate of 75% between 2000 and 2010, as compared to 2% by Puerto Ricans. In New Jersey, Mexican growth outstripped 56

Spanish use, maintenance, shift in US Table 4.1  Largest Mexican populations, by state, 2010 State

Mexican population

California Texas Arizona Illinois Colorado Florida Washington New Mexico Nevada Georgia

11,423,146 7,951,193 1,657,668 1,602,403 757,181 629,718 601,768 590,890 540,978 519,502

Puerto Rican growth by a rate of 112% to 18%. In Florida, the Cuban growth rate in that time period was 46%, while Mexicans grew at a rate of 73%, and Puerto Ricans were similar with 76% growth in the state. Overall, Puerto Ricans increased in the United States by 36% and Cubans by 44%. Dominicans grew at a rate of 84.9%, and now form the majority Latino group in Rhode Island (Figure 4.3).

Figure 4.3  Percent change in Hispanic or Latino population by county, 2000–2010 Source: United States Census Bureau [Public domain], via Wikimedia Commons

57

Devin Jenkins Table 4.2  States with Hispanic population change >100%, 2000–2010 State

2000

2010

% change

South Carolina Alabama Tennessee Kentucky Arkansas North Carolina Maryland Mississippi South Dakota

95,076 75,830 123,838 59,939 86,866 378,963 227,916 39,569 10,903

235,682 185,602 290,059 132,836 186,050 800,120 470,632 81,481 22,119

147.9 144.8 134.2 121.6 114.2 111.1 106.5 105.9 102.9

Even more notable is the geographic distribution of the growth. Most of the growth has not occurred in traditional Hispanic regions, such as New Mexico and West Texas, but rather in many areas that do not share a long Hispanic history. In fact, Figure 4.1, which shows Latino density, is virtually a negative image of Figure 4.3, which details distribution of growth. The states that have experienced the most dramatic change are found outside the Southwest, and in many cases belong to the Southeastern region of the country. Nine states more than doubled their Latino population in the decennial census period between 2000 and 2010 (see Table 4.2), and all of them, with the exception of Maryland, claim a Mexican majority among Hispanics. Even with the marked growth outside of the Southwest, the region as a whole still shows significant increases, especially when looking at raw count. The two states with the largest numerical increases are California (3,047,163) and Texas (2,791,255), followed by Florida (1,541,091), Arizona (599,532), and New York (549,339). In this case, we see a strong correlation between total growth and the total population of the state, which is not surprising; numerous studies have noted the strong relationship between numbers of Latinos and total population (cf. Bills et al. 1995; Hudson et al. 1995; Jenkins 2009a, etc.). One statistical occurrence worth noting is when a state’s Hispanic density reaches 10%. Demographic researchers often highlight this figure as it serves as a clear benchmark in noting population and linguistic growth in an area (Ennis et al. 2011; Hobbs and Stoops 2002; RiveraMills 2010). In 1980, only Colorado and the four Mexican border states counted Hispanic densities above 10%. By 2000, five more states achieved this designation—Florida, Illinois, Nevada, New Jersey, and New York (Hobbs and Stoops 2002). In the 2010 Census, 17 states counted at least a 10% Latino population. In addition to the ten previously mentioned, Idaho, Kansas, Oregon, Utah, and Washington were added in the West, and Connecticut and Rhode Island rounded out the Eastern contingent. It is clear that the Hispanic population is following a strong growth trajectory in the United States. It is the largest minority group in the country, and evidence points to its continual growth. Mexicans form the largest subset, and future studies of the Spanish language within this group must take into account total growth and geographical expansion when striving to contextualize Spanish in the United States.

Language maintenance and shift Various social factors contribute to whether or not a language is maintained in a community. What follows is a brief exploration of those factors, followed by a discussion on measuring language maintenance and shift, and finally a discussion of Spanish language use in different areas 58

Spanish use, maintenance, shift in US

of the United States, first at the city and county level, then considering large, multi-state zones where the Spanish language is undergoing different levels of growth, maintenance, or loss. As immigrant populations move to new linguistic communities, the invariable result is a shift away from the immigrant language to the dominant language in the community. The most common manifestation of language shift occurs over three generations, in which the first generation is dominant in its native language while learning the second language (English in the United States) in non-native fashion; the second generation approximates balanced bilingualism as it maintains the language of the parents in home-related domains, while acquiring English, at the very latest, upon entering school. This second generation in the United States typically receives a formal education in English and will dominate formal registers within that language. The norm for the third generation is most often complete shift to English, with very little (or no) competency in the first language of their grandparents. This pattern is not unique to Spanish, English, or to the United States, but rather is a universal trend among language immigrant populations throughout the world and throughout history (Potowski 2010; Veltman 1983). Many scholars (Grosjean 1982; Silva-Corvalán 2004; Veltman 1983; Zentella 1997, among many others) cite the absence or presence of recent and continued immigration as the most significant factor in Spanish language maintenance in the United States. Even in the face of such immigration within a community, a wide array of social factors favors assimilation and shift toward English. The structure of the community also contributes to the status of the Spanish language within that community. Concomitant to the notion of continued immigration is the size of the speech community, as communities with smaller counts tend to disfavor maintenance. The composition and geographical distribution of the speech community as a proportion of a larger community are equally as significant. Finally, the function and use of the language within the community also impacts maintenance or shift. The domain of the language (where it is used), its register and topic (how it is used), and the prescribed interlocutor (with whom it is used) go a long way toward predicting retention or loss. The social motivations toward language shift are numerous. Bills (2005) points to strong and frequent assimilatory pressures in the United States, in all areas of life: “Se siente fuertemente la preponderancia del inglés en el sistema educativo, en los avances económicos, y en todas las esferas de ‘ser americano’” (p. 66). Socioeconomic factors like education, employment, and income opportunities come to the fore in discussions on linguistic assimilation (Hudson et al. 1995). Porcel (2011) points to demographic status (e.g. group size, fertility, immigration), socioeconomic status (e.g. social stratification, economic success, instrumental value of Spanish), cultural status (e.g. linguistic enclaves, Hispanic media), and legal status (e.g. language policy and linguistic rights) as those social factors that determine maintenance or shift of a language. Grosjean (1982) lists no fewer than 18 social aspects, ranging from group demographics (e.g. geographical position and concentration) to family dynamics (e.g. social mobility, intermarriage, isolation) to individual social situations (e.g. employment, religion, activism) that contribute to the phenomenon. Beyond social factors, language attitudes have a sizeable impact on language maintenance and shift. These attitudes are manifested in different ways among different groups. In-group and out-group attitudes toward the minority language, toward the majority language, toward cultural pluralism, toward bilingualism, toward linguistic “purity,” and simply toward the “other” group all come into play (Grosjean 1982, p. 107). Self-identification and cultural identity also arise as a reflection of language attitudes (see Showstack’s chapter, this volume). Not unrelated to language attitudes are xenophobia and racism, both external and internal to Spanish speakers, as attitudes toward (other) Hispanic groups are manifested in sociolinguistic evaluations and social reactions. These reactions often contribute to language policy, language 59

Devin Jenkins

movements, and other practices that create a further impact on language use in the United States (cf. Crawford 2000; Valdés 1995). All of the above factors contribute to a shift away from Spanish in favor of English. As communities undergo these changes, measurements can be taken to observe the phenomena of maintenance and shift at the community level. Hudson et al. (1995) proposed four means of measurement for large populations. Perhaps the simplest measurement is count, which is simply a raw count of total Spanish speakers in an area, as determined by Census figures. Count is most accurately predicted by large populations. On the one hand, the count measure helps to identify large groups of Spanish speakers in the United States, but on the other, it is nearly just as predictive of large cities in general. The counterpart to count is density, which measures the Spanish-speaking percentage of a county’s total population. As one indicates a raw number and the other indicates a concentration of speakers, count and density provide two sides of an equation. Therefore, while small border communities have Spanish-speaking densities upwards of 90%, they do not represent large numbers of Spanish speakers overall. We can gain a sense of just how much the language is used and maintained in high-density areas, but density as a measure does not distinguish between Clark County, Idaho, with a density of 43% (and 780 total residents) and Los Angeles County, with a density of just 39%—and a total population of nearly 10 million. The two variables taken in tandem provide a more complete view of Spanish-speaking communities than either taken separately. A metric that, by its very definition, helps to identify language maintenance in an area is language loyalty (cf. Fishman et al. 1966). Loyalty is defined as a ratio of the number of individuals who claim Spanish as a home language (i.e. count), divided by the number of individuals claiming Hispanic or Latino ethnicity. Loyalty shows those areas in which the language is being maintained among Hispanics. It is important to note that this term does not imply any conscious decision to maintain (or not) the language among individuals or groups, neither does it suggest any converse concept of ‘disloyalty,’ but rather reflects outcomes of the many societal factors that contribute to language loss or maintenance (cf. Porcel 2011). The final method of measurement of language use and maintenance is what Hudson et al. (1995) referred to as retention. This metric purports to measure the intergenerational transmission from one generation to the next, and is measured through Census data by dividing language loyalty among 5- to 17-year-olds by loyalty among those 18 and older. While age does not always equate to generation, at the county-wide level, the age difference is enough to be significant. Communities with high counts of Spanish speakers fall into two main categories. First, as was previously mentioned, count correlates strongly with total population. The largest Spanishspeaking cities are, quite simply, the largest cities. The top five U.S. cities in Spanish-language count are, in order, New York, Los Angeles, Houston, San Antonio, and Chicago—which are the first, second, fourth, seventh, and third largest cities in the country, respectively. Of the top ten cities for Spanish language count, nine are in the top ten in total population— El Paso, ranked #19 in total population, is the lone exception, with the ninth-largest count. Communities with high counts that are not large cities are those medium-sized cities and counties that sit close to the Mexican border. The border proximity explains the higher ranking of El Paso. In border communities, high density creates a higher count, thus a relatively small city like McAllen, Texas shows a considerably higher count (586,409) than Brooklyn, New York (396,897), even though Brooklyn is 3.25 times more populated. As Spanish-speaking density has a very strong correlation with Hispanic density in a county, Figure 4.1 gives a good visual representation of areas of density in the United States. Areas with

60

Spanish use, maintenance, shift in US

the highest densities, are unsurprisingly, the Southwest, South Florida, and the Northeast, with significant densities also around Chicago. Perhaps most notable are the higher densities in the West, not confined exclusively to the Southwest. Also notable is the shading in and around North Carolina, where growth is among the highest in the country. Density holds a strong relationship with distance to the Mexican border. Of the 19 counties with the highest densities in the United States, all of them are along the border, with 17 of them in Texas and 1 each in Arizona and California. Starr County, Texas has the highest density at 95.3%. Miami-Dade, Florida, is 20th, with a density of 63.9%, and is the only county east of the Mississippi in the top 60. Miami’s bilingual community and stature in the Spanish-speaking world have been well documented (Carter and Lynch 2015; Lynch 2000; Otheguy et al. 2000). Language loyalty is a key factor when it comes to heritage speakers. Identifying where Spanish is being maintained, and where it is not, helps to show where heritage speakers are. Bills et al. (1995) showed, based on 1980 Census data, that “proximity to the border favors retention of Spanish . . . while greater distance favors shift to English” (1995, p. 25). Much of that maintenance stemmed from the fact that higher densities lend themselves to stronger support of the language within the community. In addition to opportunity for use of the language [nearer the border], there is likely to be a more favorable attitude toward Spanish deriving from community and institutional support . . . Strength in numbers may well promote a more positive evaluation of both the ethnic group and the language. p. 26 However, based on data from the 2010 Census, Jenkins (2012) found that those correlations were disappearing. A weak correlation (.29, down from .61 in 1980) still exists between distance of a city from the border and language loyalty, but at the county level (as opposed to just the city), a correlation is non-existent. This stronger language loyalty is likely a product of increased immigration and in-migration over the past 30 years, thus producing a higher percentage of first-generation speakers in areas away from the border. At an even more macroscopic level than counties and communities, dividing the Spanishspeaking areas of the country into different zones yields fundamental demographic and sociohistorical distinctions between Spanish-speaking populations. Four distinct regions of the country each have unique Spanish-language situations. Our first zone is the five states of the Southwest (Arizona, California, Colorado, New Mexico, and Texas). These states have a long-standing history with regard to the Spanish language. Many Hispanic families in these states trace their history in the region back many generations, in some cases for centuries. Many of the studies of Spanish in the Southwest have centered on language use in these states and, while the language is still being maintained in many areas, Figure 4.3 also shows that growth is not happening in some of the more traditional regions like New Mexico and West Texas. Continued immigration is a key factor in language maintenance in a community and, in the absence of a consistent in-flow of Spanish speakers, language shift occurs. A second Spanish-speaking zone is what we will call the Northwest, consisting of Idaho, Nevada, Oregon, Utah, and Washington. This zone is noteworthy for multiple reasons. First, it continues to grow. Nevada, for example, has seen a growth in count of over 1000% since the late 1990s. Additionally, each of these states claims a Hispanic density of over 10%. Finally, the Northwest provides a significant contrast from the Southwest in the sense that it is a much newer Spanish-speaking area (which explains Nevada’s inclusion in this group, in spite of its

61

Devin Jenkins

geographical position). Rather than large numbers of multiple generations of Spanish speakers, figures suggest that two or three generations would be the norm in the Northwest, and that many first-generation speakers in the families are still there. An even newer zone is what we will call the Southeast, comprising Georgia, North Carolina, South Carolina, and Tennessee. This cluster of states is where massive growth has happened, mostly from Mexico, since the late 2000s. A third-generation Spanish speaker in this area would be the exception rather than the norm. Florida is excluded from this cluster not because it is not Spanish-speaking, neither because it doesn’t have a significant number of recent Mexican immigrants, but rather because the significant Cuban population, which now counts more than 50 years since the beginning of the Cuban revolution, is in itself a contrast from these other communities. A separate study based solely on language maintenance and use among the Latino populations in Florida is warranted. A final group is the Northeast, consisting of Connecticut, New Jersey, New York, and Rhode Island. These four states were selected because each of them has a Latino population greater than 10%, and because they provide a significant contrast from the other groups in the sense that these are the only states in this study that do not have a Mexican majority among Latinos. This cluster also differs from the others due to the lack of a rural Spanish speaker, a commonality in the other three groups. Loyalty figures for the four regional clusters show that language loyalty in the Southwest and the Northwest are virtually identical, with the Southwest reporting 77.5% loyalty and the Northwest reporting 78%. The Northeast, by comparison, has an overall loyalty rate of 86%, and the Southeast reports language loyalty at a staggering 93%. The high rate of loyalty in the Southeast lends itself well to the notion that the population is so recent that most of the Spanish speakers are still first- and second-generation speakers, thus maintaining the language as a home language. At the other end of the continuum, the longitudinal presence of Hispanics in the Southwest has led to lower claiming, as third- and fourth-generation (and beyond) speakers are much more common. When looking at individual counties, the highest maintenance figures in the four clusters come from the Southeast. Of the top 20 counties in loyalty, 19 are from the Southeast cluster. Given that this is the region of the greatest new growth, the figures stand to reason. With regard to rapid growth in Colorado based on 2000 Census data, McCullough and Jenkins (2005) found that “migration of Spanish-speakers into Colorado from abroad is serving to reinforce the measures of linguistic loyalty and retention, against the tendencies to the contrary in the absence of such support.” They found that immigration was the single, most important variable with regard to Spanish language maintenance in Colorado. This appears to be the case in the Southeast, as well. Intergenerational retention varies by regional cluster as well. The Southwest reported a retention rate of 82.3%; the Northwest was once again virtually identical at 83%; and the Southeast once again had a higher rate at 86.1%. This time, the Northeast has the lowest intergenerational retention rate at 77.4%. The higher figure for the Southeast can be explained for much the same reason as the higher loyalty rate—Spanish speakers are so new to the area that even the younger children belong to the so-called “generation 1.5,” meaning that they were born in Mexico and moved here between the ages of 6 and 17—still belonging to the younger age group as defined in this study and continuing to speak Spanish as a home language. As for the lower retention rate in the Northeast, several possibilities abound. One is the low rate of growth for the existing Puerto Rican population. While the smaller Mexican population has shown a growth rate of 73% over ten years, the majority Puerto Rican population has not. In New York, for example, the Puerto Rican population grew by only 2% in the latest decennial census period. The absence of immigration within the community, then, lacks the necessary 62

Spanish use, maintenance, shift in US

reinforcement of the language from abroad. Another possibility is the absence of the rural (and the abundance of the urban) Hispanic (cf. Fishman 1972). Porcel (2011) states that: [c]ities are the type of setting exerting strong pressures toward cultural assimilation due to demands from the labor market, as well as insertion into more open and diffused social networks in which people have to engage in their daily lives. In contrast, rural dwellers usually enjoy greater isolation and self-sufficiency. p. 631 In the West and in the South, Spanish-dominant, Mexican-origin agricultural workers are found in abundance in rural areas. In the Northeast, where population densities are the highest in the country, these types of areas are much less prevalent. The dual threat of urban pressure and lack of rural occupational opportunity may be at play. The rural-urban dichotomy certainly merits further investigation.

Implications for Spanish for heritage speakers Given the widespread use of the Spanish language through the country, the reality is that communities of heritage speakers can be found in almost every corner of the nation. These communities are as diverse as they are ubiquitous. Major cities with high densities of Spanish speakers such as Miami, Los Angeles, New York, and Chicago have as many systemic, political, and educational challenges with regard to heritage speaker and/or bilingual education as smaller communities with equally high densities. Communities with lower densities (or low counts) will have to determine at what point a critical mass has been (or will be) reached, and plan accordingly. The longitudinal differences in language communities, as illustrated earlier, present a living laboratory for educational practices among heritage speakers. Of course, one size does not fit all. Identity (Potowski 2012; Rivera-Mills 2012) plays a major role in how heritage language programs are constructed. What may work in rural New Mexico may well not work in rural Idaho, just as what works in Los Angeles might be quite different from what would be effective in Chicago, yet as communities become more established, patterns for success can also vary. What would be ineffectual in Portland or Raleigh right now might be warranted a generation (or less) from now. Beaudrie (2012) identified U.S. universities that offer Spanish for Heritage Speaker (SHS) programs, and found the national rate to be at 40%. In the five Southwestern states, where SHS programs are among the oldest in the country, that figure was only 39% (64/163). The Northwest counted only 8 university programs out of 24, or 33%. Only one program (in Georgia) was identified in the Southeastern region, while Northeastern programs came in at a rate of just over half (53%). Beaudrie points out, however, that most of the Northeastern programs consist of a single course offering, suggesting that, with newer SHS programs, they simply have not had the opportunity to expand their offerings (2012, p. 210). Given the low rate in the Southwest, where densities are highest and Hispanic communities have the longest histories, there is no question that the field is ripe for growth of SHS programs. High density and/or high count areas with lower retention rates would suggest the most fertile ground for SHS programs, as they contain a critical mass of students who require educational support for language maintenance or recontact (see Lynch 2000). Communities with higher densities would, on the face of it, engender more support for such programs, given the language investment by a higher portion of the community. Conversely, higher retention rates should be fostered through bilingual educational programs, thus leading to a greater possibility of transmission to the next generation. 63

Devin Jenkins

Notes 1 Unless otherwise noted, general demographic data come from the 2010 U.S. Census, Summary File 1 (U.S. Census Bureau 2011). As the Census has simplified its format to the point that now only ten questions are asked, any figures and correlations unavailable in the 2010 Census data come from five-year estimates of the 2008–2012 American Community Survey (U.S. Census Bureau 2013). 2 The U.S. Census uses the denomination “Hispanic or Latino” for ethnic claiming. In the present study, the terms “Hispanic” and “Latino” will be used interchangeably. 3 Migration trends show that Mexicans from Northern Mexico tend to settle in those U.S. states closest to their Mexican state of origin (i.e. those from western Mexico tend to settle in California and Arizona, those from the states that border Texas settle most often in Texas), with southern Mexican emigrants showing a greater diversity of settlement due to lack of proximity to any one U.S. state. The top three destinations for immigrants from Tamaulipas, a border state with Texas, are Houston, McAllen, and Dallas, while the top destinations for veracruzanos are Raleigh, Chicago, and Atlanta (Instituto de los Mexicanos en el Exterior 2015). 4 Many native New Mexicans tend to eschew the term “Mexican,” but rather claim “Spaniard” (65,045 New Mexicans claimed accordingly in 2010), “Spanish” (57,021) or “Spanish American” (10,501) on the Census, even with little immediate connection to Spain. See Bills and Vigil (2008) for further discussion of this phenomenon.

References Amastae, Jon and Lucía Elías-Olivares (1982). Spanish in the United States: Sociolinguistic Aspects. New York: Cambridge University Press. Beaudrie, Sara M. (2012). Research on university-based Spanish heritage language programs in the United States. Spanish as a Heritage Language in the United States: The State of the Field, eds. Sara M. Beaudrie and Marta Fairclough. Washington, DC: Georgetown University Press, 203–221. Beltramo, Anthony F., Eduardo Hernández-Chávez and Andrew D. Cohen (1975). El lenguaje de los Chicanos: Regional and social characteristics of language used by Mexican Americans. Arlington, VA: Center for Applied Linguistics. Bills, Garland (2005). Las comunidades lingüísticas y el mantenimiento del español en Estados Unidos. Contactos y contextos lingüísticos: El español en los Estados Unidos y en contacto con otras lenguas, eds. Luis A. Ortiz López and Manel Lacorte. Madrid: Iberoamericana, 55–83. Bills, Garland D., Eduardo Hernández-Chávez and Alan Hudson (1995). The geography of language shift: Distance from the Mexican border and Spanish language claiming in the Southwestern United States. International Journal of the Sociology of Language 114: 9–27. Bills, Garland D. and Neddy Vigil (2008). The Spanish Language of New Mexico and Southern Colorado. Albuquerque, NM: University of New Mexico Press. Carter, Phillip M. and Andrew Lynch (2015). Multilingual Miami: Current trends in sociolinguistic research. Language and Linguistics Compass 9(9): 369–385. Crawford, James (2000). At War with Diversity: U.S. Language Policy in an Age of Anxiety. Bristol, UK: Multilingual Matters. Ennis, Sharon R., Merarys Ríos-Vargas and Nora G. Albert (2011). The Hispanic population 2010. 2010 Census Briefs. Washington, DC: U.S. Census Bureau. Fishman, Joshua A. (1972). The Sociology of Language. Rowley, MA: Newbury House. Fishman, Joshua A., Vladimir C. Nahirny, John E. Hofman and Robert G. Hayden (1966). Language Loyalty in the United States. The Hague, the Netherlands: Mouton. Grosjean, François (1982). Life with Two Languages: An Introduction to Bilingualism. Cambridge, MA: Harvard University Press. Hobbs, Frank and Nicole Stoops (2002). Demographic trends in the 20th century. Census 2000 Special Reports. Washington, DC: U.S. Census Bureau. Hudson, Alan, Eduardo Hernández-Chávez and Garland D. Bills (1995). The many faces of language maintenance: Spanish language claiming in five Southwestern states. Spanish in Four Continents: Studies in Language Contact and Bilingualism. Ed. Carmen Silva-Corvalán. Washington, DC: Georgetown University Press, 165–183. Instituto de los Mexicanos en el Exterior (2015). Retrieved from www.ime.gob.mx/mapas/circuitos/ circuitos_2010.swf.

64

Spanish use, maintenance, shift in US Jenkins, Devin L. (2009a). The cost of linguistic loyalty: Socioeconomic factors in the face of shifting demographic trends among Spanish speakers in the Southwest. Spanish in Context 6(1): 7–25. Jenkins, Devin L. (2009b). As the Southwest moves north: Population expansion and sociolinguistic implications in the Spanish-speaking Southwest. Southwest Journal of Linguistics 28(1): 53–69. Jenkins, Devin L. (2010). The state(s) of Spanish in the Southwest: A comparative study of language maintenance and socioeconomic variables. Spanish of the Southwest: A Language in Transition, eds. Daniel Villa and Susana Mills. Madrid: Iberoamericana, 133–156. Jenkins, Devin L. (2012) Language shift and shifting borders: Spanish language maintenance and distance from the Mexican border in the Southwest. Paper presented at The 41st Annual Meeting of the Linguistic Association of the Southwest. October 11–13, 2012. Ft. Wayne, IN. Lynch, Andrew (2000). Spanish-speaking Miami in sociolinguistic perspective: Bilingualism, recontact, and language maintenance among the Cuban-origin population. Research on Spanish in the United States, ed. Ana Roca. Somerville, MA: Cascadilla Press, 271–283. McCullough, Robert E. and Devin L. Jenkins (2005). Out with the old, in with the new? Recent trends in Spanish language use in Colorado. Southwest Journal of Linguistics 24: 91–110. Otheguy, Ricardo, Ofelia Garcia and Ana Roca (2000). Speaking in Cuban: The language of Cuban Americans. New Immigrants in the United States: Readings for Second Language Educators, eds. Sandra Lee McKay and Sau-Ling Cynthia Wong. Cambridge, UK: Cambridge University Press, 165–188. Porcel, Jorge (2011). Language maintenance and shift among US Latinos. Handbook of Hispanic Sociolinguistics, ed. Manuel Díaz-Campos. Malden, MA: Blackwell, 623–645. Potowski, Kim (2010). Language Diversity in the USA. Cambridge, UK: Cambridge University Press. Potowski, Kim (2012). Identity and heritage learners. Spanish as a Heritage Language in the United States: The State of the Field, eds. Sara M. Beaudrie and Marta Fairclough. Washington, DC: Georgetown University Press, 179–199. Rivera-Mills, Susana V. (2010). Latinos or Hispanics? Changing demographics, implications, and continued diversity. Southwest Journal of Linguistics 28(2): 1–20. Rivera-Mills, Susana V. (2012). Spanish heritage language maintenance: Its legacy and Its future. Spanish as a Heritage Language in the United States: The State of the Field, eds. Sara M. Beaudrie and Marta Fairclough. Washington, DC: Georgetown University Press, 21–42. Silva-Corvalán, Carmen (1994). Language Contact and Change: Spanish in Los Angeles. Oxford, UK: Oxford University Press. Silva-Corvalán, Carmen (2004). Spanish in the Southwest. Language in the USA: Themes for the Twenty-First Century. Eds. Edward Finegan and John R. Rickford. Cambridge, UK: Cambridge University Press, 205–229. U.S. Census Bureau (2011). Census 2010 Summary File 1. Retrieved from http://factfinder2.census.gov. U.S. Census Bureau (2013). 2008–2012 American Community Survey 5-Year Estimates. Retrieved from http://factfinder2.census.gov. Valdés, Guadalupe (1995). Bilingües y bilingüismo en los Estados Unidos: La política lingüística en una época antiinmigrante. Alteridades 5(10): 25–42. Veltman, Calvin (1983). Language Shift in the United States. Berlin: Mouton. Villa, Daniel and Susana V. Rivera-Mills (2010). Spanish of the U.S. Southwest: A Language in Transition. Madrid: Iberoamericana. Zentella, Ana Celia (1997). Growing up Bilingual: Puerto Rican Children in New York. Malden, MA: Blackwell.

65

5 SPANISH IN LINGUISTIC LANDSCAPES OF THE U.S. José M. Franco Rodríguez fayetteville state university, usa

Introduction As a baseline definition, Linguistic Landscape (henceforth LL) is any display of texts. LLs have been around for millennia. Hammurabi’s code of law was displayed on a stele for public view in Babylon about 3,770 years ago and graffiti was carved on exterior walls in Pompeii some 2,100 years ago. These texts were part of ancient LLs and have been scrutinized by historians for a long time to understand and depict the societies that created them. Other early publicly displayed texts such as the Rosetta Stone, a stele with a bilingual text in three different scripts displayed more than 2,200 years ago in Egypt, and the trilingual Behistun Inscription, a text carved some 2,500 years ago in current Iran, have been of special interest to philologists and linguists. The former inscription was the key to decrypting Egyptian hieroglyphs and the latter to deciphering cuneiform script. In both cases, it was the presence of more than one language that made them especially noteworthy to scholars. The coexistence of different languages has long been commonplace in LLs and is one of the leading reasons behind the establishment of LL as a field of study. The first studies strictly on LL identified the presence of different languages on outdoor signage in the 1970s. However, the term LL was not coined until the publication of a seminal work by Landry and Bourhis (1997, p. 23): “Linguistic Landscape refers to the visibility and salience of languages on public and commercial signs in a given territory or region.” The authors also pioneered a different method of LL analysis. Instead of examining the LL itself, they analyzed students’ responses to questionnaires and tests where LL was the key variable in order to gauge the correlation between LL and ethnolinguistic vitality. Shortly after Landry and Bourhis’ article, LL studies surged with an intense activity that is sustained to this day. Research has been conducted in more than 60 countries and has produced some 300 publications. The most recent collective publications on LL include five monographs and a sizable number of articles. A series of international workshops have been held yearly since 2008 (save 2011) and Linguistic Landscape, An International Journal was founded in 2015. Having language and society as common ground, this intense scholarly activity has yielded two major approaches to the object of study (Franco Rodríguez, 2011, pp. 71–72). One approach is qualitative, using mainly interviews and surveys with only anecdotal fieldwork data from the actual LL, if any. This group of studies concentrates largely on social attitudes towards the incidence 66

Spanish in linguistic landscapes of U.S.

of a language in a particular LL and the impact of language policies in the LL. Although there is no clear-cut distinction, some studies measure the effect of LL on people’s perception of a language and a linguistic community while others analyze the social effect of the current legal framework that regulates the LL. The other approach is primarily grounded in fieldwork information taken from the LL and its community members. At first, this type of research was mostly based on quantitative data, but the incorporation of qualitative data has progressively intensified. The analysis is also increasingly multimodal and combines other elements beyond language such as people, sounds, images, symbols, video, type of physical support, authorship, location, function, text’s shape and arrangement, font type, size and color, space-time configuration, and history of text. The majority of these studies examine LLs of two or more linguistic communities in contact and/or the presence of immigrant, minority, and global languages (mostly English) in LLs. In addition, various often intertwined foci have been developed including ethnic identity, inter-ethnic relationships, language vitality, effect of language policies, multilingualism, impact of immigration, minority languages, literacy, market value of languages, economy, politics, ideology, history, globalization, and social change. Researchers have used the term LL mainly to refer to language displayed on public streets. This is why LL is commonly equated with visible writing in the public domain. A closer look into the term justifies this tendency. “Landscape” denotes an expanse of scenery that can be seen in a single view and “linguistic” states the presence of language as its focal quality. That is, LLs display language that is visible (or perceptible) to others. The scope of the LL is defined by the observer, whose “single view” may comprise a wide range of physical settings (e.g. a wall, a street, a neighborhood, a city, etc.). The term LL has been consolidated and generally accepted among scholars, but the growing body of literature on LL challenges the term’s ability to capture the actual object of study. Two main issues have arisen as researchers define and expand the scope of LL studies. On the one hand, the bulk of the investigations deal with multilingual texts displayed in cities. As a result, some have suggested “linguistic cityscape” or “multilingual cityscape” as more appropriate terms. These proposals, however, have yet to gain momentum and face a shortcoming similar to LL. In effect, the suggested terms could be inadequate and more restrictive than LL if they are applied to current and potential studies unrelated to the city setting. On the other hand, what is considered LL is stretching beyond the concept of publicly visible written language in the streets. Some LL studies propose or include elements such as oral texts, texts in interior spaces, texts in cyberspace, images, symbols, and people. Most of these elements are included in the baseline definition of LL (i.e. any display of texts), but the issue is to determine if non-textual elements can be considered LL on their own. Arguably, a thinkable LL without language or an LL where language is not the focus, but rather another means to interpret or construe spaces, would be better labeled as (geo)semiotic landscape. Spanish in the LLs of the U.S. is particularly relevant to the understanding of Spanish in this country for two main reasons. First, Spanish competes with English in the LL with an apparent disadvantage. Spanish is often associated with the language of immigrants from developing countries, while English is not only the de facto official language in the U.S., but the de facto language for global communication, visible in LLs around the world. Not surprisingly, the status of Spanish in the U.S. is mainly that of a subordinate code (see Potowski, this volume). In LLs of the U.S., the use of languages is scarcely regulated even at the local level and the preservation of Spanish as a heritage language is not endorsed by language policies. In this context, the presence of Spanish in the LL springs, not solely but frequently, from local entities and, as such, the LL becomes an expression of the language’s status, public relevance, and functionality 67

José M. Franco Rodríguez

in the selected area. Second, LLs are subject to less publishing scrutiny than other written modalities and often favor the language that best relates to their potential readers. Vernacular conventions are thus often prioritized over academic guidelines, and linguistic traits of Spanish in the U.S. are expected to surface in LL texts. These circumstances make the Spanish displayed in LLs of the U.S. a unique source of sociolinguistic material not only to provide a descriptive account of Spanish, but also to assess its societal function, utility, and status, and to contribute to prospective language policies that regulate or promote Spanish as a heritage language in the public domain.

Critical issues and topics LL studies is a developing field with a methodological approach that is gaining consensus, but because it is still in the making, some core issues are currently being discussed by LL researchers: (a) LL authorship and actor, (b) definition of the unit of analysis, and (c) interpretation of quantitative data as ethnolinguistic vitality. I will briefly present summaries of these three issues before discussing Spanish LL in the U.S.

Authorship and actor Most LL studies that offer quantitative data aim to measure public vitality and status in terms of language visibility. Thus, the occurrence of a language is gauged by the number of units of analysis that contain that language. While this approach is useful to illustrate the social utility of a language in the LL, some questions arise regarding the accuracy of this type of analysis. For example, some researchers have attributed the complexity of linking linguistic traits and LL authorship to the often presumable polygenetic composition of LL texts. How many people have participated in the whole process of composing the current text? Who are those people? Why did they choose those words and language(s)? Who decided the arrangement of the text, the number of signs, the type of signs, etc? Only a few scholars have addressed these questions, and they have obtained different results that range from the owner’s full control of the text to shared decisions between the owner and the signage/advertising company. The most critical issues are, first, to differentiate between authorship and actor (i.e. the entity most reasonably associated to the text) and, second, how to determine who the actor is and what linguistic material this actor produces. This distinction is useful, for instance, to decide how much of the writing is private, institutional, or corporate; to differentiate between the use of English by locals and corporations; to report on who uses the different languages and how; and to specify the source of linguistic variations. A reliable LL analysis should associate every text to a single actor. The same actor may display many texts (different or not) in different parts of the LL, but that actor should be tallied once. That is, LL authorship is relevant for understanding the creation process and origin of LL material. However, authorship is arguably not essential in ascribing a text to a given actor. The specific persons who generate a text are not as significant as the type of entity (actor) that is associated to it and likely behind its composition. For instance, the presence of a minority language in the LL may be interpreted differently depending on the type of actor who displays it. Thus, researchers must seek evidence to determine if the actor is an international corporation, a local business, a governmental office, or an individual, regardless of the actual people behind the making of the text. There are LL texts whose authorship is very challenging to determine, but not so challenging to link them to an actor. For example, the authors of anonymous personal messages and graffiti may not be easily tracked down, but the actors can be classified as private, since they are not corporations or institutions. 68

Spanish in linguistic landscapes of U.S.

Many studies differentiate between top-down (institutional) and bottom-up (private) texts to determine the force that promotes a language or a linguistic model, but fail to specify how many actors are producing those texts. A language can have significant visibility if it is displayed in many texts, but this visibility may not equate to language vitality if there is a substantial discrepancy between the number of texts and the number of actors that produce those texts. For instance, a corporation or a state agency can inundate an LL with texts containing a given language, which may indicate the language’s institutional support or market value, but may not reflect the community’s active use of that language. A more productive classification of actors differentiates between private for individuals or small businesses; corporate for corporations, chains, or non-public organizations; and institutional for public institutions, centers, or organizations. In addition, actors are also classified as local if they originate from and have a presence in the city; regional if they are also present in other places within the region; national if they are present in areas beyond the region; and international if they have an international presence or self-identify as foreigners.

Definition of the unit of analysis One of the main challenges of producing reliable quantitative LL data is to clearly define a distinct unit of analysis. The various approaches to the notion of LL have generated different descriptions and labels for the unit of analysis. Moreover, these descriptions are being challenged by the complexity of the LL. This is the case of the two most followed methodological models: Backhaus (2006, p. 55), who uses the term “sign” and defines it as “any piece of text within a spatially definable frame”; and Cenoz and Gorter (2006, p. 71), who explain that “each establishment but not each sign was the unit of analysis.” The emphasis on the physical display in these definitions reasonably and easily evokes the commonly used term “sign.” However, it may not be enough to produce accurate quantitative analysis of LLs, as illustrated in the two scenarios below. In scenario 1, three stores display information with adhesive letters on their storefront glass doors, but the opening hours are presented differently in each of them. Store A uses the same adhesive letters with the times in English and Spanish; store B uses one piece of paper with the times in English and Spanish; and store C uses two small signs, one in English and the other in Spanish. Following Backhaus’ definition, the number of units is different for each store, despite the fact that the only major difference is the physical support. Following Cenoz and Gorter’s definition, each store counts as one single unit. In scenario 2, a clothing store that is part of a local chain in Spain uses Spanish to display the names of products in its store window, but also displays two pieces of information unrelated to clothing. One is a paper that advertises a room for rent and includes a Latin American dialectal term. The other is a sign from a credit card company that includes an English loanword. Backhaus would count three units and Cenoz and Gorter would record a single unit, even though some texts are not related to the establishment. Having the physical support as the main criterion to define the unit of analysis, even when the support is associated to an establishment, proves to be insufficient to consistently quantify languages in LLs. This is because physical support provides valuable visual information for an initial approach to the unit of analysis, but a comprehensive characterization needs to determine the relationship between content, actor, and support. It is essential to establish the correlation between the actor and the content of the text, which can be subsequently used to determine the physical boundaries of the unit of analysis. Thus, a more suitable term would be “LL text” (or just “text”) as defined by Franco Rodríguez (2013, p. 112): “Any piece(s) of writing composed by the same actor with a focal content related to that actor and displayed on a 69

José M. Franco Rodríguez

circumscribed space in the public domain.” Following this definition, all content in each store of scenario 1 is composed by one actor who displays two languages, thus one text for each store’s glass door is counted; however, there is evidence that scenario 2 texts have been composed by three different actors, thus three texts in the store window are tallied and the use of English and the Spanish regionalism are associated to different actors. These examples illustrate the importance of a well-defined unit of analysis to consistently determine the public visibility of a language and the actor behind it.

Interpretation of quantitative data as ethnolinguistic vitality Ethnolinguistic vitality refers to an ethnic group and its language’s social presence and status. Social presence includes demographic data and occurrence in media and LL. Social status comprises their influence on and representation in institutions and the economy. Although a language’s visibility in the LL can be an indicator of ethnolinguistic status, a number of studies reveal that this visibility does not necessarily correlate with the language’s vitality or its speakers’ vitality in a given area. An ethnolinguistic group can be absent or have low presence in society, but have a proportionally high language visibility. For instance, English can be ubiquitous in an LL where there are very few native English speakers, owing to its role as a global language. Alternatively, a minority language may be barely represented in the LL despite the large concentration of speakers due to the ethnolinguistic group’s low literacy rate, as is the case of Navajo. Nevertheless, the apparent lack of direct correlation between visibility and vitality should not overshadow the implications of representing or silencing an ethnolinguistic group in the LL. Visibility can still be a valuable tool to assess vitality. The challenge is to improve both data collection and data analysis for more reliable interpretations of findings. Quantitatively speaking, more solid contrasts and conclusions could be yielded if LL studies incorporate a multivariate analysis of the following variables. First, the number of texts is often used to measure language visibility, but this visibility may not reflect linguistic vitality accurately unless the analysis distinguishes between the total number of texts and the total number of different texts, since repeated texts would be tallied as different texts. In addition, if some types of texts (e.g. graffiti, moving texts, small print, etc.) are omitted for methodological reasons, the impact of omitting them can be minimized in contrastive analyses if the studies include a detailed account of the exceptions. Second, it may be inaccurate to claim robust vitality of a given language that is visible in a significant number of texts when these texts contain few words or few different words in that language. A thorough analysis needs to include the number of words, different words, and the distribution of words. Third, LL studies created to measure the presence of minority languages or global languages generally fail to report on the contexts in which they are used. Thus, stating high vitality for a significantly visible language may be inaccurate if this language is only used in one or two areas of activity/life. The use of languages needs to be categorized into domains (e.g. health care, education, real estate, etc.). The more domains in which a language is displayed, the higher public utility can be assigned to that language. Last, language can be displayed in the main section, the most prominent part of a text with larger print in a top, left, or central position, or in the informative section, a subordinate location with smaller print in a bottom, right, or marginal position that commonly complements the main section. A complete analysis needs to differentiate between these two sections. The presence of a language in both sections attests to its symbolic value and social utility, which can be interpreted as linguistic vitality. 70

Spanish in linguistic landscapes of U.S.

Research on Spanish in LLs of the U.S. References to Spanish in the LL can be found in some studies around the world that focus on other languages, or those that describe multilingual LLs in countries where Spanish is not an official language. But much of the research strictly on LLs in Spanish-speaking areas has been conducted in Spain. Catalan and Spanish were already examined by the end of the 1990s to describe language use within the bilingual autonomous region of Catalonia. This interest has continued to yield further research about the impact of language policies on the visibility of co-official languages in Spain, namely the Basque Country, Catalonia, Galicia, and Valencia. Other studies have focused on minority languages and multilingual LLs as a result of immigration and globalization in some Spanish cities, including Almeria, Barcelona, Madrid, Malaga, Mallorca, and Seville. Bibliographical reviews have been offered by Franco Rodríguez (2011), Pons Rodríguez (2012), and Castillo Lluch and Sáez Rivera (2013). Beyond Spain, there are few contributions on other LLs in Spanish-speaking countries. These are concentrated in Mexico and include topics such as the use of Oaxaca’s LL as a pedagogical tool in the teaching of English as a Foreign Language (Sayer, 2010), the occurrence of English and Yucatec Maya in business names (Pfeiler et al., 1990), the presence of English in Mexican shop names in Monterrey as part of the overall incidence of English in Mexican commerce (Baumgardner, 2006), and the impact of English in Mexico City and Monterrey’s LLs in a broader analysis of the influence of English on Mexican Spanish (Baumgardner, 2011). On the Texas-Mexico border, Martínez (2003) examines the city of Reynosa, describing the presence of English loanwords and English-influenced grammatical innovation visible in commercial signage. The studies of Spanish in LLs of the U.S. have precisely in this study an antecedent to the type of LL displayed in densely Latino areas of the southwest. The currently growing body of literature on LLs of the U.S. concentrates on areas with minority/heritage speakers of Asian languages (Chinese and Korean), Navajo, Portuguese, Russian, and Spanish. As an illustrative example of the increasing interest in Spanish, the 7th International Linguistic Landscape Workshop in 2015 included 11 papers that addressed the presence of Spanish in the LL, 7 of which were LLs of the U.S. in a wide range of settings (i.e. streets, museums, schools, online, and multimodal environments). The published studies that focus exclusively on Spanish-speaking areas in the U.S. are still few and heterogeneous, but clearly reveal LL’s potential to gauge the public status of Spanish as a minority/heritage language in the U.S. In this section, I will review 17 studies of Spanish LLs in the U.S., which as a whole suggest the apparent social value of Spanish in the LL as both a functional minority language and a conspicuous heritage symbol. This value is impacted by two contrasting forces. On the one hand, it is weakened by ideological and economic factors, which cause a disproportionately low visibility of Spanish, a preference for English and bilingual texts, and negative attitudes towards Spanish and its speakers. On the other, it is boosted by thriving ethnolinguistic vitality, which contributes to higher visibility of Spanish and its cultural heritage, greater language maintenance, and positive attitudes towards the language and Hispanics. The public utility of Spanish as the language of the community has been described by some authors. Berry (2004) used the term “business-scapes” to describe the LL produced by Latino businesses in Reno, Nevada, and surveyed 19 Hispanic-owned businesses. The author provided examples to illustrate the growing visibility of LL items associated with Latinos as marketing strategies: Spanish and bilingual signage, colors of national flags, maps, regional decorations, interior murals, and typical products. The author conducted a survey on family-owned Hispanic businesses and reported that their commercial activity covered a wide range of domains, their customers were mostly from the community, and Spanish was the customary language in 71

José M. Franco Rodríguez

business transactions. In a series of articles, Franco Rodríguez (2005, 2007, 2008) provided an evolving linguistic description of the Spanish displayed in the LL of predominantly Hispanic areas in Los Angeles and Miami-Dade counties. These studies showed how Spanish in these LLs mirrors the linguistic traits of the counties’ Spanish-speaking communities. This correlation was supported by three arguments: first, signage companies’ responses to surveys indicated no interference on the language displayed in texts; second, a quantitative analysis of the actors by type revealed that Spanish stems mostly from individuals or small businesses; and third, a comparison with other studies on Spanish in those areas validated LL text analysis in recognizing vernacular linguistic patterns. The 2008 contribution presented a systematic analysis of the LL and applied it to Los Angeles and Miami-Dade’s LLs. This analysis aimed to gauge linguistic vitality in terms of visibility and utility. Visibility was measured by the total number of texts and text placement. Utility was measured by quantifying the number of different texts, total number of words and different words, words by category, the domains of use per actor (institutional, corporate, and private), as well as the informative and symbolic functions of Spanish and English. The results revealed the communities’ bilingual condition. Spanish was amply used for practical and symbolic purposes, but the most prominent part of the text was primarily in English or bilingual, and the informative section was also mainly bilingual. English had a greater linguistic imprint in Los Angeles county’s LL, but did not mix with Spanish as much in Miami-Dade’s LL, which correlated to Latino’s higher level of education in this county. Benedict and Kent (2004) conducted a longitudinal field survey in an area of Cleveland, Ohio, to analyze the correlation between Latinos’ demographic vitality and the incidence of “semi-fixed landscape elements.” Over a period of six years, they examined signage in commercial areas and churches and other items such as house colors, distinctive Hispanic decorations, and national symbols in residential areas. The results showed a correspondence between high Latino population’s density and a greater proportion of LL items associated to Hispanics, in particular Spanish and bilingual signage. However, in other studies, the visibility of Spanish has been described as disproportionately low in relation to the size of the Hispanic population in the area. Troyer et al. (2015) combined quantitative and qualitative data analysis to examine a small town’s LL in Oregon, where one-third of its population were Spanish speakers. Quantitative examination included language choice by domain and by type of actor in relation to the number of actors, signs, and different signs. This analysis revealed that Spanish had low visibility proportionate to the percentage of the Spanish-speaking population, was visible in very few domains, and was displayed mostly by local and national businesses. Authorities at all levels employed only English, with the exception of two local texts that used Spanish to discourage illegal behavior. Qualitative analysis elicited ethnolinguistic perceptions from diverse community members: a city official and seven people who represented businesses that displayed either only English, only Spanish, or both languages. Interviewees exposed two causes that seemed to inhibit the use of written Spanish by locals in this LL: Anglo-Americans’ fear of social change fed by the Latino population’s rapid growth, and Hispanics’ fear of prejudices fed by negative immigration discourses. Some LL studies claim that the salience of Spanish in LLs within areas of dense Hispanic population is offset by an expanding English-centered ideological environment. Barker et al. (2001) discussed the impact of the English-only movement on minority languages like Spanish and on perceptions of language vitality. They argued that high visibility of Spanish in the LL could be identified in this context as a threat to the hegemony of English and therefore end up actually being detrimental to Spanish. Later, Barker and Giles (2002) conducted phone interviews to examine Anglo-Americans’ perceptions of their own community and of the Hispanic community’s vitality in Santa Barbara, California. They found that their perceptions of increasing 72

Spanish in linguistic landscapes of U.S.

Latino vitality and visibility in the LL correlated with greater support of English-only policies. Hult (2014) examined the Texan LL of San Antonio’s highway system by establishing the connection between 1) quantitative analysis or the presence of a language in the LL; 2) nexus analysis or language ideology, social status, and public function; and 3) geosemiotic analysis or language placement on the physical support, content expressed in that language, and other visual elements. The specified focus was advertising and commercial signage at the local level. The results showed that the vast majority of this LL consisted of monolingual English signs, which the author considered a reflection of the English-dominant language ideology (i.e. educational policies and language activism), not the actual city’s bilingual makeup. Hult further analyzed the presence of Spanish on three non-local signs connected to an international corporation, a national corporation, and a state agency. These signs depicted Spanish in this LL as a marginalized language that was associated with the working class and hardly recognized as the local communities’ functional code. A similar negative view is portrayed by Martínez (2014), who scrutinized Spanish-speaking youth’s perceptions of LLs displayed within a wide variety of health care facilities in Hidalgo County, Texas. He used an innovative approach in LL studies called photovoice methodology, which consisted of three steps. First, members of the community took pictures to capture evidence of a given issue. Second, those pictures were presented in community meetings to discuss any needs for action. And third, the outcomes of the discussion were communicated to the appropriate institution so that action was taken based on the narratives presented. This study was mostly qualitative, but some quantification was carried out: code preference (i.e. more English than Spanish), error analysis (i.e. many more in Spanish than in English), physical support (i.e. three times more makeshift signage in Spanish), and content (i.e. more information about health in English and more procedural/instructional and regulatory in Spanish). Study participants perceived a message of contempt towards Spanish-speakers and expressed disapproval of this LL. They felt that Spanish speakers were not as appreciated as English speakers. The author argued that the LL reflected the subordination of Spanish in an English-dominant environment, which was a result of the evident unbalanced power relations between the two ethnolinguistic groups. Such English-only policies, however, seem to more greatly impact institutions and non-local texts. Yanguas (2009) examined the public language use of English, Spanish, or both in the LLs of two Washington, DC Hispanic neighborhoods in relation to two other factors: language ideology (i.e. institutional support and community beliefs) and language management (i.e. language policies and local choices). In light of an expanding official-English ideology, the author underscored the relevance of this type of analysis to demonstrate how practices at the local level differed, or not, from practices at the state and federal levels, as well as to determine how much and for what purpose a language was used in a community. Some examples were provided to illustrate, first, how this LL evidenced the public usefulness and vitality of Spanish in the area and, second, the apparent conflict between local actors’ beliefs and DC’s English-centered standpoint. While local actors supported bilingual practices, DC endorsed a de facto language policy that disregarded the community’s linguistic reality by using only English in the LL. The author concluded that English-centered ideology was showcased in the LL and therefore affected Hispanics and non-Hispanics’ language attitudes alike. Similar conclusions can be found in the comparison made by Franco Rodríguez (2013, pp. 127–129) between the LL of Almeria, Spain, and Los Angeles and Miami-Dade counties. This comparison indicated a slightly higher incidence of English in those two counties, where English was used by many more local actors. The reported data shows that Spanish was actually utilized by significantly fewer corporations and public institutions. This finding suggests that Spanish, despite its public utility, enjoyed less institutional support and market value in these two counties than in the officially monolingual city of Almeria. 73

José M. Franco Rodríguez

Other authors have examined the impact of Spanish in the LL on attitudes towards Spanish and Hispanics. Mitchell (2010) compared attitudes towards Hispanics’ vitality with their actual occurrence in the LL of a Hispanic neighborhood in Pittsburgh. The author discussed a newspaper article that expressed “fear” of the growing Latino presence in the neighborhood, despite being less than 5% of the population. One of the possible reasons given for this mistaken perception is how people experience a “semiotic landscape” or an LL that includes both visual and oral elements (even smells). Since all institutional signs were in English, Mitchell focused on private texts and quantified businesses, signs, languages in signs, and languages heard. The results showed a proportionally lower percentage of Spanish texts than Latino residents, and a balance between English and Spanish displayed on Hispanic businesses. However, the misperception of high Latino incidence in this LL could be explained by non-written phenomena: the clustering of a few Hispanic businesses and the high percentage of spoken Spanish in the selected area. Some reasons for those negative attitudes have been explained by Benedict and Kent (2004). Their study reinforced four relevant issues for LL studies: even a very low number of LL items can create a perception of an ethnic neighborhood; results may be different in areas with a higher percentage of more acculturated second- and third-generation Latinos; English is the overall most salient language regardless of demographics; and better understanding of LLs can be gained by combining field surveys with interviews about the motivations behind displaying LL items. But the impact of the LL on attitudes is not always negative. Barker and Giles (2002) found out that, contrary to their expectations, less contact with Spanish in the LL was linked to a greater support for English-only policies. The authors presumed, though, that results would be reversed if they included change as a factor by asking about perceptions regarding a future LL with more Spanish. Dailey et al. (2005) analyzed high school students’ attitudes towards Anglo- and Hispanic-accented speakers of English and towards the LL in a broader sense (i.e. 28 oral and written language items indoors, outdoors, and in the media). The multivariate analysis revealed that Anglos’ contact with English or Spanish in the LL did not affect their perceptions of either type of speaker; however, in the case of Latinos, the higher their exposure to English the more favorably they rated Anglo speakers and, conversely, the higher their exposure to Spanish the less favorably. A greater exposure to Spanish attenuated favoring Anglos’ traits over those of Hispanics, which led the authors to infer a potential correlation between Spanish in LL, attitudes, and language maintenance. Finally, participants in the study conducted by Troyer et al. (2015) offered some examples of how the LL can be an asset for ethnolinguistically diverse communities. In their opinion, Spanish and Latinos in the LL enrich the local cultural environment, exposure to Spanish and Hispanics increase tolerance, and authority’s LL awareness can promote more thoughtful, integrative LL policies. While language policies to promote Spanish in LLs of the U.S. are still to be developed, some authors have described Spanish in the LL as a means for and evidence of language maintenance and heritage preservation. Franco Rodríguez (2011) contrasted data from Los Angeles and Miami-Dade counties’ LLs with their sociocultural context. Although Spanish was negatively affected by education policies, generational language attrition, lower social status, and official-English movements, contextualized LL data analysis revealed the persistent maintenance of Spanish as a dynamic minority language in these counties. Seals (2012) examined the LL created by a demonstration in Washington, DC. Although Spanish or Hispanics were not the focus of this study, both became a major part of the analysis. Both were examined in an uncommonly studied setting, a transient LL, and from a multimodal perspective that included writing, flags, people, and videos. The results illustrated the Hispanics’ bilingual makeup and their high visibility as an ethnolinguistic group. García et al. (2013) studied two murals and the surrounding signs displayed in the largely Hispanic neighborhood of East Harlem in Upper 74

Spanish in linguistic landscapes of U.S.

Manhattan, New York City. These murals were characterized as contestation tools that the Latino community used to express the condition of a struggling ethno-cultural conglomerate, something they could not fully convey in English or Spanish. The authors argued that Englishcentered education and low Spanish literacy in this area cornered Spanish to the oral domain. Unsurprisingly, English was the most visible language and the only one used by institutions in the LL. Some bilingual and Spanish texts were displayed by businesses as a marketing tool. There was limited Spanish, but this research presented two indicators of linguistic endurance in the LL: the Spanish used by the community for social activism, and the apparent link between Spanish revitalization and local gentrification afoot.

Recommendations and future directions LL studies provide a unique angle through which to examine Spanish in the U.S. due to the complexity of the object of study. The LL is a collective text in which numerous actors converge and participate along with other elements such as languages, spaces, sounds, speech, smells, images, symbols, objects, people, ideologies, interests, and beliefs. For this reason, future analyses of Spanish in LLs of the U.S. should be increasingly more comprehensive: multilayered, multivariate, multimodal, quantitative and qualitative, synchronic and diachronic, comparative, and anchored in shared solid methods. As more comprehensive studies develop, more conclusive answers will be provided to the issues presented in the previous section. Studies on Spanish in the LLs of the U.S. reveal a discrepancy between the high concentration of Hispanics and the low visibility of Spanish in the LL, especially in texts. The main reason suggested behind this phenomenon is the subordinate status of Spanish in an increasingly English-centered environment. This may be the case, but further research is needed. Data on language should specify the number of Spanish speakers, their oral proficiency, domains of language use, and their literacy in Spanish and English. Second- and third-generation heritage speakers may use some Spanish at home, but have limited knowledge of and formal exposure to it, especially when Hispanic youth have been educated in the U.S. and speak English well. Oral interviews and surveys distributed to Hispanic local actors who display English could provide additional information on this issue. English might be the code they know best, the potential customers’ code, or a plausible step in the acculturation process. In addition, quantitative data on institutional texts used to assess official support of Spanish must be interpreted in light of relevant demographic data. Comparative analyses could determine whether or not Spanish has less social recognition in areas that experience a sudden Hispanic influx than in areas considered traditional Latino enclaves. This hypothesis would shed light on why the initial and sole institutional support of Spanish would occur at the local level in newly formed Hispanic communities. Spanish is likely to first have the pragmatic backing of local officials who are in direct contact with the ethnolinguistic reality. Interviews could provide useful insights on local agencies’ use of Spanish, but also on determining when and why state and federal agencies opt to use Spanish in the LL. This apparent slow or absent institutional reaction to a Latino community’s ethnolinguistic makeup contrasts with the corporations’ promptness, whose alacrity can be explained by their interest in the commodification of languages to generate profit (Leeman & Modan, 2009). Yet a detailed account of the corporate and local businesses that display Spanish, how much, and the reasons behind their use of Spanish can reveal significant information about the language’s economic value, which can be interpreted as ethnolinguistic vitality. Future studies can attempt to connect linguistic landscapes to other theoretical work, such as the interaction between the organization of urban space and the political economy (Harvey, 2009). 75

José M. Franco Rodríguez

Last but not least, LL studies in the U.S. present two highly worthwhile applications. One is their potential role in the teaching about Spanish and Latinos in the U.S. LL research findings could be incorporated in courses that cover aspects such as current widespread vernacular traits, the vitality of Spanish and Hispanics as a minority ethnolinguistic group, the impact of English as a global and majority language, and diachronic sociocultural changes in Hispanic communities. The other application is the suitable use of LL studies to inform and influence language policies that foster both Spanish visibility in the LL and ethnolinguistic awareness. This application will benefit from interviews that assess perceptions and attitudes and from the photovoice methodology described earlier, both of which can raise LL consciousness among community members and policymakers. Further and more comprehensive research will be essential to validate current findings and build a case for the promotion of Spanish in the LL. The main findings to confirm are those that elicit Hispanics’ feelings of contempt: scarce presence of institutional texts with Spanish, content that is mainly regulatory and prohibitive, and poorly written texts that are flooded with errors and do not convey the right meaning or are incomprehensible. Thus, this type of LL studies becomes a means of language activism.

Further reading Gorter, D. (2013). Linguistic landscapes in a multilingual world. Annual Review of Applied Linguistics, 33, 190–212. [This article provides a synopsis of LL studies development, discussions on terminology, current theoretical frameworks and methodological approaches.]

References Backhaus, P. (2006). Multilingualism in Tokyo: A look into the linguistic landscape. International Journal of Multilingualism 3(1), 52–66. Barker, V., & Giles, H. (2002). Who supports the English-only movement? Evidence for misconceptions about Latino group vitality. Journal of Multilingual and Multicultural Development 23(5), 353–370. Barker, V., Giles, H., Noels, K., Duck, J., Hecht, M., & Clement, R. (2001). The English-only movement: A communication analysis of changing perceptions of language vitality. Journal of Communication 51(1), 3–37. Baumgardner, R. J. (2006). The appeal of English in Mexican commerce. World Englishes 25(2), 251–266. Baumgardner, R. J. (2011). Mexico’s ‘Commission for the Defense of the Spanish Language’ (1981–82). International Journal of Linguistics 3(1), E20. Retrieved from www.macrothink.org/journal/index.php/ ijl/article/view/736/pdf. Benedict, A., & Kent, R. B. (2004). The cultural landscape of a Puerto Rican neighborhood in Cleveland, Ohio. In D. D. Arreola (Ed.), Hispanic Spaces, Latino Places: Community and Cultural Diversity in Contemporary America (pp. 187–205). Austin, TX: University of Texas Press. Berry, K. A. (2004). Latino commerce in northern Nevada. In D. D. Arreola (Ed.), Hispanic Spaces, Latino Places: Community and Cultural Diversity in Contemporary America (pp. 225–238). Austin, TX: University of Texas Press. Castillo Lluch, M., & Sáez Rivera, D. M. (Eds.). (2013). Paisajes lingüísticos en el mundo hispánico. Revista Internacional de Lingüística Iberoamericana XI(1), 7–152. Cenoz, J., & Gorter, D. (2006). Linguistic landscape and minority languages. International Journal of Multilingualism 3(1), 67–80. Dailey, R. M., Giles, H., & Jansma, L. L. (2005). Language attitudes in an Anglo-Hispanic context: The role of the linguistic landscape. Language and Communication 25, 27–38. Franco Rodríguez, J. M. (2005). El español en el Condado de Los Ángeles desde la señaléctica comercial y urbana. Hispania 88(4), 825–833. Franco Rodríguez, J. M. (2007). El español en el Condado de Miami-Dade desde su paisaje lingüístico. Lingüística en la Red V, 1–29. Franco Rodríguez, J. M. (2008). El paisaje lingüístico del Condado de Los Ángeles y del Condado de Miami-Dade: Propuesta metodológica. Círculo de Lingüística Aplicada a la Comunicación 35, 3–43.

76

Spanish in linguistic landscapes of U.S. Franco Rodríguez, J. M. (2011). Linguistic landscape and language maintenance: The case of Los Angeles and Miami-Dade counties. In M. A. Morris (Ed.), Culture and Language: Multidisciplinary Case Studies (pp. 69–119). Frankfurt, Germany: Peter Lang. Franco Rodríguez, J. M. (2013). An alternative reading of the linguistic landscape: The case of Almería. Revista Internacional de Lingüística Iberoamericana XI(1), 109–134. García, O., Espinet, I., & Hernández, L. (2013). Las paredes hablan en El Barrio: Mestizo Signs and Semiosis. Revista Internacional de Lingüística Iberoamericana XI(1), 135–152. Harvey, D. (2009). Social Justice and the City. Athens, GA: University of Georgia. Hult, F. M. (2014). Drive-thru linguistic landscaping: Constructing a linguistically dominant place in a bilingual space. International Journal of Bilingualism 18(5), 507–523. Landry, R., & Bourhis, R. Y. (1997). Linguistic landscape and ethnolinguistic vitality: An empirical study. Journal of Language and Social Psychology 16(1), 23–49. Leeman, J., & Modan, G. (2009). Commodified language in Chinatown: A contextualized approach to linguistic landscape. Journal of Sociolinguistics 13(3), 332–362. Martínez, G. A. (2003). Signs of the times: Globalization and the linguistic landscape along the US-Mexico border. Rio Bravo New Series 2(1), 57–68. Martínez, G. A. (2014). Vital signs: A photovoice assessment of the linguistic landscape in Spanish in healthcare facilities along the U.S.-Mexico border. The International Journal of Communication and Health 4, 16–24. Mitchell, T. D. (2010). ‘A Latino community takes hold’: Reproducing semiotic landscapes in media discourse. In A. Jaworski & C. Thurlow (Eds.), Semiotic Landscapes: Language, Image, Space (pp. 168–186). New York: Continuum. Pfeiler, B., Franks A., & Briceño Martin E. (1990). El maya y el inglés en la nomenclatura de los comercios meridanos. Revista de la Universidad Autónoma de Yucatán 174, 49–54. Pons Rodríguez, L. (2012). El paisaje lingüístico de Sevilla: Lenguas y variedades en el escenario urbano hispalense. Sevilla, Spain: Diputación de Sevilla. Sayer, P. (2010). Using the linguistic landscape as a pedagogical source. English Language Teachers Journal 64(2), 143–154. Seals, C. A. (2012). Creating a landscape of dissent in Washington, DC. In C. Hélot, M. Barni, R. Janssens & C. Bagna (Eds.), Linguistic Landscapes, Multilingualism and Social Change (pp. 127–138). Frankfurt, Germany: Peter Lang. Troyer, R. A., Cáceda, C., & Giménez Eguíbar, P. (2015). Unseen Spanish in small-town America: A minority language in the linguistic landscape. In S. Ben Said & R. Rubdi (Eds.), Conflict, Exclusion, and Dissent in the Linguistic Landscape (pp. 52–76). London: Palgrave Macmillan. Yanguas, I. (2009). The linguistic landscape of two Hispanic neighborhoods in Washington D.C. Revista Electrónica de Lingüística Aplicada 8, 30–44.

77

6 LINGUISTICS AND LATINO STUDIES Intersections for the advancement of linguistic and social justice Lourdes Torres depaul university, usa

Introduction In a 2014 article in the journal Latino Studies, Ana Celia Zentella calls on Latino studies scholars to conduct research on Latinos and language. She states: Latino studies scholars have a leading role to play in opening the nation’s linguistic and cultural borders. Instead of ignoring the role of language in the varied topics we study, or viewing languages as finite essences that define particular identities, we must encourage an expanded view of the linguistic repertoire of “America” in general. Our work should enable our fellow citizens to view the nation’s languages as a tapestry that weaves together the rich colors of transculturation with the creative breaks of codeswitching and the powerful threads of hybridity. We must discourage the negative connotations of bilingualism and bidialectalism as impure mixes, as inappropriate or as indicative of inauthenticity, and view them as strengths instead. Our goal is nothing less than the end of linguistic discrimination as essential for the preservation of democracy, following in the footsteps of African American efforts, and counting on their support. 2014, p. 632 This is not the first time such a call had been made in the pages of the Latino Studies journal. In 2004, Antonia Darder made a similar appeal: “Given the increasingly tenuous state of culturally democratic ideals in this country, it is imperative that Latino studies scholars heed Antonio Gramsci’s advice and put the question of language at the forefront of our inquiry” (2004, p. 231). Both scholars underscore the intersection of culture, language, and power, and the fact that language and language rights are inextricably linked to democracy, social justice, and citizenship. For example, the role of language is pivotal in conceptualizations of belonging in the U.S., as speaking English or having or not having an accent is often used as a measuring stick to determine the rights of immigrants and their worthiness of citizenship. 78

Linguistics and Latino studies

Likewise, English-only politics and attempts to eradicate bilingual education are symptoms of the power dynamics in the U.S., and discrimination based on language is just one dimension of the racialization Latinos experience. To what extent has Latino studies taken on the challenge described by Zentella and others? While there is much research exploring the structure of Spanish varieties spoken in the U.S. and how they evolve over time in diverse communities (see Escobar & Potowski 2015, chapters 3, 4, and 5), many of these are quantitative studies that are not grounded in community language practices. That is, most studies have not examined the intersection of language and society and how language really functions in the life of Latino communities. This chapter discusses some pivotal studies that have explored the intersections between sociolinguistics and Latino studies. My intention is not to present a historical overview of the numerous studies focused on the ways of speaking of U.S. Latinos, but rather to discuss some of the trends in this work. I explore how sociolinguistics and Latino studies scholarship has responded to prejudicial, status quo notions of Latino ways of speaking, and has sought to bring awareness to the discriminatory repercussions inherent in such ideology since the 1980s. I then turn to more recent scholarship since the start of the 2000s exploring the incursions of both linguistic anthropology and discursive analysis in issues of Latino language. I conclude by suggesting future directions towards understanding the role of Spanish and English in Latino lives and examples of productive advocacy that sociolinguists and Latino studies scholars have undertaken to advance language rights in the Latino community.1

Latino studies and sociolinguistics, and their shared investment in the study of Latino language Latino studies had its roots in the evolution of Puerto Rican and Chicano studies in the 1960s and 1970s (Aparicio 1999, 2003; Cabán 2003; Flores 1997). One important aim of this field is to produce interdisciplinary scholarship about the historical and present day realities of Latinas and Latinos and their struggles for social justice in the U.S., with scholars generating analyses of local, national, and transnational factors that condition the lives of Latinos in the U.S. As an interdisciplinary field located across education, sociolinguistics, sociology, history, and others, Latino studies has had an impact across the social sciences and humanities and has contributed to and created new areas of knowledge. The rise of the field is very much tied to the struggle for civil rights including bilingual education and linguistic rights. For the purposes of this chapter, I review research that connects Latino studies and linguistics in that the work focuses on Latino ways of using language from one of these fields located across the social sciences. Work in sociolinguistics theoretically deals with social aspects of language – that is, the ways in which social factors are connected to linguistic features. Through the development of sophisticated quantitative methodologies, sociolinguistics has made many strides in describing how linguistic and social variables influence language use. However, the work often provides descriptions of what people are doing linguistically without considering the sociocultural context of such usage or the relevance of the language use to the speaker or their community. As Romaine (1996, p. 101) puts it, sometimes even sociolinguists treat language “as an object having an existence apart from the social agents who use it.” A number of sociolinguists (Dittmar 1996; Hymes 1996, etc.) have made a case for a critical sociolinguistics (Singh 1996) that has as its focus the social meaning of language. They argue for the need to go beyond quantitative accounts of language use and language change to study the social meaning of language and its importance for its speaker. They remind us of the limitations of studies that provide descriptions 79

Lourdes Torres

of variation and correlations of social and linguistics variables without a discussion of social meaning. For example, it is not enough to know that a particular sound is produced in a certain way by X group of people (according to their age, gender, socioeconomic status, ethnoracial group, etc.) and in a different way by Y group of people. A critical sociolinguist, on the other hand, would provide us with an analysis of language that leads to an understanding of the relationship between language and social practices – that is, how members of relevant communities view X and Y productions of that sound, and what such varieties communicate to locals about the people who produce them, which ultimately can lead to an important social impact on the communities themselves. Linguistic work on Spanish in the U.S. has a long history. In an early attempt to describe the state of linguistics research on Latinos, Teschner, Bills, and Craddock (1975) review studies produced up to the mid-1970s. In his introduction to the volume, Bills points out that most of the work can be classified as falling into the “Hispanic tradition” of linguistics, which he describes as characterized by an “interest in the accumulation of speech fragments with little concern for linguistic or sociological content,” and “an almost exclusive interest in deviations from standard Spanish” (Bills 1975, pp. vi–vii). While subsequent sociolinguistic research in Spanish in the U.S. over the decades became more sophisticated and engaged more modern theoretical and methodological frameworks, many studies were arguably still focused on deviations from standard Spanish (see, for example, Klein 1980; Klein-Andreu 1986; Lavandera 1981) and less concerned with the impact of the research on language on Latino communities (see Torres 1991 for a review of this work). Even more currently, many studies still fall prey to what Rosaura Sánchez describes in Chicano Discourse as “meaningless quantitative studies, giving us proportions and numerous statistics indicating the number of times a particular variant appeared in the speech of one group or another.” She calls for a more integrative approach where linguistic features are analyzed within the social, political, and economic context of the community. Sanchez suggests that when studying “minority languages” one has to be attentive to how such studies can be used to perpetuate stereotypes of inferior and deficient varieties and, by extension, speakers. In that regard, it is clear that some of the work on Spanish in the U.S. has been more concerned with theoretical and descriptive linguistic questions rather than with issues of language justice within a specific sociocultural context or the implications of this research for Latino communities. The Spanish in the U.S. conferences offer a useful overview of the variety of approaches taken by scholars. Since the first conference in Chicago in 1980 initiated by Lucía Elías Olivares, there have been twenty-five conferences dedicated to the study of Spanish in the U.S. Many of these have led to the publication of conference proceedings (see http://spanishintheus.org/ for a list of conferences and volumes). The papers presented at these conferences and the publications that followed range from those that focused on phonological, morphological, and syntactic variation, to processes like borrowing and code-switching, to those that engage language ideology, attitudes, language planning, and policy issues, and from primarily descriptive, to studies that attempt to link structure to language use and social relevance.

Early Latino studies approaches to Spanish in the U.S. Language has always been a site through which the institutionalization of racist and classist practices is played out and thus should be a central concern of Latino studies. Before the 1980s, Latino language was analyzed primarily using a deficit model (Acosta-Belén 1975; Teschner, Bills, & Craddock 1975). The boundaries of Spanish and English were rigorously maintained and enforced, and the bilingualism of working class Latino children was seen as a problem that needed to be eradicated. Both the Spanish and English spoken by Latinos were judged to be 80

Linguistics and Latino studies

deficient, and language mixing or code-switching was deemed to be particularly noxious and damaging. In the 1980s, the work of scholars such as Shana Poplack, Rosaura Sánchez, and Gloria Anzaldúa interrupted this disparaging narrative. While Poplack relies on a sociolinguistics approach, and Sanchez and Anzaldúa adopt a more multidisciplinary methodology, they were all concerned with challenging deficit models of Spanish in the U.S. Although disciplinarily a linguist and not a self-described Latino studies scholar, Poplack’s groundbreaking and intriguingly titled study, “Sometimes I’ll start a sentence in Spanish y termino en español: Toward a typology of code-switching” (1980) revolutionized how sociolinguists and other scholars understood the practice of code-switching. Previously seen as unstructured, random behavior, and by many parents and educators as a sign of confusion and lazy language practices (Acosta-Belén 1975), Poplack demonstrated that code-switching was in fact systemic and rule governed, and most surprisingly for many, that it was precisely those who had the strongest command of both English and Spanish who most frequently engaged in code-switching (see Toribio’s contribution to the present volume about code-switching). Pousada and Poplack (1979) published another controversial work, “No case for convergence: The Puerto Rican Spanish verb system in a language contact situation,” which posited that the Spanish spoken by Puerto Ricans in New York was not converging toward English as commonly believed, but rather possessed a verb system that was similar to other non-contact Spanish varieties. While the years have produced a far more complex view of the results of language contact, these early studies backed by empirical research challenged the pervasive idea that Puerto Rican children were “semi-lingual” with corrupted language. They led to a new perspective on the language practices of Latinos and rigorous methodologies more grounded in empirical research than in the impressionistic studies that until then dominated the field. Pousada and Poplack conducted their research as part of their work with the Language Policy Task Force (LPTF) of the Center for Puerto Rican Studies at Hunter College in New York. The Task Force produced a number of studies in the 1980s (1982, 1988), including Intergenerational Perspectives on Bilingualism: From Community to Classroom (1982). This comprehensive study sought to assess the Spanish and English language resources of the Puerto Rican community in East Harlem in order to suggest necessary policy changes in the education of the children in the community. The LPTF was at the forefront of conducting studies that described the language practices of the Puerto Rican community and advocated for policies that enhanced the educational opportunities afforded to Puerto Rican children. This was a healthy change from previous studies that blamed students themselves, instead turning the focus toward schools that failed to adequately understand the linguistic reality of the children and the bilingual skills they brought to the classroom (Flores 2005; García 2010). Outside of linguistics but within the same timeframe as the studies just discussed, analyses of Chicano language practices by Rosaura Sánchez (1983) and Gloria Anzaldúa (1987) identified language as a key component of the oppression of Latinos as well as a site of innovation and potential liberation. Sánchez’ Chicano Discourse (1983) provides a sociopolitical framework for analyzing Chicano language. The book begins with a Marxist history of the Chicano people in the U.S. and considers how historical and economic factors shape their language use in rural and urban spaces. While its evidence is primarily impressionistic, the study was among the first to link ways of speaking to the material conditions of the community and made the case that language cannot be studied without considering the speakers’ socio-historic background and context. Likewise, Anzaldúa’s Borderlands: La Frontera (1987) is a Latino studies classic that revolutionized thinking about hybridity and the politics of identity. The chapter “How to tame a wild tongue” provides a powerful meditation on “linguistic terrorism” or the discrimination from all sides faced by Chicanos for their ways of speaking: native English speakers find fault 81

Lourdes Torres

with Chicano English, while native Spanish speakers criticize Chicano Spanish, and just about everyone finds code-switching objectionable. Anzaldúa establishes a strong link between ways of speaking and people’s identity, stressing that when you criticize a person’s language you are essentially demeaning the worth of the individual. While neither Sánchez nor Anzaldúa were linguists, their exploration of language in the Chicano community offered pointed assessments of the linguistic situation of Latinos in the U.S. In fact, Anzaldúa’s analysis of Latino language practices, firmly grounded in sociopolitical reality, powerfully captures the essence of the Latino studies linguistic project of the last twenty-five years, seeking to accurately describe the linguistic situation of Latinos and shed light on and challenge an ideology that devalues the ways of speaking of marginalized groups, preserves normative ways of speaking, and upholds white supremacy.

Linguistic anthropology Scholars have since expanded these analyses and, importantly, have offered new analytic tools to explore Latino language that underscore the importance of studying language in its socioeconomic and political context. The most productive research has emerged from the field of linguistic anthropology. For example, Urciuoli (1997) detailed how racialization is enacted through linguistic judgments with devastating effects. Racialization is the process by which people are categorized according to attributes such as skin color or language, and then subjected to stigmatization and marginalization. Urciuoli unpacks how both the Spanish and English varieties spoken by Puerto Rican New Yorkers are heavily monitored and negatively stereotyped by white people and other gatekeepers. For example, although many of Urciuoli’s participants were native speakers of English, they noted that they were often told by whites that they spoke “broken English,” “bad English,” or had “a heavy accent” Urciuoli (1997, p. 2). Such assessments about the varieties these speakers control clearly result from their racial and class position in society: Urciouli documents how even when working class Puerto Ricans speak English, their varieties are judged as inadequate by the mainstream. She argues that the devaluation of Puerto Rican language, since both their Spanish and English are deemed inferior, cannot be divorced from the disrespect and discrimination Puerto Ricans face in the U.S. Her analysis, based on ethnographic research, employs an intersectional analysis to illuminate how race and class shape linguistic production as well as the responses to it. For decades, Ana Celia Zentella has remained at the forefront of exploring how language intersects with sociopolitical issues to condition the lives of Latinos. Like Urciuoli, Zentella’s work is based in linguistic anthropology, but goes one step further by insisting that the work linguists do must be self-consciously political. While some scholars shy away from defining their projects as political, Zentella expressly terms her framework “anthro-political linguistics” (1997, 2002, 2005). Prioritizing the study of language in its sociopolitical context, she argues that language is inherently implicated in socioeconomic life, so that to ignore the analysis of how language works in society is to continue to mask the injustices that are perpetuated on people of color and other marginalized groups. She calls on scholars to make it their business to shed light on these injustices and to work toward linguistic policies that may promote change. Zentella has influenced a generation of scholars that continue in the tradition she advanced. Her 1997 ethnographical work Growing Up Bilingual: Puerto Rican Children in New York painstakingly analyzes the speech of children in the same Puerto Rican community where Poplack conducted her study. She provides a close analysis of the speech of five girls that is firmly grounded in their community and their interactions with the adults and children in their world. Based on her ten-year corpus of the informal, conversational speech from the children, she reaffirms the free-morpheme and syntactic constraint rules proposed by Poplack (1980), providing 82

Linguistics and Latino studies

further evidence of the rule-governed nature of code-switching. In Zentella’s quest to make the research relevant for those working with Latino children, her subsequent work (2005, 2009) aims to help parents and educators understand bilingual Latino children’s language and work with them to develop strong linguistic and literary skills in both English and Spanish. For example, Building on Strength: Language and Literacy in Latino Families and Communities (Columbia Teacher’s College Press, 2005) brings together cutting-edge articles on language socialization and outlines policies that parents, schools, and communities can take to build on the cultural richness inherent in bilingual communities. Finally, Mendoza-Denton’s Homegirls: Language and Cultural Practice Among Latina Youth Gangs (2008) offers a critical ethnography of the ways of speaking of a subculture of Latina girls who belong to gangs in California. The author undertakes the study of linguistic forms within their sociocultural context, focusing not only on words and sounds but also other semiotic practices (i.e. tattoos, makeup, clothing) that capture how the girls express themselves. MendozaDenton unpacks the divisions and affinities between the Spanish-speaking Sureña girls and the more Americanized English preferring Norteña girls in California, and connects them to larger structures such as nationalism, racial and gender identity, and class consciousness.

Nativist anti-immigrant discourses Discourse analysis of mainstream media, government documents, and everyday language has helped us understand how white supremacy and the degradation of Latinos is enacted and perpetuated in public discourse. Santa Ana (2004) uses critical discourse analysis to focus on how Latino stereotypes are perpetuated though the language of the media. He argues that metaphors of Latinos as animals, invaders, and outsiders in the media serve to reinforce and perpetuate negative perceptions of Latinos in the U.S. He finds that such language creates anxiety and hostility in the mainstream population and leads to discriminatory policies against Latinos. For example, Proposition 187, which severely limited social, educational, and health benefits available to undocumented immigrants, was passed in California in 1994 after a sustained campaign that demonized Latinos. After another fierce campaign filled with hateful rhetoric, in 1997 Californians passed Proposition 227 which effectively ended bilingual education in the state. Likewise, Chavez (2013a, 2013b) has explored how the discourse around the Latino threat narrative and illegality serve to disparage Latinos, undermine their human rights, and keep them in vulnerable positions. This research can be seen as an extension of earlier work that detailed how the rhetoric around English Only laws and policies served to marginalize Latinos and question their allegiance to the U.S. As of 2015, 31 states have declared English their official language, and there are consistent calls to make English the official language nationally. Although outward hostility toward Latinos in many modern circles is deemed unacceptable, the Spanish language has become a proxy that can be attacked without seeming racist (Hill 2008). Thus, the rhetoric surrounding English-only policies demonstrates that language has become an acceptable way to express displeasure about immigrants and their growing numbers in the U.S. (Chavez 2013a; Santa Ana 2002). In this discourse, whiteness is associated with standard English and deemed the unmarked norm, while everything else is judged deviant. Latinos are cautioned against using Spanish through English-only and restrictive legislation and repression,2 yet Anglos feel no compunctions about using Spanish in public places and in fact exert their superiority and dominance by attempting to render Spanish insignificant. Hill’s (1998) work on “Mock Spanish” documents how Latinos are kept in their place through the trivialization and distortions of Spanish words. She argues that through the use of anglicized pronunciation of Spanish words and the creation of mock Spanish words by adding the vowel –o at the end (i.e. “no problemo,” 83

Lourdes Torres

“mucho bookos,” “el cheapo”), Anglos exert control and dominance. Likewise, Zentella’s (2002) work on the “chiquitafication” of Latinos and Latino languages – strategies to discredit Latinos through claims that they refuse to learn English – and the resulting stigmatization of the Spanish varieties that Latinos speak and the labeling of second-generation bilinguals as semi- or a-linguals exemplify other common strategies aimed at maintaining white supremacy.

New directions in understanding the role of Spanish and English in Latino lives In this section, I briefly present four overarching areas of scholarship that have potential to uncover additional complex instances of the disenfranchisement of Latinos that is enacted through primarily linguistic means. These include the use of engaged scholarship, a focus on Latino communities of indigenous origin, the power of language brokering and translanguaging, and the importance of examining intra-Latino relations as related to Spanish.

Engaged scholarship Across many decades, linguists and Latino studies scholars have been actively engaged in the struggle for language rights for Latinos. This advocacy is very much in the spirit of the work done by linguists such as William Labov and Geneva Smitherman, who responded to the uninformed and degrading depictions of Black English in the 1960s and 1970s, and helped educators and parents develop strategies to legitimize the language of African American children. Likewise, sociolinguists such as Shana Poplack and Ana Celia Zentella, and education scholars including Guadalupe Valdés and Ofelia García, have been active in fighting for racial equality and linguistic diversity in the Latino context. In their copious academic writings, passionate teaching, and memorable public speaking engagements, scholars such as these consistently advocate for respect for Latino communities and their language practices. Their groundbreaking work on bilingualism in the U.S., bilingual education, code-switching, language socialization, “Spanglish,” and “English-only” policies has not only shaped the perspective and approach of emerging scholars in multiple fields, but has also had a significant impact in the public sphere. They struggle to educate people in multiple fields and worlds about how language functions in social contexts and how scholars and students have a responsibility to participate in public discourse to challenge discrimination and injustice. Sociolinguists and Latino studies scholars have been at the forefront of just about every language rights issue that impacts Latinos. From the early struggle to reinforce the notion that bilingualism should be treated as a valuable asset rather than as a deficit, and that code-switching is an innovative and skillful verbal strategy, to the fight against English-only laws and toward a society that values its multilingual population, sociolinguists have fought through their rigorous, engaged scholarship and fierce participation in public debate. Some recent examples of this activism are good models for the type of engagement that continues to be necessary. The Language and Social Justice Committee of the American Anthropology Association has led the charge in a number of grueling battles and has effected substantive change. For example, they challenged the Census Bureau to change the language used to describe speakers of immigrant languages other than English, and as a result the insulting term “linguistically isolated” was eliminated as a category for individuals who do not speak English “very well.” In another example, Ana Celia Zentella and Jose Del Valle led the charge for accountability of the Real Academia Española, an organization notorious for its disrespect for U.S. Latino Spanish varieties. Although in 2012 the Academia began including the descriptor term 84

Linguistics and Latino studies

estadounidismo in its Diccionario de la lengua española (DRAE) to refer to Spanish words and usages of the U.S., the dictionary still did not include the term Spanglish. After many years and pressure from the Academia Norteamerica de la Lengua Española,3 in 2014 the Academy finally included a definition of Spanglish in the DRAE. Unfortunately, but not surprisingly, the definition was insulting: “Modalidad del habla de algunos grupos hispanos de los Estados Unidos, en la que se mezclan, deformándolos, elementos léxicos y gramaticales del español y del inglés.” Rather than recognize that Spanish in the U.S., like all varieties of Spanish, is always undergoing change, Spanglish is singled out as a deformed variety. In a cogently argued petition signed by hundreds of linguists and educators, Zentella and Del Valle wrote: Por otro, la referencia a los fenómenos asociados con el contacto español-inglés en EE.UU. como «deformaciones» muestra que, a pesar de los notables esfuerzos de las academias por limpiarse los mocos del purismo y mostrar una cara limpia y abierta, es la propia institución académica (y no los hablantes de espanglish) la que, al excluir las conclusiones de los trabajos de investigación que muestran la sistematicidad de estas prácticas lingüísticas, perpetra escandalosas tergiversaciones de la naturaleza y funcionamiento del lenguaje. www.elcastellano.org/ns/edicion/2014/abril/carta.html The petition and pressure were successful, and the Real Academia dropped “deformándolos” from their definition of “espanglish” in the DRAE. Another example is the “Drop the I-word” campaign promoted by linguistic anthropologists, which has been urging news organizations like the New York Times and the Associated Press to stop using the term “illegal” when referring to undocumented people in the U.S., since the language we use to frame a debate has a powerful effect on the debate itself. “Illegal” is a dehumanizing term that serves to create distance and demonize not only undocumented people but also all Latinos by extension. The “Drop the I-word” campaign has successfully challenged the language used in conversations around immigration in local and national newspapers, blogs, on Facebook, and other social media. This is a significant victory for as Jonathan Rosa (2012) reminds us, “While language change is not necessarily equivalent to social change, struggles over representations of immigration make it possible to imagine and enact an alternative politics of inclusion in this nation of immigrants.” Finally, “linguistic profiling” is the phrase Zentella (2014) uses to describe the restrictions placed on Spanish-speaking in public spaces. The establishment of English-only-laws has given people and institutions permission to attempt to restrict the speaking of Spanish in public places. For example, in the workspace, the number of complaints against Spanish-speakers has increased dramatically. Zentella (2014, p. 623) states: In the first 4 years after the Equal Employment Opportunity Commission (EEOC) began tracking job-related accent and language discrimination charges in 1996, there was a 500 percent increase in cases (1996: 91 vs. 2000: 447) (US Commission on Civil Rights, 2011); the annual average remains circa 460; approximately 88 percent of the English-only complaints are made by Hispanics. Clearly, this is yet another case of discrimination against Latinos. There is ample evidence that a hostile climate exists for Spanish speakers in the U.S. and that this chauvinistic attitude materializes in myriad ways (Feagin & Cobas 2014). A number of studies have examined the meaning and impact of these laws on Latino communities in the workplace 85

Lourdes Torres

and beyond (Callahan & Gándara, 2014; Cameron, 1998; Weinstein 2012). As Zentella suggests in the quote that introduces this chapter, there is a crucial role for Latino studies scholars and linguists interested in advancing language rights and human rights. Zentella and many others have provided a model of how we can insert our research and our voices in important policy debates that have significant repercussions on the lives of Latinos in the U.S.

Indigenous communities An emerging area in Latino studies is the focus on Latin American-origin Indigenous communities in the U.S. An important component is how the Indigenous family adapts linguistically to an unfamiliar milieu across generations and the extent to which younger generations are maintaining Indigenous language skills (Cassanova et al. 2016). Indigenous communities striving to maintain their languages in the U.S. face extraordinary obstacles (Makar 2012; Velasco 2010) in part because they have even fewer resources to support these languages than do Spanish speakers. Some scholars have analyzed the efficacy of bilingual education programs for Indigenous children (Collins 2012; Velasco 2010), while others have studied how Mayan speakers rely on their Indigenous languages for expressions of solidarity and to preserve their heritage (Kramsch & Whiteside 2008; Whiteside 2009, 2013). Cassanova et al. (2016) usefully analyze the linguistic decisions that Indigenous families make according to a language ecology framework that considers factors such as the multiple language ideologies Indigenous families encounter both in Mexico and the U.S., local and national beliefs about language, labor market realities, and material resources available to Indigenous families that many desire to transmit their native languages. Clearly, future work needs to focus on the dynamics of Indigenous language maintenance and shift in these growing communities.

Language brokering and translanguaging Recent studies on Latino children and their evolving language practices offer innovative ways of thinking about what it means to grow up as a Latino child with access to more than one language. This work is an excellent corrective to earlier studies that considered Latino children to be impoverished since they were often raised in a bilingual context. The work of Robert Weisskirch (2005) and others (Morales & Hanson 2005; Olmedo 2005) on language brokering finds that Latino children often serve as translators in both oral and written contexts. In a study based on self-reports, Weisskirch (2005) finds that Latino adolescents express positive feelings about the translation services they perform for their families. In fact, those who engage in language brokering tend to have more positive feelings about their Latino identity then do children who do not engage in this practice. More recent work on language brokering (Katz 2014), documents how important it is to examine the practice from a family systems approach that sheds insight on how these practices have repercussions beyond the individual. Based on ethnographic work in numerous settings in a Los Angeles community, Katz documents the positive and negative repercussions of language brokering. For instance, children can connect the immigrant family to a host of resources and introduce them to media and technology at the same time that brokering can require much time and effort at the expense of homework, which can have long-term effects on children’s education. Similarly, other work finds that when children are asked to translate sensitive or complicated medical/financial information for their families that is beyond their cognitive capacity, they can experience significant amounts of stress. This has led to requirements such as Executive Order 13166, signed by President Clinton in 2000, requiring all agencies that receive federal funding to provide translation and interpretation services as necessary. 86

Linguistics and Latino studies

In the field of bilingual education, recent work has advocated for developing the “translanguaging” skills of students. This concept, which connects linguistics, education, and Latino studies, privileges the ways of speaking of Latino communities and the richness of their linguistic repertoires, encouraging children to freely utilize all of the linguistic resources available to them. Ofelia García (2009a, 2009b, 2011) has argued for bilingual education that focuses on acknowledging and building on all the tools that children bring to school, arguing that in order to be educated for the 21st century, rather than prioritizing English and insisting on bounded, separate languages, students should be encouraged to embrace and develop all the languages in their repertoire and the multiple language practices they and their family engage in.

Inter- and intra-Latino relations Recent work concerning language and Latino communities has taken some innovative directions. Studies recognize that neither Spanish nor English, especially in their normative forms, can be said to represent the entirety of Latino communities. This research stream interrogates the identification of Spanish as emblematic of Latino identity. As De Genova and Ramos-Zayas (2003), Zentella (2002), and Rosa (2014) argue, Spanish often becomes a point of difference among Latinos of different ethnolinguistic groups due to the complexity of Spanish varieties, varying proficiencies, and the diverse valuing of varieties that exists across the Americas. In actuality, both Spanish and English are deployed by Latinos to express aspects of their identities and to forge both solidarity and difference with Latinos from other ethnolinguistic groups, as well as with U.S. white Americans, African Americans, and others. A number of studies are examining the interactions of Latinos from different ethno-nationalities and tracing how language emerges as an important tool to negotiate their relationships. De Genova and Ramos Zayas’ chapter (2003) on the language practices and ideologies of Puerto Ricans and Mexicans was one of the first to look at the role of language in intraLatino relations. They argued that rather than being a point for unification for Latinos, ideologies around the varieties of Spanish language actually serve as a way of magnifying differences between the two groups. They found in Chicago that both Puerto Ricans and Mexicans rely on hierarchizations and stereotypes of Spanish, English, and bilingualism to racialize each other and create distance. For example, some Mexicans say that Puerto Ricans do not speak Spanish well, while some Puerto Ricans criticize Mexicans for not speaking English. Puerto Ricans’ English is also found to be unacceptable as it is associated with Black English. Rosa (2014, 2015), also in Chicago, explores language ideologies in a public high school and argues against a simplistic assimilation process. He looks at how Puerto Rican and Mexican students perform their ethnolinguistic identities through linguistic practices. They also use these particular ways of speaking to create borders and boundaries around their diasporic identifications. Rosa develops the idea of “inverted Spanglish,” the use of Spanish lexical items with English phonology, to explore the ways that second- and third-generation Puerto Rican and Mexican students experiment with and mix Spanish and English to make a statement about who they are, in defiance of the ways they are expected to speak. Through the creation of inventive speech strategies, students reject the marginalization and stigmatization that is attached to their identities and language, and validate their particular variety, while marking it as different from other varieties in their environment. Rosa notes that high school students have internalized the idea that accented English is inferior, so they work to prove that they can speak unaccented English at the same time that they signal that they can understand Spanish. He finds that students who are not first-generation Spanish speakers privilege English and that, “They must signal their Latina/o identities by always sounding like they could speak Spanish in English, while carefully 87

Lourdes Torres

preventing too much Spanish from seeping into their English” (2014, p. 52). He concludes that these strategies are examples of how Puerto Rican and Mexican young people are reimagining and creating Latino panethnicity. While acknowledging that Spanish varieties can be used to signal intra-Latino conflict, Negrón (2014) is interested in examining how Latinos from different backgrounds deploy varieties of Spanish and English to express both solidarity and potential cooperation. One way they do this is by emphasizing the importance of Spanish as a potential basis for panethnic identification. Other strategies speakers use to invoke solidarity include code-switching, style-shifting, using Spanish pronunciation for Latino names, and the use of English in particular contexts to index a nonspecific Latino identity. This is an especially useful strategy in those instances when speaking in a specific ethnolinguistic Spanish variety may be taken as a sign of exclusivity. Negrón notes that “we can think of Latino ethnolinguistic repertoire as being composed of nested repertoires with features that, depending on the context and speakers’ objectives, can serve to index a specific ethnonational identity or a broader panethnic Latino identity” (2014, p. 114). Rather than just assuming it is one or the other, her microanalysis of conversations between Latinos reminds us that speakers can chose to emphasize a common Latinidad or signal exclusive identities depending on the context and the intent of the speakers. Potowski (2016) examines the ethnolinguistic identities and ideologies of three groups of Latinos in Chicago: Mexicans, Puerto Ricans, and mixed ethnicity “MexiRicans.” She finds that first-generation Puerto Ricans and Mexicans hold the most negative opinions about the other group, while second-generation individuals tend to produce negative linguistic judgments about each other, and the third generation was the least likely to offer either negative social or linguistic assessments of the other group. Also, reflecting internalized stigmatization of their own variety, the people most critical of Puerto Rican Spanish were Puerto Ricans themselves across all three generations. MexiRican participants, as embodiments of features of both Mexicans and Puerto Ricans, were also less likely to reproduce negative linguistic judgments about the speakers of any particular linguistic variety. Across the three generations, there seems to be a decline in the belief that one variety was better than any other. Potowski (2016) asserts that subsiding negative linguistic assessments might be due to the reality that for these speakers, as suggested by Zentella (2002), the Spanish language is not an essential component of Latino identity. In an earlier piece, Potowski (2012) summarized some of the ways in which Latinos discriminate against each other based on the kind of Spanish they use, how proficient they are in it, and indeed whether they speak Spanish at all. Finally, recent studies on language and identity are investigating the links between language and other dimensions of identity between different groups of Latinos. For example, Cashman (2015; Cashman & Trujillo in Chapter 9 this volume) analyzes ethnographic interviews with bilingual queer individuals from Phoenix, Arizona, finding intersections of Latino ethnic identity, sexual identity, and language practices. Specifically, she discusses how the act of coming out or maintaining silence about sexual identity impacts speakers’ access to the Spanish language and thus conditions their language practices. She argues that since it is the first-generation migrants who tend to speak Spanish and are most likely to have issues with homosexuality, if Latinos come out and are alienated from their Spanish dominant family members, this likely promotes a shift to English. While emerging Latino studies and sociolinguistic research begins to respond to the challenge put forth by Zentella (2014), many more studies are needed, especially at a time when Latinos and immigrants continue to be vilified and marginalized. As we have seen, the best of such linguistic and Latino studies research can serve to assess community language practices within a socio-political context, unmask linguistic injustices, and advance liberatory linguistic policies. 88

Linguistics and Latino studies

Notes 1 Two additional language-based areas with relevance to Latino studies, but not described here, include the maintenance and loss of Spanish among U.S. Latinos and the teaching of Spanish as a heritage language. These issues are addressed in Chapters 4 and 20 (respectively) of the present volume. 2 Recent cases include drivers being issued tickets for “not speaking English” and employees hired for knowing Spanish and then fired for speaking it on the job (see http://potowski.org/resources/repression for a list). 3 However, this organization is also guilty of unapologetically shaming U.S. Latinos for their ways of speaking Spanish; see http://potowski.org/hablando_bien, and Lynch and Potowski (2014).

References Acosta-Belén, E. (1975). Spanglish: A case of languages in contact. In Burt, M., & Duly, H. (Eds.) New Directions in Second Language Learning, Teaching and Bilingual Education, Washington, DC: TESL, 151–158. Anzaldúa, G. (1987). Borderlands: La frontera. San Francisco, CA: Aunt Lute. Aparicio, Frances R. (2003). Latino cultural studies. Critical Latin American and Latino Studies 12, 3–31. Aparicio, Frances R. (1999). Reading the ‘Latino’ in Latino studies: Toward re-imagining our academic location. Discourse 21(3), 3–18. Bills, Garland. (1975). Linguistic research on United States Hispanic: State of the art. In Teschner, R. V., Bills, G. D., & Craddock, J. R. (Eds.) Spanish and English of United States Hispanos: A Critical, Annotated Linguistic Bibliography. Arlington, VA: Center for Applied Linguistics. Cabán, P. A. (2003). Moving from the margins to where? Three decades of Latino/a studies. Latino Studies 1(1), 5–35. Callahan, R. M., & Gándara, P. C. (2014). Contextualizing bilingualism in the labor market: New destinations, established enclaves and the information age. In Callahan, R. M., & Gándara, P. C. (Eds.) The Bilingual Advantage: Language, Literacy and the US Labor Market (Bilingual Education and Bilingualism) 99. Bristol, UK: Multilingual Matters. Cameron, C. D. R. (1998). How the Garcia cousins lost their accents: Understanding the language of Title VII decisions approving English-only rules as the product of racial dualism, Latino invisibility, and legal indeterminacy. La Raza LJ 10, 261–307. Cashman, H. (2015). Intersecting communities, interwoven identities: Questioning boundaries, testing bridges, and forging a queer latinidad in the US Southwest. Language and Intercultural Communication 15(3), 424–440. Cassanova, S., O’Connor, B. H., & Anthony-Stevens, V. (2016) Ecologies of adaptation for Mexican Indigenous im/migrant children and families in the US: Implications for Latino studies. Latino Studies 14(2), 192–213. Chavez, L. R. (2013a). The Latino threat: Constructing immigrants, citizens, and the nation. Stanford, CA: Stanford University Press. Chavez, L. R. (2013b). ‘Illegality’ across generations: Public discourse and the children of undocumented immigrants. In Menjívar, C., & Kanstroom, D. (Eds.) Constructing Immigrant ‘Illegality’: Critiques, Experiences, and Responses. Cambridge, UK: Cambridge University Press, 84–110. Collins, J. (2012). Migration, sociolinguistic scale, and educational reproduction. Anthropology & Education Quarterly 43(2), 192–213. Darder, A. (2004). The politics of language: An introduction. Latino Studies 2(2), 231–236. De Genova, N., & Ramos-Zayas, A. Y. (2003). Latino Crossings: Mexicans, Puerto Ricans, and the Politics of Race and Citizenship. New York: Routledge. Dittmar, N. (1996). Descriptive and explanatory power of rules in sociolinguistics. In Singh, R. (Ed.) Current Issues in Linguistic Theory (Towards a Critical Sociolinguistics, Vol. 125). Philadelphia, PA: John Benjamins Publishing Company, 115–150. Escobar, A. M., & Potowski, K. (2015). El español de los Estados Unidos. Cambridge, UK: Cambridge University Press. Feagin, J. R., & Cobas, J. A. (2014). Latinos facing racism: Discrimination, resistance, and endurance. Boulder, CO: Paradigm Publishers. Flores, B. (2005). The intellectual presence of the deficit view of Spanish-speaking children in the educational literature during the 20th century. In Pedraza, P., & Rivers, M. (Eds.) Latino Education: An Agenda for Community Action Research. New York: Routledge, 75–98.

89

Lourdes Torres Flores, J. (1997). Latino studies: New contexts, new concepts. Harvard Educational Review 67(2), 208–222. García, O. (2009a). Bilingual Education in the 21st Century: A Global Perspective. Malden, MA and Oxford, UK: Wiley-Blackwell. García, O. (2009b). Education, multilingualism and translanguaging in the 21st century. In Mohanty, A., Panda, M., Phillipson, R., & Skutnabb-Kangas, T. (Eds.) Multilingual Education for Social Justice: Globalising the Local. New Delhi, India: Orient Blackswan (formerly Orient Longman), 128–145. García, O. (2010). Monoglossic ideologies and language policies in the education of U.S. Latinas/os. In Murillo, E. G. Jr., Villenas, S. A., Galván, R. T., Muñoz, J. S., Martínez, C., & Machado-Casas, M. (Eds.) Handbook of Latinos and education: research, theory and practice. New York: Routledge, 157–169. García, O. (2011). From language garden to sustainable languaging: Bilingual education in a global world. Perspectives 34(1), 5–9. Hill, J. H. (1998). Language, race, and white public space. American Anthropologist 100(3), 680–689. Hill, J. (2008). The Everyday Language of White Racism. New York: Wiley-Blackwell. Hymes D. (1996). Report from an underdeveloped country: Toward linguistic competence in the United States. In Singh, R. (Ed.) Current Issues in Linguistic Theory (Towards a Critical Sociolinguistics, Vol. 125). Philadelphia, PA: John Benjamins Publishing Company, 150–195. Katz, V. S. (2014). Kids in the Middle: How Children of Immigrants Negotiate Community Interactions for Their Families. New Brunswick, NJ: Rutgers University Press. Klein, F. (1980). A quantitative study of syntactic and pragmatic indications of change in the Spanish of bilinguals in the US. In Labor, W. (Ed.) Locating Language in Time and Space. New York: Academic Press, 69–82. Klein-Andreu, F. (1986). La cuestión del anglicismo: apriorismos y métodos. Thesaurus: Boletín del Instituto Caro y Cuervo 40, 1–16. Kramsch, C., & Whiteside, A. (2008). Language ecology in multilingual settings. Towards a theory of symbolic competence. Applied Linguistics 29(4), 645–671. Language Policy Task Force. (1982). Intergenerational Perspectives on Bilingualism: From Community to Classroom. New York: Centro de Estudios Puertorriqueños. Language Policy Task Force. (1988). Speech and Ways of Speaking in a Bilingual Puerto Rican Community. New York: Centro de Estudios Puertorriqueños. Lavandera, B. (1981). Lo quebramos, but only in performance. In Duran, R. P. (Ed.) Latino Language and Communicative Behavior. Norwood, NJ: Ablex, 45–67. Lynch, A., & Potowski, K. (2014). La valoración del habla bilingüe en los Estados Unidos: Fundamentos sociolingüísticos y pedagógicos en Hablando bien se entiende la gente. Hispania 97(1), 32–46. Makar, C. (2012). Building communities through bilingual education: The Asociación Tepeyac of New York. In García, O., Zakharia, Z., & Otcu, B. (Eds.) Bilingual Community Education and Multilingualism: Beyond Heritage Languages in a Global City (Vol. 89). Bristol, UK: Multilingual Matters, 45–59. Mendoza-Denton, N. (2014). Homegirls: Language and Cultural Practice among Latina Youth Gangs. New York: John Wiley & Sons. Morales, A., & Hanson, W. E. (2005). Language brokering: An integrative review of the literature. Hispanic Journal of Behavioral Sciences 27(4), 471–503. Negrón, R. (2014). New York City’s Latino ethnolinguistic repertoire and the negotiation of latinidad in conversation. Journal of Sociolinguistics 18(1), 87–118. Olmedo, I. M. (2005). The bilingual echo: Children as language mediators in a dual-language school. In Farr, M. (Ed.) Latino Language and Literacy in Ethnolinguistic Chicago. New York: Routledge, 135–153. Poplack, S. (1980). Sometimes I’ll start a sentence in Spanish y termino en español: Toward a typology of code-switching. Linguistics 18(7–8), 581–618. Potowski, K. (2012). SHL identity and heritage learners: Moving beyond essentializations. In Beaudrie, S. M., & Fairclough, M. (Eds.) Spanish as a Heritage Language in the United States: The State of the Field. Washington, DC: Georgetown University Press, 161–178. Potowski, K. (2016). Inter-Latino language and identity: MexiRicans in Chicago. New York: John Benjamins. Pousada, A., & Poplack, S. (1979). No case for convergence: The Puerto Rican Spanish verb system in a language contact situation (Vol. 5). New York: Language Policy Task Force, Centro de Estudios Puertorriqueños, CUNY. Romaine, S. (1996). The status of sociological models and categories. In Singh, R. (Ed.) Current Issues in Linguistic Theory (Towards a Critical Sociolinguistics, Vol. 125). Philadelphia, PA: John Benjamins Publishing Company, 99–114.

90

Linguistics and Latino studies Rosa, J. (2012). Contesting representations of immigration. Anthropology News, www.anthropology-news. org/index.php/2012/10/09/contesting-representations-of-immigration/. Rosa, J. (2014). Learning ethnolinguistic borders: Language and diaspora in the socialization of US Latinas/ os. In Rolón-Dow, R., & Irizarry, J. G. (Eds.) Diaspora Studies in Education: Toward a Framework for Understanding the Experiences of Transnational Communities. New York: Peter Lang, 39–60. Rosa, J. (2015). Nuevo Chicago? Language, diaspora and Latino/a panethnic formations. In Reiter, R. M., & Rojo, L. M. (Eds.) A sociolinguistics of diaspora: Latino practices, identities, and ideologies (Vol. 6). New York: Routledge. Sánchez, R. (1983). Chicano Discourse. Rowley, MA: Newbury House Publishing. Santa Ana, O. (2002). Brown tide rising: Metaphors of Latinos in contemporary American public discourse. Austin: University of Texas Press. Singh, R. (Ed.) (1996). Current Issues in Linguistic Theory (Towards a Critical Sociolinguistics, Vol. 125). Philadelphia: John Benjamins Publishing Company Teschner, R. V., Bills, G. D., & Craddock, J. R. (1975). Spanish and English of United States Hispanos: A Critical, Annotated Linguistic Bibliography. Arlington, VA: Center for Applied Linguistics. Torres, L. (1991). The study of US Spanish varieties: some theoretical and methodological issues. Sociolinguistics of the Spanish-Speaking World: Iberia, Latin America, United States. Tempe, AZ: Bilingual Press/Editorial Bilingüe, 255–270. Urciuoli, B. (1997). Exposing prejudice: Puerto Rican experiences of language, race and class. Boulder, CO: Westview Press. Velasco, P. (2014). The Language and Educational Ideologies of Mixteco-Mexican Mothers. Journal of Latinos and Education, 13(2), 85–106. Weinstein, L. M. (2012). The role of labor law in challenging English-only policies. Harvard Civil RightsCivil Liberties Law Review (CR-CL), 47, 219–249. Weisskirch, R. S. (2005). The relationship of language brokering to ethnic identity for Latino early adolescents. Hispanic Journal of Behavioral Sciences, 27(3), 286–299. Whiteside, A. (2009). ‘We Don’t Speak Maya, Spanish, or English’: Yucatec Maya-Speaking Transnationals in California and the Social Construction of Competence. The Native Speaker Concept: Ethnographic Investigations of Native Speaker Effects, ed. N. Doerr, 215–231. Whiteside, A. (2013). Research on transnational Yucatec Maya-speakers negotiating multilingual California. Journal of Applied Linguistics and Professional Practice, 1(1), 103–112. Zentella, A. C. (1997). Growing up bilingual: Puerto Rican children in New York. New York: Wiley-Blackwell. Zentella, A. C. (2002). Latin@ languages and identities. In Suárez-Orozco, M. and Páez, M. (Eds.) Latinos! An agenda for the 21st century. Berkeley: University of California Press, 321–338. Zentella, A. C. (Ed.). (2005). Building on strength: Language and literacy in Latino families and communities. New York: Teachers College Press. Zentella, A. C. (2009). Multilingual San Diego: Portraits of language loss and revitalization. San Diego, CA: University Readers. Zentella, A. C. (2014). TWB (Talking while Bilingual): ‘Linguistic profiling of Latina/os, and other linguistic torquemadas. Latino Studies, 12 (4), 620–635.

91

7 SPANISH AND IDENTITY AMONG LATIN@S IN THE U.S. Rachel Showstack wichita state university

Introduction Jessica describes herself as a ‘Spanglish’ person. In her home community in Houston, she speaks Spanish to represent her Mexican identity and align with other Spanish speakers. At home in conversation with her Mexican parents and U.S.-born sister, she switches back and forth between Spanish and English depending on the person to whom she is speaking. In her Spanish class for heritage speakers, she only speaks Spanish and insists that her classmates do the same, but on the university campus outside of the classroom, she almost always interacts in English, even with her friends who speak Spanish. The value and social meaning associated with Spanish and Spanish-English bilingual practices is different in each of the contexts in which Jessica uses language, and Jessica brings with her to these contexts a history of how others have evaluated or responded to her language use in other similar contexts. For example, at a conference on diversity that she attended when she was in high school, Jessica was introduced to a girl from Spain who laughed at the way she spoke and called it ‘broken Spanish.’ When she talks with her cousins who live in Mexico, they say she talks like an American. Back home in Houston, she does not receive these kinds of evaluations of her language—her use of Spanish allows her to represent a desired linguistic identity in that context. Jessica’s desire to speak ‘correctly’ motivates her in her formal study of the language, but may affect her language choice in other contexts outside of her home community. Research on language and identity among U.S. Latin@s has addressed the value and social meaning of ‘Spanish’ and ‘Spanglish’ in a range of social contexts in diverse bilingual communities in many parts of the country. This chapter explores the development of a theoretical approach to language and identity, the different types of research that have been conducted on Spanish and identity among heritage speakers, and the findings of such research in both community and educational contexts. I also consider the implications of this research for the education of Spanish-English bilinguals in the U.S., and I argue that there is a need for research that examines heritage speakers’ individual histories and considers each context as a place where multiple discourses intersect. 1

Historical perspectives The social meaning and value of Spanish in the U.S. has changed as a result of social and historical developments over the last two centuries. During the colonial period, Spanish was a dominant 92

Spanish and identity among Latin@s in U.S.

language in the area that is now considered the Southwest. After 55% of what was Mexican territory became part of the U.S. in the Treaty of Guadalupe Hidalgo in 1848, English became the official language of this region. Bilingualism was tolerated during an initial period, but a ‘subtractive’ view of Americanization, which emphasized complete conformity to U.S. culture, began to develop during World War I. This led to an English-only mentality in public schools, and Latin@ children were often punished for speaking Spanish (Blanton, 2004). During the 1960s and 1970s, inspired by the civil rights movement and the unionizing efforts in California led by labor activist César Chávez, many Latin@s in the U.S. began to fight for their political representation and social acceptation. What came to be known as the Chicano movement focused on the rights of the working class, the maintenance of Mexican culture, and resistance to exploitation and political domination (Orozco, 2009). During this period, the Spanish language held important symbolic value in Chicano activist groups. For example, in El Movimiento Estudiantil Chicano de Aztlán (MEChA), Spanish symbolized a cultural unity tied to Chicano indigenous heritage (Limón, 1982). In her seminal essay How to Tame a Wild Tongue (first published in 1987), Chicana activist Gloria Anzaldúa explored the connection between language and identity for Spanish speakers from the U.S. Southwest. For Anzaldúa, speakers may use ‘Chicano Spanish,’2 to represent a Chicano identity when interacting with other Chicanos, but may select from a range of other varieties of Spanish and English depending on the interlocutor and the identity they choose to represent. Anzaldúa also pointed out that because dominant societal discourses devalue Chicano Spanish, some speakers feel uncomfortable or even shameful when speaking that variety, and their interlocutors often make judgments about them based on the way they speak (Anzaldúa, 2007). During the period of the Chicano movement, the U.S. Census Bureau also constructed a link between the Spanish language and Latin@ identity; the classification of ‘Spanish mother tongue’ was used, generally in conjunction with surnames and place of birth, to identify the ‘Hispanic origin’ population (Leeman, 2013). While this change in Census procedures reflected an interest in more accurate representation of Latin@s in U.S. Census data, it is reminiscent of an essentialized (or overgeneralized) link between the Spanish language and Latin@ identity that has prevailed in public discourses, allowing xenophobia to take the form of anti-multilingualism. While national educational discourses mirrored the Movimiento’s focus on a link between Spanish and Latin@ identity in the 1960s and 1970s, global political and economic developments in the 1980s and 1990s led to a view of Spanish as a commodity for success in business and international relations (Leeman & Martínez, 2007). Leeman and Martínez (2007) found that these two discourses are reflected in the Spanish heritage language (HL) textbooks published during each period, respectively. They point out that in the context of teaching Spanish as an HL, the dominant discourses of both periods represented limitations to the social value of Spanish in the U.S.; in the early period, Spanish was linked to the home, denying the public and political nature of the language, while in the later period a focus on Spanish as an international language led to a delegitimization of local linguistic practices. Both the essentialized link between the Spanish language and Latin@ identity, and the view of Spanish as an international (and not local) language, have helped to fuel the ‘English-only’ movement, which has particularly targeted U.S. Latin@s (Dueñas González, 2000). The 1980s and 1990s saw a wave of legislation restricting language in official contexts in the U.S. The ‘English-only’ movement began in 1983 when Senator S.I. Hayakawa of California founded an organization called ‘U.S. English’ to advocate for an amendment to the U.S. Constitution that would make English the ‘official language’ of the U.S. Since 1983, legislation to establish English as the official language of the U.S. has been proposed several times in Congress; the 93

Rachel Showstack

most recent version of such legislation, the ‘English Language Unity Act’ was written in 2007 and has been proposed in every Congress since then (Tamasi & Antieau, 2014; U.S. English, 2015). Such an amendment has not been passed, nor is it likely that it will be passed at the federal level. Nonetheless, Senator Hayakawa’s proposal has inspired state and municipal versions, and 31 states have adopted official English legislation since 1983. In 1998, California passed Proposition 227, which legislated that instruction be given only in English in classrooms, effectively ending the bilingual education system in public schools. The ‘English-only’ movement reflects a series of myths about bilingualism in the U.S., including the idea that the use of foreign languages will fragment and divide the nation, and an assumption that minority groups refuse to learn English. Zentella (1997b) argues that the movement also reflects and promotes a negative portrayal of Spanish speakers in the national public discourse. (See Chapter 3 this volume for a discussion of the current linguistic climate in the U.S.) Contemporary perspectives on language and identity can help researchers understand how these dominant discourses play into the ways that Latin@ individuals position themselves through language use and represent themselves as particular kinds of language users. In the next section, I discuss the critical issues and topics that have emerged in scholarly research on language and identity in general and with respect to Spanish speakers in the U.S.

Critical issues and topics Joshua Fishman, who pioneered research on minority languages, language and ethnicity, and language contact situations, is known for a view that all individuals have an inherent emotional and spiritual connection with their native language or the language of their immediate ancestors (Myhill, 1999). This perspective, which Myhill (1999) calls the ‘language-and-identity ideology,’ is evident in Fishman’s later writing: “This soul (the essence of a nationality) is not only reflected and protected by the mother tongue, but, in a sense, the mother tongue is itself an aspect of the soul, a part of the soul, if not the soul made manifest” (Fishman, 1989:276, emphasis in original, as cited in Myhill, 1999). Myhill points out that much research in the sociology of language has equated language with ethnocultural identity. The language-and-identity ideology is also evident in work within the social-psychological paradigm (e.g. Tajfel, 1974, 1981). For example, Giles and Byrne (1982) propose a framework that assumes a strong correlation between bilinguals’ identification with particular ethnolinguistic groups and their level of language maintenance. The problem with the correlational approach, as He (2010) sees it, is that it tends to evaluate complex and evolving constructs such as motivation, attitude, ethnic identity, proficiency, and literacy in terms of numerical values, and leads one to think that these sociocultural traits are essential, ‘built-in,’ and unchanging qualities (2010:71). He (2010) criticizes these approaches because they ascribe to a perspective that views identity as something that is much more dynamic and contextually relevant. Contemporary approaches in linguistics, anthropology, and educational research view identity as a form of social action that is constantly being created and recreated, shaped by both the backgrounds that people bring with them to a social situation, and on the context of the social situation itself (Bucholtz & Hall, 2004; Pavlenko & Blackledge, 2004). In this view, identity is constituted by the values and assumptions of particular social contexts, and by societal belief systems, which offer positions of power to certain categories of individuals and not to others, and at the same time, it is negotiated and created through interaction (Davies & Harré, 1990). To understand the interplay between the actions of individuals and the structure of communities, societies, and institutions, researchers draw from a range of theoretical frameworks. In this section, I will discuss four areas of theory that often contribute to current studies on language 94

Spanish and identity among Latin@s in U.S.

and identity: language ideologies, identity construction and negotiation, investment and imagined communities, and hybridity and transcultural practices.

Language ideologies People’s identities can be shaped in part by assumptions that associate certain types of social meaning with the use of a certain language, or with certain types of language use. These assumptions, called ‘language ideologies’ (Kroskrity, 2004; Woolard, 1998), can include dominant societal discourses, or prevailing societal beliefs and academic assumptions (i.e. the language-and-identity ideology), and they can also describe assumptions about language that are shared within a particular group of speakers. In the case of Spanish in the U.S., dominant discourses represent Spanish as inferior to English and at the same time devalue the Spanish spoken by U.S. Latin@s, giving second language (L2) learners more credit for their Spanish skills than bilinguals receive for theirs (Pomerantz, 2002). This is because of prevailing discourses in Spanish language education that construct Spanish as a ‘foreign’ language despite the long-standing presence of Spanish in the U.S., and devalue the language varieties commonly spoken by U.S. Latin@s (Leeman, 2014; Pomerantz & Schwartz, 2011; Valdés et al., 2003), and a societal orientation that positions the linguistic skills of minority language speakers as a ‘problem’ (Ortega, 1999; Ruiz, 1984). Textbooks used in Spanish L2 courses have traditionally acknowledged U.S. Spanish speakers only minimally (Gutiérrez & Fairclough, 2006), and research on the perspectives of Spanish L2 learners in the U.S. demonstrates that these learners often position themselves as linguistically superior to the native Spanish speakers with whom they are acquainted in domestic contexts (Pomerantz, 2010; Pomerantz & Schwartz, 2011). On the other hand, some communities of Spanish speakers in the U.S. associate the use of Spanish or Spanish-English code-switching with group membership (Anzaldúa, 2007; Mendoza-Denton, 2008; Zentella, 1997a). Research on language and identity often draws on Bourdieu’s notion of ‘symbolic power,’ a power to influence other people’s understandings of the world, which is shaped by the value that certain dispositions hold in a particular context, or ‘symbolic marketplace’ (Bourdieu, 1977, 1991). Several studies exploring the relationship between language ideologies and the Spanish used by Latin@s in the U.S. have drawn from Bourdieu in their analyses (e.g. Lowther Pereira, 2010; Showstack, 2012). For example, in a study on symbolic power in the HL classroom, Showstack (2012) explored the language ideologies implicit in classroom interaction and examined how such discourses played into the ways students represented themselves as Spanish speakers and Latin@s in classroom language use. The analysis focused on apparent relations of power as reflected in the content of participants’ utterances, and assumptions that the participants articulated about the identities of particular types of speakers. She found that some of the participants oriented toward an elite identity in which speaking Spanish ‘correctly’ was valued, claiming symbolic power in their own Spanish, and criticized people who did not fit into that category, positioning them as less powerful. In a later study, Showstack (2015a) found that some HL students who claimed symbolic power in the language skills they had learned in the classroom also identified monolingual Spanish speakers as the best judges of their own language skills, reproducing hegemonic discourses that position Spanish as a ‘foreign’ language, while other students constructed positions of expertise outside of the dominant paradigm. Showstack’s data suggest that students’ individual histories of language use played an important role in whether they positioned themselves as legitimate Spanish speakers in the symbolic marketplace of the classroom and in other contexts. García and Torres-Guevara (2010) criticize language policies in the education of Latin@s in the U.S. for promoting a view of language as “an autonomous skill that functions independently 95

Rachel Showstack

from the context in which it is used” (2010:182). This perspective is part of what the authors call a ‘monoglossic language ideology,’ which values monolingualism and ignores the bilingual repertoires and practices of U.S. Latin@s. Monoglossic language ideologies in bilingual and HL education contradict current theory in applied linguistics and sociolinguistics that describes language as a social practice that emerges in specific contexts (Blommaert, 2005). In the classroom, monoglossic ideologies may limit opportunities for the creation of social meaning through hybrid language practices that reflect the multilingual lives of U.S. Latin@s.

Identity construction and negotiation Ideologies like those described earlier provide conversation participants with frames for understanding the meaning of language use in a particular context, and they can also shape the ways that Spanish speakers represent their identities and the identities of others when they narrate their experiences with language and use language in interaction. On the other hand, individuals can also exercise individual agency and challenge prevailing assumptions about identity through their language use. When people use language in a way that others associate with a certain type of identity, they are ‘indexing’ that identity with their language use. ‘Indexicality’ refers to the use of a sign, such as a linguistic form, to point directly or indirectly to a particular behavior, point of view, attitude, or social position (Ochs, 1993). For example, a student who makes an effort to enforce the use of ‘Standard Spanish’ in the context of the HL classroom, correcting others’ language when they produce calques and Anglicisms, may be indexing an expert identity. However, understanding the ways that people index social identities is not as simple as associating certain linguistic forms with certain types of people; the indexing of identity may happen indirectly (He, 2004; Ochs, 1993). In the case of the HL classroom, an expert identity is associated with a particular set of linguistic and interactional practices. In order for interlocutors to understand the indexical reference to an expert identity, they must have an understanding of how the expert identity is structured within the discursive practices of that classroom. Individuals do not always represent social positions by reflecting the existing assumptions about language and identity; they can also challenge these assumptions through ‘performance.’ Although this term has been used with diverse meanings in linguistic anthropology, it often refers to a highly deliberate social display that challenges or subverts dominant ideologies (Bucholtz & Hall, 2004). One way that this happens is through a process called ‘denaturalization,’ in which “identities come to be severed from or separated from claims to ‘realness’” (Bucholtz & Hall, 2004:386). For example, Showstack (2012, 2016) demonstrates how Spanish heritage speakers in HL courses denaturalize essentialized constructs of bilingual identity by displaying exaggerated representations of the classroom monolingual standard norm of language use (e.g. by translating the name of a Texas grocery store to Spanish when speaking Spanish with classmates who live in Texas). The notion of ‘performance’ described here differs from Butler’s (1990) concept of ‘performativity,’ which focuses on the production of gender as a reiteration of hegemonic practices. In addition to representing themselves as particular types of people, individuals can also ascribe social positions to others. For example, instructors in multilingual contexts can ascribe novice positions to students by correcting and evaluating their language use, and they can also position them as experts by acknowledging the knowledge and experiences they bring with them to the classroom (He, 2004; Palmer et al., 2014; Showstack, 2015b). The identities that people represent and ascribe to others are often negotiated in interaction, or ‘co-constructed’ (Jacoby & Ochs, 1995). 96

Spanish and identity among Latin@s in U.S.

Investment and imagined communities The ways in which language ideologies influence the identities and practices of individuals are related to not only the cultural contexts of language use but also the histories and life situations of individual leaners. Peirce (1995) proposed that language learners who feel that using the target language will bring them social return have ‘investment’ in the language and are therefore more likely to represent themselves as ‘legitimate speakers’ by choosing to use the language in certain situations in which dominant power relations might not position them as legitimate. This perspective has shed light on the language practices of Spanish-English bilingual students in dual language bilingual programs, a context in which both Spanish and English language skills hold a certain value and language learning is an important goal for some students. Research on language use in dual language classrooms has suggested that students’ language choice in different contexts and their willingness to take risks with language reflect their social investments, including not only investment in learning an additional language, but also their investment in achieving a range of different types of social positions in the classroom (Mateus, 2014; Potowski, 2004). In subsequent work, Norton points out that the ways language learners perceive communities of speakers of the target language with which they are not directly affiliated can affect their learning trajectories (Kanno & Norton, 2003; Norton, 2001). Originally proposed by Anderson (1991:241) the concept of ‘imagined communities’ refers to “groups of people, not immediately tangible and accessible, with whom we connect through the power of the imagination” (ibid.). While Anderson’s discussion of imagined communities focused on the concept of ‘nation,’ Kanno and Norton (2003) expand the concept to include a broader range of conceptualizations of possible future communities for language learners. For example, a learner’s imagined community could include a professional community into which she or he hopes to become integrated, or a transnational community. Norton (2010) points out that the communities represented by language instructors are often disconnected from the ways in which learners see themselves using the target language in the present and the future, and argues that this disconnect may affect learners’ investment in the language. For example, in an ethnographic study on a Spanish HL class in a charter high school in the U.S. Southwest, Helmer (2013, 2014) found that students resisted the ways instructors positioned them within imagined communities through classroom interaction and choice of pedagogical materials. One instructor chose to use foreign language teaching materials, which did not seem to acknowledge the students’ identities as target language insiders (Helmer, 2014); another instructor, attempting to let students know that she valued them as a linguistic resource, characterized them as ‘Pachuco,’ unintentionally ascribing to them a ‘gangster’ identity (Helmer, 2013:280).

Hybridity and transcultural practices ‘Hybridity’ is a theoretical concept that has been important in contemporary research on identity in multilingual contexts. The term is often used to describe the ways that speakers use language to simultaneously draw from multiple social practices to represent new and alternative identities and, in educational contexts, create new opportunities for learning (Gutiérrez et al., 1999). The combination of two languages, or multiple varieties of a language, has been identified as an important aspect of identity and social meaning among Latin@s in the U.S. (Martínez, 2013; Zentella, 1997a). The hybrid language practices of Spanish-English bilinguals have been described in the seminal works of Anzaldúa, who wrote about her personal experiences as a Chicana growing up in the Texas borderlands (Anzaldúa, 2007), and of Zentella, who 97

Rachel Showstack

did an ethnographic study on bilingualism as a social practice in a Puerto Rican neighborhood in New York, among others (Zentella, 1997a). Current research on language and globalization that addresses the ways in which people negotiate identities across several cultures and geographical locations simultaneously often uses terms such as ‘transnational’ and ‘transcultural’ that highlight the fluidity of languages and cultures in a globalized world (Back, 2015). Farr (2006) studied a community of Mexican-origin families living in both Michoacán, Mexico and Chicago and examined the ways in which they used language to represent local identities. While Farr uses the term ‘transnational’ to describe this community, Back (2015) chooses the terms ‘translingual’ and ‘transcultural’ to describe a group of Ecuadorian musicians who travel abroad to perform music. Back’s participants draw from multiple languages and cultures to perform gendered and ethnic identities, drawing on different ‘historicities,’ or shared cultural histories, depending on their interlocutors and audiences. Drawing on the work of scholars from applied linguistics such as Canagaraja (2013), Back points out that the term ‘transnational’ is often used to make reference to a concept of two or more static cultures, while ‘translingual’ and ‘transcultural’ better describe the ongoing production and negotiation of multiple identities among her participants.

Language and identity in specific contexts Research on the social meaning of Spanish among Latin@s in the U.S. has addressed language practices and perceptions in a variety of contexts. This section reviews research that has been conducted in family, community, and work contexts and discusses the role of HL and bilingual education in transforming Latin@ students’ linguistic identities.

Family, community, and work contexts Zentella’s (1997a) ethnographic study of the linguistic practices within a Puerto Rican community in New York called El Bloque demonstrated how individual bilingual speakers create social meaning through different kinds of language use. Zentella’s participants tended to use more Spanish with adults in the community and more English with their peers, but they frequently practiced code-switching when interacting with other bilinguals. One of the primary reasons that her participants used code-switching was to construct an in-group identity. The study also demonstrated that the juxtaposition of English and Spanish in particular moments of interaction was used for a range of interactional purposes, and that participants chose from a variety of different language varieties (and different kinds of language mixing) depending on the social context. Among Latin@ gangs in California, Spanish and particular uses of English can represent not only Latin@ identity but also gang membership. In an ethnographic study on the linguistic and bodily practices (i.e. dress and gesture) of young Latin@ high school students who belonged to two different gangs, the Norteños and the Sureños, in northern California, Mendoza-Denton (2008) found that her participants indexed membership in each gang through the use of Spanish and English and the production of particular linguistic features, such as discourse markers and the pronunciation of certain vowels. Spanish was associated with the Sureños and English was associated with the Norteños, but members of each group made conflicting claims about the other group’s language practices and ideologies. MendozaDenton observed that, while both groups claimed to produce monolingual speech in one language or the other, all of her study participants produced language that exhibited features of bilingual speech. 98

Spanish and identity among Latin@s in U.S.

In contrast to the urban contexts of Zentella’s and Mendoza-Denton’s studies, Torrez (2013) examined the relationship between language and identity among working-class Mexicandescent families in rural Michigan. She investigated the meaning and value that members of these families associated with a variety they called Mexicano (similar to Chicano Spanish) and found that Mexicano legitimated the experiences and traditions of members of Mexicano communities, setting them apart from recent immigrants, and it served as a resource for communicating with and supporting fellow farmworkers. In addition to creating solidarity within a specific Spanish-speaking community, a speaker’s choice to use Spanish, and his or her choice of linguistic features, can be a move to represent a sense of solidarity with other Latin@s of different ethnolinguistic backgrounds. Research on the language use of Central Americans in the U.S. shows that Salvadorans and Hondurans often avoid using the personal pronoun ‘vos’ when interacting with others who are not firstgeneration immigrants from Central America (Raymond, 2012; Woods & Rivera-Mills, 2012). Negrón (2014) examined the language use of eleven Latin Americans from Queens, New York and found that they used linguistic and discursive strategies to align with their Latin@ interlocutors by downplaying ethnic differences and constructing a sense of solidarity. She focused on one conversation between two of her participants, ‘Roberto’ and ‘William,’ to demonstrate how they used diverse strategies, including the production of features from different varieties of Spanish and code-switching between English and Spanish, to align with each other and negotiate their shared Latin@ identity. However, Spanish speakers do not necessarily need to draw from multiple dialects of Spanish in order to identify with more than one Latino group. In a study on Latin@s with one Mexican parent and one Puerto Rican parent in Chicago, Potowski (2014) found that the majority of her participants spoke a variety of Spanish that was phonologically either Mexican or Puerto Rican, and yet they claimed to be equally Mexican and Puerto Rican and challenged others who questioned their membership in either group. Another area that has been addressed in studies on language and identity among Latin@s in community contexts is the relationship between language proficiency and ethnic identity. Research indicates that some Latin@ groups do perceive language proficiency as linked to ethnic identity. For example, in a study on the language socialization practices of Mexican-descent families in California, Pease-Alvarez (2002) found that for Mexican immigrant parents, teaching Spanish to children and socializing them to use the language at home was a way to overcome threats to their children’s Mexican identity. However, several studies have suggested that certain groups of bilinguals in the U.S. do not consider a high level of proficiency in Spanish to be necessary for the representation of Latin@ identity or membership in a particular Latin@ group (Pedraza, 1985; Toribio, 2000; Zentella, 1997a). This perspective is reflected in Koike and Graham’s (2006) analysis of a Spanish-language political debate between Dan Morales and Tony Sánchez, two candidates who were competing for the Texas Democratic Party’s nomination for the governor of the state of Texas in 2002: both candidates were heritage speakers of Spanish, but while Sánchez emphasized his fluency in Spanish, Morales emphasized his identity as an English-speaking Hispanic and appeared to represent a view that people with a dual Hispanic and U.S. identity do not have to speak Spanish to be Hispanic. DuBord (2014) points out that it is not always desirable for Spanish speakers to accept the dominant discourses that link Spanish to Latin@ identity. In a study on the interactions that took place at a day labor center in Arizona, she found that day laborers used English as part of a set of discursive practices to perform the identity of a ‘good worker,’ simultaneously resisting dominant discourses that position Latin@ immigrant workers as a “faceless mass of unskilled Spanish speakers” (2014:119). 99

Rachel Showstack

Transforming identities through HL education Educational contexts differ significantly from the contexts described earlier in terms of how Latin@s construct and understand the relationship between the Spanish language and social meaning because of the institutional discourses that shape classroom practices. These discourses may include the ideologies described earlier that subordinate the Spanish language and its speakers and position Spanish as a ‘foreign’ language. Latin@s represent and orient toward perspectives on the value of Spanish in particular educational contexts in different ways, depending on how language ideologies are represented in institutional discourse and on the perspectives and experiences that individuals bring with them from their home language practices (Showstack, 2013). Dual language immersion has been identified as a context in which the social meaning and value of Spanish and English are under constant negotiation (Palmer, 2007). Research on bilingual education has demonstrated that English is often favored over Spanish because it is associated with greater status and power (Palmer, 2009; Potowski, 2004). However, recent studies have also explored the ways in which students from Spanish-speaking households and their instructors counteract dominant societal discourses, claiming symbolic power in the use of Spanish and in hybrid language practices (Martínez, 2013; Mateus, 2014; Palmer et al., 2014). Like bilingual education programs, ethnically diverse university campuses are also contexts in which multiple discourses intersect and are negotiated, leading to a range of possible associations between the Spanish language and different kinds of social value. Urciuoli (2008) demonstrated that Hispanic bilingual university students in campus multicultural organizations at one U.S. university constructed their linguistic skills both as an ‘added value’ and as a ‘deficit.’ When students oriented to the ‘deficit model,’ they described themselves as deficient speakers, rather than acknowledging their home or community language varieties as legitimate genres appropriate for certain contexts. The sense that one’s home language practices are somehow flawed can be particularly acute for Spanish heritage speakers who take university-level courses focused on language, especially when the instructors of those courses are not trained to teach HL learners. A groundbreaking study by Potowski (2002) revealed that many Spanish HL learners in Spanish courses designed for L2 learners classified their own Spanish as bad and were often corrected when producing linguistic features typical of U.S. Spanish in the classroom. In a related study, Achugar and Pessoa (2009) found that members of the Bilingual Creative Writing Graduate Program at the University of Texas, El Paso valued the use of Spanish in the academic context, but expressed negative attitudes toward the use of local varieties. Coryell et al. (2010) examined perspectives of HL learners taking an on-line Spanish HL course, and found that all of the students constructed a ‘culturally proper ideal metanarrative,’ a view that they needed to acquire ‘formal Spanish’ or ‘true Spanish’ in order to be culturally proper. (See Chapter 33 this volume for further discussion on heritage speakers’ attitudes toward Spanish dialects.) Research has investigated the construction and negotiation of Spanish HL identity and related language ideologies in a range of different types of Spanish language courses, demonstrating that each classroom represents a unique set of discursive practices and available identity categories (Abdi, 2011; Harklau, 2009; Helmer, 2013, 2014; Lowther Pereira, 2010; Showstack, 2012, 2013, 2015b). While earlier studies highlight the ways in which HL learners either become marginalized by classroom discourse (e.g. Abdi, 2011; Harklau, 2009) or construct discourses that marginalize others (Showstack, 2012), there is also a need to understand what happens when HL instructors make an effort to counteract dominant ideologies. Language educators have done a great deal of work in recent years to counteract the presence of dominant discourses about Spanish in the U.S., and research conducted on HL 100

Spanish and identity among Latin@s in U.S.

learner identity at the university level has revealed how HL education can contribute to empowerment and the construction of expert identities among students in the Spanish HL classroom (Leeman, 2005; Leeman et al., 2011; Sánchez-Muñoz, 2013; Showstack, 2015b). Sánchez-Muñoz (2013) investigated the effects of a Spanish HL course on the linguistic confidence of the students in the class. Her participants expressed that their identities as Latin@s were closely related to their Spanish language proficiency. Many of the students reported greater linguistic confidence, especially in their writing skills, after having taken the course. Sanchez-Muñoz points out that Spanish HL courses play an important role not only in Spanish language maintenance but also the development of linguistic self-esteem and a sense of community belonging. Leeman et al. (2011) suggest that the expert identities students construct in the classroom context do not necessarily correspond to positions of agency and empowerment outside of the classroom. They developed a service-learning program for their advanced Spanish students that included as one of its goals the promotion of expert identities for their HL learners outside the classroom. After participating in the service-learning program, their students reflect a sense of expertise in using the language in contexts in which some had not positioned themselves as experts previously.

Future directions and recommendations The studies described in this chapter focus on the value of Spanish and on the relationship between Spanish and identity, either in homes and communities or in formal educational contexts, and on the ways in which such discourses can come into contact through educational practices. In order to better understand moments of intersection between the different discursive practices in which Spanish heritage speakers participate, more research is needed to understand the connections between individual Spanish heritage speakers’ histories—the discourses about language to which they have been exposed and how others have positioned them throughout their lives—and the ways in which they negotiate and make sense of their identities as Spanishspeakers and as multilinguals in specific moments of interaction (see Young, 2014). Thus, there is a need for further case studies and linguistic life histories on individual heritage speakers (He, 2014), building on the work of Zentella (1997a) and drawing on more recent research that addresses the construction of identity in interaction and in narrative. In the field of applied linguistics, such case studies will contribute to an understanding of the dialectic between learner histories and classroom language learning practices (Young & Astarita, 2013). In the field of Spanish in the U.S., case studies on individual HL learners will contribute to an understanding of the ideological contexts of Spanish language use in the U.S. Finally, an understanding of how HL learners develop their roles as participants in a wide range of discursive practices over time can lead to a reconceptualization of HL learning. While it has been established in the field of HL education that instructors need to go beyond comparing HL students to the imagined ‘native speaker,’ such advancement depends on a nuanced understanding of alternative perspectives on HL linguistic identity. In order to promote HL investment in classroom language learning, pedagogical practices must acknowledge both the communities of practice in which the students use Spanish outside of the classroom and the ways they envision themselves using the language in the future. Further research on individual learners will allow pedagogues and curriculum developers to avoid representing the kind of essentialized perspectives on language and identity that can alienate HL learners even in classroom contexts that aim to legitimize minority language practices. 101

Rachel Showstack

Research on the relationship between Spanish, language value, and identity reveals a wide range of diversity in the meaning and value of Spanish and bilingual practices in different community and educational contexts. In particular, the symbolic value of Spanish in educational contexts is often quite different from its value within Spanish-speaking families and communities. Gutiérrez et al. (1999) suggest that the classroom can be a place where home and community discourses meet institutional discourses, allowing for the creation of what they call a ‘third space.’ They show how one instructor in a bilingual classroom engaged students in a class discussion by drawing on local knowledge, acknowledging its value, while at the same time teaching new concepts. Since Gutiérrez et al.’s study, a great deal of scholarly work has explored the ways in which HL and bilingual education can engage students by recognizing family and community discursive practices and identities. Two types of curriculum design that acknowledge and explore the identities associated with home and community language practices have been applied in the field of HL education: critical language awareness (Leeman, 2005; Martínez, 2003) and critical service learning (Leeman et al., 2011). By exploring the social value of language and the relationship between language and identity in different contexts, students not only develop identities as language experts, they also become empowered to counteract dominant ideologies that subordinate U.S. Spanish speakers. (See Chapter 22 for a further discussion of critical approaches.) In addition to developing curriculum to address HL students’ affective needs, it is important for teachers of bilingual students to realize that the relationship between language and identity is embodied in classroom interactional practices. Studies combining classroom ethnography with discourse analysis in both HL and bilingual education have demonstrated how teachers reflect particular ideologies about the relationship between language and identity (Palmer et al., 2014; Showstack, 2015b). Palmer et al. (2014) propose a set of ‘translanguaging’ strategies through which dual language instructors can draw from both Spanish and English to recognize students’ bilingual practices in the classroom, countering the traditional enforcement of a ‘monolingual norm’ in bilingual education. Showstack (2015b) points out that HL instructor training should also address the ways that language ideologies are embodied in teacher stance-taking practices, including practices that index stances of expertise and authority, stances toward particular linguistic forms associated with U.S. Spanish, and stances toward the expectations for language use in the classroom context.

Notes 1 A pseudonym. 2 A variety used widely in the Southwest that includes features of rural Mexican Spanish as well as forms that have resulted from contact with English.

Further reading Leeman, J., & Martínez, G. (2007). From identity to commodity: Ideologies of Spanish in heritage language textbooks. Critical Inquiry in Language Studies, 4(1), 35–65. This study analyzes the discourses on the value of Spanish represented in Spanish language textbooks for heritage speakers in two different historical periods. Norton, B. (2013). Identity and language learning: Extending the conversation. Bristol, UK: Multilingual Matters. Drawing on a case study on immigrant women in Canada, Norton presents her influential theory of investment, imagined communities, and language learning (originally published in 2000). Zentella, A. C. (1997). Growing up bilingual: Puerto Rican children in New York. Malden, MA: Blackwell. This study is an ethnography of bilingualism as a social practice in a Puerto Rican neighborhood in New York.

102

Spanish and identity among Latin@s in U.S.

References Abdi, K. (2011). ‘She really only speaks English’: Positioning, language ideology, and heritage language learners. Canadian Modern Language Review, 67(2), 161–189. Achugar, M., & Pessoa, S. (2009). Power and place: Language attitudes towards Spanish in a bilingual academic community in southwest Texas. Spanish in Context, 6(2), 199–223. Anderson, B. (1991). Imagined communities: Reflections on the origin and spread of nationalism (Revised ed.). New York: Verso. Anzaldúa, G. (Ed.) (2007). How to tame a wild tongue. In Borderlands la frontera: The new mestiza (3rd ed., pp. 75–97). San Francisco, CA: Aunt Lute Books. Back, M. (2015). Transcultural performance: Negotiating globalized indigenous identities. Language and globalization. New York: Palgrave. Blanton, C. K. (2004). The strange career of bilingual education in Texas, 1836–1981. College Station, TX: Texas A&M University Press. Blommaert, J. (2005). Discourse: A critical introduction. Cambridge, UK: Cambridge University Press. Bourdieu, P. (1977). The economics of linguistic exchanges. Social Science Information, 16(6), 645–688. Bourdieu, P. (1991). Language and symbolic power. Cambridge, MA: Polity Press. Bucholtz, M., & Hall, K. (2004). Language and identity. In A. Duranti (Ed.), The handbook of linguistic anthropology (pp. 369–394). Malden, MA. and Oxford, UK: Blackwell. Butler, J. (1990). Gender trouble: Feminism and the subversion of identity. New York: Routledge. Canagaraja, A. S. (2013). Translingual practice: Global Englishes and cosmopolitan relations. London: Routledge. Coryell, J. E., Clark, M. C., & Pomerantz, A. (2010). Cultural fantasy narratives and heritage language learning: A case study of adult heritage learners of Spanish. Modern Language Journal, 94(3), 453–469. Davies, B., & Harré, R. (1990). Positioning: The discursive production of selves. Journal for the Theory of Social Behavior, 20(1), 43–63. DuBord, E. (2014). Language, immigration and labor: Negotiating work in the U.S.-Mexico borderlands. Language and globalization. New York: Palgrave. Dueñas Gonzáles, R. (2000). Introduction. In R. Dueñas Gonzales (Ed.), Language ideologies: Critical perspectives on the Official English movement. Vol. 1. (pp. xxvii–xlvii) Urbana, IL: National Council of Teachers of English. Farr, M. (2006). Rancheros in Chicagoacán: Language and identity in a transnational community. Austin, TX: University of Texas Press. Fishman, J. A. (1989). Language and nationalism: Part II. In J. A. Fishman (Ed.), Language and ethnicity in minority sociolinguistic perspective (pp. 269–367). Bristol, UK: Multilingual Matters. García, O., & Torres-Guevara, R. (2010). Monoglossic ideologies and language policies in the education of U.S. Latinas/os. In E. G. Murillo Jr., S. A. Villenas, R. T. Galván, J. S. Muñoz, C. Martínez, & M. Machado-Casas (Eds.), Handbook of Latinos and education: Theory, research and practice (pp. 183–193). New York, NY: Routledge. Giles, H., & Byrne, J. L. (1982). An intergroup approach to second language acquisition. Journal of Multilingual and Multicultural Development, 3(1), 17–40. Gutiérrez, K. D., Baquedano-López, P., & Tejada, C. (1999). Rethinking diversity: Hybridity and hybrid language practices in the third space. Mind, Culture and Activity, 6(4), 286–303. Gutiérrez, M. J., & Fairclough, M. (2006). Incorporating linguistic variation into the classroom. In R. Salaberry & B. Lafford (Eds.), The art of teaching Spanish (pp. 173–191). Washington, DC: Georgetown University Press. Harklau, L. (2009). Heritage speakers’ experiences in new Latino Diaspora Spanish classrooms. Critical Inquiry in Language Studies, 6(4), 211–242. He, A. W. (2004). Identity construction in Chinese heritage language classes. Pragmatics, 14(2–3), 199–216. He, A. W. (2010). The heart of heritage: Sociocultural dimensions of heritage language learning. Annual Review of Applied Linguistics, 30, 66–82. He, A. W. (2014). Heritage language development and identity construction throughout the life cycle. In T. G. Wiley, J. C. Peyton, D. Christian, S. C. K. Moore, & N. Liu (Eds.), Handbook of heritage, community, and Native American languages in the United States: Research, policy, and eductional practice (pp. 324–332). New York: Routledge. Helmer, K. A. (2013). A twice-told tale: Voices of resistance in a Borderlands Spanish Heritage Language Class. Anthropology & Education Quarterly, 44(3), 269–285. Helmer, K. A. (2014). ‘It’s not real, it’s just a story to learn Spanish’: Understanding heritage language learner resistance in a Southwest charter high school. Heritage Language Journal, 11(3), 186–206.

103

Rachel Showstack Jacoby, S., & Ochs, E. (1995). Co-construction: An introduction. Research on Language and Social Interaction, 28(3), 171–183. Kanno, Y., & Norton, B. (2003). Imagined communities and educational possibilities: Introduction. Journal of Language, Identity and Education, 2(4), 241–249. Koike, D. A., & Graham, C. P. (2006). Who is more Hispanic? The co-construction of identities in a U.S. Hispanic political debate. Spanish in Context, 3(2), 181–213. Kroskrity, P. V. (2004). Language ideologies. In A. Druanti (Ed.), A companion to linguistic anthropology (p. 496–517). Malden, MA: Blackwell. Leeman, J. (2005). Engaging critical pedagogy: Spanish for native speakers. Foreign Language Annals, 38(1), 35–45. Leeman, J. (2013). Categorizing Latinos in the history of the US Census: The official racialization of Spanish. In J. Del Valle (Ed.), A Political History of Spanish (pp. 305–323). Cambridge, UK: Cambridge University Press. Leeman, J. (2014). Critical approaches to teaching Spanish as a local/foreign language. In M. Lacorte (Ed.), The handbook of Hispanic applied linguistics (pp. 275–292). New York: Routledge. Leeman, J., & Martínez, G. (2007). From identity to commodity: Ideologies of Spanish in heritage language textbooks. Critical Inquiry in Language Studies, 4(1), 35–65. Leeman, J., Rabin, L., & Román-Mendoza, E. (2011). Identity and activism in heritage language education. The Modern Language Journal, 95(4), 481–495. Limón, J. (1982). El meeting: History, folk Spanish, and ethnic nationalism in a Chicano student community. In J. Amaste & L. Elías-Olivares (Eds.), Spanish in the United States: Sociolinguistic aspects (pp. 301–322). Cambridge, UK: Cambridge University Press. Lowther Pereira, K. A. (2010). Identity and language ideology in the intermediate Spanish heritage language classroom. PhD dissertation. The University of Arizona, Tucson. Martínez, G. (2003). Classroom-based dialect awareness in heritage language instruction: A critical applied linguistic approach. Heritage Language Journal, 1(1), 1–14. Martínez, R. A. (2013). Reading the world in Spanglish: Hybrid language practices and ideological contestation in a sixth-grade English language arts classroom. Linguistics and Education, 24, 276–288. Mateus, S. G. (2014). She was born speaking English and Spanish! Bilingual status in a kindergarten twoway dual language classroom. Texas Papers in Foreign Language Education, 16(1), 59–74. Mendoza-Denton, N. (2008). Homegirls: Language and cultural practice among Latina youth gangs. Malden, MA: Blackwell. Myhill, J. (1999). Identity, territoriality and minority language survival. Journal of Multilingual and Multicultural Development, 20(1), 34–50. Negrón, R. (2014). New York City’s Latino ethnolinguistic repertoire and the negotiation of latinidad in conversation. Journal of Sociolinguistics, 18(1), 87–118. Norton, B. (2001). Non-participation, imagined communities and the language classroom. In M. P. Breen (Ed.), Learner contributions to language learning: New directions in research (pp. 159–171). Harlow, UK: Pearson Education Limited. Norton, B. (2010). Language and identity. In N. Hornberger & S. McKay (Eds.), Sociolinguistics and language education (pp. 349–369). Bristol, UK: Multilingual Matters. Ochs, E. (1993). Constructing social identity: A language socialization perspective. Research on Language and Social Interaction, 26(3), 287–306. Orozco, C. (2009). No Mexicans, women, or dogs allowed: The rise of the Mexican American civil rights movement. Austin, TX: University of Texas Press. Ortega, L. (1999). Language and equality: Ideological and structural constraints in foreign language education in the U.S. In T. Huebner & K. A. Davis (Eds.), Sociopolitical perspectives on language policy and planning in the U.S. (pp. 244–266). Amsterdam: John Benjamins. Palmer, D. K. (2007). Building and destroying students’ ‘academic identities’: The power of discourse in a two-way immersion classroom. International Journal of Qualitative Studies in Education, 21(6), 1–24. Palmer, D. K. (2009). Middle-class English speakers in a two-way immersion bilingual classroom: ‘Everybody should be listening to Jonathan right now . . . ’. TESOL Quarterly, 43(2), 177–202. Palmer, D. K., Martínez, R. A., Mateus, S. G., & Henderson, K. (2014). Reframing the debate on language separation: Toward a vision for translanguaging pedagogies in the dual language classroom. Modern Language Journal, 98(3), 757–772. Pavlenko, A., & Blackledge, A. (Eds.). (2004). Negotiation of identities in multilingual contexts. Clevedon, UK: Multilingual Matters.

104

Spanish and identity among Latin@s in U.S. Pease-Alvarez, L. (2002). Moving beyond linear trajectories of language shift and bilingual language socialization. Hispanic Journal of Behavioral Sciences, 24(2), 114–137. Pedraza, P. (1985). Language maintenance among New York Puerto Ricans. In L. Elías-Olivares, E. A. Leone, R. Cisneros, & J. R. Gutiérrez (Eds.), Spanish Language and Public Life in the United States (pp. 59–71). New York: Mouton. Peirce, B. N. (1995). Social identity, investment, and language learning. TESOL Quarterly, 29(1), 9–31. Pomerantz, A. (2002). Language ideologies and the production of identities: Spanish as a resource for participation in a multilingual marketplace. Multilingua, 21, 275–302. Pomerantz, A. (2010). Speaking Spanish outside the foreign language classroom: An analysis of learner narratives. Critical Inquiry in Language Studies, 7(1), 1–27. Pomerantz, A., & Schwartz, A. (2011). Border talk: Narratives of Spanish language encounters in the US. Language and Intercultural Communication, 11(3), 176–196. Potowski, K. (2002). Experiences of Spanish heritage speakers in university foreign language courses and implications for teacher training. ADFL Bulletin, 33(3), 35–42. Potowski, K. (2004). Student Spanish use and investment in a dual immersion classroom: Implications for second language acquisition and heritage language maintenance. The Modern Language Journal, 88(1), 75–101. Potowski, K. (2014). Ethnolinguistic identities and ideologies among Mexicans, Puerto Ricans, and ‘MexiRicans’ in Chicago. In R. M. Reiter & L. M. Rojo (Eds.), A sociolinguistics of diaspora: Latino practices, identities, and ideologies (pp. 13–30). New York: Routledge. Raymond, C. W. (2012). Reallocation of pronouns through contact: In-the-moment identity construction amongst Southern California Salvadorans. Journal of Sociolinguistics, 16(5), 669–690. Ruiz, R. (1984). Orientations in language planning. NABE Journal, 8(2), 15–34. Sánchez-Muñoz, A. (2013). Identidad y confianza lingüística en jóvenes latinos en el sur de California. In D. Dumitrescu & G. Piña-Rosales (Eds.), El español en Estados Unidos: E pluribus unum? (pp. 217–232). New York: Academia Norteamericana de la Lengua Española. Showstack, R. E. (2012). Symbolic power in the heritage language classroom: How Spanish heritage speakers sustain and resist hegemonic discourses on language and cultural diversity. Spanish in Context, 9(1), 1–26. Showstack, R. E. (2013). Ideology and identity in Spanish heritage language classroom discursive practices. PhD dissertation. The University of Texas at Austin, Austin, TX. Showstack, R. E. (2015a). Expertise, imagined communities and Spanish heritage language teaching. Paper presented at the International Society for Language Studies, Albuquerque, NM. Showstack, R. E. (2015b). Institutional representations of ‘Spanish’ and ‘Spanglish’: Managing competing discourses in heritage language instruction. Language and Intercultural Communication, 15(3), 341–361. Showstack, R. E. (2017). Stancetaking and language ideologies in heritage language learner classroom discourse. Journal of Language, Identity and Education, 16(5), 271–284. Tajfel, H. (1974). Social identity and intergroup behavior. Social Science Information, 13, 65–93. Tajfel, H. (1981). Human groups and social categories. Cambridge, UK: Cambridge University Press. Tamasi, S. and Antieau, L. (2014). Language and linguistic diversity in the U.S.: An introduction. New York: Routledge. Toribio, A. J. (2000). Language variation and the linguistic enactment of identity among Dominicans. Linguistics, 38(5), 1133–1159. Torrez, J. E. (2013). Somos mexicanos y hablamos mexicano aquí: Rural farmworker families struggle to maintain cultural and linguistic identity in Michigan. Journal of Language, Identity, and Education, 12, 277–294. Urciuoli, B. (2008). Whose Spanish? The tension between linguistic correctness and cultural identity. In M. Niño-Murcia & J. Rothman (Eds.), Bilingualism and identity: Spanish at the crossroads with other languages (pp. 257–277). Philadelphia, PA: John Benjamins. U.S. English. 2015. 19 November 2015. Retrieved from www.us-english.org/view/310. Valdés, G., González, S. V., García, D. L., & Márquez, P. (2003). Language ideology: The case of Spanish in departments of foreign languages. Anthropology & Education Quarterly, 34(1), 3–26. Woods, M. R., & Rivera-Mills, S. V. (2012). El tú como un ‘mask’: Voseo and Salvadoran and Honduran identity in the United States. Studies in Hispanic and Lusophone Linguistics, 5(1), 191–216. Woolard, K. A. (1998). Introduction: Language ideology as a field of inquiry. In B. B. Schieffelin, K. A. Woolard, & P. V. Kroskrity (Eds.), Language ideologies: Practice and theory (pp. 3–47). New York: Oxford University Press.

105

Rachel Showstack Young, R. F. (2014). ‘Don’t know much about history’ and it’s NOT a wonderful world. Paper presented at the Conference on Interactional Competence and Institutional Practices, University of Neuchâtel, Switzerland. Young, R. F., & Astarita, A. C. (2013). Practice theory in language learning. Language Learning, 63(Supplement 1), 171–189. Zentella, A. C. (1997a). Growing up bilingual: Puerto Rican children in New York. Malden, MA: Blackwell. Zentella, A. C. (1997b). The Hispanophobia of the official English movement in the U.S. International Journal of the Sociology of Language, 127, 71–86.

106

8 SPANISH AS A HERITAGE LANGUAGE AND THE NEGOTIATION OF RACE AND INTRA-LATINA/O HIERARCHIES IN THE U.S. Rosalyn Negrón university of massachusetts boston

Introduction Latina/os have sociohistorical roots in continents including North and South America, Europe, Africa, and Asia, making them a pan-ethnic,1 multiracial, and multicultural people. At least 19 national origin dialects of Spanish are spoken by different U.S. Latina/o groups (but see Erker, this volume, for a critique of the concept of dialect), which combined with the multiple dialects of English they speak2 render them among the most linguistically diverse of U.S. groups. Latina/o groups (e.g. Cubans, Ecuadorians, Salvadorians) have experienced distinct immigration histories and trajectories within the U.S., which partially explains their different social statuses. This diversity confounds the U.S. racial system, which has historically categorized the U.S. population into white and non-white and conferred status on groups in relation to the white dominant majority. What happens when the realities of Latina/o diversity and the inflexibility of the U.S. racial system meet, and how does this translate into the everyday of social and linguistic practice? I argue that Spanish as a heritage language (SHL) must be examined within a broad repertoire of linguistic options available to Latina/os as they negotiate the U.S. racial hierarchy. Latina/os’ linguistic repertoires may include multiple dialects of English and Spanish, different styles, standard and non-standard forms, as well as hybrid forms resulting from language contact (Bailey, 2000, 2001; Zentella, 1997; Durán and Toribio, this volume) and this variety enables the creative and flexible negotiations of multiple dimensions of difference, including race. Consider for example the following conversation between two Dominican American high school students in Rhode Island (Bailey, 2000:195–196): Excerpt 1: [Isabella and Janelle are sitting on steps outside of the main school building at the end of their lunch period. Isabella has returned from eating lunch at a diner near the school, and she has been describing the generous size of the turkey club sandwich she has just eaten.] 107

Rosalyn Negrón

J:  Only with that turkey thingee      // ya yo (es)toy llena.                  “I’m already full” I:                 //Two dollars and fifty cent. J: That’s good. That’s like a meal at    //Burger King. 1:                 //That’s better than going to Burger King, you know what I’m saying? And you got a Whopper, French fries,                    //and a drink. And = J:                  //Yeah I: = the French fries cost a dollar over there. J: For real? I: Sí, sí ¿Como no? “Yes, really.” Mírale el ombligo. Miralo. Se le ve, ya se lo tapó. ((looking at a passerby)) “Look at J:  her belly button. Look. You can see it, she already covered it.” (.5) I: Seguro porque se lo enseñó. ((laughing)) “She must have showed it.” (1.5) J: //( ) I: //But it’s slamming, though, oh my God, mad [“a lot of”] turkey she puts in there. J: That’s one thing 11-, I love the way como l- = I:                “how th-” J:  = the American [“white Americans”] be doing sandwich, they be rocking [“are excellent”], them things, yo, they put everything in there yo. Bailey (2000: 197) points out that “the variety and juxtaposition of linguistic resources by Janelle and Isabella in the above exchange reflect their specific life experiences and aspects of their social world.” This includes: 1) their bicultural socialization as exemplified by their alternation between English and Spanish; 2) their sustained contact and identification with African American youth evident in their use of African American Vernacular English (AAVE) (e.g. habitual be in “the American be doing sandwich, they be rocking them things, yo”); and 3) a “we” / “they” contrast between themselves and “Americans” (“white people”), suggesting an identification as non-white and non-American. Even in a brief moment of mundane conversation it is evident that Latina/o identity, language, and race interact in complex ways that flout any attempts to draw a direct link between SHL use and discrete Latina/o identities. Because of the rigid black-white binary imposed by the U.S. racial system, for many Latina/ os using (or avoiding) Spanish can serve to circumvent the limitations set by race. For example, some Latina/os have been found to use Spanish to differentiate themselves from African Americans (Bailey, 2000; Toribio, 2003). Still others may speak in AAVE to affiliate as/with African Americans, as shown in Excerpt 1. Latina/os may use English to deflect an immigrant identity in an anti-immigrant climate where the use of Spanish is a marker of foreignness. Others may wish to de-emphasize an indigenous heritage by using a dialect of Spanish that invokes a euro-Latina/o identity (Negrón, 2011). In other words, race, Latina/o identity, and SHL overlap in sometimes mutually reinforcing but sometimes contradictory ways: Spanish can be a vehicle for circumventing racialization as black, but its use also upholds a racial classification as non-white. Also important is the fact that U.S. racial formation processes are different from those in Latin America.3 Many immigrants to the U.S. bring with them Latin American racial concepts, which can create tensions between U.S. Latina/o subgroups in ways that put Spanish language ideologies directly at play. Consider Excerpt 2, translated into English from an interview I conducted in Spanish with a 48-year-old white Colombian woman who migrated to the U.S. in her thirties.

108

Race, intra-Latina/o hierarchies in U.S.

Excerpt 2: I feel proud to be Colombian. I don’t feel sorry. I tell everyone that I’m Colombian because independently of thinking that Colombians work with drugs, Colombians are very educated, very educated. Manners of the Colombians are better than that of other cultures, it seems to me; at the table, in everything. For example, Dominicans, they eat with their mouth open, they talk when they’re eating, they don’t greet, and they don’t say goodbye. We Colombians are very educated. And over there in Colombia, those manners come from Europe, from the Spaniards. Listen, I’m working with a Colombian gentleman in real estate. What an educated man, that man, it terrifies me, that man you can just see that he is Colombian. As soon as he started talking to me, like that, so educated, so unhurried, so calm to talk, I said to myself, this man is Colombian, and he was Colombian. Alma highlights verbal and nonverbal communicative acts that in her mind distinguish Colombians and Dominicans. The direct link she draws between Colombians and Europe (while ignoring the Dominican Republic’s own European connections) has racial overtones: she references a Latin American historical racial narrative that equates whiteness (read Colombian) with nobility and purity, and blackness (read Dominican) with the uncultured and crude. In fact, during fieldwork in New York City, it was not unusual to hear my Latina/o informants describe Colombian Spanish as more formal, proper, and closer to its Iberian origins compared to Dominican Spanish. To the extent that such intra-Latina/o hierarchies signal differential social and economic rewards, intra-Latina/o fissures complicate efforts to promote SHL institutionally (c.f. Bedolla, 2003). In this chapter, I analyze examples that I argue constitute attempts by Latina/os to negotiate racial categorization through language. I consider the ways that members of different Latina/o subgroups use language and language ideologies in both affiliative and differentiating ways, the latter often couched in racial terms. Critical to this discussion is an examination of the status of specific Latina/o subgroups within the U.S. racial hierarchy, which is closely tied to the status of Spanish language varieties spoken by U.S. Latina/os. Moreover, Spanish generally indexes a non-white identity and thus Spanish is often stigmatized in public domains (Urciuoli, 1996). This has important implications for the everyday use of Spanish and may play a role in Spanish language loss among the children of Latina/o immigrants.

What is race? Smedley (1998: 693) defines race as “the organization of all peoples into a limited number of unequal or ranked categories theoretically based on differences in their biophysical traits.” Racial ideology is particularly pernicious because these superficial biophysical differences are thought to reflect meaningful differences in the fundamental nature – and indeed, the value – of human groups. In fact, geneticists have dismissed the notion that people from different so-called “races” are biologically distinct. They find that the genetic differences within racial groups are in fact greater than the differences between them (American Anthropological Association, 1998). Nevertheless, in the U.S., the concept of race was reified as a system of legal, social, and economic control of blacks and indigenous peoples. The roots of race were established in the colonial era, further entrenched and reproduced to justify the enslavement of blacks, and carried on through de jure racism during the Jim Crow era. Then and now, racial categories signify a binary of white/non-white. This binary was codified by the rule of hypodescent, which

109

Rosalyn Negrón

conferred a subordinate black racial status to anyone with even “one drop” of African blood. Whites, as the primary beneficiaries of the American racial system, are deemed normative, the standard by which other racial groups are evaluated, whereas all persons with any non-white blood are considered non-white. This racial binary has been used as a heuristic for assigning worth, capacity, and morality. Latin America shares with the U.S. a history of European colonialism, African and indigenous slavery, and centuries of racial mixing. In both the U.S. and Latin America, race has been strongly tied to class. From its inception in the colonial past, racial classification and control was a means by which the labor of non-white groups was exploited for the benefit of whites in power. In both locations, the economic logic of race in the 18th and 19th centuries posited that non-white groups were unfit for any economic activity that required more than brute bodily effort. Pseudo-scientific racist studies affirmed the animal-like nature of non-white people, which claimed that their low intelligence and lack of discipline justified their exploitation (Gould, 1996). This racist economic logic, which had blacks as its main target, would later be applied to other groups who were needed for their labor, especially immigrants from Asia and Latin America. In addition to the ways that slavery and systemic racial discrimination has constrained the accumulation of wealth for racialized groups, the persistence of the racist economic logic has relegated many non-white groups, including Latina/os, to labor force positions that limit their economic mobility. In tandem with a new racism (Barker, 1981) that conceives of cultural difference as almost genetic in nature, the class positions of racialized groups are explained by their culture rather than in terms of the structural consequences of race. Despite these broad similarities, in contrast to the U.S. black-white binary, Latin American categorization systems have historically had more intermediate categories between black and white. For example, in Puerto Rico, the categories indio, trigueño, and jabao each pertain to different combinations of skin color and hair form types (Gravlee, 2005).4 The concepts of mestizaje (racial mixing) and blanqueamiento (racial whitening) point to some unique aspects of Latin American’s racial history, the legacies of which are experienced by many Latina/os today. As a nation-building ideology, mestizaje advances a sort of post-racial society in which racial and cultural mixing between white, black, and indigenous peoples produces a nation of one people (Wade, 2005). However, Stutzman (1981) argues that throughout Latin America the inclusive ideal of mestizaje in reality excluded blacks and indigenous people. Whiteness as the national ideal was inscribed into policy in several Latin American countries, including Argentina, Brazil, the Dominican Republic, and Cuba, all of which implemented measures to accelerate blanqueamiento through increased European immigration.

Modern-day U.S. Latina/o racial self-identification Given these histories, the racial concepts that Latina/o immigrants bring with them to the U.S. do not neatly match the U.S. racial categorization system. This is evident in the way that Latina/os respond to questions about race in the U.S. Census. It is important to note that the U.S. Census, after increasing pressure, eliminated “Hispanic/Latina/o” as a racial category in 2010, leaving Hispanics/Latina/os to choose between white, black, and other categories. The 2010 Census shows that 53% of U.S. Latina/os identify as white (U.S. Census, 2011a). In contrast, only 2.5% of Latina/os in U.S. identified as black. In writing about Puerto Rico’s gradual but persistent “whitening,” Duany (2010) suggests that the percentage of Puerto Ricans identifying as white (approximately 75%) belies the islands long history of racial mixing. Importantly, after white, “Some other race” was the next most common response to the race question on the U.S. Census, with 36.7% of Latina/os identifying in this way. This category includes instances in which 110

Race, intra-Latina/o hierarchies in U.S.

Latina/o respondents used “Hispanic” or “Latina/o” as a racial identification. Grieco and Cassidy (2001) found that 97% of respondents who identified as “Some other race” on the 2000 Census were Latina/o. This in turn may be a reflection of the long history of mestizaje in Latin America. Results of the Racial and Ethnic Target Test (RAETT), which checked for the effect of question format on racial and ethnic self-identification, suggests that when race and ethnicity questions are asked separately, Latina/os are more likely to skip the race question (Hirschman et al., 2001). The RAETT further found that a larger proportion of Latina/os reported more than one race than other groups. Also, it appears that across Latina/o groups there is significant variation in racial self-identification, particularly when Hispanic/Latina/o is not given as an option for racial self-classification (Tucker et al., 1996). These studies suggest that the phenoty­ pical conception of race in the U.S. does not correspond with folk concepts of race and ethnicity among Latina/os. In general, although U.S. Latina/os show a preference for white identification, they reject the black/white binary options in favor of the Latina/o category, suggesting that Hispanic/Latina/o should once again be considered a racial category on the Census. Indeed, many U.S. Latina/os view Latina/o as part of their racial identity (Pew Research Center, 2015). For U.S. Latina/os, the incompatibilities between U.S.-based and Latin American race concepts promote strategies that serve to bridge, manipulate, or reject racial categorization systems. I later provide examples of the ways that language serves as an important tool for such negotiations.

Race and language among U.S. Latina/os Given that race is frequently conflated with class, the language varieties of racialized groups are frequently evaluated in terms of their economic value. Given the primary black-white U.S. dichotomy, varieties of African American English reflect discourses about African Americans in the U.S. (see, for example, Lippi-Green, 2012). Spanish, too, has become racialized and seen itself subject to contradictory demands. On the one hand, it is conceived in terms of its costs to economic mobility. The persistence of Spanish use among heritage speakers is taken as evidence that U.S. Latina/os cannot be full economic players in the American nation, where English is the principal language of the economy (Urciuoli, 1996). On the other hand, Spanish holds value for non-heritage speakers as the growing Latina/o population and their purchasing power, which reached $1.3 trillion in 2015, has expanded opportunities for marketing to and providing services for Latina/os (Nielsen, 2016). Nevertheless, the use of Spanish in public space has been subject to control and rejection, tied in part to fears about the end of the U.S. way of life conceived by the dominant white majority (Cobas and Feagin, 2008). The fierce debates that surrounded the English-Only legislative initiatives of the 1990s revealed that what was at stake was not merely questions of official expedience or educational progress, but rather the integrity of the nation itself (Urciuoli, 1996). Popular and official discourse about the growth of Latina/o immigration to the U.S. was couched in terms of a threat to the fundamental character of the (white) American nation (c.f. Huntington, 2004). Advocacy for the support and maintenance of Spanish was seen as a threat to the expected process of assimilation, by which immigrants would become economically productive Americans. As seen in the dismantling of bilingual education programs throughout the country in the past 20 years,5 schools have been a key battleground for competing views about the value of bilingualism versus the importance of English for social and economic mobility.6 Urciuoli (1996: 26) argues that the use of Spanish in public arenas was framed as contrary to U.S. national ideology, which deemed assimilation, social mobility, and the use of English as hallmarks of citizenship, and the use of Spanish “as signs of illiteracy and laziness, which people are morally obliged to control through education.” Working class Puerto Ricans’ bilingual 111

Rosalyn Negrón

practices in the public sphere, for example, was a basis for their racialization, as their “broken” English, their accents, and their language mixing was taken by white gatekeepers as evidence of their subordinate status. Urciuoli further contends that race has always been about restricting who could be deemed part of the nation and thus subject to equal rights, protection, and opportunities. In the U.S. case, the normative American, Urciuoli (1996: 16) writes, “is white, English-speaking, middle-class . . . [and] if you don’t fit into this view of the generic American, then your membership to the nation is provisional.” Membership in the U.S. nation entails the assimilation of cultural tools needed to claim the promise of economic mobility enshrined in “American dream” ideology. Though the U.S. does not have an official language, English is viewed as the rightful language of the U.S. nation and a crucial vehicle for economic mobility. Members of the lower classes, and by conflation, the racialized groups that occupy vulnerable class positions, are unfit citizens in their inability to advance economically. Languages other than standard English are discouraged and controlled in the public spheres of economic activity and socialization, such as workplaces and schools. Urciuoli (1996) and Cobas and Feagin (2008) show that such control may entail Latina/os being directly asked not to speak Spanish in public by white bystanders, Latina/os having their Spanish accents ridiculed when speaking English, or Latina/os experiencing pressures to speak “good English.” Through such tactics, Spanish language-use affirms Latina/os’ racialization as non-white and as un-American.

Critical issues In the remainder of this chapter I outline two main areas in which the study of race is critical for understanding SHL use trends in the U.S. First, I discuss how Latina/os use language to negotiate race in their interactions with other Latina/os. This can entail downplaying racial and other differences by invoking a collective Latina/o identity, or latinad. It may also involve highlighting intra-Latina/o differences. Second, I examine how Latina/os resist racial categorization through their language practices. The processes entailed in these two areas are not totally distinct from each other. That is, invoking a Latina/o identity through language is also a way in which Latina/os resist racial categorization.

Spanish and race within Latina/o pan-ethnicity Between U.S.-raised Latina/o SHL speakers in particular, a switch from English to Spanish during a conversation can serve to affirm a shared Latina/o identity. Consider Excerpt 3 between William and Roberto, two Latinos who had previously not known each other and who were unsure about each other’s ethno-racial backgrounds, namely they were unsure if each was Latino. Both Roberto and William have a white racial appearance. But Roberto in particular has certain stylistic features (e.g. clothing and hair style and bodily stances) that make it difficult to unambiguously categorize him as non-Latino white. They met when Roberto (Venezuelan) walked into William’s (Puerto Rican) cell phone shop in Queens, New York. From the beginning of their conversation the men spoke only in English; that is, until a shared Latina/o identity because relevant for their respective business interests: Excerpt 3: William: Yeah, I, I own an online magazine called Cuchifrito7 for Thought,8 it’s been around for 8 years. Roberto: Ok. 112

Race, intra-Latina/o hierarchies in U.S.

William: Ahm, I’m working with a company called Asamblea Latina [“Latina/o Assembly”]. They did something really big in, ah, Flushing Meadow Park last year. Roberto: No me digah [“You don’t say?”] William: Yeah and it’s all Latino, and from 21 countries. Roberto: Oh, coño, ehtá bien [“damn, that’s good”]. Negrón (2014: 100) Seconds before this exchange, William read a business card with Roberto’s unambiguously Spanish first and last name. This made it possible for William to categorize Roberto as Latino despite Roberto’s white racial ambiguity. Once he established that Roberto was Latino, William introduced his Latina/o-oriented business. Roberto in turn affirmed William’s instantiation of a shared Latino identity by switching to Spanish. The rest of their bilingual exchange unfolded within a frame of shared latinidad. As a collective project that unfolds through both discursive and institutional practices, latinidad reduces racial difference by exalting the Latina/o pan-ethnicity as a multiracial people. William subscribed strongly to this notion. His website – a blog for Latina/o-oriented news, creative writing, opinions, and information – shows how keenly he believes in cultivating latinidad, connecting with other Latina/os, and advancing Latina/o interests. In a video that he once posted to the website, he evocatively described in Spanish Latina/ os’ indigenous, African, and Spanish origins, three racial and cultural ancestries that he saw as common links for all Latina/os. William’s use of the phrase ‘all Latina/o’ above is consistent with the philosophy of Latina/o inclusivity that he strongly espouses. Thus, latinidad emphasizes the shared aspects of the Latina/o experience in the U.S., even as it encompasses significant heterogeneity across multiple domains of social organization: demographic, linguistic, racial, socio-economic, and so on. Though Spanish use tends to signify Latina/o identity, its use and proficiency among Latina/ os vary significantly. In fact, Spanish may actually serve to foreground differences between Latina/os who speak it fluently and those who do not (De Genova and Ramos-Zayas, 2003; Ghosh Johnson, 2006; Mendoza-Denton, 2014; Zentella, 2007). De Genova and Ramos-Zayas (2003) show that between Mexicans and Puerto Ricans in Chicago, differences in Spanish dialectal features and in levels of bilingualism were sources of division. Through criticisms of Puerto Rican Spanish, Mexicans demoted Puerto Ricans to a racialized status “approaching” blackness, invoking racial stereotypes about laziness and welfare-dependency, while simultaneously undermining Puerto Ricans’ authenticity as Latina/os. This was a way for Mexicans to bolster their own relative, tenuous position in the U.S. racial order. This tenuous position is partly tied to common perceptions of Mexicans as immigrant and therefore not American. In turn, Puerto Ricans in De Genova and Ramos-Zayas’ study reproduced mainstream discourses about Mexicans as “illegal,” “foreign,” and “Third World.” Just as intra-Latina/o divisions exist in ways that encourage language strategies of differentiation, so too can members of the same Latina/o subgroup mark difference. In his work referenced earlier, Bailey (2000) describes distinctions between recent Dominican immigrants and Dominicans who had spent most of their life in the U.S. Complaining about the perceived overuse of Spanish or the “hick” nature of particular forms of Dominican Spanish were some ways that Dominican American youth marked a contrast between themselves and their recent immigrant peers. Similarly, Mendoza-Denton (2014) describes how the use of Spanish and indigenous appearance lent a measure of Mexican authenticity to the working-class Piporra youth gang she studied in northern California. However, Piporra’s white, middle-class counterparts, Fresas, used those markers of Mexican authenticity to reproduce Latina/o American racial ideology, which positioned indigenous Mexicans at the bottom of the racial and socioeconomic 113

Rosalyn Negrón

hierarchy. Competing ideologies about Spanish use are implicated in broader debates about what it means to be Latina/o, who can and cannot make claims to a Latina/o identity, as well as the relative position of Latina/o subgroups within racial hierarchies. Thus, speaking a heritage language requires much more than communicative competence. It also requires picking up on the politics of linguistic identities and negotiating often contradictory language ideologies. Thus, even as all Latina/os are subject to racialization in relation to the white U.S. majority, Latina/o subgroups are themselves subject to a hierarchy of value that has salience within the Latina/o pan-ethnicity. The intra-Latina/o hierarchy develops through the interplay between two related processes. First, the binary U.S. racial system imposes rigid categories that limit Latina/os’ options for racial self-identification. Second, the racial categorization systems that operate in individual Latin American countries exert their own meanings and influence. In the U.S., black and indigenous Latina/os experience a racial double-whammy. They are discriminated against by their lighter compatriots and fellow Latina/os, as well as subject to disadvantage in the U.S. institutionalized racial hierarchy (Córdova and Cervantes, 2010). In noting that black and indigenous Latina/os have been victims of a sin of omission in media and marketing representations of Latina/os, Torres-Saillant (2002) shows that in important ways the systems of racial categorization in the U.S. and Latin America are not wholly incompatible. Yet, while the U.S. racial system is binary, with white and non-white as mutually exclusive, in Latin America an expanded set of ranked categories forms a continuum in which racial categories may blend into each other, reflecting the particular histories of racial mixing in Latin America. Mendoza-Denton (2014) points out that through the use of cultural markers like language, Latina/os may move along this continuum to become more or less white, black, or indigenous (indio) across contexts. And regarding language, in interactions with co-ethnics, Latina/os may be guided by the Latin American racial ideology that equates standard Spanish with superior cultural traits. Roberto, whom I introduced earlier in this chapter, echoes this ideology in his description of his use of standard forms of Spanish with certain Spanish speakers: “I pronounce the s’s, I pronounce the r’s, I pronounce everything; something I thank my dad for, man.” Roberto’s father was born in Spain but had lived many years in Venezuela, where he was part of the working-class. Roberto associated his father’s Spanish ancestry with strong values and education: “My father was born in Spain, he was very educated”; and “My father is that type of Spaniard that, he brought me up with values!” The flip side of this is that certain varieties of Spanish are deemed less prestigious. Specifically, Latina/os with dark skin and other racialized physical features are tied to less prestigious language varieties. As a case in point, Dominicans’ status in the Latina/o social order has acute racialized overtones, given that approximately 90% of Dominicans have African ancestry (Torres-Saillant, 1998). Along with Puerto Ricans, Dominicans are among the poorest and least educated Latina/os in the U.S. They are also among the darkest phenotypically. Zentella (1990) reported that Dominican’s disadvantaged position in NYC’s hierarchy has consequences for how the Dominican dialect is perceived and used. She found that while few Latina/os used words associated with the Dominican dialect, Dominicans (whom she notes are overall darker and poorer in NYC), in contrast, were the only subgroup that adopted from all other groups without exception. She suggests that Colombians, Cubans, and Puerto Ricans contribute to Dominican linguistic insecurity through their widespread rejection of Dominican Spanish. In their later work on NYC Spanish, Otheguy and Zentella (2012) present weakened coda /s/, as a less prestigious linguistic feature stereotypically tied to speakers of Caribbean varieties of Spanish.9 When we look at the distribution of languages spoken at home by Latina/os from different groups, intriguing questions arise about the relationship between race and SHL maintenance. Table 8.1 (from Rumbaut, 2006) shows the linguistic profiles of foreign-born and second- and later- (“second +”)generation Latina/os in the 2000 Census. 114

5,780,867 31.3 984,087 6.3 6.1 8.4 6.2 5.7 4.9 6.8 6.1 5.1 7.9 18.8 6,764,954 19.9

355,778 1.9 14,422,751 93.0 93.6 91.2 93.4 94.0 94.4 92.6 93.4 94.2 89.9 55.9 14,778,529 43.5

12,313,265 66.7 105,506 0.7 0.3 0.5 0.4 0.3 0.6 0.6 0.5 0.7 2.2 25.2 12,418,771 36.6

202,899,620 95.5 5,787,871 36.1 34.5 37.1 32.9 11.7 14.2 32.7 22.2 24.3 35.1 49.9 208,687,491 91.4

3,110,915 1.5 10,167,640 63.5 65.2 62.4 66.4 88.0 85.2 66.8 76.8 74.9 63.4 49.1 13,278,555 5.8

Spanish

English only

All other languages

English only

Spanish

Second+-generations (U.S.-born)

First-generation (foreign-born)

For Puerto Ricans, “foreign-born” includes those born on the island of Puerto Rico, and “U.S.-born” refers to those born on the mainland.

N % N % % % % % % % % % % % N %

Source: 2000 U.S. Census, 5% PUMS.

a

Mexican Puerto Ricana Cuban Dominican Salvadoran, Guatemalan Central American, other Colombian Peruvian, Ecuadorian South American, other Other Spanish, Hispanic, Latino Total

Hispanic

Not Hispanic

Ethnic identity

Table 8.110  Language spoken at home of foreign-born and U.S.-born Hispanics and non- Hispanics, 2000 (persons 5 years and older)

6,380,528 3.0 66,913 0.4 0.2 0.5 0.6 0.4 0.6 0.5 0.9 0.8 1.5 1.0 6,447,441 2.8

All other languages

Rosalyn Negrón

We notice that among first-generation Latina/o immigrants, rates of Spanish use at home are similarly high. However, in the second and later generations we see marked differentiation in the linguistic profiles of SHL speakers. Second + -generation Dominicans have the highest rate of Spanish use (88%) and the lowest rate of English-only use (11.7%). Other Latina/o groups with high rates of Spanish use include Salvadorans and Guatemalans. Puerto Ricans (62.4%), South Americans (63.4%), and Mexicans (65.2%) had the lowest rate of SHL use. For Puerto Ricans and Mexicans, the higher rates of English use and relatively lower rates of SHL can be explained in large part by their multi-generational history in the U.S. and centurieslong encounter with U.S. imperialism. Dominicans, Salvadorans, and Guatemalans are more recent immigrants. Certainly, the immigration histories of individual Latina/o groups play an important role in the linguistic profiles of U.S. Latina/os. But is immigration history sufficient as an explanation? South Americans had the highest rate of English-only use and nearly the lowest rate of SHL use despite the fact that South American Latina/os from countries like Argentina, Chile, and Venezuela, like Dominicans, are relatively recent immigrant groups in the U.S. One hypothesis is that dark-skinned immigrants are more subject to raced-based discrimination and neighborhood segregation that limits their contact with native English speakers. Another possibility is that white Latina/o immigrants may enjoy other advantages (e.g. higher social class and education) that position them for professional and educational opportunities in which mastery of English is requisite. Clearly the picture is complex, and race alone cannot explain the patterns presented in Table 8.1. My point is to show that considering race in relation to SHL can cast a potentially useful light on the processes of SHL maintenance.

Resisting racialization through linguistic practice There is evidence that the meeting of U.S.-based racial categorization schemes with those of Latin America yield new, more complex systems that influence the lived experiences of Latina/ os in the U.S. For example, Bailey (2001) argues that second-generation Dominicans resist both the outright rejection of black identity that characterizes older generations of Dominicans (see, for example, Duany, 1998; Torres-Saillant, 1998) and the rigid U.S. racial binary. Instead, many young Dominicans develop ethno-linguistic repertoires that reflect their daily interactions with both black and white Americans as well as other Latina/os (as shown in Excerpt 1). In my work in New York City (Negrón, 2011) I show that through linguistic flexibility and cross-cultural competence, Latina/os creatively switch between ethnic and even racial categories in their everyday discourse. Young second-generation immigrant New Yorkers, Latina/os among them, are opting more and more for cosmopolitan identification that does not easily tie them to static ethnoracial categories and related behavioral expectations (Kasinitz et al., 2008). Julia, a 19-year-old Colombian-American woman whom I interviewed, grappled with the words to describe this cosmopolitan orientation (Excerpt 4): Excerpt 4: Julia: And I like being that way. I like having a little bit of aspect from each other person. Like I said my nationality is Colombian. But my culture is . . . you know? Is how I’m living right now. My culture’s . . .  RN: If you were to put a label on it, what would it be? Julia: I don’t even know (Laughs).

116

Race, intra-Latina/o hierarchies in U.S.

Hip-hop dash . . . because you know it has to go to music. Hip-hop dash, Hispanic in general dash . . . I don’t know. Very ethnic, I know little things about different cultures. Yeah, but I don’t know a lot about one specific one. You know? . . . Because I really don’t, I can’t emphasize on one culture because I don’t know about it. I don’t even know much about my own. I don’t, I can’t . . . if you ask me some history on Colombia, I don’t know. You know. Ask me geographics . . .  RN: But is that necessary in order for you to have a sense of being Colombian? Julia: No, not at all. I know what I am. I’m proud of what I am. But I don’t see it as though I have to be labelled just as that. You know? I wasn’t raised in it. You know what I’m saying? It’s different if I was actually brought up in [Colombia] or with parents or in a situation that they thought that was an important thing to emphasize. My parents didn’t emphasize on where I was from. I was raised around . . . the majority of my mom’s friends were all Puerto Rican. My mom used to be tired as hell coming home from work, I was raised on Puerto Rican food. RN: Oh, really? Julia: You know, pasteles [“patties”] or pastelones [“sweet plantain casserole”], you know habichuelas con eso [“beans with that”], you know, I was raised on that. So it’s different. I’m, I . . . I don’t know. It’s different. For Julia, this reluctance to label herself as just Colombian also had implications for negotiations of her racial identity. As she explained in another point of our interview (Excerpt 5): Excerpt 5: RN: So describe the first time you had an awareness of your ethnicity or your race, or your culture, you know, that you were different. Julia: I always knew that, I knew that none of my friends spoke Spanish, I knew that, you know. I came home and I ate different things. I always knew that plus it’s like skin tone. I’m light as hell and my father is dark as shit. You know? RN: So, basically they would call you like the “white” girl? Julia: Nah, I was never called white girl! “Oh, you are a Spanish girl” (laughing). RN: Oh, Spanish girl. Julia: . . . They would be “See you, sweetie, see you sweetie, how do you say that? How do you say this?” That was it. Like, there was a lot of other Spanish people too, there was a lot of Dominicans, Puerto Ricans a lot of Africans but I never felt at home with them because, I don’t know, like I never felt like that at home with like the Dominicans and the Puerto Ricans because to them I wasn’t Spanish enough because I would still hang out, I’ve always been around black people [African Americans]. Yeah, but I don’t know . . . like most of [the] Puerto Ricans whatever were real fair skin and stuff like that like, speak Spanish constantly. You know, it’s easier for me to speak English. I think in English I’m gonna speak in English. I only speak Spanish to my parents. I wasn’t Spanish enough. As a phenotypically white person who identified strongly with African American culture, she recognized that because she was a Spanish-speaker, others would not automatically label her as white. Nevertheless, her preference for English and her greater comfort with her black friends meant that she did not readily identify with other Latina/os.

117

Rosalyn Negrón

The realities of race in the U.S. constrain options and establish incentives for racial selfidentification. Yet we know that racialized people are not without agency, nor subjects of an unchanging caste system (Perlmann and Waldinger, 1997). And we know that race has valence: the meaning and significance of skin color can shift in interaction with other markers of ethnicity like language and cultural knowledge and performance. Through language, Latina/os exercise a number of racial options, which can include tactics of distancing from undesirable racial identities or highlighting desirable ones, as contextually useful to them. At times, such tactics also include crossing to or passing as a member of racial outgroups. The racial status hierarchies indicated by such choices point to the fissures that complicate the forging of a cohesive Latina/o collectivity. For white Latina/os, for example, Spanish-language use can be a way to discursively “become” more black in order to take on a racial minority identity. Consider for example, the following story told to me by Roberto, the Venezuelan first-generation immigrant described earlier, who, because of his light skin and blue eyes, was often assumed to not be Latina/o (Excerpt 6): Excerpt 6: You know one thing when you are in a black neighborhood, right? You don’t want these motherfucking molletos to think you’re white-white! Fuck that! Me hago boricua [“I make myself Puerto Rican”] . . . instantly! Like I remember the last time I got high, I was on my way to cop [buy], and I knew these niggas was not even gonna look at me [and assume I was an undercover cop]. You know what I did? I turned the phone to vibrate so it won’t ring, and I had the thing y me pongo hablar [“and I start to talk”], “Mira que si este, que si lo otro, cla, cla, cla . . . [Look, this and that, blah, blah, blah]. Hablando una conversación con el aire [Having a conversation with the air]! Pero en español [But in Spanish]. En boricua [In Puerto Rican]. Y los tipos ahí [And the dudes there]: Bueno [‘Well’], you’re not white!” Note that it is Puerto Rican Spanish, and not his native Venezuelan Spanish, that Roberto uses to keep the strangers around him from thinking that he was white of the European, normative sort. Through his use of a language variety associated with a marginalized and highly racialized Latina/o subgroup (Puerto Ricans), Roberto reproduces a racial script in which Puerto Ricans as drug buyers are unmarked actors. Roberto also uses the Spanish racial epithet, molleto, to refer to blacks. The racial ideologies operating in his linguistic practices seem to undermine his desire to deflect a white racial identity. Underlying all of this is an attempt to wrest some control over the ascription of race. It is a messy business, one where even as subjects of racialization, Latina/ os reproduce the racial worldview. Often, linguistic acts that circumvent racial categorization aim to deflect racialized stigmas. Abel, an Ecuadorian satellite TV subscription salesman, provides an example of the ways that a devalued indio-Latina/o ancestry led him to distance himself linguistically and discursively from that identity. Upon his arrival in the U.S., Abel worked in construction in New Jersey, where his experiences deeply influenced his racial and ethnic self-presentation: “When I lived in New Jersey, I was ashamed to say I was Ecuadorian,” he admitted. In Abel’s view, Ecuadorian identification was laden with negative connotations, given the predominantly indio and mestizo Ecuadorian population in the New York City area. He subscribed to a racio-geographic folk categorization scheme common in Ecuador in which people from the Ecuadorian coastal cities like Guayaquil (costeños) were distinguished from those who hailed from the highlands (serranos). According to Abel and other Ecuadorians whom I interviewed, highlanders were closely linked to an indigenous or indio identity. While he avoided the categorization, Abel also had described 118

Race, intra-Latina/o hierarchies in U.S.

himself to me as indio and pointed to certain physical characteristics such as his brown skin and nose shape as marking his indigenous heritage. While he sometimes romanticized a certain indio lifestyle, he identified so strongly as a costeño that he used derogatory labels like “tira flecha” (“arrow thrower”) and “cholo” to describe Latina/os with indigenous heritage and even himself (“Yo soy cholo” [“I am a cholo”]). To mitigate the impact of his indio appearance, Abel opted for guayaquileño as his primary identification. Particularly relevant to this discussion of Spanish language use and race is that in addition to distancing himself from indio identity through the use of ethnic labels, Abel sometimes used a Colombian dialect to assume a more white, non-indio identity (Excerpt 7). Excerpt 7: Abel: I didn’t say that I was Ecuadorian. They have to treat me the same as them, the same as other people. I said I was Colombian. RN: And they believed you? People believed you? Abel: Because I changed my accent. Abel’s wish to distance himself from the associations made between Ecuadorian identity and indigeneity, compounded by his own indigenous heritage and status as an undocumented immigrant, motivated him to pass as “anything but Ecuadorian” in an effort to expand his options for mobility in the face of factors that would otherwise limit them. Latina/os’ use of Spanish in general is also a way to shape the racial assumptions people make of them. For example, Toribio (2003: 6) describes a linguistic dilemma experienced by Daniel, a 46-year old Dominican living in the U.S. While he expressed a desire to lose his Spanish accent when speaking in English, this same feature afforded him a way to distinguish himself from African Americans: I went to college two years for English, y no se me ha perdido el acento. ( . . . and I haven’t lost the accent . . . ) I have an accent that is very deep . . . When I’m in the woods over there, they see me, I’m not white, so what I am? I’m minority . . . (They know I’m not black) when they hear me speak. How can Spanish work as a way to resist racialized stigma, if Spanish itself bolsters Latina/o’s non-white categorization? Crucial here is that the negotiation of race for many Latina/os entails creating a space between black and white in the U.S. racial binary. Recall that, while more than half of Latina/os identified as white in the 2010 U.S. Census, a significant number insisted on identifying as neither black nor white. With both blackness and whiteness as reference points, Latina/os use the tools available to them to secure for themselves a favorable racial position. Language is one such tool. The prevailing notion is that within white normative (monolingual) contexts, Spanish has lower status in relation to standard American English.11 But the preceding discussion suggests that the racial valence of Spanish changes across contexts and across speakers. Understanding the contexts in which Spanish is or is not stigmatized, or the cases when Spanish does not neatly signify a Latina/o identity, can improve our understanding of whether the U.S. racial binary is indeed under transformation.

Recommendations for practice Spanish as a heritage language must be viewed within the context of a broader repertoire of linguistic options that Latina/os use to invoke or de-emphasize particular Latina/o identities. The status of specific Latina/o subgroups within the U.S. racial hierarchy affects the status of 119

Rosalyn Negrón

Spanish language varieties spoken by U.S. Latina/os. Thus, applying a racial lens to the study of SHL directs our attention to multiple varieties of Spanish, each potentially subject to different assimilation pressures and patterns of proficiency. The restraints imposed on the use of Spanish in public spheres (Cobas and Feagin, 2008; Urciuoli, 1996) along with U.S.’s rigid racial order, has contributed to “a wide variety of identity configurations and levels of Spanish proficiency among U.S. Spanish speakers” (Potowski, 2012: 183). Quite simply, for some Latina/os, Spanish language use improves their racial position, while for others it worsens it. It is likely that levels of Spanish proficiency are at least partly tied to race, in that a desire for an improved racial position has long been an important motivating force for assimilation. Additionally, spatial segregation along racial lines may reinforce Spanish language maintenance by reducing the level of contact between Spanish and English speakers (c.f. Stevens, 1992). The notion that racialized minority languages and cultures are best kept to the private sphere – rather than accommodated in public, civic, and policy realms – is particularly relevant for heritage language studies. Heritage languages bind individuals to social groups in ways that are personal and familiar (Fishman, 2001). SHL studies have as a key aim the maintenance of the Spanish language in the U.S., particularly in the face of assimilation pressures, and as we have seen, in the face of racialization processes that devalue Spanish. Yet Spanish language maintenance among U.S. Latina/os is tied to the acceptance and accommodation of Spanish in the public spheres. SHL, a language of the familiar and of personal space, must be legitimized in the public space through policy that at its core recognizes the value of linguistic diversity in the U.S. and views language discrimination as perpetuating racial injustice. Increased acceptance of language diversity undermines the myth of the monolingual nation.12 By extension, language diversity inscribed in policy would counter the deeply embedded, historical notion that to be a rightful part of the U.S. nation one must be white and English-speaking. Schools are key sites for such policies. In the domain of education, SHL studies can push against the historical role that schools in the U.S. have played as places where minority languages and cultures went to die. Indeed, schools have supported racial formation processes in multiple ways, which include advancing assimilation and minimizing non-white contributions to building the U.S. nation. For example, schools were an instrumental part of governmental policy to acculturate Native American children as a way to erase Native American cultures (Adams, 1995). More recently, Arizona and Texas have implemented policies to prohibit ethnic studies in high school curricula or to block efforts by Latina/o educators to include more Latina/o historical figures in social studies curricula (McKinley, 2010). The fact that black and Latina/o students are most likely to attend low performing schools is a further example of how schools reproduce race and class positions (Orfield and Lee, 2005). Policy and curricular initiatives that promote SHL maintenance in schools, and that do so with an explicit awareness of Spanish dialectal diversity, are a way to protect and expand language diversity in the U.S. Such campaigns must be part of a wider movement to transform discourses about language diversity in the U.S., binding the interests of heritage language speakers to discourses that frame bi-/multilingualism as essential to global citizenship. But efforts to promote SHL maintenance would be bolstered by coordinated efforts among activists, policy-makers, and educators mobilizing on behalf of Latino/a communities. To this end, transformation is also needed in inter-racial understanding among Latino/as, given the intra-Latino/a racial divisions described in this chapter. Latino/a and other SHL educators should examine the racial dimensions of any biases and assumptions they may have about different Spanish varieties, and integrate material on race, and on Spanish language diversity into their curricula. SHL educators can also play an important role in fostering dialogue among student heritage speakers about language and race, and challenging racializing attitudes that shape language preferences, practices, and beliefs. 120

Race, intra-Latina/o hierarchies in U.S.

Last, as the material realities of race make clear, any movement that frames linguistic diversity as a matter of racial justice must address the political and economic logics that preserve racial inequality. Here too policy is critical, and the political advocacy of SHL researchers and educators ever more important. Linguistic research that illuminates the links between language discrimination and racial discrimination must be communicated to state and government officials through statements and policy briefs, and through coordinated efforts with racial justice organizations.

Notes 1 A panethnicity is an umbrella ethnic group that includes multiple nationalities and cultures that share a sense of common ancestry. 2 We could also include multiple indigenous languages such as Maya or Quechua in this linguistic diversity. However, not all indigenous groups from Latin America identify as Latina/o. 3 By Latin America I refer to the Spanish-speaking countries of South, Central, and North America and the Caribbean. 4 While the black-white binary has been a persistent organizing principle in the U.S. racial order, there are some common U.S. folk categories like “mixed,” “light-skinned,” and “brown,” which suggest that the differences between Latin American and U.S. racial categorizations are not as stark as traditionally thought. 5 The recent passing of Proposition 58 in California, which reinstates bilingual instruction in public schools, may foretell a nationwide shift away from English-only education policies. 6 These debates miss that bilingualism and language acquisition are not mutually exclusive; immersion in multiple languages actually promotes English acquisition (Ramírez et al., 1991; Lindholm-Leary, this volume). 7 Cuchifrito, which are fried foods that include pork and various sorts of patties, are associated with Caribbean Latina/o cultures, in particular with Puerto Rican food. 8 The name of the magazine has been altered to protect William’s anonymity. I chose a replacement name that retained certain cultural connotations in the original. 9 Weakened coda /s/ is present in both Caribbean and mainland varieties of Spanish (Otheguy and Zentella, 2012). 10 Reprinted with permission from the National Academies Press, Copyright 2006, National Academy of Sciences. 11 Hispanicized English forms like Chicano English are also racialized and stigmatized in relation to standard American English (Fought, 2003). 12 This is crucial because despite the linguistic diversity in the U.S., pressures against bilingualism are strong being that 80% of the U.S. population reports speaking only English at home (U.S. Census, 2011b).

Acknowledgements I would like to credit the American Anthropological Association, publisher of the Journal of Linguistic Anthropology; the National Academies Press, publisher of Hispanics and the Future of America; and Bilingual Review Press at Arizona State University, publisher of Bilingual Review|La Revista Bilingüe, for their content used this chapter.

Further reading Bucholtz, M. (2001). From mulatta to mestiza. In K. Hall & M. Bucholtz (Eds.), Teaching modern foreign languages: A handbook for teachers (351–373). New York: Routledge. Describes how women with ambiguous or multi-ethnic identities use language to negotiate multiple ethnic and racial categories. Davis, T. Y., & Moore, W. L. (2014). Spanish not spoken here: Accounting for the racialization of the Spanish language in the experiences of Mexican migrants in the United States. Ethnicities, 14(5), 676–697. The authors show how the use of Spanish by first-generation Mexican immigrants functions as a racial marker that signals Latina/o immigrants’ racially subordinate status in the U.S.

121

Rosalyn Negrón Mendoza-Denton, N. (1999). Sociolinguistics and linguistic anthropology of U.S. Latinos. Annual review of Anthropology, 28, 375–395. With a focus on Chicana/os and mainland Puerto Ricans in the U.S., the author provides a comprehensive review of the microstructural characteristics of Spanish language varieties and their sociolinguistic use. Pimentel, C. (2011). The color of language: The racialized educational trajectory of an emerging bilingual student. Journal of Latinos and Education, 10(4), 335–353. Examines race and language ideologies that shaped educators’ perceptions of Latina/o students’ abilities, through the case of a Latino student’s progress in bilingual education programs. Schmidt, R. (2002). Racialization and language policy: The case of the USA. Multilingua, 21(2/3), 141–162. Based on an analysis of the discursive social context of language policy proposals, the author argues that pluralist language policies along with policies that address economic inequalities are better poised than English-only policies to address the injustices of racialization in the U.S.

References Adams, D. W. (1995). Education for Extinction: American Indians and the Boarding School Experience, 1875–1928. Lawrence, KS: University Press of Kansas. American Anthropological Association (1998). American Anthropological Association Statement on Race. Understanding Race, May, 17. Retrieved from www.aaanet.org/stmts/racepp.htm. Bailey, B. (2000). The language of multiple identities among Dominican Americans. Journal of Linguistic Anthropology, 10(2), 190–223. Bailey, B. (2001). Dominican-American ethnic/racial identities and United States social categories. International Migration Review, 35(3), 677–708. Barker, M. (1981). The new racism: conservatives and the ideology of the tribe. Junction Books. Bedolla, L. G. (2003). The identity paradox: Latino language, politics and selective dissociation. Latino Studies, 1(2), 264–283. Cobas, J. A., & Feagin, J. R. (2008). Language oppression and resistance: The case of middle class Latinos in the United States. Ethnic and Racial Studies, 31(2), 390–410. Córdova, D., & Cervantes, R. C. (2010). Intergroup and within-group perceived discrimination among US-born and foreign-born Latino youth. Hispanic Journal of Behavioral Sciences, 32(2), 259–274. De Genova, N., & Ramos-Zayas, A. Y. (2003). Latino crossings: Mexicans, Puerto Ricans, and the politics of race and citizenship. New York: Routledge. Duany, J. (1998). Reconstructing racial identity: Ethnicity, color, and class among Dominicans in the United States and Puerto Rico. Latin American Perspectives, 25(3), 147–172. Duany, J. (2010). Whitening (Blanqueamiento). El Nuevo Día, May, 13. Retrieved from www.prdream. com/wordpress/topics/2010/05/2135/. Fishman, J. (2001). 300-Plus years of heritage language education. In J. K. Peyton, D. A. Ranard, & S. NcGinnis (Eds.), Heritage languages in America: Preserving a national resource in the United States, 87–97. McHenry, IL: Delta Systems Inc. and Washington, DC: Office of Educational Research and Improvement. Fought, C. (2003). Chicano English in context. New York: Palgrave Macmillan. Ghosh Johnson, S. E. (2006). Mexiqueño? Issues of identity and ideology in a case study of dialect contact. Doctoral dissertation, University of Pittsburgh, PA. Gould, S. J. (1996). The mismeasure of man. New York: WW Norton & Company. Gravlee, C. C. (2005). Ethnic classification in southeastern Puerto Rico: The cultural model of ‘color’. Social Forces, 83(3), 949–970. Grieco, E., & Cassidy, R. (2001). Overview of race and Hispanic origin, Census 2000 Brief, C2KBR/01–1. Washington, DC: U.S. Census Bureau. Hirschman C., Alba R., & Farley R. (2001). The meaning and measurement of race in the U.S. census: Glimpses into the future. Demography, 37(3), 381–393. Huntington, S. P. (2004). The Hispanic challenge. Foreign Policy, 141(2), 30–45. Kasinitz, P., Mollenkopf, J., Waters, M. C., & Holdway, J. (2008). Inheriting the city: The children of immigrants come of age. New York: Russell Sage Foundation. Lippi-Green, R. (2012). English with an accent: Language, ideology, and discrimination in the United States, 2nd edition. New York: Routledge. McKinley Jr., J. C. (2010). Texas conservatives win curriculum change, The New York Times, March 12. Retrieved from www.​nytimes.​com/​2010/​03/​13/​education/​13texas.​html.

122

Race, intra-Latina/o hierarchies in U.S. Mendoza-Denton, N. (2014). Homegirls: Language and cultural practice among Latina youth gangs. New York: John Wiley & Sons. Negrón, R. (2011). Ethnic identification among urban Latinos: Language and flexibility. El Paso, TX: LFB Scholarly Publishing. Negrón, R. (2014). New York City’s Latino ethnolinguistic repertoire and the negotiation of latinidad in conversation. Journal of Sociolinguistics, 18(1): 87–118. Orfield, G., & Lee, C. (2005). Why segregation matters: Poverty and educational inequality. K-12 integration and diversity. Retrieved from http://escholarship.org/uc/item/4xr8z4wb. Otheguy, R., & Zentella, A. C. (2012). Spanish in New York: Language contact, dialectal leveling, and structural continuity. Oxford, UK: Oxford University Press. Perlmann, J., & Waldinger, R. (1997). Second generation decline? Children of immigrants, past and present-a reconsideration. International Migration Review, 31(4), 893–922. Pew Research Center (2015). The many dimensions of Hispanic racial identity. Multiracial in America: Proud, diverse and growing in numbers, June. Washington, DC. Potowski, K. (2012). Identity and heritage learners. Moving beyond essentializations. In S. M. Beaudrie & M. Fairclough (Eds.), Spanish as a heritage language in the United States: The state of the field, 179–199. Washington, DC: Georgetown University Press. Ramírez, J., Yuen, S., Ramey, D., & Pasta, D. (1991) Final report: Longitudinal study of structured English immersion strategy, early-exit and late-exit bilingual education programs for language-minority children, Vols. I & II. Prepared for the U.S. Department of Education. San Mateo, CA: Aguirre International. Rumbaut, R. G. (2006). The making of a people. In National Research Council (US) Panel on Hispanics in the United States; M. Tienda and F. Mitchell (Eds.) Hispanics and the Future of America (16–65). Washington, DC: National Academies Press. Retrieved from www.ncbi.nlm.nih.gov/books/ NBK19896/. Smedley, A. (1998). ‘Race’ and the construction of human identity. American Anthropologist, 100(3), 690–702. Stevens, G. (1992). The social and demographic context of language use in the United States. American Sociological Review, 57(2), 171–185. Stutzman, R. (1981). El mestizaje: An all-inclusive ideology of exclusion. In N. Whitten (Ed.), Cultural transformation and ethnicity in Modern Ecuador (45–94). Urbana, IL: University of Illinois Press. The Nielsen Company. (2016). From the ballot box to the grocery store: A 2016 perspective on growing Hispanic influence in America report. Retrieved from www.nielsen.com/us/en/insights/reports/2016/from-theballot-box-to-the-grocery-store-hispanic-influence-in-america.html. Toribio, A. J. (2003). The social significance of Spanish language loyalty among Black and White Dominicans in New York. Bilingual Review/La Revista Bilingüe, 27(1), 3–11. Torres-Saillant, S. (1998). The tribulations of blackness: Stages in Dominican racial identity. Latin American Perspectives, 25(3), 126–146. Torres-Saillant, S. (2002). Problematic paradigms: Racial diversity and corporate identity in the Latino community. In M. Suárez-Orozco and M. Páez (Eds.), Latinos: Remaking America (435–455). Berkeley, CA: University of California Press. Tucker C., McKay R., & Kojetin B. (1996). Testing methods of collection racial and ethnic information: Results of the current population survey supplement on race and ethnicity. BLS Statistical Notes 40. Washington, DC: Bureau of Labor Statistics. Urciuoli, B. (1996). Exposing prejudice: Puerto Rican experiences of language, race, and class. Long Grove, IL: Waveland Press. U.S. Census Bureau. (2011a). Overview of race and Hispanic origin: 2010. 2010 Census Briefs. Retrieved from www.census.gov/prod/cen2010/briefs/c2010br-02.pdf. U.S. Census Bureau. (2011b). Language use in the U.S. American community survey. Retrieved from www.census.gov/content/dam/Census/library/publications/2013/acs/acs-22.pdf. Wade, P. (2005). Rethinking mestizaje: Ideology and lived experience. Journal of Latin American Studies, 37, 239–257. Zentella, A. C. (1990). Lexical leveling in four New York City Spanish dialects: Linguistic and social factors. Hispania, 73(4), 1094–1105. Zentella, A. C. (1997). Growing up bilingual: Puerto Rican children in New York. New York: Wiley-Blackwell. 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.), A companion to Latina/o studies (25–38). Oxford, UK: Blackwell Publishing.

123

9 QUEERING SPANISH AS A HERITAGE LANGUAGE Holly Cashman university of new hampshire, usa

Juan Antonio Trujillo oregon state university, usa

Introduction In this chapter, we examine materials and practices in SHL programs that constitute discrimination or that otherwise contribute to negative affect for learners who identify as both Latinx1 and queer—a term we will use interchangeably with LGBTQ+2 to refer to people whose sexual and/or gender identity do not align with social norms that collectively fall under the label of heteronormativity. Our examination centers on an analysis of the representation and social positioning of queerness and of LGBTQ+ people in SHL textbooks and is supported by discussions about the ways queer-identified Latinxs relate to the Spanish language and Spanish-speaking social networks and by reflections on the role and status of queer faculty in the SHL classroom. We conclude with thoughts about ways that contesting heteronormativity constitutes good pedagogical practice and ways to construct classroom environments that are more supportive of LGBTQ+ identities. Heteronormativity refers to the complex of socially privileged norms established by heterosexuals that impose a strict male/female sexual binary, compel adherence to distinct gender roles, and constrain sexual or romantic relationships to opposite-sex couples. Research shows that the enforcement of heteronormativity generates a hostile climate in school settings for queer youth (see Kosciw et al., 2014; Rankin, 2006). Warner (1991), queer theorist who first popularized the term, decried heterosexual culture’s “totalizing” tendency and its “exclusive ability to interpret itself as society” (p. 8), highlighting the especially negative consequences for young people whose identities stray beyond heteronormative parameters: Heterosexual ideology, in combination with a potent ideology about gender and identity in maturation, therefore bears down in the heaviest and often deadliest way on those with the least resources to combat it: queer children and teens. p. 9 Enforcement of heteronormative ideology in school settings may take many forms, including exposure to biased language, harassment, restrictions on participation in activities, prohibition 124

Queering Spanish as a heritage language

of LGBTQ+ content, and policing of gender identity or expression (Kosciw et al., 2014). Our chapter directly addresses the ways that SHL materials and practices may represent this type of hostile ideological enforcement. This chapter adopts models of sexual identity articulated by queer theorists (e.g. Butler, 1990; Foucault, 1982) who espouse an expansive, poststructuralist view of social constructivism—one that regards sex, gender, and sexual attraction as separate constructs, not fixed, biologically determined classifications. Sexual orientation, sex, and gender are understood to be fluid and subject to interaction with other socially constructed facets of identity such as race and ethnicity. It is the intersection of identity markers that defines one’s positionality within systems of power and privilege. The intersection of queerness and Latinx identity, for example, creates potential for the racialization of a sexual identity, which in turn leads to compounded marginalization within most domains in U.S. society. That said, important implications of a broad view of social construction and intersectional theory are that the identity “crossroads”—to borrow Anzaldúa’s (2012) metaphor—can be a position of power and that reconstruction can bring greater social and political self-realization for queer-identified individuals. Although we have adopted terminology and concepts from mainstream contemporary queer theory, queer of color criticism challenges the Eurocentrism of this approach and its failure to address issues like race and class that form part of the complex identities present in the SHL environment. As Hames-García (2011) notes: Indeed transformative understandings of the relationships among race, capitalism, gender and sexuality are probably stronger precisely when elaborated from outside of Eurocentric frameworks and intellectual genealogies rather than from within them. p. 42 We argue that adequate representation of queer Latinx identities in the SHL setting means going beyond the discourse of marriage equality, adoption, and open military service. These concepts are arguably heteronormative, white-centric, and “homonationalist”— tied to maintaining nationalist/capitalist structures (Puar, 2007)—and fail to address issues such as racialized violence, economic injustice, and racial privilege that impact queer communities of color. Indeed, elements of this analysis fall outside the life experience or academic training of many queer-identified language faculty and textbook authors.

Queer Latinxs, language, and identity We do not have evidence that LGBTQ+ Latinx language proficiency and language practices are qualitatively different from those of non-queer Latinxs. Queer Latinxs appear to fall on the same linguistic continuum described in current literature (e.g. Valdés, 2001), with speakers that occupy all points of a spectrum between monolingual English and monolingual Spanish extremes. LGBTQ+ Latinxs are born in the U.S. and immigrants who arrived as adults; they are DREAMers and they are citizens; they are out of the closet and they are silent about their sexuality; they live in cities and in rural areas. In other words, the diversity of people in the U.S. who are both Latinx and LGBTQ+—in terms of language proficiency, language practices, life trajectory, immigration status, and degree of openness about their sexuality—is such that making generalizations is impossible. What is certain, however, is that there are queer Latinx students and teachers in SHL classrooms across the country whose experience merits attention. Motivation to learn a language may be impacted by affective factors associated with the language, culture, and speakers. Fishman (2001) defines heritage languages by learners’ personal 125

Holly Cashman and Juan Antonio Trujillo

and cultural connection to these factors. Valdés (2001) notes that this gives them a motivation for learning language that differs from that of non-heritage learners. Although there is research on the relationship between Latinx students’ attitudes and motivations to study SHL (reviewed in Ducar, 2012, for example), there is no exploration of the language attitudes of LGBTQ+ Latinxs, their experiences as heritage language learners (HLLs), or their motivations to study the language. We hope that this brief discussion underscores a significant gap in the literature and a need to know more about LGBTQ+ HLLs. In her exploration of language maintenance and social ties among LGBTQ+ Latinxs, Cashman (2015) describes the interconnectedness of language, ethnic identity, and sexuality. She notes that: It is most often the case for bilingual Latin@s in the US that their parents, grandparents, and members of their generations are the most consistent Spanish language influences; yet it is precisely these family ties that are most threatened by when queer Latin@s come out . . . In other words, it is precisely the people who are most proficient in Spanish who are most likely to react negatively toward LGBTQ+ people, such as a son or daughter or other relative coming out. If Spanish-speaking network ties tend to promote the maintenance of Spanish, as has been argued elsewhere, then it follows that coming out might threaten Spanish language maintenance among LGBTQ+ Latin@s. p. 73 Although Cashman uses the term “bilingual,” many of the participants in her ethnographic research do not consider themselves bilingual for a variety of reasons, and in interviews many revealed complicated relationships with the Spanish language. LGBTQ+ Latinxs are not necessarily more alienated from Spanish than their heterosexually identified peers due to their sexuality or gender identity and expression. However, we believe that sexuality and gender identity may interact with other factors in the lives of LGBTQ+ Latinxs to promote alienation from Spanish with a resulting impact on their relationship with the language and/or their orientation toward SHL classes and curricula. Although we do not have specific information about the experiences of queer Latinxs in language classes, we do know a great deal more generally about the experiences of LGBTQ+ and non-binary students. Data from the 2013 National School Climate Survey (Kosciw et al., 2014) published by the Gay, Lesbian & Straight Education Network (GLSEN) tell us that over half of LGBT students felt unsafe at school due to their sexual orientation, and over a third of students felt unsafe due to their gender expression. More than seven of ten LGBT students heard “gay” used negatively frequently or often at school, nearly two-thirds heard other homophobic remarks frequently or often, and over half reported hearing homophobic remarks from teachers and staff. Against this backdrop, the survey found that LGBT students in schools with Gay-Straight Alliances, inclusive curricula, and supportive faculty and staff were less likely to hear harassing or homophobic speech, less likely to report feeling unsafe at school, less likely to report missing school because they felt unsafe or uncomfortable, more likely to report feeling connected to their school community, had higher GPAs, and were more likely to continue their education. Among the five recommendations made by the survey’s authors, is “Increasing student access to appropriate and accurate information regarding LGBT people, history, and events through inclusive curricula and library internet resources” (Kosciw et al., 2014, p. 14). The GLSEN survey focuses on student experiences in the K-12 setting, but there has been less research of this kind at the post-secondary level (Renn, 2010). Rankin (2006), in a survey of 126

Queering Spanish as a heritage language

the literature on the experiences of LGBTQ+ students on college campuses, summarizes that, “These studies suggest high rates of victimization of LGBT people in the form of verbal harassment, physical assault, and negative campus climate” (p. 114), and she emphasizes that “this is particularly true for minorities within the LGBT community such as students with disabilities, transgender students, and students of color” (pp. 113–114). She noted further that many studies across different social groups indicate a connection between perception of discriminatory environment and academic attainment, while recognizing that “students from different social identity groups experience, or at least perceive, campus climates differently” (p. 113).

Heteronormativity in SHL textbooks and materials As we described earlier, heteronormativity prescribes a social framework rooted in acceptance of biologically determined binary sexes leading to couple relationships between oppositesex people. This framing of heterosexual couples as a socially privileged model is reinforced through legal institutions such as marriage and through cultural representation in literature, media, and visual arts. Despite the rapidly changing legal status of same-sex relationships in many places around the world (e.g. marriage equality at the national level in the U.S. and in many Latin American countries), queer folk, and particularly queer and trans(gender) people of color (QTPOC), face real and continued discrimination in housing, employment, health care, and other arenas. Avoiding heternormativity requires going beyond positive representations of same-sex marriage and coming out. This is especially important to understand in the SHL environment where participants are not only queer, but overwhelmingly QTPOC, whose life experience has not been well represented in mainstream queer theory or by national LGBTQ+ political organizations (Hames-García, 2011). For example, anti-trans(gender) violence soared while the nation’s attention was focused on events leading to the landmark marriage decision in the Supreme Court. The Human Rights Campaign (2015) reports that of the 53 known transgender murder victims from 2013–2015, at least 87% were people of color. SHL materials have been critiqued with regard to Spanish language ideologies (Leeman, 2014, 2012, 2005; Leeman & Martínez, 2007), Spanish language variation (Beaudrie, Ducar, & Relaño-Pastor, 2009; Ducar, 2009; Martínez, 2003), and representation of Latinx cultures (Ducar, 2006), but not for their treatment of LGBTQ+ identities. This lack of attention by researchers appears to follow a silencing of queerness by textbook authors as well. Despite a growing awareness of the need for greater racial and ethnic diversity, and the elimination of even covert racism and sexism in the curriculum, inclusion of LGBTQ+ identities appears to lag far behind. This invisibility in teaching materials can negatively impact queer students’ experiences in the classroom (Gray, 2013; Nelson, 2008). Liddicoat (2009) asserts that “For gay and lesbian students, the language classroom can be an environment that silences much that is central to their lives” (p. 191). Nelson (2006) describes the language teaching profession’s collective notion of a “monosexual community of interlocutors” in which “straight people are interacting exclusively with other straight people” (p. 1), leading to one-dimensional understandings of the language student and language communities. Attempts to assert non-heterosexual identities are often treated as examples of linguistic failure (Liddicoat, 2009). ESL and EFL researchers began to ask about the experience of gay and lesbian learners and the invisibility of queer people in language textbooks and curricula over two decades ago (see Nelson, 1993; Snelbecker, 1994), but even relatively recent research indicates that ESL and EFL textbooks still fail to be inclusive of LGBTQ+ individuals (Goldstein, 2015; Paiz, 2015; Pawelczyk & Pakuła, 2015; Sunderland & McGlashan, 2015). Discussions of sexual diversity in the teaching of languages other than English, including SHL, have lagged behind ESL/EFL. 127

Holly Cashman and Juan Antonio Trujillo

To our knowledge, there are no studies of representations of LGBTQ+ people in Spanish language textbooks designed for native English speakers, so our discussion of SHL textbooks is necessarily lacking a comparison to L2-centered Spanish language textbooks. Even in very recently published overviews of the state of SHL theory and practice (Beaudrie, Ducar, & Potowski, 2014; Beaudrie & Fairclough 2012), there is no reference to queer or LGBTQ+ identities in the SHL classroom or curriculum. SHL professionals might find themselves pondering the following questions that we intend to address in the analysis that follows: Are LGBTQ+ identities represented in the most recent SHL textbooks and materials? If so, how? To what extent are any apparent efforts at inclusion reinforcing a heteronormative ideological frame rather than opening up the conversation to include queer and feminist of color critiques of the existing social order? Do the materials create space for the negotiation of a QTPOC identity within an Anglo-dominated environment?

Data and analysis In order to examine the treatment of LGBTQ+ identities in SHL textbooks, we chose the following nine texts (see reference section for full bibliography): Entre mundos, 2nd ed. (Alonso-Lyrintzis & Zaslow, 2004) Conozcámonos! (Mrak & Aponte, 2007) ¡Sí se puede! (Carreira & Geoffrion-Vinci, 2008) Español escrito, 6th ed. (Valdés, Teschner, & Enríquez, 2008) Conversaciones escritas (Potowski, 2011) Nuestro idioma, nuestra herencia (García, Carney, & Sandoval, 2011) Manual de gramática y ortografía para hispanos, 2nd ed. (Francés & Benítez, 2012) La lengua que heredamos, 7th ed. (Márquez, 2012) Nuevos mundos: lectura, cultura y comunicación; curso de español para bilingües, 3rd ed. (Roca, 2012) We limited our examination to textbooks published from 2004 to present and only included books aimed at the post-secondary level, since discussions of sexuality or sexual diversity in K-12 curricula remain limited. Following Gray (2013), which explores LGBTQ+ invisibility in English language teaching materials in the UK, we examined the textbooks closely for their representation of LGBTQ+ people and compared this with their representation of heterosexuality. There were five textbooks that included no representation of LGBTQ+ people or topics: ¡Conozcámonos!, ¡Sí se puede!, Entre mundos, Manual de gramática y ortografía para hispanos, and Nuestro idioma, nuestra herencia. Manual de gramática focused almost exclusively on grammar rather than culture, thus the majority of the textbook consists of grammar, vocabulary, and spelling/accentuation exercises with little representation of any type of sexuality. In Entre mundos, ¡Conozcámonos!, Nuestro idioma, nuestra herencia, and ¡Sí se puede!, however, the total lack of representation of LGBTQ+ people was accompanied by some degree of representation of ‘traditional’ families and heterosexual relationships or desire. We found the most extreme and remarkable example of this in ¡Conozcámonos!, where the textbook’s readings and grammar activities all revolve around the Guerra family, a Mexican American family of five whose

128

Queering Spanish as a heritage language

family tree even appears on the inside front cover of the textbook for reference. Topics of readings and cloze grammar and vocabulary activities include how the family celebrates holidays, for example, and marriage difficulties between the family’s oldest son and his wife emerge as a topic. The diversity in the Latinx community is represented through the opposite-sex mates each of the adult children is paired with—from El Salvador, Miami (Cuban), and New York (Puerto Rican). Through the use of the trope of the Guerra family, ¡Conozcámonos! roots itself in a framework that privileges heterosexuality, and this privileging is underscored by the complete erasure of queer people. Other textbooks’ heteronormative frameworks are more subtle and varied than that of ¡Conozcámonos! In Entre mundos, for example, we find activities that are quite typical instances of heteronormativity in language textbooks. In one activity, students fill in the blanks in a paragraph reportedly describing the results of a survey on what girls (chicas) look for in a man (hombre). In a follow-up activity, students must write a paragraph on the topic of what men look for in a woman, after conducting an informal survey of their classmates. This same idea is echoed in chapter 9 in a game show-style activity in which students have to attempt to guess the top five things a man looks for in the ideal woman. Even more problematic is a writing activity in which students are compelled to describe what they look for in an ideal mate. This type of activity, while innocuous at first glance, creates a situation where students are forced to out themselves (which they might not be prepared to do), put on an inauthentic heterosexual identity (which no one should be forced to do), or silence themselves, opting not to participate actively in the activity (which goes against the very purpose of this kind of activity—to encourage students to use the language, engage with the material, and participate actively in class). Entre mundos also has two chapters on families, a site where heteronormativity is likely to be found. Here, however, the first chapter on families seems to be concerned mainly with already problematic tasks of contrasting Anglo and Latinx families, while the second focuses on the question of divorce, which is presented as a kind of social problem. In both chapters, the normativity is clear; they are talking about heterosexual parent-headed families. All these activities and topics reinforce the invisibility of queer lives, creating a potentially hostile classroom environment for queer students and instructors. In both ¡Sí se puede! and Nuestro idioma, nuestra herencia the family is generally presented as headed by a heterosexual couple with children. In a discussion of naming practices, for example, in the first chapter of ¡Sí se puede!, there is a discussion of what happens when women marry men and have children, and students are asked to brainstorm male and female names, normalizing the gender binary and heterosexual coupling practices. Where LGBTQ+ people, facts/events, or contributions could be mentioned, they generally are not, such as in a list or profile of writers or artists at the end of the chapter, or in readings on contemporary issues (e.g. why not include LGBTQ+ tourism in the extensive list of tourism types in chapter 4 of Nuestro idioma, nuestra herencia). Where LGBTQ+ folks are mentioned—like ¡Sí se puede! references to Frida Kahlo (pp. 24–25) and Ricky Martin (p. 118), or Nuestro idioma, nuestra herencia discussing Pedro Almodóvar in profiles of Penelope Cruz and Antonio Banderas (pp. 183–184) and Ricky Martin in a short reading about Shakira (p. 28)—their sexuality is not discussed. Of course, in Nuestro idioma, nuestra herencia, none of the profiles mentions anything about the (hetero)sexuality, gender identity, or romantic/sexual relationships of any individuals, and the textbook’s creative structure avoids the usual traps (such as a chapter on the family). This does not mean, however, that heteronormativity is not woven into the book in more subtle ways. In spelling and grammar exercises and some readings as well, the traditional family headed by a mother and father with children comes into play again and again, as do sentences involving heterosexual desire:

129

Holly Cashman and Juan Antonio Trujillo

1 2 3

Mi abuelo me (18. explicar) que las historias de entonces eran sencillas y las actrices María Félix y Dolores del Río (19. hacer) que valiera la pena ir al cine. (p. 49) El actor de la nueva película es admirado por todas las chicas. (p. 125) Los señores Ramos quieren que su hija estudie en la Universidad de Sevilla. (p. 175)

A reader aggressively looking for any kind of queer content might jump on the mention of Paul Leduc (1983), among a list of more respected Mexican cineastes from the 1980s for his direction of Frida: naturaleza viva, an artistic film about the bisexual artist in which her sexuality is explored, or the later Hollywood film Frida, directed by Julie Taymor (2002) and including music from the iconic Mexican lesbian singer Chavela Vargas. In one activity in this same chapter, Y tu mamá también and Temporada de patos are mentioned as possible films to investigate further. These examples, however, require background knowledge that the student likely would not have, and they are incredibly subtle compared to the explicit and regular inclusion of heterosexuality.

Challenges of “inclusion” Four of the nine textbooks we reviewed included some degree of explicit inclusion of LGBTQ+ people or themes: La lengua que heredamos, Nuevos mundos, Conversaciones escritas, and Español escrito. In La lengua que heredamos, we see a minimal effort to include one gay celebrity, singer Ricky Martin, in a chapter about Puerto Rico. The inclusion takes the form of a reading on Ricky Martin’s career, his becoming a father, and his decision to come out of the closet. The reading is relatively superficial but mentions the potential threat that coming out could pose to his success and the criticism that he has received. The reading (pp. 119–121) is followed by two sets of questions, the first comprehension-related and the second reflection-oriented. The reflection questions ask students to give their opinion on homosexuality and beliefs about whether homosexuality is genetic or learned, whether it’s important to know artists’ sexual orientation, whether knowing a friend was gay would change your affection or respect for them, or if they are in favor of laws to protect against discrimination: 1 2 3 4 5

¿Qué opina usted de la homosexualidad? ¿Cree que es algo congénito, o una orientación aprendida? ¿Es importante para usted saber la orientación sexual de los artistas? ¿Sabe cuál es la palabra apropiada para referirse a una mujer homosexual? Si usted descubre que un familiar suyo o una amistad es homosexual, ¿cambiaría en algo su afecto o respeto por esa persona? ¿Está a favor de las leyes que protegen contra todo tipo de discriminación, o hace alguna excepción? Márquez, 2012, p. 121

Later in the chapter, one of three options for reflection and writing are the “asuntos . . . muy controversiales” of same-sex marriage and adoption by same-sex couples (p. 123). The questions, although they may appear neutral, are skewed toward a rejection of LGBTQ+ people and queer visibility. The use of “homosexual” rather than “gay” or “queer” when both of these Englishlanguage words are used extensively in Spanish-language contexts, already slants the questions toward a pathologizing orientation and brings along the issues of power, legitimacy, and the policing of sexual identities. The framing of the issues of LGBTQ+ rights as controversial is not an objective description of a landscape of public opinion that is much more complicated. Putting LGBTQ+ identities up for debate is a strategy for inclusion seen in multiple textbooks, but the frequency of its use should not numb us to the alienating effect that it could have on the queer 130

Queering Spanish as a heritage language

student or faculty member. The profile of Ricky Martin may be a positive example (albeit obvious and totally devoid of nuance—a use of the “good gay” token described by Bell & Binnie, 2000), but any benefit of this inclusion is immediately undercut by how the inclusion is framed. In Nuevos mundos, more than in any other textbook reviewed, there is a sustained effort to include LGBTQ+ people and themes, from the mention of a character’s sexuality in El beso de la mujer araña as part of a profile of Puerto Rican actor Raúl Juliá or a photo of a tuxedo-clad, male couple and a mention of gay rights and legalization of same-sex marriage in an article about contemporary Spain, to a reading on marriage equality in Argentina in a chapter on human rights. The activities and questions around the inclusion seem carefully designed to avoid making LGBTQ+ rights a topic of debate while at the same time fostering discussion. This difficult balancing act can be exemplified by the story about Ricky Martin, one of several celebrities featured in the chapter on Puerto Rico and Puerto Ricans. The article, which is the same as that featured in La lengua que heredamos, is followed by questions that differ markedly from those discussed above. Although the pre-reading questions focus on popular interest in the lives of celebrities and the personal vs. public divide, a question about how coming out might help to combat stereotypes about “hombres y mujeres gay” is included. The questions following the reading include:   1 ¿Por qué cree que pudiera ser tan difícil para algunas personas salir del closet? ¿Qué cree que hace que una persona no se acepte a sí mismo tal y como es y que se niegue a sí mismo y a otros acerca de su identidad o sentimientos verdaderos?   2 ¿Qué otros grupos en la historia del mundo han tenido que luchar por sus derechos? ¿Cómo ha sido la lucha? Dé ejemplos y comenten qué luchas se han ganado y cuáles quedan por ganarse.   3 ¿Sabe usted cuál es la situación legal de los gays y lesbianas en los Estados Unidos? ¿En su estado? ¿Se pueden casar legalmente? En los estados en que existe una unión civil o un registro de domestic partner, ¿cómo es diferente a un matrimonio entre heterosexuales?   4 ¿Conoce algunos países donde los gays se pueden casar por ley civil y tienen los mismos derechos que los heterosexuales? Mencione algunos y, si tiene tiempo, investigue para la próxima clase cuál es la situación en algunos de estos países (Canadá, Argentina, España, Gran Bretaña, Bélgica, por ejemplo) para compararlos con los Estados Unidos.   5 ¿Por qué razones cree usted que en estos países se logró legalizar el matrimonio gay mientras que en los Estados Unidos no se ha logrado uniformemente a nivel nacional?   6 ¿Qué ocurre en países como China, Irán o Uganda? Investigue si le interesa averiguar y compartir la información en clase.   7 ¿De qué manera cree que la salida del clóset de Ricky Martin puede ayudar a otras personas a aceptarse a sí mismo o a si misma?   8 ¿Cree que los ataques del cardenal Luis Aponte Martínez contra Ricky Martin son injustos? ¿Y qué cree que quiere decir Ricky Martin con su contestación cuando dice que él sólo se enfoca en el amor y en la igualdad . . . y dice «Yo solo sumo, no resto y aquí lo único que se habla es del amor».   9 Ricky Martin es padre de dos niños, Matteo y Valentino, y hay gente que dice que los niños deben tener un padre y una madre. Sin embargo, en el mundo hay muchos tipos distintos de familias. Algunos niños crecen sin madre o sin padre. Otros sin los dos y se crían con un familiar o en una institución. Cuente anónimamente de personas que usted conoce que son de una familia no convencional. 10 ¿Por qué cree que a veces hay personas que esconden ser gay o se lo niegan a sí mismos(as)? ¿Qué cree que sería lo mejor para esas personas? Roca 2012: p. 121 131

Holly Cashman and Juan Antonio Trujillo

In stark contrast to the questions in La lengua que heredamos, these questions prompt students to reflect on why self-acceptance and coming out might be difficult for people, and they contextualize the LGBTQ+ movement within a civil rights framework. Rather than asking students to judge queer people and weigh in on whether or not they are deserving of equal rights, these questions attempt to put students in the place of an LGBTQ+ person, to try to imagine the world through their eyes, and to ponder both the inequality that exists as well as the benefits and contributions LGBTQ+ people make. Finally, question 9 attempts to engage students in a critique of heteronormativity, as seen in the dominance of the “traditional” family. Despite this careful, more nuanced inclusion, several problems remain: Why does the gay celebrity not just get to be featured for his work, artistic contributions, and/or family? Why does the profile choose to focus on coming out, marriage equality, and adoption? Although the questions are worded carefully to avoid asking students to give their personal opinion, this subtlety could easily be lost on students and even instructors, leading to the same debate purposely provoked by the first treatment in La lengua que heredamos. Another challenge of inclusion that we see in Nuevos mundos is the stereotype. When we think “gay,” too often we think only of the gay man and let him stand in for all queer people. This is a problem in Nuevos mundos, where, despite the serious attempt at thoughtful inclusion, only gay men are represented. Other than a brief mention in a question following the Ricky Martin reading, the only mention of lesbians is a problematic one in a short reading by Rosa Ontiveros about women’s rights organizing: “Cuando las mujeres nos agrupamos para defendernos entonces las cosas cambian, y sólo nos agrupamos porque, de repente, somos feas, lesbianas, amargadas” (p. 292). This negative reference to lesbians is part of a larger sexism that the author is trying to critique, yet the necessary unpacking is lacking. Although Nuevos mundos endeavors to be inclusive, there is no inclusion of lesbians, people who identify as (gender)queer, or people who identify as trans(gender), and the inclusion takes a normalizing form with a focus on coming out, a process and practice that has been critiqued by QTPOC (Acosta, 2011; Yon-Leau & MuñozLaboy, 2010), and marriage, which is viewed by some as a capitulation to “respectability politics” through which many queer people are marginalized (LaSala 2007), although it remains an important equality issue for others. In Conversaciones escritas,3 inclusion takes the form of debate, but in a more extended way than a few questions following a reading as in La lengua que heredamos. Conversaciones escritas takes an argumentative writing approach that focuses on contemporary topics in order to help students use their language skills in “a broad array of contexts, ranging from personal experience to public policy debates” (p. iii). In chapter 6, this debate approach focuses on gender and sexuality. The first three readings examine the relationship between sex and gender, and students are led to debate whether gender differences are real or whether gender is a social construct. Reading four explores the concept that “Cambiar para los homosexuales es posible”, which puts the question of whether sexuality is biological or socially created and whether, therefore, it’s possible to overcome same-sex desire through therapy; the chapter’s final reading is a debate on “El matrimonio entre gays” (pp. 214–232). Although debate and argumentative writing are frequently used activities in language learning, this set-up is problematic when the debate is centered squarely on the topic of students’ marginalized sexual identities in the heteronormative classroom environment. Focusing solely on the last two readings, we find a number of serious issues. First, although we recognize that the book intends to be provocative in its selection of texts, the inclusion of National Association for Research & Therapy of Homosexuality (NARTH)4 crosses a line. This is a fringe group whose therapy is completely discredited. It promotes longdebunked junk science that has harmed countless people, and its inclusion in a textbook gives it more weight and credibility than it deserves. Given the higher rates of suicide, self-harming, and 132

Queering Spanish as a heritage language

drug and alcohol abuse among LGBTQ+ individuals, particularly QTPOC (O’Donnell, Meyer, & Schwartz, 2011), it is irresponsible to give NARTH a platform even when followed by afterreading activities that may nudge students toward suspicion of their claims. Two after-reading exercises ask students to write something in which they argue for or against an issue related to LGBTQ+ rights or identities: the first is to write a letter to an imagined condo association weighing in on whether a gay couple should be allowed to sublet an apartment in the building, and the second is a thesis of their choice related to the topic. Given the set-up, it is quite likely that students may argue against equal rights or for the possibility of therapy, which might make for a particularly difficult peer editing exercise for a queer classmate, whether they are out or closeted, struggling with their identity or not. Even in the grammar explanations we can find surprisingly problematic examples such as: “¿Has visto «Brokeback Mountain»? Yo no pienso ver ninguna película que vaya en contra de mi moral.” or “Ellos han aceptado que su hijo es gay, pero ella lo ve como un trastorno” (p. 221). Examples such as these validate exclusionary discourse in what should be a welcoming and nurturing environment for all students regardless of sexual identity. A final example of problematic inclusion in Conversaciones escritas is found in a grammar activity where incorrect or offensive language is included. In one sentence, Juan Gabriel is described as “un putazo,” which is highlighted in a textbox as an offensive term but is nevertheless disturbing to see in a grammar activity. In another sentence, a trans person is described as a person who “was born a man but later transformed into a woman” (our translation), an essentializing discourse that is generally rejected by trans communities. Perhaps no SHL textbook exemplifies compulsory heterosexuality more than Español escrito. The elementary-level textbook in its sixth edition is centered on two characters: Fernando González, an engineering student from El Paso, Texas, and Marisela Suárez, a nursing student from the Dominican Republic who lives in New York City. Both Fernando and Marisela are from large, nuclear families—mom and dad, with six or seven children. Each of the textbook’s chapters begins with a reading about some member of one of the main characters’ families. The readings are artificial at best, and, at worst, stereotypical, offensive, and highly problematic. Space limitations prevent us from analyzing in depth the gender stereotypes that are pervasive in these readings or the ideologies of race and class that are woven into the stories. We will focus instead on the heteronormativity that permeates the text, where the heterosexual relationship and the so-called traditional family structure are imposed at every turn. No queer family members are found in this fictional family, but LGBTQ+ characters and references do arise. In chapter 9, in a story about overcoming obstacles, Elvira, Marcela’s aunt, tells us about how she managed to become the principal of an elementary school in New York after having only six years of formal education in the Dominican Republic. After losing her job working for a family in New York City, Elvira says she did two things quickly in order to secure her future in the U.S. The second thing—getting her GED—is in line with the chapter’s education theme, but the first thing Elvira is reported to have done is quite surprising: [m]e casé con un americano homosexual que quería darles gusto a sus padres quienes exigían que “por fin se portara como un macho” (él era amigo de esa amiga de una amiga con la que pasé mi primera semana en Nueva York y con quien nunca había perdido contacto), así que yo ya tenía la ciudadanía asegurada si sabía guardar bien las apariencias matrimoniales (vivimos juntos por cinco años). p. 121 After achieving her goal of becoming an elementary school principal, Maricela happily reports that Elvira could finally divorce her husband, whose parents had died. Other than the 133

Holly Cashman and Juan Antonio Trujillo

comprehension question “¿Cómo se hizo ciudadana Elvira?” (p. 122), there is no follow-up to this outrageous element of the story. The writing prompts that follow focus on education, bilingualism, and job opportunities, not on the marriage nor even the attempted rape that Elvira survives in her first job as a domestic in the Dominican Republic. Although marriages such as this do occur, it is problematic having this as the only reference of any kind to LGBTQ+ persons in this textbook. There are additional items of interest in terms of LGBTQ+ representation in Español escrito. In chapter 17, the topic of women’s role in society is examined. The main reading is from Marisela’s grandmother, who argues that the only role for women is to be wives and mothers, and that artificial insemination is a scandal and degeneracy. In the second set of questions that follows, students are asked to reflect in writing on the following questions: ¿Crees que es necesario que todos nos casemos? ¿Qué opinas de los solteros? ¿De las solteras? ¿Qué derecho tiene la mujer soltera de ser madre? Si un hombre no se casa y no se hace sacerdote de la iglesia católica (que no permite que los sacerdotes se casen), ¿qué significado tiene eso para ti? pp. 268–269; emphasis added While this question is certainly open to interpretation, it clearly constructs men and women who do not marry a person of the opposite sex as a suspicious aberration that requires explaining. It is almost impossible to imagine that the answer to the final question would be anything other than the presumption of a non-conforming sexual identity. The final chapters involve Marisela and Fernando writing diary entries about their ideal partner and then, predictably, meeting and falling in love. Fernando begins his reflection on his ideal woman, writing: Ahora que me falta solo un año para graduarme, he decidido empezar a salir seriamente con mujeres. Bueno, no es que haya salido yo con hombres, pero no comenzaron a gustarme mucho las mujeres—bueno, las muchachas—hasta que tenía quince años. p. 318; emphasis added This alienating joke serves to underscore the heteronormative orientation of the text and highlight the heterosexual assumption that permeates the language classroom. This assumption that students are heterosexual is made explicit in the writing task that follows Marisela and Fernando’s reflections: (Pregunta para mujeres:) ¿Qué es lo que tú quieres en un hombre? (Pregunta para hombres:)  ¿Qué es lo que tú quieres en una mujer? ... ¿Deben tener el derecho de casarse las mujeres lesbianas además de los hombres homosexuales? ¿Deben tener el derecho de tener hijos propios o de adoptarlos? ¿Deben recibir las parejas homosexuales los mismos beneficios que los matrimonios heterosexuales? Explica tu punto de vista. (p. 320) The first two questions make it virtually impossible for a queer student to participate. It is made perfectly explicit that women are the only people who have the right to describe what they look for in a man, and only men are allowed to describe what they look for in a woman. 134

Queering Spanish as a heritage language

The final question in the activity puts LGBTQ+ rights up for debate, an activity whose marginalizing effect we have discussed in more detail earlier. In summary, the textbooks examined overwhelmingly normalize heterosexuality. In some cases, the heteronormativity takes an aggressively anti-queer tone either by erasing LGBTQ+ identities or by the use of pathologizing language and situations. In other cases, there is an attempt to be inclusive by referring to celebrities with broad appeal or to unavoidable political discussions, although in the latter case queer lives are treated as controversial. We found that Entre mundos came closest to challenging heteronormativity in a meaningful and consistent manner, and at least one book we analyzed—Conversaciones escritas—has undergone major revisions resulting in a new edition that addresses concerns raised in our analysis. However, no textbook we analyzed merits an unqualified endorsement. Even in the best books, no effort was made to engage any of the issues explored in contemporary queer Latinx scholarship such as the negotiation of anticolonial queer identity, the relationship between queer of color communities and white gay cultural institutions, the disproportionate victimization of queer and trans people of color, or transnational queer identities (e.g. the Undocuqueer movement). The vast cultural production of queer Latinxs, central to contemporary Chicanx Studies, was also completely absent.

Queering the SHL environment Current SHL materials fall demonstrably short in their representation of queer identities and perspectives, often asserting heterosexual privilege in ways that feel overtly hostile to LGBTQ+ students and faculty and ignoring issues of concern to QTPOC. We now turn to a discussion of how the SHL learning environment might be made more welcoming and inclusive. Given the shortcomings of published materials, we find that the greatest possibility for meaningful and timely change lies in cultivating a knowledgeable faculty—teachers who can bring their lived experiences to the SHL classroom, implement practices drawn from critical pedagogy, and, when necessary, engage in what Reagan and Osborn (2002) refer to as “curricular nullification,” a strategy for challenging curriculum elements that are at conflict with social justice principles.

Queer Latinx faculty One clear way to improve the SHL learning environment for queer students is to attend to the diversity of the teaching staff and, in particular, support the needs of queer Latinx faculty who can serve as role models and repositories of subaltern knowledge that is lacking in the official curriculum. We do not have data specifically about LGBTQ+ Latinx SHL faculty, but there are some important conclusions to be drawn from general literature on “minority” or “diverse” faculty in U.S. colleges and universities (e.g. Fryberg & Martínez, 2014). These are faculty who must navigate their own complex identities while dealing with personal and institutional discrimination aimed at them from multiple directions. One concern is the pattern of bias that emerges in both formal and informal evaluations of teaching of queer and non-white faculty. DiPietro and Faye (2005; cited in Huston, 2006) found significant bias against Latinx faculty in student evaluations of teaching. Reid (2010) found that “students evaluated racial minority faculty more negatively than White faculty across a variety of factors” on a popular internet faculty rating site (p. 145). Anderson and Kanner (2011) report that students use different criteria to assess faculty objectivity based on their perception of the professor’s sexuality, attributing a political agenda to lesbian and gay faculty but not to straight faculty when analyzing identical course materials. 135

Holly Cashman and Juan Antonio Trujillo

Pressure to reduce the visibility of performances of LGBTQ+ or other Latinx identities in exchange for approval continues outside of the faculty/student relationship as well. Anecdotally, it is not unusual for language faculty from an SHL background to face challenges to their linguistic and cultural authenticity from colleagues—often non-Latinx L2 learners themselves—who defend Eurocentric language norms and Octavio Paz-inspired notions of gender roles in Latin American society.5 In addition to having to constantly assert sexual agency and defend themselves from charges of inauthenticity, queer Latinx faculty are in short supply at most colleges and universities, which leads to extraordinary service demands. Little time or energy remains for producing scholarship once the diversity needs of the institution have been met.

Critical pedagogy The critical pedagogy movement was launched by Paulo Freire in his seminal Pedagogy of the Oppressed, originally published in 1968, a book that came as a direct response to dehumanizing practices he observed in educational settings (Freire, 2000). The call to decenter power in the classroom and to engage in student-centered dialogic inquiry around relevant social issues made by critical pedagogy advocates (Giroux, 2005; hooks, 1994; Shor ,1997) could not be more on point for environments like the SHL classroom where participants—queer or not—are targets of multiple systems of oppression. The applicability of critical pedagogy to language acquisition is further confirmed in Norton and Toohey (2004), Reagan and Osborn (2002), and especially Osborn (2006), which carefully articulates critical pedagogy principles with the World-Readiness Standards or “Five Cs” (The National Standards Collaborative Board, 2015) that form the foundation of much of today’s language curriculum. Even more specific advocacy for the use of critical pedagogy in the SHL classroom comes from Leeman (2005, 2012, 2014), Trujillo (2009), and Villa (2010), among others. Critical theory calls for an educational experience that invites personal and social transformation—an education that asks participants (both students and teachers) to recognize their role in systems of oppression and to engage in dialogue with their world and their lived experiences to upset the unfair distribution of power. Critical theory, above all, requires educators to interact with students in a way that is humanizing and that enhances the agency and subjectivity of learners. This is in clear alignment with the underlying ethos of queer communities of color, and it is what we have shown to be lacking in most SHL materials. Most of the concrete recommendations we offer to both classroom practitioners and textbook authors are rooted in Freire’s (2000) call to engage fully with students—not just intellectually, but emotionally—as co-constructors of knowledge. This requires an explicit rejection of language and practices that objectify and marginalize individuals, particularly those who are targets of discrimination from the dominant culture. Cammarota and Romero (2006; cited in Sálazar, 2013) specify that the right kind of connection between faculty and student happens when there are “reciprocal opportunities to share their lives” (p. 128), when dehumanizing experiences shared with each other are met with compassion, and when learning is situated in issues relevant to marginalized participants. When we as faculty apply these criteria to the LGBTQ+ issues, specific solutions to the dehumanizing elements found in SHL materials begin to emerge: ••

We approach non-heteronormative experiences of sex, gender and sexual orientation as facets of human existence worthy of our curiosity and interest rather than topics of debate or controversy, never assuming that all of our learners are heterosexual. 136

Queering Spanish as a heritage language

•• •• ••

We acknowledge the way certain terms resonate within queer communities—from the pathologizing force of words like “homosexual” to the deep violence of hate speech (e.g. fag, puto, marimacha, etc.)—and we choose better ways to describe queerness. We hold space for those who have been targeted so that their experiences can—on their terms—be shared, acknowledged, and validated. We celebrate the diverse forms and expressions of gender and sexual identity that help constitute the individuality of our learners, and we question deficit-based tropes and stereotypes that serve only to objectify, categorize, and dehumanize.

Freire (2000) explains that a primary goal of “rehumanization” in critical pedagogy is to lead both teacher and learner to a shared status of “Subject”; full agents in the process of co-creating knowledge in dialogue with each other and the world (pp. 68–69). Sálazar (2013) describes the negative impact of denying agency over the expression of marginalized identities to Spanish speakers: “I found that when Mexican immigrant students are denied access and use of their mother tongue in academic spheres, these students begin to devalue their native language and denigrate their culture” (p. 130) and are left with their minds in a colonized state. By extension, requiring queer Latinx learners to suppress expressions of sexual and gender identity contributes to a similar dehumanization and disempowerment. Respect for learner agency over sexual identities suggests additional good SHL classroom practices: •• •• ••

Create conditions that allow queer students to feel safe disclosing their non-heteronormative identities (but that never force them to do so). Construct activities that do not coerce LGBTQ+ students into performances of heterosexual identity (such as role-playing that requires binary gender roles or heterosexual relationship patterns). Provide access to linguistic and cultural tools that enable learners to perform their identity in a variety of contexts (e.g. vocabulary, variant gender agreement, pronouns).

These recommendations admittedly require acknowledging aspects of our learners’ lives that are traditionally excluded from classroom discourse—our students are not just minds to mold, they are human beings with emotions and desires that can and should be engaged as part of the learning process. Osborn (2006) deals explicitly with the application of critical pedagogy principles in the language classroom. He embraces the dialogic approach of Freire and operationalizes it in what he refers to as the critical inquiry cycle (CIC), a framework that involves examining issues of relevance to learners in four phases: 1) conduct informed investigation; 2) undertake inductive analysis; 3) formulate tentative conclusions; and 4) engage in mutual critical reflection. This process is intended to be used with questions that are not “overly specific” (p. 33) or geared toward reaching some predictable outcome, and it is expected to involve stakeholders beyond the classroom if possible. Osborn proposes 15 social justice-based themes that lend themselves to world language and culture education through critical inquiry. These include identity, affiliation, conflict and discrimination, rights, and resistance and marginalization, all of which can be approached effectively through structured examination of sexuality and gender. The examination of queer-themed discourse has been demonstrated effective with students as early as fifth grade (Moita-Lopes, 2006). Finally, curricular nullification can be employed as a form of resistance against inappropriate or inadequate queer representation in SHL materials. Curricular nullification is any action undertaken by educators with the intent of subverting the message of the official curriculum. 137

Holly Cashman and Juan Antonio Trujillo

Reagan (2016) points out that the “hegemony of the textbook” (p. 176) can be challenged by either subtractive nullification (removing or minimizing content) or, more frequently and perhaps more ethically, by additive nullification involving the introduction of materials or conversations that reframe problematic content. A faculty member required to teach from any of the books analyzed previously might find it useful to take the criticisms we identified directly to the students and make them part of the lesson or challenge the message of the textbooks less conspicuously by bringing in supplemental materials by queer Latinx writers and artists that present a different perspective.

Conclusion Our analysis of specialized SHL materials reveals that although one or two textbooks show a commitment to fair representations of queer identities, across the board there is a pattern of erasure, stereotyping, and dehumanizing discourse. Current research on the learning climate for younger LGBTQ+ students indicates that this problematic representation is part of a larger system of discrimination that leads to lower academic achievement and, tragically, to higher rates of depression and lower self-esteem for learners (Kosciw et al., 2014). We propose that the solution lies first in supporting the needs of queer faculty, particularly queer faculty of color, and capitalizing on their expertise in the production of SHL materials as well as welcoming their presence in the SHL classroom. Regardless of sexual or ethnic/ racial identity, we see language faculty as key agents for transforming the learning environment through humanizing pedagogical practice and an ongoing commitment to challenge and reframe oppressive discourse. There is always some risk involved in transformative education, not just for LGBTQ+ faculty of color but for allies as well. But as one teacher reported when asked about his choice to risk including queer-themed materials in his Spanish classroom, “I think I do it primarily for the gay or questioning students so that they don’t feel excluded in what otherwise is a very heterosexist curriculum.” We believe that there is great fulfillment to be found in creating an environment where faculty and student are all fully present and fully human, with all elements of their identity intact. It is hard to deny the value of deliberate LGBTQ+ inclusiveness once an educator has shared a tearful moment with a queer Latinx student who never believed he would see lives like his honored in the curriculum or witnessed another disclose his gay identity to classmates for the first time as part of a class activity. For us, the risk is worth it.

Notes 1 “Latinx” is used here as a non-binary, gender-inclusive alternative to the universal masculine (i.e. “Latino”) or the gender-binary reinforcing (“Latina/o” or “Latin@”). In other words, “Latinx” is a term meant to include all people across the spectrum of gender identities, people who identify as genderqueer or gender fluid, people who identify as trans(gender), people who identify as intersex, and people who identify as cisgender men and women. 2 LGBTQ+ is a common variant of the initialism LGBTQ, conceived as an inclusive and non-stigmatizing way to refer to minoritized sexual and gender identities. The initials stand for lesbian, gay, bisexual, trans(gender), and queer or questioning. The plus sign acknowledges a vast array of additional identities such as intersex, asexual, two spirit, and straight ally. 3 This analysis is of the first edition (2011) of the textbook Conversaciones escritas. An updated, second edition of the text that addresses the concerns expressed in this chapter was published in 2017, and we acknowledge the work of the author and the publisher in their willingness, even eagerness, to thoroughly revise it.

138

Queering Spanish as a heritage language 4 In 2014, NARTH established an umbrella organization called the Alliance for Therapeutic Choice and Scientific Integrity and re-branded itself as the NARTH Institute. Conversion therapy for minors has been banned in several states and municipalities in the U.S. 5 In the “Máscaras mexicanas” essay from Laberinto de la soledad, Paz reinforces the Madonna-whore concept for women and lays out a framing of male homosexuality centered on essentializing roles of penetrator and penetrated in the act of anal sex.

References Acosta, K. L. (2011). The language of (in)visibility: Using in-between spaces as a vehicle for empowerment in the family. Journal of Homosexuality, 58, 883–900. Alonso-Lyrintzis, D., & Zaslow, B. (2004). Entre mundos, 2nd ed. Upper Saddle River, NJ: Pearson Prentice Hall. Anderson, K. J., & Kanner, M. (2011). Inventing a gay agenda: Students’ perceptions of lesbian and gay professors. Journal of Applied Social Psychology, 45(6), 1538–1564. Anzaldúa, G. (2012). Borderlands/La frontera: The new Mestiza, 4th ed. San Francisco, CA: Aunt Lute Books. Beaudrie, S., Ducar, C., & Potowski, K. (2014). Heritage language teaching: Research and practice. New York: McGraw Hill. Beaudrie, S., Ducar, C., & Relaño-Pastor, A. M. (2009). Curricular perspectives in the heritage language context: Assessing culture and identity. Language, Culture and Curriculum, 22(2), 157–174. Beaudrie, S., & Fairclough, M. (Eds.). (2012). Spanish as a heritage language in the United States: The state of the field. Washington, DC: Georgetown University Press. Bell, D., & Binnie, J. (2000). The sexual citizen: Queer politics and beyond. Cambridge, UK: Polity Press. Butler, J. (1990). Gender trouble: Feminism and the subversion of identity. London: Routledge. Cammarota, J, & Romero, A. (2006). A critically compassionate intellectualism for Latina/o students: Raising voices above the silencing in our schools. Multicultural Education, 14(2), 16–23. Carreira, M., & Geoffrion-Vinci, M. C. (2008). ¡Sí se puede! Boston, MA: Houghton Mifflin. Cashman, H. (2015) Queer Latina/o networks in the city: Languages, identities and the ties that bind. In R. Márquez Reiter & L. Martín Rojo (Eds.), A sociolinguistics of diaspora: Latino practices, identities and ideologies (pp. 66–80). London: Routledge. DiPietro, M. & Faye, A. (2005). Online student-ratings-of-instruction (SRI) mechanisms for maximal feedback to instructors. Paper presented at the 30th Annual Meeting of the Professional and Organizational Development Network; Milwaukee, WI. Ducar, C. (2006). (Re)Presentations of U.S. Latinos: A critical discourse analysis of Spanish heritage language textbooks. Unpublished doctoral dissertation. University of Arizona, Tucson, AZ. Ducar, C. (2009). The sound of silence: Spanish heritage textbooks’ treatment of language variation. In M. Lacorte & J. Leeman (Eds.), Español en Estados Unidos y otros contextos de contacto: Sociolingüística, ideología y pedagogía (pp. 347–368). Madrid: Iberoamericana. Ducar, C. (2012). SHL learners’ attitudes and motivations: reconciling opposing forces. In S. Beaudrie & M. Fairclough (Eds.), Spanish as a heritage language in the United States: The state of the field (pp. 161–178). Washington, DC: Georgetown University Press. Fishman, J. (2001). 300-plus years of heritage language education in the United States. In J. K. Peyton, D. A. Ranard, & S. McGinnis (Eds.), Heritage languages in America: Preserving a national resource (pp. 81–97). Washington, DC: Center for Applied Linguistics. Foucault, M. (1982). The archaeology of knowledge and the discourse on language. New York: Pantheon. Francés, M. E., & Benítez, R. (2012). Manual de gramática y ortografía para hispanos, 2nd ed. Upper Saddle River, NJ: Pearson Education. Freire, P. (2000). Pedagogy of the oppressed, 30th anniversary ed. New York: Continuum. Fryberg, S. A., & Martínez, J. E. (2014) The truly diverse faculty: New dialogues in American higher education. New York: Palgrave Macmillan. García, H. A., Carney, C., & Sandoval, T. (2011). Nuestro idioma, nuestra herencia. New York: McGraw-Hill. Giroux, H. (2005). Border crossings: Cultural workers and the politics of education, 2nd ed. New York: Routledge. Goldstein, B. (2015). LGBT invisibility in language learning materials. Language Issues: The ESOL Journal, 26(2), 35–40. Gray, J. (2013). LGBT invisibility and heteronormativity in ELT materials. In J. Gray (Ed.), Critical perspectives on language teaching materials (pp. 40–63). Basingstoke, UK: Palgrave.

139

Holly Cashman and Juan Antonio Trujillo Hames-García, M. (2011). Queer theory revisited. In M. Hames-García & E. J. Martínez (Eds.), Gay Latino studies: A critical reader (pp. 19–45). Durham, NC: Duke University Press. hooks, b. (1994). Teaching to transgress: Education as the practice of freedom. New York: Routledge. Human Rights Campaign (2015). Addressing anti-transgender violence: Exploring realities, challenges and solutions for policymakers and community advocates. Washington, DC: Human Rights Campaign. Retrieved from http://hrc-assets.s3-website-us-east-1.amazonaws.com//files/assets/resources/HRCAntiTransgenderViolence-0519.pdf. Huston, T. A. (2006). Race and gender bias in higher education: Could faculty course evaluations impede further progress toward parity? Seattle Journal for Social Justice, 4(2), 591–611. Kosciw, J. G., Greytak, E. A., Palmer, N. A., & Boesen, M. J. 2014. The 2013 National School Climate Survey: The experiences of lesbian, gay, bisexual and transgender youth in our nation’s schools. New York: GLSEN (Gay, Lesbian & Straight Education Network). Retrieved from www.glsen.org/article/ 2013-national-school-climate-survey. LaSala, M.C. (2007). Too many eggs in the wrong basket: A queer critique of the same-sex marriage movement. Social Work, 52(2), 181–183. Leduc, P. (Director). (1983). Frida: Naturaleza viva [Motion picture]. Mexico: Clasa Films Mundiales. Leeman, J. (2005). Engaging critical pedagogy: Spanish for native speakers. Foreign Language Annals, 38(1), 35–45. Leeman, J. (2012). Investigating language ideologies in Spanish as a Heritage Language. In S. Beaudrie & M. Fairclough (Eds.), Spanish as a heritage language in the United States: The state of the field (pp. 43–59). Washington, DC: Georgetown University Press. Leeman, J. (2014). Critical approaches to teaching Spanish as a local/foreign language. In M. Lacorte (Ed.), The Routledge handbook of Hispanic applied linguistics (pp. 275–292). New York: Routledge. Leeman, J., & Martínez, G. (2007). From identity to commodity: Ideologies of Spanish in heritage language textbooks. Critical Inquiry in Language Studies, 4(1), 35–65. Liddicoat, A. J. (2009). Sexual identity as linguistic failure: Trajectories of interaction in the heteronormative language classroom. Journal of Language, Identity, and Education, 8, 191–202. Márquez, S. (2012). La lengua que heredamos, 7th ed. Hoboken, NJ: John Wiley & Sons. Martínez, G. (2003). Classroom based dialect awareness in heritage language instruction: A critical applied linguistic approach. Heritage Language Journal, 1(1), 1–14. Moita-Lopes, L. P. (2006). Queering literacy teaching: Analyzing gay-themed discourses in a fifth-grade class in Brazil. Journal of Language, Identity, and Education, 5(1), 31–50. Mrak, N. A., & Aponte, E. K. (2007). ¡Conozcámonos! Boston, MA: Cengage Learning. Nelson, C. (1993). Heterosexism in ESL: Examining our attitudes. TESOL Quarterly, 27, 143–150. Nelson, C. D. (2006). Queer inquiry in language education. Journal of Language, Identity, and Education, 5(1), 1–9. Nelson, C. D. (2008). Sexual identities in English language education: Classroom conversations. New York: Routledge. Norton, B. & Toohey, K. (2004). Critical pedagogies and language learning. Cambridge, UK: Cambridge University Press. O’Donnell, S., Meyer, I. H., & Schwartz, S. (2011). Increased risk of suicide attempts among Black and Latino lesbians, gay men and bisexuals. American Journal of Public Health, 101(6), 1055–1059. Osborn, T. A. (2006). Teaching world languages for social justice: A sourcebook of principles and practices. Mahwah, NJ: Lawrence Erlbaum Associates. Paiz, J. M. (2015). Over the monochrome rainbow: Heteronormativity in ESL reading texts and textbooks. Journal of Language and Sexuality, 4(1), 77–101. Pawelczyk, J., & Ł. Pakuła. (2015) Constructing gender and sexuality in the EFL classroom: Textbook construction and classroom negotiation? In S. Mills and A. S. Mustapha (Eds.), Gender representation in learning materials: International perspectives (pp. 193–211). New York: Routledge. Potowski, K. (2011). Conversaciones escritas. Hoboken, NJ: John Wiley & Sons. Puar, J. (2007). Terrorist assemblages: Homonationalism in queer times. Durham, NC: Duke University Press. Rankin, S. R. (2006). LGBTQ+A students on campus: Is higher education making the grade? Journal of Gay & Lesbian Issues in Education, 3(2/3), 111–117. Reagan, T. G. (2016). Language teachers in foreign territory: A call for a critical pedagogy-infused curriculum. In L. Cammarata (Ed.), Content-based foreign language teaching: Curriculum and pedagogy for developing advanced thinking and literacy skills (pp. 173–191). New York: Routledge.

140

Queering Spanish as a heritage language Reagan, T. G., & Osborn, T. A. (2002) The foreign language educator in society: Toward a critical pedagogy. Mahwah, NJ: Lawrence Erlbaum Associates. Reid, L. D. (2010). The role of perceived race and gender in the evaluation of college teaching on ratemyprofessors.com. Journal of Diversity in Higher Education, 3(3), 137–152. Renn, K. A. (2010). LGBT and queer research in higher education: The state and status of the field. Educational Researcher, 39(2), 132–141. Roca, A. (2012). Nuevos mundos: lectura, cultura y comunicación; curso de español para bilingües, 3rd ed. Hoboken, NJ: John Wiley & Sons. Sálazar, M. (2013). A humanizing pedagogy: Reinventing the principles and practice of education as a journey toward liberation. Review of Research in Education, 37, 121–148. Shor, I. (1997). When students have power: Negotiating authority in a critical pedagogy. Chicago, IL: University of Chicago Press. Snelbecker, K. A. (1994). Speaking out: A survey of lesbian, gay and bisexual teachers of ESOL in the U.S. Unpublished master’s thesis. School for International Training, Brattleboro, VT. Retrieved from http://files.eric.ed.gov/fulltext/ED375680.pdf. Sunderland, J., & M. McGlashan. 2015. Heteronormativity in EFL textbooks and in two genres of children’s literature (Harry Potter and same-sex parent family picturebooks). Language Issues: The ESOL Journal, 26(2), 17–26. Taymor, J. (Director). (2002). Frida [Motion picture]. United States: Miramax Films. The National Standards Collaborative Board. (2015). World-readiness standards for learning languages. 4th ed. Alexandria, VA: Author. Trujillo, J. A. (2009). Con todos: Using learning communities to promote intellectual and social engagement in the Spanish curriculum. In J. Leeman & M. Laporte (Eds.), Español en Estados Unidos y otros contextos de contacto: Cuestiones sociolingüísticas, ideológicas y pedagógicas (pp. 369–395). Madrid: Iberoamericana/ Vervuert. Valdés, G. (2001). Heritage language students: Profiles and possibilities. In J. K. Peyton, D. A. Ranard, & S. McGinnis (Eds.), Heritage languages in America: Preserving a national resource (pp. 37–77). Washington, DC: Center for Applied Linguistics. Valdés, G., Teschner, R., & Enríquez, H. (2008). Español escrito, 6th ed. Upper Saddle River, NJ: Pearson Education. Villa, D. (2010). ¿¡Cómo que Spanglish!? Creating a service learning component for a Spanish heritage language program. In S. Rivera-Mills & J. A. Trujillo (Eds.), Building communities and making connections (pp. 120–135). Newcastle upon Tyne, UK: Cambridge Scholars Publishing. Warner, M. (1991). Introduction: Fear of a queer planet. Social Text, 29, 3–17. Yon-Leau, C., & Muñoz-Laboy, M. 2010. ‘I don’t like to say that I’m anything’: Sexuality and cultural critique among sexual minority Latino youth. Sexuality Research and Social Policy, 7, 105–117.

141

Q Taylor & Francis Taylor & Francis Group

� http://taylorandfrancis.com

PART II

Linguistic studies

Q Taylor & Francis Taylor & Francis Group

� http://taylorandfrancis.com

10 MORPHOLOGY, SYNTAX, AND SEMANTICS IN SPANISH AS A HERITAGE LANGUAGE Silvina Montrul university of illinois at urbana-champaign, usa

Introduction Heritage speakers of Spanish are bilingual individuals in places where Spanish is a minority language (Montrul 2016a; Rothman 2009), and the focus of this chapter is on grammatical aspects of their heritage language. The grammatical development of a heritage language cannot be understood without taking into account the minority status of the language and its perceived prestige and value. For example, Lambert (1977) introduced the terms additive and subtractive bilingualism to explain how language status affects the type and degree of bilingualism developed (see also Cummins & Swain 1986). Additive bilingualism results when the individual’s first language is a societally dominant and prestigious language and the individual learns another language, typically voluntarily, which can be another official language or a second language. As the bilingual develops full command of her native language, she adds knowledge of another language. In this case, the acquisition of the other language, even if it happens in childhood, does not interfere with the healthy development of the native language. The opposite of additive bilingualism is subtractive bilingualism. Subtractive bilingualism occurs most often when the first and native language is a minority language with little social prestige and value. Due to social pressure, negative attitudes, and often lack of educational opportunities, the acquisition of the societal language, especially at school, contributes to the gradual weakening or even replacement of the native minority language by the second majority language. The vast majority of heritage speakers find themselves in a subtractive situation by which they often become highly competent speakers of the majority language at the expense of the development of the heritage language. Research on heritage languages and their speakers since the late 1990s has been particularly concerned with understanding more fully the particular characteristics of such uneven development (Montrul 2016a). Heritage speakers’ degree of bilingualism is highly variable. By now it is widely accepted that bilinguals with perfectly balanced command of two languages are a myth, and that the majority of bilinguals use their languages in different contexts and for different purposes (Grosjean 2008). Thus, the linguistic competence of all bilinguals is fluid, changes along the lifespan, and is often unequal, as captured by the concept of dominance. The dominant language is the language that the bilingual uses the most or more often, and it also happens to 145

Silvina Montrul

be psycholinguistically stronger, with larger vocabulary, wider structural variety, and fluent production (see contributions in Silva-Corvalán & Treffer-Dallers 2016). The non-dominant language is used less often or in more restricted contexts and is psycholinguistically weaker, ranging from reduced fluency in production with high interference from the stronger language to merely receptive knowledge of it (i.e., comprehension without productive ability). Heritage speakers can exhibit functional proficiency in both productive and receptive abilities, or in receptive abilities only (Sherkina-Lieber 2011). Many Spanish heritage speakers in the United States are dominant in English, and consistent with the concept of subtractive bilingualism, Spanish is often their weaker language, as judged from patterns of language use reported in language background questionnaires, self-rating scales and self-reports, speech samples, and standardized or researcher-made vocabulary and proficiency measures (Montrul 2016b). The majority of studies of heritage speakers of Spanish in the United States have reported this type of subtractive bilingualism, especially studies whose participants attended either transitional bilingual programs or English-only elementary schools. Several factors contribute to the observed variability in heritage language grammatical proficiency, such as the age at which the individual was exposed to the majority language (Montrul 2008), the degree to which the majority language was spoken at home together with the heritage language, whether the heritage language is spoken only by the parents or by other family members including siblings (size and density of social networks) (Hurtado & Vega 2004), access to the heritage language at school, in the public domain, and the size of the speech community beyond the home (Tse 2001), among many others. As the minority language weakens, by the time heritage language speakers reach young adulthood, their linguistic systems often display structural differences in many grammatical areas when compared to the linguistic systems of heritage speakers who are dominant in the heritage language, monolinguals or bilinguals raised in a majority language situation, and Spanishdominant bilinguals of the parental generation (adult immigrants). Just as heritage speakers display dissociations in proficiency by language skill, they also exhibit uneven mastery of different aspects of the grammar. This suggests that different aspects of grammar may require different thresholds of input to develop as they do in monolinguals, perhaps along the lines outlined by Yang’s (2016) variational learning model (space limitations do not allow further elaboration). Understanding how grammatical outcomes are impacted by reduced input during childhood carries great practical significance. In the United States, many heritage speakers attending university seek to relearn or further develop their heritage language in a formal environment and often share the classroom with second language learners with less experience with the language and culture. It is important to know what classes for heritage speakers should be building on at the level of language structure. This chapter summarizes the results of experimental studies that investigated the linguistic knowledge of Spanish heritage speakers in the United States. Although in recent years there has been significant emphasis on aspects of grammar that do not reach full development and end up not fully mastered—such as inflectional morphology—not all areas of grammar eventually stabilize at non-target levels. In fact, we will see that some areas of the heritage grammar seem to be acquired at native-like levels,1 including complex structures. In the rest of the chapter I focus on these areas to illustrate the characteristics of the grammatical systems of many heritage speakers in morphology, syntax, and the related interfaces of semantics and pragmatics.

Inflectional morphology Inflectional morphology is a very vulnerable area in heritage language grammars, and Spanish is no exception. Inflectional morphology carries grammatical information and is the locus of 146

Morphology, syntax, and semantics

crosslinguistic variation driving syntactic differences between languages. Mature native speakers who grew up in a situation where Spanish was a majority language usually produce the appropriate morphological form of nouns, verbs, pronouns, and noun phrases in required contexts. When language learners do not produce morphological forms in all required contexts consistently (i.e., there are omissions) or use different forms and in variable ways, it is common in language acquisition to refer to these inconsistent forms as developmental “errors.” Inflectional morphology develops early in language acquisition, and although monolingual Spanish-speaking children make developmental errors, these errors eventually go away. By a certain age in childhood, depending on the developmental schedule of specific morphology, children eventually comprehend and produce inflectional morphology like adults. However, inflectional morphology is often partially acquired, without reaching full mastery, in young adult Spanish heritage speakers. Some morphological patterns displayed by heritage speakers are very similar to the patterns displayed by second language learners (Au et al. 2002).

Nominal morphology Spanish nouns are marked by gender (masculine, feminine) and number (singular, plural). Gender and number are two affixal morphemes (word markers) (libr-o-s, malet-a-s). Determiners (el, un, este), subject pronouns (él, ella), and object pronouns (lo, la, le) are free morphemes, but can also contain gender and number information, as shown in (1) and (2).  (1) Ellas    tienen unos    libros    nuevos.    ¿Los    quieres? they-fem.pl have some-masc-pl book-masc.pl new-masc.pl. them-masc.pl want-2nd p. sg “They have some new books. Do you want them?”  (2) Ellos   vendieron estas     maletas   viejas.   ¿Las   ves? they-masc.pl sold   these-fem.pl. suitcases.fem.pl old.fem.pl. them-fem.pl see.2nd. sg. “They sold those old suitcases. Do you see them?” In Spanish masculine gender is considered the default (Harris 1991). Although about 90% of Spanish nouns have a regular ending for gender, there are irregular nouns that do not end in the canonical masculine –o and canonical feminine –a ending. They end in non-transparent endings like –e or a consonant. Gender, number, and case in nouns are mastered at an early age by monolingual Spanish-speaking children (Montrul 2004a). Children produce gender marking by age 3 or 4 with almost 95–100% accuracy, with the exception of irregular, less frequent, and marked forms (Pérez Pereira 1991). Interestingly, inflectional morphology is noticeably affected in heritage language grammars, with non-target gender agreement being very common. There are several studies on gender assignment and agreement conducted with Spanish heritage speakers in the United States, both with adults (Alarcón 2011; Montrul, Davidson, de la Fuente, & Foote 2014; Montrul, de la Fuente, Davidson, & Foote 2013; Montrul, Foote, & Perpiñán 2008a) and with children (Anderson 1999; Montrul & Potowski 2007; Mueller Gathercole 2002a). All these studies found that, unlike monolingual gender assignment, it is as variable and inconsistent for heritage speakers with low to intermediate proficiency in Spanish, as it is among second language learners. When Spanish heritage speakers make gender errors, these are most frequent with feminine nouns and with nouns with non-canonical or non-transparent word endings (lápiz 147

Silvina Montrul

“pencil,” nube “cloud,” papel “paper”). If masculine gender is the default and feminine is the marked form, clearly Spanish heritage language grammars also show simplification of marked forms and overapplication of the masculine default. Higher proficiency heritage speakers can achieve native like levels with gender assignment and agreement in production and processing (Alarcón 2011; Montrul et al. 2013). Another grammatical area substantially affected in heritage grammars is case, which is also subject to systematic simplification. Case is a morphological category that marks the syntactic function of noun phrases in the sentence. Typically, in nominative-accusative languages like Spanish, subjects are marked with nominative, direct objects with accusative, and indirect objects with dative case. Nominative case is unmarked. Dative case is present in clitic pronouns and marked with the preposition “a,” as in (3) and (4), and so are some instances of accusative case with definite, specific animate direct objects, known as DOM marker as in (5).   (3) Juan le dio un abrazo a Pedro. Juan him.dat gave a hug to Pedro “Juan gave Pedro a hug.”   (4) A María le gusta ir al cine. to María her.dat like to go to the movies “María likes to go to the movies.”   (5) Miguel visitó a su prima. Miguel visited DOM his cousin “Miguel visited his cousin.” Like gender agreement, erosion of case marking has been found in several heritage languages including Spanish in the United States (Montrul 2004b; Montrul & Bowles 2009; Pascual y Cabo 2013), and Spanish in Neuchatel, Switzerland (Grosjean & Py 1991). Case marking with animate, specific direct objects, as in (5), is highly susceptible to partial acquisition and non-native mastery in heritage language grammars. The only available study on monolingual Spanish-speaking children (Rodríguez-Mondoñedo 2008), which analyzed spontaneous production samples from the CHILDES database, indicated that by age 3.00 and earlier, young children mark animate specific direct objects with “a” with 85% accuracy. However, Ticio (2015), who examined the development of DOM in simultaneous bilingual children from the CHILDES database, found that by age 3.6, the bilingual children had not acquired DOM to the level of the monolingual children, reaching barely 26% accuracy. Montrul and Sánchez-Walker (2013) investigated production of DOM marking in child (mean age 11) and young adult heritage speakers (mean age 21), and found that the bilingual children reached 60% accuracy and the young adult heritage speakers about 80%. The adults were also consistent with previous studies of adult heritage speakers by Montrul (2004b) and Montrul and Bowles (2009), who also found that regardless of proficiency levels, young adult heritage speakers produced unmarked animate specific direct objects in oral production (for example, Sara vio ∅ la mujer) and judged them as acceptable in grammaticality judgment tasks. Montrul (2014, 2016c), Montrul, Bhatt and Girju (2015), and Montrul and Bowles (2009), showed that, when the dative marking of indirect objects, the dative marking of gustar-type verbs and the marking of animate, specific direct objects are taken into account, heritage speakers make fewer errors marking indirect objects, 148

Morphology, syntax, and semantics

which is assumed to be an instance of structural case (Butt 2006). However, when case is lexical or semantic—as with gustar-type verbs and animate, specific direct objects—it is more vulnerable to erosion in Spanish heritage speakers’ grammars. To summarize, nominal morphology in Spanish, such as gender agreement and case, are very vulnerable and subject to morphological variability in heritage language grammars.

Verbal morphology The verbal domain also shows differential levels of acquisition in heritage language speakers, although overall verbal morphology (subject verb agreement, complex tenses, aspect, and mood) appears to be more stable than nominal morphology. Agreement: Spanish marks person and number agreement on the verb, which agrees with the subject (e.g., Yo sé “I know,” tú sabes “you know,” él sabe “he knows,” ellas/ellos saben “they know”). Montrul (2006) showed that intermediate and advanced proficiency heritage speakers of Spanish in the United States were highly accurate on subject verb-agreement in an oral narrative task (above 96%). Foote (2011) tested knowledge of verbal agreement in heritage speakers and L2 learners with intermediate and advanced proficiency in Spanish in language production, and also found very high accuracy by the experimental groups on agreement. Tense: Tense locates the event in the time axis and signals the difference between present, past, and future. In general, there are few if any reports of errors with tense in heritage grammars. Silva-Corvalán (1994, 2014), who studied oral samples from first-generation Mexican immigrants and second- and third-generation Spanish heritage speakers in the United States, did not record errors with tense and temporality. The heritage speakers Silva-Corvalán interviewed used all the simple tense forms (present, preterit, imperfect, future), and distinguished between past, present, and future. However, with respect to the future, they used predominantly the periphrastic form (ir a + infinitive “go to”) instead of the simple future synthetic form (-ending in –r- as in ama-r-é “I will love,” teme-r-é “I will fear,” vivi-r-é “I will live”), a pattern also found in Latin American varieties. Unlike the first-generation speakers in Silva-Corvalán’s study, the heritage speakers did not have productive use of the complex compound tenses, e.g., pluperfect indicative and subjunctive (hubo/hubiera visto “had seen”), future and conditional perfect (habrá/ habría visto “would have seen”), and analytic forms in Spanish (future llegaré “I will arrive”) (Silva-Corvalán 1994, p. 30). The general observation is that errors with tense are rare with simple tenses, but many heritage speakers do not actually develop productive use of many complex forms, at least as judged by production data. Aspect: Aspect is another verbal category concerned with the internal temporal constituency of a situation (state or event), such that these can be regarded as having an endpoint or not. Aspect can be expressed lexically by the inherent lexical semantics of the verb and its interaction with direct and indirect arguments and adjuncts (Dowty 1986; Verkuyl 1994). This is called lexical aspect (Smith 1991) and represents the way humans perceive and categorize situations. Lexical aspect is expressed in the verb or predicate. Some predicates are telic or perfective (with an inherent endpoint), such as achievements like llegar “arrive” and accomplishments like leer un libro “read a book,” while activities like correr “run” and states like ser “be” are atelic or imperfective. Aspect can also be expressed grammatically through the use of inflectional morphology on the verb, and in Spanish this morphology combines with accomplishments, achievements, activities, and states. In Spanish, grammatical aspect is marked by the preterit and imperfect: Ayer Juan salió de vacaciones “Yesterday Juan left on vacation” vs. A esta hora ayer Juan salía de vacaciones 149

Silvina Montrul

“At this time yesterday Juan left on vacation.” In the sentence with the preterit, the event is conceived as finished and bounded, whereas the sentence with the imperfect depicts the event in progress at the time and unbounded. Aspectual morphology is quite vulnerable in heritage language grammars. Using very different methodologies and theoretical approaches, both Silva-Corvalán (1994, 2014) and Montrul (2002, 2009) found that young adult Spanish heritage speakers in the United States confuse aspectual distinctions between perfective and imperfective forms. Spanish heritage speakers use preterit for imperfect forms and vice versa in oral production, and have been shown to have difficulties interpreting the meaning of preterit and imperfect morphology in experimental tasks involving felicity judgments (Montrul 2002, 2009). A recent study of child and adult Spanish heritage speakers by Cuza et al. (2013) showed that compared to the monolingual comparison groups, the child and adult heritage speakers had lower production of imperfect and overextension of preterit to imperfect contexts, and that compared to the preterit, the imperfect is underdeveloped and incompletely acquired in Spanish heritage speakers. The imperfect is also mastered later than the preterit in Spanish monolingual children (Hodgson 2005). Similarly, Potowski (2007) found that heritage speakers in eighth grade attending a dual language school were slightly less accurate than native speakers in their distribution of preterite and imperfect morphology by aspectual category. Mood: The verbal category that is most affected in Spanish as a heritage language is mood. All languages express modalities, but not all languages have mood. Mood is the grammatical expression of modality, marked by verbal morphology on verbs. Although monolingual Spanishspeaking children start using subjunctive forms with a restricted set of verbs that subcategorize for subjunctive and in negative command by age 3.00 (Gallo Valdivieso 1994), semantically and pragmatically conditioned uses of subjunctive, and subjunctive in relative and in adverbial clauses are not mastered until about age 12 (Blake 1983). First-generation Spanish-speaking immigrants in the United States retain the subjunctive in all these contexts, but second- and third-generation heritage speakers tend to replace indicative for subjunctive in contexts where subjunctive is required or strongly preferred by monolinguals, or misuse the subjunctive to signal different semantic and pragmatic meanings of a given expression based on context (Lynch 1999; Martínez-Mira 2009; Montrul 2009; Silva-Corvalán 1994). For example, Silva-Corvalán (1994) found that low proficiency speakers did not produce subjunctive forms, using the indicative exclusively in both obligatory and in variable contexts, as in (6) and (7) (Silva-Corvalán 1994, p. 42) (PI = present indicative, PS = present subjunctive).   (6) *I hope que no me toca (PI) la misma problema. (= toque PS) “I hope I don’t run into the same problem.”   (7) Quizás vengo mañana (= venga (PS)). “Maybe I come tomorrow.” In some cases, the use of indicative or subjunctive depends on meaning and implicatures. The indicative implies a fact whereas the subjunctive implies a hypothetical situation. In a study with intermediate and advanced proficiency Spanish heritage speakers, Montrul (2007) found high error rates with subjunctive in a written task of morphological recognition and little discrimination between the semantic implicatures of indicative and subjunctive morphology in variable contexts in a written sentence meaning judgment task, such as with relative clauses (Busco a una profesora que enseña-indic./enseñe-subj. francés. “I am looking for a teacher who teaches/would 150

Morphology, syntax, and semantics

teach French”). The indicative implies that the teacher exists whereas the subjunctive implies that such teacher might not exist. Thus, the simplification of the subjunctive in adult heritage speakers appears to extend to written comprehension as well. Thus, comparing grammatical and morphological categories within the verbal domain, mood is significantly more affected than aspect. Montrul (2009) tested knowledge and use of tense/ aspect and mood in Spanish heritage speakers of three proficiency levels: low, intermediate, and advanced, as well as a comparison group of monolingually raised native speakers. Although the heritage speakers used preterit/imperfect and indicative/subjunctive in two oral tasks, many of the errors evidenced in production also showed up in written recognition and in tasks of semantic discrimination. Montrul’s results reflect the same trends reported by Silva-Corvalán (1994). That is, many of the Spanish heritage speakers who exhibited unstable knowledge of mood displayed better command of grammatical aspect. The heritage speakers’ knowledge of grammatical aspect appears more solid with prototypical grammatical aspect-predicate type combinations, such as achievements and accomplishments (the telic classes) in the preterit or states and activities (the atelic classes) in the imperfect. By contrast, those conditions where the lexical semantic features of the verb and the semantic features of the aspectual form clash, such as achievements in the imperfect and states in the preterit, proved more problematic for the low proficiency speakers as well. Therefore, the results of Montrul (2009) are consistent with the attrition effects observed with tense/aspect and with mood in childhood (Merino 1983; Silva-Corvalán 2003, 2014), even when all these studies used very different research methodologies and tasks. Hypotheticality: Naturally, the simplification of subjunctive, future, and conditional forms has consequences for the expression of hypothetical discourse, an area of difficulty for heritage speakers with low proficiency in the language. If . . . then conditional sentences are complex sentences, requiring specific combinations of tenses and moods depending on the degree of factuality or hypotheticality. In Spanish, as in many other languages, there are three types of conditional sentences that vary in their hypotheticality and (counter-)factuality. The first type is the simplest and takes simple present in the if clause (protasis) and simple future in the then clause (apodosis), as in (8).   (8) Si llueve mañana no regaré las plantas. if it rains-pres tomorrow not I water-fut the plants “If it rains tomorrow I will not water the plants.” The other two types, shown in (9) and (10), are more complex, and represent irrealis and hypothetical meanings. In Spanish, they require subjunctive and conditional forms, although there is dialectal variation in the use of the conditional in both clauses (Lavandera 1984), or the use of imperfect indicative instead of subjunctive (Silva-Corvalán 1985) in monolingual varieties:   (9) Si tuviera tiempo, terminaría de leer este libro hoy. if I had-subj time, I finish-cond reading this book today “If I had time, I would finish reading this book today.” (10) Si hubiera sabido que venías a las 4, te habría esperado. if I had known-pluperf subj that you come-imp at 4, I you would have waited-cond perf “If I had known you were arriving at 4 I would have waited for you.” 151

Silvina Montrul

Several studies (Fairclough 2005; Gutiérrez 1996; Lynch 1999; Silva-Corvalán 1994) have found that while first-generation Spanish-speaking immigrants produce the three types of clauses with the most typical tenses as shown in (8–10), second- and third-generation heritage speakers show considerable variation in the use of verbal paradigms with the conditional clauses requiring conditional and subjunctive verbal forms, as in (9) and (10). The general tendency is to replace subjunctive and conditional forms with the indicative, and the compound tenses with simple tenses. Silva-Corvalán (1994) observed that the first-generation immigrants had a complex system of verb morphology that allowed them to convey different degrees of possibility, assertiveness, predictive certainty, etc. By contrast, second-generation heritage speakers exhibited a more restrictive set of choices, using almost exclusively indicative morphology to convey a strong degree of assertiveness and predictive certainty, without differentiating morphologically between more or less possible situations in the hypothetical world created. The two third-generation children studied longitudinally from ages 1 to 6 (Silva-Corvalán 2014), hardly used the future and the conditional forms, but retained the past tense. To summarize the patterns of verbal morphology, the degree of erosion and simplification observed in different speakers seems to be related to the degree of proficiency of the heritage speakers and the complexity of the verbal forms. In general, heritage speakers develop and retain solid knowledge of agreement and tense, but the categories that interface with semantics and pragmatics (aspect and mood) are more prone to simplification, and more so if they require complex syntax, like subjunctive and conditional forms that must occur in complex sentences.

Morphosyntax-semantics-pragmatics interface As just noted, interfaces are the most affected areas of heritage speaker grammars. In this section I summarize work on determiners, subject pronouns, and object clitics.

Determiners Spanish has definite and indefinite articles inflected for gender and number, as in (11) and (12), which precede the noun. Definite determiners have anaphoric properties, referring to items introduced in discourse, and different semantic interpretations, which have to do with genericity and specificity. Montrul and Ionin (2010) investigated the interpretation of definite articles with plural noun phrases in Spanish heritage speakers in the United States. The syntactic and semantic distribution of definite and indefinite articles in Spanish and English is largely similar, but the two languages differ in the expression and interpretation of plural noun phrases. Consider the English examples and their Spanish equivalents in (11–14). Spanish plural noun phrases with definite articles can express generic reference, as in (11), or specific reference, as in (12). English plurals with definite articles can only have specific reference, as in (14), while generic reference is expressed with bare plural noun phrases, as in (13). Bare nouns (nouns without articles) are ungrammatical in Spanish when they are in subject position, as in (15), unless they are modified. Bare nouns in object position are grammatical, as in (16). (11) Los elefantes tienen colmillos de marfil. (generic reference) (*the) elephants have tusks of ivory (12) Los elefantes de este zoológico son marrones. (specific reference) the elephants of this zoo are brown 152

Morphology, syntax, and semantics

(13) Elephants have ivory tusks.      (generic reference) (14) The elephants in this zoo are brown.   (specific reference) (15) *Elefantes tienen colmillos de marfil. (16) Vi elefantes con colmillos de marfil “I saw elephants with ivory tusks.” Misuse and omission of articles with noun phrases in subject position as in (15) are very common in Spanish heritage speakers with intermediate to advanced proficiency in Spanish, especially in written production. Montrul and Ionin (2010) used a written truth value judgment task, and an acceptability judgment task to test the generic/specific interpretation of plural determiners in Spanish. They found that the Spanish heritage speakers incorrectly accepted bare plural subjects with generic reference in Spanish (*Cebras tienen rayas “Zebras have stripes”) and interpreted definite articles as having specific rather than generic reference, as in English. While Spanish heritage speakers retain the determiner system of Spanish, they seem to adopt the determiner system of English, the majority language.

Subject pronouns A phenomenon observed in Spanish and in a variety of heritage languages (Greek, Italian, Russian, Hindi, Arabic, and many others) concerns the licensing of null and overt subjects. In null subject languages, expressed and unexpressed subject pronouns are grammatical because the person and number information is recoverable from the agreement morphology, as in (17). (17) Ella/ Ø llegó de Madrid. She arrived-3rd sg from Madrid “She arrived from Madrid.” However, the distribution of null and overt subjects is licensed by discourse-pragmatic factors, such as topic continuation, topic shift, or switch reference. Overt subjects are strictly required when new information is introduced, and a contrast is established in discourse, as examples (18) and (19) show. The answer to question (18) requires a focused noun phrase and must be expressed by an overt subject. In (19), the overt subject él is a topic if unstressed, but it can also be a contrastive focus if it is stressed. The pronoun in this case can co-refer with the subject of the previous clause because it emphasizes the subject. (18) ¿Quién vino?    El/Mario/*Ø vino.                focus “Who came?    He/Mario/*Ø came.” (19) El periodistai dijo que éli/ÉLi no había escrito ese reporte.      topic/contrastive focus the journalist said that he (himself) not had written that report “The journalist said that he had not written that report.” The use of null and overt subjects is also relevant to establish reference in discourse. For example, when there is no switch in reference between a series of sentences, null subjects are appropriate when there is topic continuity, whereas overt subjects are pragmatically infelicitous, as in (20). 153

Silvina Montrul

By contrast, overt subjects are appropriate when there is topic shift and a different referent is introduced, as in (21). Null subjects are infelicitous in these switch reference contexts (examples from Silva-Corvalán 1994, p. 148). (20) Pepe no vino hoy a trabajar. *Pepe/?él/Ø estará enfermo.        same reference Pepe no came today to work Pepe/?él/Ø will be sick “Pepe did not come to work today. He must be sick.” (21) Hoy no fui a trabajar. Pepe/él/*Ø pensó que estaba enferma.       switch reference today I no went to work Pepe/él/*Ø thought that I was sick “Today I did not go to work. Pepe/he thought I was sick.” While the distribution of overt and null subjects in English may also be regulated by discoursepragmatic factors, null subjects are rare in English (Torres Cacoullos & Travis 2015). Heritage speakers tend to use more overt subjects in pragmatically illicit contexts than monolingual speakers or speakers with a stronger command of Spanish than of English. Higher uses of overt subjects in lower proficiency heritage speakers may be related to the simplification of agreement morphology and more frequent use of canonical (SVO) word order. Sometimes, the features of the null pronoun cannot be reliably and consistently retrieved from overt agreement on the verb. However, the rate of null and overt subjects is also affected in heritage speakers who do not show problems with agreement and word order: lower rates of null subjects and overuse of overt subjects have been documented in several studies (Montrul 2004a; Montrul & Sánchez-Walker 2015; Otheguy & Zentella 2012; Silva-Corvalán 1994, 2014). It seems that the pragmatic features that regulate the use of null subjects erode, resulting in higher rates of overt pronouns in discourse contexts where a null subject would be more felicitous than an overt subject/pronoun (Sorace 2000). This phenomenon has been observed in production and in interpretation tasks (Keating, VanPatten, & Jegerski 2011). Finally, similar difficulties with the pragmatic constraints on null and overt subjects in null subject languages have been reported in studies of bilingual children of pre-school and school-age (Montrul & Sánchez-Walker 2015; Paradis & Navarro 2003; Silva-Corvalán 2014). This suggests that the pragmatic distribution of subjects is vulnerable to partial and incomplete acquisition during childhood.

Object clitics Object pronouns in Spanish are clitics. Third person accusative pronouns like las and los, see examples (1) and (2), carry gender and number agreement, whereas the dative pronoun le/les only carries number marking. There is also the impersonal and reflexive clitic se, used with some verbs and some constructions. Clitics are syntactically and morphologically independent words, but unlike strong pronouns (él, ella) that have their own stress, clitics depend phonologically on a stressed host, like a verb. In Spanish, clitics precede finite verbs, as in (22). With imperatives, infinitives and participles (non-finite verbs) clitics appear after the verb, as in (23). When there is a sequence of a finite auxiliary and a participle or an infinitive, clitics stay low next to the nonfinite form or rise before the conjugated auxiliary, as in (24). Omission of clitics or null objects is usually not grammatical in Spanish (unlike in Brazilian Portuguese, for example), except in very restricted contexts like indefinite objects, as in (25). Furthermore, Spanish is a clitic doubling language, especially with datives, and the clitic can co-occur with the noun phrase it refers to. With indirect objects, clitic doubling is optional, as in (26), but in clitic left 154

Morphology, syntax, and semantics

dislocations—when the object is fronted as in (27)—dative and accusative clitic doubling is obligatory. Finally, there are other constructions that require obligatory dative clitic doubling, like inalienable possession, as in (28). (22) María los compró. Maria them-masc.pl. bought “María bought them.” (23) hazlo/   hacerlo/haciéndolo do.imp it/do.inf. it/doing it “do it/doing it” (24) María quiso venderlo./ María lo quiso vender. Maria wanted to sell it/Maria it wanted to sell “María wanted to sell it.” (25) Trajiste dinero? Sí, traje. brought 2nd sg money? yes brought “Did you bring money? Yes, I brought (some).” (26) María (le) dio un regalo a Pedro Maria him-dat. gave a gift to Pedro “Maria gave Pedro a gift.” (27) Las carpetas las dejó Juan en la oficina. the-fem.pl folder-fem.pl them.fem.pl. left Juan in the-fem. office-fem “The folders, Juan left them in the office.” (28) Le quebraron la nariz a Pedro. Him.dat broke the-fem nose-fem to Pedro “They broke Pedro’s nose.” Unlike Spanish, English does not have clitics, but this structural difference does not seem to affect Spanish as a heritage language. Silva-Corvalán (1994) found that, overall, clitics were very strong and resilient in heritage language grammars. All the participants produced clitics, the clitics were generally well placed with respect to the verb, and ungrammatical clitic omissions were very infrequent. Silva-Corvalán found that second-generation heritage speakers omitted less than 1% of clitics (0.8%), and the third generation, those with the weakest competence in Spanish, omitted 2.7% dative and accusative clitics and 6.6% reflexive clitics in different constructions. Montrul (2004b) analyzed the oral narratives of 24 heritage speakers and found that they produced accusative and dative clitics like the baseline group, at a rate of 30% of all objects. Only one participant exhibited three ungrammatical omissions of clitics. All the heritage speakers produced clitic doubling with indirect objects like the native speakers, and knew where to place the clitics with respect to the finite verb. Lower proficiency heritage speakers, however, did not produce clitic doubling in inalienable possession constructions (like example 28) and produced possessive pronouns as in English (su nariz “his nose”). In another study, Montrul 155

Silvina Montrul

(2010) investigated knowledge of clitics in a grammaticality judgment task and a comprehension task. Sentences with clitics and different word orders, including clitic left dislocations as in (27) and preverbal objects, were the target of this study. Montrul found that, in general, the heritage speakers tested had more-native-like knowledge of all these structures compared to the L2 learners of similar proficiency in Spanish, confirming that knowledge of clitics is quite robust in Spanish heritage speakers. Leal Méndez, Rothman, and Slabakova (2015) tested the semantics of clitic left dislocations in Spanish heritage speakers of advanced and intermediate proficiency, finding that the heritage speakers did not differ from the group of native speakers to which they were compared. In general, it seems that the morphosyntax of clitics, clitic placement and clitic doubling, and the referential properties of clitics, are very resilient in Spanish heritage language grammar and acquired at native-like levels.

Complex syntax In this section I discuss aspects of complex syntax, like relative clauses, passives, and questions, which involve movement of constituents and referential dependencies. These structures are acquired later in the acquisition process, after age 3. Many of these structures are also more frequent in written discourse, especially verbal passives, and it is likely that heritage speakers do not fully master these structures, especially if they were not schooled in Spanish. Pascual y Cabo (2013), who included passive sentences with transitive verbs and unaccusative verbs in a test whose main objective was to test the syntactic properties of psych verb constructions, found that adult heritage speakers generally exhibited monolingual-like judgments, demonstrating knowledge of the syntactic and semantic restrictions of the passive voice in Spanish. For example, they accepted passive sentences with canonical transitive verbs and class II psych-predicates (La casa fue diseñada por los arquitectos “The house was designed by the architects,” *La niña fue asustado [sic] por los perros “The girl was frightened by the dogs”) and were generally sensitive to ungrammatical sentences with unaccusative predicates in passive constructions (*El libro fue desaparecido por Pedro “*The book was disappeared by Pedro”). Another complex structure is relative clauses, which are adjectival clauses that modify a noun phrase. Relative clauses involve long-distance dependencies because they connect the noun it modifies with a relative pronoun. Depending on the syntactic analysis, this is accomplished through movement of constituents and a gap, or through a binding relationship between the head of the relative clause and the referent of the relative clause. O’Grady, Lee, and Choo (2001) tested comprehension of subject and object relative clauses by Korean heritage speakers, and Polinsky (2011) conducted a similar study with Russian heritage speakers using a picture sentence-matching task. Both studies found that heritage speakers had more difficulty with the comprehension of object relative clauses (The cat that the dog is chasing) than with the comprehension of subject relative clauses (The dog that is chasing the cat). Sánchez-Walker (2013) tested written comprehension of subject and object relative clauses with inanimate objects and subjects in young adult Spanish heritage speakers in the United States. Subject relative clauses in English have V-O order within the relative clause, as in (29). Spanish has the English order, as in (30a), but also has object-verb inversion and displays O-V, as in (30b). (29) The submarine that sank the boats (V-O) (30) a El submarino que hundió los barcos (V-O) b El submarino que los barcos hundió (O-V) 156

Morphology, syntax, and semantics

In object relative clauses, the complementizer that is optional in English, and the word order within the relative clause is S-V, as in (28). In Spanish, the complementizer que is not optional, and both S-V and V-S word orders are possible within the relative clause, as in (32a, 32b). (31) The submarine (that) the boats sank. (S-V) (32) a El submarino que los barcos hundieron. (S-V) b El submarino que hundieron los barcos. (V-S) Sánchez-Walker (2013) found that, in general, the Spanish heritage speakers were quite accurate comprehending relative clauses, probably because they were of higher proficiency than the heritage speakers of Russian and Korean tested in other studies. At the same time, she found that the Spanish heritage speakers were very accurate with the sentences whose word order was like in English, yet inaccurate interpreting the sentences with the word order that is different from that of English (30b and 32a). Question formation with wh-movement is another complex syntactic structure that has received attention in Spanish as a heritage language, especially because there are some syntactic differences with English regarding subject-verb inversion and the use of complementizers. The subject wh-questions in (33) and (34) differ in the grammaticality of the complementizer. In English, complementizers are not required and are actually ungrammatical if expressed, as in (33a, 33b). In contrast, the complementizer is required in Spanish, as in (34a, 34b). In some syntactic approaches this difference is called the that-t effect. (33) a *Who do you think that came? b Who do you think came? (34) a ¿Quién crees que vino? who you-think that came?

b ¿*Quién crees vino? who you-think came?

Mueller Gathercole (2002b) tested bilingual school-age children in Miami on their knowledge of the that-t effect (examples 33 and 34) with a grammaticality judgment task. She found that the children performed better with the grammaticality judgment task in Spanish than in English, and recognition of ungrammatical sentences improved from 2nd grade to 5th grade. The children who attended dual immersion schools and came from homes where Spanish was the only language spoken performed better than children attending English-only schools and children of homes where English and Spanish were spoken. Montrul et al. (2008b) investigated knowledge of wh-movement, subject-verb inversion, and the use of complementizers in heritage speakers of intermediate proficiency in Spanish and monolingually raised native speakers of Spanish. The two groups rated grammatical and ungrammatical sentences with subject and object questions in a written grammaticality judgment task. Despite significant differences between native and heritage speakers, the heritage speakers were quite accurate with subject-verb inversion and complementizers even though Spanish and English differ in this regard. Therefore, this area of syntactic knowledge shows developmental effects and improvement with age when bilingual children and heritage speakers are compared. 157

Silvina Montrul

Cuza and Frank (2011) investigated the syntax-semantics interface by focusing on the meaning associated with embedded interrogatives. Embedded interrogatives have two complementizers in Spanish differ and from statements which have only one. Consider the sentences in (35) and (36). (35) María le dijo a Juan adonde fueron los niños.    Statement “Maria told Juan where the children went.” (36) Maria le dijo a Juan que a dónde fueron los niños.  Question “Maria asked Juan where the children went.” In (35) the sentential complement [adonde fueron los niños] is interpreted as a reported assertion whereas in (36) [a dónde fueron los niños] is a reported question introduced by the complementizer que “that.” The main verb is the same in both sentences (decir “say”) and the complementizer que introduces the question. Of course, the same sentence in (36) could be introduced with the verb preguntar “ask,” in which case the complementizer que would be optional. In English, two complementizers are not possible, and the meaning of a complex sentence as assertion or indirect question is marked by the meaning of the main verb. Cuza and Frank tested 32 lowadvanced proficiency heritage speakers of Spanish on their knowledge of the meaning of these sentences. The main tasks were a written sentence completion task, an acceptability judgment task, and a preference task. The results showed that the heritage speakers used indirect questions with two complementizers very infrequently in the completion task (23%) compared with the comparison group of fluent speakers of Spanish (60%). In the acceptability judgment task, unlike the native speakers, the heritage speakers were not sensitive to the semantic difference between structures with two complementizers and structures with one complementizer. A few heritage speakers with advanced proficiency showed sensitivity to the semantic distinction between the two types of sentences in the preference task, but they were still less inclined to accept these options than the comparison group of native speakers. Therefore, this aspect of syntax-semantics appears to not be fully developed in Spanish heritage speakers.

On the nature of grammatical development Heritage languages develop from childhood and into adulthood, like any other language in native speakers. In every language, different aspects of the language have developmental schedules. For example, phonetics and phonology develop in utero and during the first year of life, while morphosyntax emerges at about age 2, and complex syntax after age 3 or 4. Early acquisition and mastery refers to the pre-school period (up to age 4), while later language development is considered from age 4–5 and later (late childhood). This period coincides with the onset of schooling, and many heritage speakers do not receive schooling in the heritage language (see Lindholm-Leary, this volume). The acquisition of heritage languages deals with the study of the developmental stages and outcome of learning a heritage language from childhood and into adulthood, as well as the wax and wane of the heritage language in response to input factors (Montrul 2016a). As a sociopolitically minority language, the heritage language is acquired as a first language during the first years of life, as in sequential bilinguals, or simultaneously with the majority language since birth, as in simultaneous bilinguals. The morphosyntax of the Spanish language is vast and complex, and the descriptive review presented here is hardly exhaustive. The aim of this chapter was to 158

Morphology, syntax, and semantics

synthesize most recent research on the morphological and syntactic abilities of Spanish heritage speakers. As this chapter shows, some grammatical areas have been studied more than others, other areas need to be investigated in more depth since the review here is based only one study, and other areas have not yet been investigated. Nonetheless, some preliminary conclusions can be reached about the vulnerability or resilience of different aspects of morphology, syntax, and semantics in the context of established developmental schedules for Spanish. Table 10.1 presents a summary of the main findings. Table 10.1 shows that even though Spanish is the weaker language in heritage speakers of Spanish, different aspects of the structure of Spanish show differential levels of mastery by the time heritage speakers reach young adulthood. In general, we see that inflectional morphology, which is highly dependent on input factors and frequency, is vulnerable to simplification and loss, especially in heritage speakers of lower proficiency in Spanish. Yet, within this broad area, there are differences. In general, nominal morphology (gender, case) is more affected than verbal morphology. Within verbal morphology, there are very few studies on agreement, but it looks like agreement and tense morphology are more resilient than aspectual and mood morphology, which are also more related to meaning and syntactically more complex. Mood emerges very early in child acquisition (about age 2 with commands and verbs that subcategorize the subjunctive (Valdivieso 1994), but is not mastered until age 12 in monolingual children (Blake 1983). It is therefore not surprising that insufficient input and use of the heritage language during late childhood affects this verbal category more than any others. In terms of morphosyntax that interfaces with semantics and pragmatics, there are also important dissociations. We see that clitics, which do not exist in English, are acquired robustly by child and adult heritage speakers and used in the appropriate syntactic positions depending on verb finiteness. Heritage speakers also know and use clitic doubling at native levels. Clitic omissions are minimal, and the errors observed have to do with case and agreement mismatches, but not with the syntax or semantics of clitics. In contrast, the use of definite determiners is Table 10.1  Morphosyntactic abilities of Spanish heritage speakers Grammatical area

Developmental schedule

Inflectional morphology Nominal morphology gender agreement early case early Verbal morphology agreement early tense early aspect early mood early emergence, late mastery Morphosyntax-semantics-pragmatics interface definite articles early object clitics early subject pronouns early Complex syntax-semantics verbal passive late childhood relative clauses late childhood wh-questions late childhood interrogatives late childhood

159

Vulnerable or resilient?

vulnerable vulnerable quite resilient resilient vulnerable very vulnerable vulnerable semantics very resilient vulnerable pragmatics quite resilient quite resilient quite resilient vulnerable semantics

Silvina Montrul

highly influenced by the majority language (English) at the semantic level, and so is the distribution of overt subjects in Spanish, which appear to not respect pragmatic constraints on their use. The different developmental schedules and degree of vulnerability of different structures are less related to time of acquisition and appear to depend on the grammatical complexity of the particular category as a function of the architecture of language (syntax, morphology, semantics, discourse) and its frequency and transparency in the input. Finally, compared to the study of inflectional morphology and overt subject distribution, the acquisition of complex syntax has been much less studied. The few studies covered in this chapter show that the syntax of verbal passives, relative clauses, and wh-questions are not in general problematic for heritage speakers of Spanish, while the semantics of embedded interrogatives, a quite subtle and perhaps infrequent phenomenon, is not properly mastered. However, these conclusions are based on a handful of studies and should be further pursued in future research. At the same time, there are still many areas of morphosyntactic knowledge that have not been studied in heritage speakers that will give us a more nuanced picture of how the overall grammar is affected under fluctuating input conditions. For example, heritage speakers appear to show robust knowledge of clitic pronouns at the syntactic level, but do they also develop full knowledge of their referential properties in discourse, as in the difference between pronouns and reflexives (for example, in sentences like Sara vio a Maria y la saludó. Sara se/la sorprendió, who do la and se refer to?) Although the review presented here suggests that complex sentences are not very much affected in heritage grammars, more experimentation with different structures involving complementation and subordination and using tests and techniques targeting comprehension, processing, oral and written production are in order, since many of these structures in particular are fully mastered, integrated, and used productively during the school age period.

Conclusion Investigating the morphosyntactic abilities of Spanish heritage speakers is important for theoretical views of language and language acquisition and for practical considerations related to pedagogical applications, language therapy interventions, and language policies. Theoretically, we need a better understanding of how different aspects of morphosyntax are partially acquired or mastered depending on quantity of input and use. Unfortunately, space limitations do not allow me to elaborate further on the theoretical significance of these findings (but see Montrul 2016a, in press). Results from studies of morphosyntax can inform pedagogical practices as to what to focus on in the classroom and how to support the development of particular structures. Beyond the classroom, they can inform language policies to promote academic support of the heritage language throughout the entire school age period until the end of high-school, not just elementary school (for those schools that offer bilingual education). For speech pathologists, these results can offer guidelines of what to expect typically developing bilinguals to achieve by certain ages and how to understand what children with speech delays may need help with. As we have seen, many semantic and pragmatic aspects of morphosyntax, which are crucial to extract meaning and implications from texts, are vulnerable to incomplete acquisition when the language is not used enough and not supported at school. Professional use of the language, which is a desired goal of many Spanish heritage speakers who seek to improve their academic proficiency in the language, requires command of complex syntax, semantics, and pragmatics.

160

Morphology, syntax, and semantics

Note 1 All native speakers of English, for example, acquire that the rule for the past tense is the addition of the morpheme –ed to a verb stem and do so productively with above 90% accuracy in obligatory contexts. They reach ceiling performance and they converge on that performance with little variation, as opposed to L2 learners and young monolingual children who make errors with the application of this rule during development. Ceiling performance with an obligatory rule would be considered native ability in morphology, for example.

References Alarcón, I. (2011). Spanish gender agreement under complete and incomplete acquisition: Early and late bilinguals’ linguistic behavior within the noun phrase. Bilingualism: Language and Cognition 14, 332–350. Anderson, R. (1999). Noun phrase gender agreement in language attrition. Preliminary results. Bilingual Research Journal 23, 318–337. Au, T., Knightly, L., Jun, S.-A., & Oh, J. (2002). Overhearing a language during childhood. Psychological Science 13, 238–243. Blake, R. (1983). Mood selection among Spanish-speaking children, ages 4 to 12. The Bilingual Review 10, 21–32. Butt, M. (2006). Theories of case. Cambridge, UK: Cambridge University Press. Cummins, J. & Swain, M. (1986). Bilingualism in education: Aspects of theory, research and practice. London: Longman. Cuza, A., & Frank, J. (2011). Transfer effects in the syntax-semantics interface: The case of double-que questions in heritage Spanish. The Heritage Language Journal 8, 66–89. Cuza, A., Pérez-Tattam, R. Barajas, E., Miller, L., & Sadowski, C. (2013). The development of tense and aspect morphology in child and adult heritage Spanish: Implications for heritage language pedagogy. In J. Schwieter (Ed.), Innovative research and practices in second language acquisition and bilingualism, pp. 193–220. Amsterdam: John Benjamins. Dowty, D. (1986). The effects of aspectual class on the temporal structure of discourse: semantics or pragmatics? Linguistics and Philosophy 9, 37–61. Fairclough, M. (2005). Spanish and heritage language education in the United States. Struggling with hypotheticals. Frankfurt, Germany: Verveurt Iberoamericana. Foote, R. (2011). Integrated knowledge of agreement in early and late English-Spanish bilinguals. Applied Psycholinguistics 32, 187–220. Gallo Valdivieso, P. (1994). Adquisiciones gramaticales en torno al imperativo: Lo que se aprende dando órdenes. In S. López Ornat (Ed.), La adquisición de la lengua española, pp. 47–58. Madrid: Siglo XXI. Grosjean, F. (2008). Studying bilinguals. Oxford, UK: Oxford University Press. Grosjean, F., & Py, B. (1991). La restructuration d’une première langue: L’intégration de variantes de contact dans la compétence de migrants bilingues. La Linguistique 27, 35–60. Gutiérrez, M. (1996). Tendencias y alternancia en la expresión de condicionalidad en el español hablado en Houston. Hispania 79, 568–577. Harris, J. (1991). The exponence of gender in Spanish. Linguistic Inquiry 22, 27–62. Hodgson, M. (2005). Children’s production and comprehension of grammatical aspect. In R. Gess & E. Rubin (Eds.), Theoretical and experimental approaches to romance linguistics, pp. 125–142. Amsterdam: John Benjamins. Hurtado, A., & Vega, L.A. (2004). Shift happens: Spanish and English transmission between parents and their children. Journal of Social Issues 60(1), 137–155. Keating, G., VanPatten, B., & Jegerski, J. (2011). Who was walking on the beach? Anaphora resolution in Spanish heritage speakers and adult second language learners. Studies in Second Language Acquisition 33, 193–222. Lambert, W. (1977). Effects of bilingualism on the individual: Cognitive and socio-cultural consequences. In P. A. Hornby (Ed.), Bilingualism: Psychological, social and educational implications (pp. 15–28). New York: Academic Press. Lavandera, B. (1984). Variación y significado. Buenos Aires: Hachette. Leal Méndez, T., Rothman, J., & Slabakova, R. (2015). Discourse-sensitive clitic-doubled dislocations in heritage Spanish. Lingua 155, 85–97.

161

Silvina Montrul Lynch, A. (1999). The subjunctive in Miami Cuban Spanish. Bilingualism, contact and language variability. Unpublished doctoral dissertation, University of Minnesota, MN. Martínez-Mira, M. I. (2009). Spanish heritage speakers in the southwest: Factors contributing to the maintenance of the subjunctive in concessive clauses. Spanish in Context 6, 105–126. Merino, B. (1983). Language loss in bilingual Chicano children. Journal of Applied Developmental Psychology 4, 277–294. Montrul, S. (2002). Incomplete acquisition and attrition of Spanish tense/aspect distinctions in adult bilinguals. Bilingualism: Language and Cognition 5, 39–68. Montrul, S. (2004a). The acquisition of Spanish. Amsterdam: John Benjamins. Montrul, S. (2004b). Subject and object expression in Spanish heritage speakers: A case of morphosyntactic convergence. Bilingualism: Language and Cognition 7, 125–142. Montrul, S. (2006). Bilingualism, incomplete acquisition and language change. In C. Lefebvre, L. White, & C. Jourdens (Eds.), L2 Acquisition and Creole genesis. Dialogues, pp. 379–400. Amsterdam: John Benjamins. Montrul, S. (2007). Interpreting mood distinctions in Spanish as a heritage language. In K. Potowski & R. Cameron (Eds.), Spanish in contact. Policy, social and linguistic inquiries, pp. 23–40. Amsterdam: John Benjamins. Montrul, S. (2008). Incomplete acquisition in bilingualism: Re-examining the age factor. Amsterdam: John Benjamins. Montrul, S. (2009). Incomplete acquisition of tense-aspect and mood in Spanish heritage speakers: Special issue of The International Journal of Bilingualism 13(2), 239–269. Montrul, S. (2010). How similar are L2 learners and heritage speakers? Spanish clitics and word order. Applied Psycholinguistics 31, 167–207. Montrul, S. (2014). Searching for the roots of structural changes in the Spanish of the United States. Lingua 151, 177–196. Montrul, S. (2016a). The acquisition of heritage languages. Cambridge, UK: Cambridge University Press. Montrul, S. (2016b). Dominance and proficiency in second language acquisition and bilingualism. In C. Silva-Corvalán & J. Traffers-Dallers (Eds.), Measuring dominance in bilingualism, pp. 15–35. Cambridge, UK: Cambridge University Press. Montrul, S. (2016c). Losing your case? Dative experiencers in Mexican Spanish and heritage speakers in the United States. In Diego Pascual y Cabo (Ed.), Advances in Spanish as a heritage language, pp. 99–123. Amsterdam: John Benjamins. Montrul, S. (in press). The bottleneck hypothesis and heritage language acquisition. In J. Cho, T. Judy, & T. Leal Mendez (Eds.), Volume in Honor of Roumyana Slabakova. Amsterdam: John Benjamins. Montrul, S., Bhatt, R., & Girju, R. (2015). Differential object marking in Spanish, Hindi and Romanian as heritage languages. Language 91, 3, 564–610. Montrul, S., & Bowles, M. (2009). Back to basics: Differential object marking under incomplete acquisition in Spanish heritage speakers. Bilingualism: Language and Cognition 12, 363–383. Montrul, S., Davidson, J., de la Fuente, I., & Foote, R. (2014). Early language experience facilitates gender agreement processing in Spanish heritage speakers. Bilingualism: Language and Cognition 17, 118–138. Montrul, S., Foote, R., & Perpiñán, S. (2008a). Gender agreement in adult second language learners and Spanish heritage speakers: The effects of age and context of acquisition. Language Learning 58, 503–553. Montrul, S., Foote, R., & Perpiñán, S. (2008b). Knowledge of wh-movement in Spanish L2 learners and heritage speakers. In J. Bruhn de Garavito & E. Valenzuela (Eds.), Selected Papers from the 8th Hispanic Linguistics Symposium, pp. 93–106. Somerville, MA: Cascadilla Press. Montrul, S., de la Fuente, I., Davidson, J., & Foote, R. (2013). The role of experience in the acquisition and production of diminutives and gender in Spanish: Evidence from L2 learners and heritage speakers. Second Language Research 29, 1, 87–118. Montrul, S., & Ionin, T. (2010). Transfer effects in the interpretation of definite articles by Spanish heritage speakers. Bilingualism: Language and Cognition 13, 449–473. Montrul, S., & Potowski, P. (2007). Command of gender agreement in school-age Spanish bilingual children. International Journal of Bilingualism 11, 301–328. Montrul, S., & Sánchez-Walker, N. (2013). Differential object marking in child and adult Spanish heritage speakers. Language Acquisition 20, 109–132. Montrul, S., & Sánchez-Walker, N. (2015). Subject expression in school-age Spanish-English bilingual children. In A. Carvalho, R. Orozco, & N. Shin (Eds.), Subject pronoun expression in Spanish: A crossdialectal perspective, pp. 231–248. Washington, DC: Georgetown University Press.

162

Morphology, syntax, and semantics Mueller Gathercole, V. (2002a). Grammatical gender in monolingual and bilingual acquisition. A Spanish morphosyntactic distinction. In D. K. Oller & R. Eilers (Eds.), Language and Literacy in Bilingual Children, pp. 207–219. Bristol, UK: Multilingual Matters. Mueller Gathercole, V. 2002b. Monolingual and bilingual acquisition. Learning different treatments of the that-trace phenomena in English and Spanish. In D. K. Oller & R. Eilers (Eds.), Language and Literacy in Bilingual Children, pp. 220–254. Bristol, UK: Multilingual Matters. O’Grady, W., Lee, M., & Choo, M. (2001). The acquisition of relative clauses by heritage and nonheritage learners of Korean as a second language. A comparative study. Journal of Korean Language Education 12, 283–294. Otheguy, R., & Zentella, A. C. (2012). Spanish in New York. New York: Oxford University Press. Paradis, J., & Navarro, S. (2003). The use of subjects by a Spanish-English bilingual child, crosslinguistic influence or the influence of the input? The Journal of Child Language 30, 371–393. Pascual y Cabo, D. (2013). Agreement reflexes of emerging optionality in heritage speaker Spanish. Unpublished doctoral dissertation, University of Florida. Pérez Pereira, M. (1991). The acquisition of gender: What Spanish children tell us. Journal of Child Language 18, 571–590. Polinsky, M. (2011). Reanalysis in adult heritage language: A case for attrition. Studies in Second Language Acquisition 33, 305–328. Potowski, K. (2007). Tense and aspect in the oral and written narratives of two-way immersion students. In D. Eddington (Ed.), Selected Proceedings of the 6th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages, pp. 123–136. Somerville, MA: Cascadilla Proceedings Project Rodríguez-Mondoñedo, M. (2008). The acquisition of differential object marking in Spanish, Probus 20, 111–145. Rothman, J. (2009) Understanding the nature and outcomes of early bilingualism: Romance languages as heritage languages. International Journal of Bilingualism 13(2),155–163. Sánchez-Walker, N. (2013). Comprehension of subject and object relative clauses in Spanish heritage speakers and L2 learners of Spanish. Qualifying doctoral paper, University of Illinois at UrbanaChampaign, IL. Sherkina-Lieber, M. (2011). Comprehension of Labrador Inuttitut functional morphology by receptive bilinguals. Unpublished doctoral dissertation, University of Toronto, ON. Silva-Corvalán, C. (1985). Modality and semantic change. In J. Fisiak (Ed.), Historical semantics: Historical word formation. Berlin: Mouton. Silva-Corvalán, C. (1994). Language contact and change: Spanish in Los Angeles. Oxford, UK: Clarendon. Silva-Corvalán, C. (2003). Sociolingüística y pragmática del español. Washington, DC: Georgetown University Press. Silva-Corvalán, C. (2014). Bilingual language acquisition: Spanish and English in the first six years. Cambridge, UK: Cambridge University Press. Silva-Corvalán, C., & Traffers-Dallers, J. (Eds.) (2016). Measuring dominance in bilingualism. Cambridge, UK: Cambridge University Press. Smith, C. (1991). The parameter of aspect. Dordrecht, the Netherlands: Kluwer Academic Press. Sorace, A. (2000). Differential effects of attrition in the L1 syntax of near-native L2 speakers. Proceedings of the 24th Boston University Conference on Language Development pp. 719–725. Sommerville, MA: Cascadilla Press. Ticio, E. (2015). The acquisition of differential object marking in Spanish-English early bilinguals. Linguistic Approaches to Bilingualism 5(1), 62–90. Torres Cacoullos, R., & C. Travis. (2015). Subject pronoun realization in Spanish and English: assessing inter-linguistic functional equivalence via intra-linguistic inherent 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. Tse, L. (2001). Heritage language literacy: A study of U.S. biliterates. Language, Culture and Curriculum 14(3), 256–268. Verkuyl, H. (1994). A theory of aspectuality: The interaction between temporal and atemporal structure. Cambridge, UK: Cambridge University Press. Yang, C. (2016). The price of productivity. Cambridge, MA: MIT Press.

163

11 HERITAGE SPANISH PHONETICS AND PHONOLOGY Rebecca Ronquest north carolina state university, usa

Rajiv Rao university of wisconsin-madison, usa

Introduction Research in the area of bilingual speech production and perception has a long history dating back decades, in which researchers have established that bilingual speakers’ past experiences with their languages, including age of acquisition of one or both languages, age of arrival to the country of immigration, and types of exposure to and usage of the first (L1) and second (L2) language, have a significant impact on shaping their phonetic and phonological systems (see Flege 1995; Flege & Eefting 1988, among many others). Until recently, however, very little work has directly examined the phonetic and phonological systems of heritage speakers, and more specifically, heritage speakers of Spanish (hereafter HSS) as they have been linguistically and culturally defined (see Potowski, this volume). The goal of the present chapter is to review investigations that examine the phonological production and perceptual patterns of HSS who were either born in the U.S. or moved here prior to the age of 6 and have experienced and continue to experience contact between Spanish and English. In the past several decades, considerable literature in the field of linguistics has examined different components of the linguistic system of HSS. Morphosyntax and pedagogy have received considerable attention, most likely because they are associated with academic forms of language, particularly writing, which HSS typically do not develop as children. However, studies on the heritage Spanish phonetic and phonological system are limited in number, perhaps due to impressionistic observations that HSS often “sound” like native speakers of Spanish (NSS). Researchers such as Benmamoun, Montrul, and Polinsky (2010), however, have suggested the existence of a “heritage accent,” although at present it is not entirely clear what that might entail. As a point of departure, some valid questions one may ask about the heritage Spanish sound system are: How are HSS similar to and different from NSS and age-matched L2 learners who acquired Spanish later in life? How might such similarities and differences contribute to HSS “sounding” like NSS? Or, conversely, how might they contribute to the perception of a “heritage accent?” This chapter is dedicated to addressing these and other related questions by

164

Heritage Spanish phonetics and phonology

reviewing studies on the production of segmental (i.e., individual consonant and vowel sounds) and prosodic (i.e., intonation, rhythm, and stress) features in heritage Spanish, often in comparison to native Spanish and English trends. Where relevant, investigations of speech perception are also incorporated. Finally, the chapter concludes by discussing the theoretical importance of the body of work covered, the utility of laboratory methods, and specific challenges posed by research on HSS.

Consonants Much of the initial work on the heritage Spanish consonantal system has focused on the voiceless and voiced stop consonants /ptk/ and /bdg/, respectively, as these groups of sounds have key phonetic and phonological differences in English and Spanish. The voiceless stops /ptk/, which are produced with a complete blockage of airflow and an absence of vocal cord vibration, differ with respect to voice onset time (VOT): the duration between the separation of articulatory organs and the initiation of vocal cord vibration associated with the following segment (Lisker & Abramson 1964). The time lapse between these two moments is typically longer in English than in Spanish, especially voiceless stops in word-initial position, resulting in English speakers realizing aspirated allophones [phthkh] with VOT > 50 milliseconds and a noticeable expulsion of air from the mouth upon production, and unaspirated allophones [ptk] typically with VOT < 50 milliseconds in other positions (Lisker & Abramson 1964). Spanish, however, is characterized by the presence of unaspirated allophones, and consequently shorter VOT with values ranging from 0–35 milliseconds in all positions. Regarding the voiced stops /bdg/, Spanish and English differ in their realization of these sounds in two central ways. In English, vocal fold vibration usually begins at the time of the release of the stop, while Spanish voiced stops are often referred to as prevoiced because voicing is initiated before the release of the stop (Lisker & Abramson 1964). A second important difference concerning /bdg/ relates to the presence of weakened variants in most varieties of Spanish. NSS produce these as weakened variants in some positions, but as true stops in others. While some traditional literature (e.g., Quilis 1993) refers to these variants as “fricatives” (i.e., sounds produced with muscular tension that produces audible friction), more acoustically grounded research (e.g., Eddington 2011) argues that they should be classified as “approximants” (i.e., where organs approach one another with less tension and friction). The approximants, represented by the symbols [βðγ], typically occur between vowels, both within and across word boundaries. The stop allophones [bdg], in contrast, are manifested after pauses, nasals (i.e., /m/ and /n/) and /l/ (only in the case of /d/). In English, such weakening between vowels does not take place and stop allophones typically occur across the board. Of additional importance, the phoneme /b/ in Spanish typically corresponds with two written forms, and (i.e., botar “to bounce” and votar “to vote” have different meanings but are both phonemically /botar/), while in English /b/ and /v/ are separate phonemes (i.e., “ban” and “van” are pronounced differently and have different meanings). The differences highlighted above between English and Spanish related to VOT and the weakening of /bdg/ have inspired what has been discovered about HSS’ consonants to date. Some of the earliest impressionistic (i.e., not acoustic) studies to compare HSS to L2 learners of Spanish revealed that HSS’ realizations of the voiced and voiceless stops more strongly resembled native speech and were judged as sounding more native-like than those produced by L2 learners (Au et al. 2002), and that /b/ and /d/ were judged as sounding more native-like than /g/ (Knightly et al. 2003). A related study (Au et al. 2008) expanded upon previous work by

165

Rebecca Ronquest and Rajiv Rao

distinguishing HSS as active (i.e., childhood speakers) or receptive (i.e., childhood overhearers) users of Spanish as children, but the overall results still showed that HSS performed more native-like than L2 learners. More recently, studies on stop consonants have emphasized acoustic analysis utilizing software such as Praat (Boersma & Weenink 2017). Such techniques have allowed for a closer analysis of sounds, revealing additional similarities and differences between HSS and NSS as well as evidence of how language dominance may influence fine-grained phonetic detail. Kim (2011), for example, noted that while HSS and native speaker controls patterned similarly with respect to their perception of natural and manipulated VOT stimuli, their production of the voiced and voiceless stops differed. When compared to NSS, the HSS, who were Englishdominant, produced the voiceless stops with significantly longer VOT and the voiced stops with significantly less prevoicing (both of which are more English-like). Conversely, their production of these same sounds in English did not differ significantly from the native English control group’s productions. In conjunction, these results not only demonstrate a clear influence of language dominance on speech production, but also indicate a mismatch between production and perception for this particular group of HSS. Furthermore, Amengual’s (2012) analysis of the VOT of /t/ showed that the primary VOT distinction between HSS and NSS occurred when /t/ appeared in words possessing a corresponding cognate in English, such as teléfono or total. That is, the HSS’ VOT increased and thus trended toward English-like aspiration in words that were more similar to English. Additional studies focusing on voiced stops have revealed that linguistic and extralinguistic factors influence how HSS produce these sounds. Rao (2014) examined the production of Spanish and English intervocalic /b/ in reading and picture description tasks. The analysis was based on articulatory tension-based distinctions (Martínez Celdrán 1985, among others), which were used to phonetically classify productions as pure approximants (i.e., the most weakened), tense approximants (i.e., more organ tension than the pure form, resulting in a sound closer to the fricative [v]), and stops. The statistical analysis demonstrated that a lesser degree of past and present use of Spanish with family and in social networks, stressed syllables, reading tasks, and word boundaries all significantly decreased pure approximant realizations. These factors increased stop productions, while in the case of , they raised the rate of tense approximants. Subsequently, Rao (2015) expanded upon Rao (2014) by analyzing all three voiced stops (/bdg/). The statistical outcomes of this second study conveyed that /b/ is unique when compared to /dg/: it showed distinctly higher rates of less-weakened forms. In sum, the overarching implications of Rao’s two studies are that visual access to orthography, English phonological processes/orthographic conventions, and prominent phonological contexts promoting articulatory strengthening (i.e., stressed syllables and word boundaries) all combined to reduce pure approximants, though increased past and present exposure to Spanish clearly minimized non-native-like realizations. Although much of the research examining the pronunciation of consonants by HSS has focused on stop consonants, rhotics (i.e., “r-like” sounds) are also of interest given the different phonemic inventories and realizations of these sounds in Spanish and English. The tap rhotic /ɾ/, which is characterized by one vibration resulting from the tongue tip’s brief contact with the alveolar ridge, is a phoneme in Spanish that contrasts only between vowels with the trill rhotic /r/, which occurs through a series of moments of brief contact between the tongue tip and the alveolar ridge (Hualde 2005). The trill is subject to considerable dialectal variation in Spanish, which, in some regions, is realized as a velarized variant (see Lipski 1994 for a review). The tap is produced in English (e.g., in “better”), but is not a phoneme, and the trill does not exist in English. 166

Heritage Spanish phonetics and phonology

Studies analyzing the production of rhotics in heritage Spanish have revealed a large degree of individual variation (Henriksen 2015) as well as a potential influence of social and dialectal factors (O’Rourke & Potowski 2016). Henriksen’s (2015) analysis of first-generation bilingual immigrants and second-generation Chicago area HSS of Mexican descent showed a vast array of rhotic realizations across speakers. While the rhotic contrast was generally maintained, it was done via modifications to segment duration rather than number of vibrations. The author suggests that this finding could be attributed to the large degree of phonetic variability associated with the production of the sounds in question, and not necessarily the result of English influence. O’Rourke and Potowski (2016) also assessed the production of rhotics in Chicago as produced by different generations of heritage speakers of Mexican, Puerto Rican, and MexiRican (i.e., mixed parentage) descent. They reported that, as a group, speakers of Puerto Rican and MexiRican heritage produced fewer velarized trills – a variant found almost exclusively in Puerto Rican Spanish – when communicating with a speaker of Mexican descent, but more frequently produced this variant in conversations with another Puerto Rican speaker. The authors concluded that this was further evidence of the stigmatized value of the velarized variant and, as such, it was utilized as a means to signal in-group membership. While the majority of the investigations described have focused on how HSS produce consonants, understanding how they perceive the sounds of the heritage and/or majority language is equally important, given that the way in which sounds are perceived may influence how they are produced. Recall that Kim (2011) noted differences only in production, but not perception, of the voiced and voiceless stops when comparing HSS and native controls. Boomershine (2014) and Mazzaro, Cuza, and Colantoni (2016), however, suggest that while HSS often pattern closely with NSS in terms of their perception/discrimination of some consonantal and vocalic contrasts, they may not pattern identically. In an analysis of the perception of [d], [ð] and [ɾ], Boomershine (2014) found that HSS, native Spanish bilinguals (i.e., L1 Spanish – late L2 learners of English) and monolingual Spanish controls all rated [d]/[ɾ] and [ɾ]/[ð] as sounding “more different.” HSS’ perception of [d] and [ð], however, differed from the monolingual controls: monolingual Spanish speakers found these sounds to be very similar, likely because they are variant productions of the phoneme /d/ in Spanish, whereas HSS and native Spanish bilinguals rated them as sounding slightly more different, most likely because they are separate phonemes in English. Mazzaro, Cuza, and Colantoni’s (2016) comparison of HSS’, long-term immigrants’, and recent arrivals’ (i.e., control group) discrimination of the voiced and voiceless stops and front and back vowels revealed that HSS were just as accurate as the control group in discriminating /p/ and /b/, /t/ and /d/, and vocalic contrasts, but patterned with the longterm immigrants in being less accurate with respect to /k/ and /g/. Perception research on HSS therefore supports the notion that early exposure to the heritage language facilitates the formation of phonetic categories, although factors such as place of articulation and the phonological role of sounds in the heritage or majority language are clearly influential. As described by Kim (2011), there is also the potential for mismatches between perception and production; that is, even though contrasts might be perceived accurately, sounds may not necessarily be realized faithfully. To recap, previous research on the production of consonants by HSS has demonstrated that variables such as the presence/absence of orthographic representation in tasks, cognate status, and strength of ties to English/Spanish affect the production of consonants, as do linguistic variables such as syllable stress and position relative to word boundaries. In particular, the presence of orthography, cognates, stressed syllables, and word boundaries have all led to the production of less native-like stops, especially in English-dominant HSS. Despite these non-native productions, comparing the overall body of work in this section to previous L2 studies on the same 167

Rebecca Ronquest and Rajiv Rao

topic (e.g., Díaz-Campos 2004; Zampini 1994) illustrates that HSS’ consonants trend more strongly toward those of NSS. This idea further supports the previously mentioned positive role that exposure to Spanish during childhood plays in more closely approximating a native-like phonology (Knightly et al. 2003). However, while the study of consonantal production has drastically increased in recent years, there is still plenty of room for further exploration. Additional perception-based studies are needed in order to better understand the potential contribution of consonants to a “heritage accent.” Finally, while the discussion of consonants to this point has been enlightening, further work is needed on classes beyond stops and rhotics, including focusing on /s/ voicing or the production of /l/.

Vowels Research on the heritage Spanish phonetic/phonological system has also included studies of how HSS produce and perceive Spanish vowels. The Spanish vowel system consists of five contrastive vowels /ieaou/, which are organized fairly symmetrically in the acoustic space (Martínez Celdrán 1995). In contrast, the English vowel inventory is fairly large, consisting of between 11 and 15 phonemic vowels that are characterized by considerable dialectal and social variation (Ladefoged 2005; Maddieson 1984). While Quilis and Esgueva (1983) noted some cross-dialectal differences in vowel production, most researchers agree that Spanish vowel systems do not vary as widely across geographic regions as do those of English (Hualde 2005; Ladefoged 2005; but see Lope Blanch 1972 and Delforge 2008 for dialectal variation). In addition, unstressed vowels in English are typically reduced and produced as a central vowel known as schwa (Delattre 1969; Ladefoged 2005). A representation of the English and Spanish vowel systems, adapted from Bradlow (1995), is presented in Figure 11.1. The approximate location of schwa is also included, although these values represent a synthetic schwa based on a male vocal tract (Johnson 2003) and not an actual speaker’s production. Investigations of vowel production have reported a number of similarities and differences between the systems of HSS and NSS, most notably an asymmetrical acoustic distribution, considerable reduction of unstressed vowels, and significant effects of speech style on production. In comparison to the monolingual Spanish vowel system, HSS’ Spanish vowels have been found to exhibit a distinct organization within the acoustic space. Willis’s (2005) findings of a lowered /o/ and fronted /u/ in New Mexican Spanish were also reported in Boomershine (2012) and Ronquest (2012) for HSS in North Carolina and Chicago, respectively. The latter two studies also described a greater dispersion of the front vowel space relative to the back vowel space, as well as a backed and raised /e/. Such distributional differences suggest that the Spanish vowel system may not be as triangular or symmetrical as traditionally described for some HSS. There is also some potential that the more anterior position of Spanish /u/ in heritage vowel systems is the result of English contact, given that English /u/ is typically fronted relative to its Spanish counterpart (cf. Ladefoged 2005 for English; Quilis & Esgueva 1983 for Spanish). One of the most consistent findings across studies of heritage Spanish vowels, and a similarity they share with L2 learners’ productions (see Menke & Face 2010), is the greater tendency for unstressed vowels (such as otras and encontrar) to centralize than what has been described for monolingual varieties of Spanish. Alvord and Rogers (2014), Boomershine (2012), and Ronquest (2012, 2013) all reported that HSS’ unstressed vowels occupied a more centralized portion of the acoustic space when compared to their stressed counterparts. Additionally, Ronquest (2013) found that atonic vowels were significantly shorter than tonic vowels, but that the observed differences in vowel quality (i.e., centralization) were not the result of the shorter duration of atonic vowels. The prevalence of centralization, or movement of unstressed 168

Heritage Spanish phonetics and phonology

English and Spanish vowel systems 2400

2200

2000

1800

1600

1400

F2 (Hz) 1200 1000

800

600

400

200

0 0

200 i

u e

ɪ

e

ʊ o

ə ɛ æ

a

u 400

o

600

ɔ

ʌ ɑ

F1 (Hz)

i

800

1000 Spanish (Bradlow 1995)

English (Bradlow 1995)

schwa (Johnson 2003)

Figure 11.1 Spanish and English vowels produced by four male speakers of American English and four male speakers of Madrid Spanish (adapted from Bradlow 1995). The approximate location of schwa (i.e., F1 = 500 Hz, F2 = 1500 Hz) is based on values presented in Johnson (2003) for a synthetic schwa.

vowels toward the center of the vowel space, is particularly interesting in the context of bilingual speakers. Many traditional descriptions of Spanish vowels argue that stressed and unstressed vowels are similar with respect to their quality and duration (Delattre 1969; Hualde 2005). English unstressed vowels, in contrast, are frequently shortened and reduced to schwa (Delattre 1969; Ladefoged 2005). Although the reduction observed in these studies could be attributed, in part, to extensive English contact, Alvord and Rogers’s (2014) first-generation speakers, who were the most Spanish-dominant of the three groups analyzed, also reduced unstressed vowels. Their findings suggest that phonetic reduction may not be unique to the L2 and heritage vowel systems, but rather can emerge due to even very minimal contact with English and, therefore, merits further investigation. Despite their differences, however, the vowel production of HSS and NSS seems to be similarly affected by speech style. Vowels produced in informal tasks (e.g., sociolinguistic interviews, narratives) were more condensed within the acoustic space than those produced in formal, controlled word lists and carrier phrases in both Alvord and Rogers (2014) and Ronquest (2012). These findings are consistent with research on monolingual Spanish varieties, which exhibit similar tendencies (Harmegnies & Poch-Olivé 1992; Poch, Harmegnies, & Martín Butragueño 2008). That is, vowel space expansion in controlled speech relative to more naturalistic/less controlled styles has been documented in both monolingual and bilingual varieties, even though the exact distribution and location of specific vowels within the acoustic space is not necessarily identical across groups. Of additional importance, style-induced variability seems to be fairly consistent across distinct groups of HSS, with Alvord and Rogers’s (2014) Miami Cuban bilinguals exhibiting similar patterns to those reported by Ronquest (2012) for Mexican and Puerto Rican HSS in Chicago. 169

Rebecca Ronquest and Rajiv Rao

In addition to examining how HSS produce Spanish vowels, two recent investigations have assessed the perception and discrimination of Spanish and English vowel contrasts. Mazzaro, Cuza, and Colantoni (2016) observed that heritage speakers, long-term immigrants, and recent immigrants did not differ significantly with regard to their discrimination of the front and back vowels (/i/ and /e/ and /o/ and /u/, respectively). Boomershine (2013), however, reported differences between monolingual speakers of Spanish and bilinguals (i.e., HSS and advanced, late L2 learners of English and Spanish) concerning their perception of English front vowel contrasts. Similarity ratings of pairs such as bait/bit and beet/bit revealed that HSS and the two bilingual speaker groups rated these pairs as “more different,” whereas the monolingual Spanish speakers perceived them as “more similar.” Combined, these two studies confirm that language exposure and dominance shape the perceptual system, resulting in sensitivity to vocalic contrasts in both languages. In summary, research on the heritage Spanish vowel system, although still in its early stages, suggests that HSS resemble NSS in their production of Spanish vowels in some respects but differ in others. Stylistic variation, for example, appears to be consistent across HSS from different U.S. geographic regions and monolinguals alike. While the distributional differences and the presence of unstressed vowel reduction may suggest influence from English,1 direct comparisons across studies and to NSS should be done with caution. Research on Spanish vowels is limited compared to studies on consonants, with many investigations only examining Peninsular speakers’ production of stressed vowels in open syllables. Thus, the differences described may not be the result of contact with English, but rather representative of purely phonetic, dialectal, or methodological differences. In addition, as HSS are exposed to a highly variable input variety that might already contain some contact-induced features, direct comparisons with NSS may be somewhat misleading. Assessing similarities between HSS and their caretakers or, as pointed out by Menke and Face (2010), to other groups of bilingual speakers, may offer more useful insight into bilingual vowel production. Moving forward, careful selection of comparison groups and the addition of perceptual measures, along with those of production, may help determine which aspects of vowel production and perception are characteristic of HSS in general.

Prosody Prosody is perhaps the least explored area of the phonetic and phonological systems of HSS. It is divided into three main areas: intonation, which refers to modifications to pitch that influence the interpretation of meaning, for example, in statements versus questions and when conveying different attitudes and emotions (Hualde 2005); stress, which deals with the cuing and perception of relative prominence in speech (Hualde 2005); and rhythm, which generally addresses the distribution of stressed and unstressed moments in speech across time (Arvaniti 2012). Regarding rhythm, it has historically been associated with at least a two-way division based on duration: stress-timed languages (e.g., English) have relatively similar durations between stressed syllables, while syllable-timed languages (e.g., Spanish) tend to exhibit relatively similar durations of stressed syllables themselves (Pike 1945; see Arvaniti 2009, 2012 for an in-depth discussion of the topic of rhythm/timing and the problematic aspects of it). Specifically with regard to intonation, the few existing phonological studies on HSS are couched in the Autosegmental-Metrical (AM) model of intonational phonology (Ladd 2008; Pierrehumbert 1980). This model proposes that mental representations of intonation, or phonological targets above the level of individual segments, are comprised of a single high/low tone associated with a pitch peak or valley, respectively, or sequences of these tones. These tones or tonal sequences, occurring at the word-level and associating with stressed syllables, are deemed 170

Heritage Spanish phonetics and phonology

pitch accents. Such monotonal or bitonal targets can also be boundary tones tied to the edges of phrases, which are ways in which prosody chunks portions of discourse into units. The pitch accent and boundary tone sequences at the ends of utterances are particularly important for pragmatic purposes; for example, pitch rises, falls, or rise-fall movements corresponding with the end of an utterance in Spanish can be used to distinguish a question from a statement or a neutral statement from a more emphatic statement. Previous studies comparing the intonational systems of HSS to those of bilingual and monolingual native speakers have uncovered significant influences of factors such as utterance type, contact between varieties of Spanish and English, experience with and attitude toward Spanish, and task type. In Alvord’s (2006, 2010a, 2010b) examination of the intonation of three generations of Miami Cuban bilinguals, the third of which was comprised of HSS, the presence of both final rises and final falls across the data was the most noteworthy finding. A statistical look at question patterns revealed that first- and third-generation speakers preferred final falls, while those of the second generation clearly favored rises. Regarding the third generation in particular, analyses suggested that both the affirmation of Cuban identity and potential influence from English motivated the prevalence of final falls. Along similar lines, Henriksen’s (2012) work on intonation in Chicago area HSS and NSS also reported more key intergroup differences in questions as compared to statements, where HSS produced a wider range of configurations. The author posited that increased contact between English and Spanish could have yielded an innovative inventory of intonational strategies. Robles-Puente’s (2014) study on prosody in Los Angeles incorporated two groups of HSS that differed in age: adults either born in Los Angeles or who emigrated from Mexico as very young children, and adolescents and children born in Los Angeles to immigrant parents from Mexico. HSS’ trends in English and Spanish did not differ significantly; the same tonal configurations occurred in both of their languages. In terms of intergroup differences, while both groups of HSS exhibited phonological targets used by the NSS control group, the younger group produced intonational trends more closely approximating those of the controls. Rao’s (2016) study is smaller in scale but similar to that of Robles-Puente (2014); however, Rao’s study focused more on details concerning participants’ past and present use of Spanish, as well as demographic information about both the participants and their parents. Like Alvord (2006, 2010a) and Henriksen (2012), Rao also noted increased variation in question intonation and suggested that exposure to unique types of input at the individual speaker level plays a role in the observed intonational differences across the data (see also Pascual y Cabo & Rothman 2012). Finally, another recent investigation on HSS intonation, which is the first to emphasize task type, is Colantoni, Cuza, and Mazzaro (2016), who examined the statement intonation of HSS and older immigrants in reading and narrative tasks. They found significant differences in the pitch accents of both groups only in the reading task, which supports the orthography-related findings described in the previous section. The authors suggest that the lack of formal education, and thus of literacy skills in Spanish, led to the production of read samples that were less natural. As such, it was recommended that researchers beware when implementing reading tasks with HSS. Lastly, the authors extrapolated the results to the pedagogical level by providing some strategies for implementing prosody practice in the heritage Spanish classroom. Moving onto stress, two notable studies of HSS have revealed perception versus production discrepancies and the need to look at prosody’s interaction with other areas of linguistic competence. Kim (2015) compared the perception and production of lexical stress by HSS, NSS, and 4th semester L2 learners using verb minimal pairs (e.g., paso/pasó; stressed vowels are bolded). The HSS patterned with the NSS in their accurate perception of penultimate (e.g., paso) and final (i.e., pasó) stress, whereas the L2 learners were not able to distinguish between the 171

Rebecca Ronquest and Rajiv Rao

two stress patterns because they were biased toward the penultimate pattern. In the production experiments, however, the HSS and L2 learners often lengthened unstressed vowels in paso-type words, effectively producing a pasó-type word (i.e., final stress) instead. The latter tendency was not observed in the NSS’ productions. In a different approach to stress, Hoot (2012) revealed that HSS and Mexican NSS shifted stress leftward in a similar fashion when communicating presentational focus/new information (e.g., Mi mamá compró un carro rather than Compró un carro mi mamá was judged as acceptable in response to ¿Quién compró el carro?), suggesting that, at least for the sample of speakers in question, prosodic prominence in Spanish is flexible and does not necessarily have to fall in rightmost position. This analytical approach is unique in that it discussed the specific interface of prosody and syntax, extended the discussion of interfaces between linguistic areas to a general level, and argued for the inclusion of more interfaces in future examinations of heritage grammars. Regarding differences related to vowel duration, which is a common metric used to quantify speech rhythm, Robles-Puente (2014) conducted an analysis to see if HSS’ speech was more stress-timed or syllable-timed. It was shown that adult HSS produced English trends when speaking either language, whereas younger HSS born in LA accommodated their measures to the favored pattern of the language being spoken. In sum, when taking into account the differences between the findings for issues related to timing and those tied to intonation, RoblesPuente concluded that components classified within the general area of prosody seem to exhibit distinctions in bilingual grammars. After reviewing the relatively small body of studies on the prosodic system of HSS, it is clear that HSS pattern similarly to NSS in some respects, but the majority of studies report considerable differences, namely regarding question intonation, stress production, and rhythm. The cumulative findings have shed light on the contribution of prosodic factors to differences between heritage and native grammars above the level of individual sounds. Such differences could be prosodic features contributing to a “heritage accent.” However, once again, there is a lack of work from the perspective of perception, which, as seen at the segmental level, is begging to be addressed for prosody as well. Additionally, since intonation in particular exhibits a tremendous degree of dialectal variation (see Prieto & Roseano 2010), examining this specific part of prosody in HSS descending from a wider range of Spanish-speaking countries beyond Mexico and Cuba is a future research direction deserving of attention. Finally, stress is an important prosodic concept to continue investigating because it has pedagogical implications; that is, conscious awareness of stress in speech could assist many HSS with writing accent marks in Spanish.

Future research considerations The investigations summarized in this chapter have revealed a number of important similarities and differences between HSS and NSS (and, to a lesser degree, adult L2 learners of Spanish) with regard to their production and perception of Spanish segmental and prosodic features. This section expounds upon the theoretical importance of this work, speaking specifically to the merit of studies of speech perception and the utility of perceptual models in understanding the heritage Spanish phonetic/phonological system. In addition, it comments on the usefulness of laboratory approaches in examining pronunciation and makes suggestions on how researchers might tailor their methodologies based on participants’ experience with formal, written Spanish. Finally, the section concludes by discussing the challenge of addressing individual variation in heritage populations and revisiting the questions posed in the introduction. The majority of the works summarized in this chapter have focused on investigations of speech production, which have served to characterize and provide acoustic-phonetic details 172

Heritage Spanish phonetics and phonology

about HSS’ pronunciation of certain individual sounds, syllables, phrases, and utterances. Studies of speech perception, however, are equally important in augmenting our understanding of the underlying phonological system(s) of HSS and offering insight into why the production of certain sounds may prove difficult and/or differ from previously established norms. Models of speech perception (e.g., Speech Learning Model (SLM), Flege 1995; Perceptual Assimilation Model (PAM), Best 1995; Native Language Magnet (NLM), Kuhl 1993) have been widely employed in research examining cross-language perception and L2 phonological acquisition. Although the specific tenets of these models differ and will not be described in detail here, in very general terms, each posits that the way in which L2 and/or non-native speech sounds will be perceived, and in the case of the SLM, subsequently produced, depends on how they are processed with reference to L1 phones, articulatory gestures, or exemplar prototypes. Drawing upon one or more of these models to explore the sound systems of HSS, and including experiments of speech perception as well as production (as done by Kim 2011, 2015), will allow researchers to: (1) understand how HSS’ unique experience with the heritage and majority languages has shaped their perceptual system, and (2) move from purely descriptive characterizations of speech sounds to predicting how and why a particular sound or group of sounds may vary based on their perceptual (dis)similarity with competing sounds of the majority language. In terms of methodological approaches, the use of “natural” or “spontaneous” speech versus “laboratory” speech samples in the field of phonetics has received much attention in recent years. Although laboratory speech has been criticized as “unnatural” and not representative of how speakers may use language in informal settings, Xu (2010) argues in favor of these approaches, describing some of the advantages of analyzing speech “sampled under experimental control” (p. 330). Speech obtained in naturalistic settings, for example, may be either too poor in quality to analyze acoustically or may be lacking in a sufficient number examples of the segment under investigation. Employing controlled speech tasks therefore allows researchers to isolate the features under study, effectively removing many potentially confounding factors that perhaps could not be eliminated from data obtained under less-controlled conditions. One of the first steps in understanding what might constitute a “heritage accent” is isolating and characterizing individual segments or suprasegmental features – an undertaking that can very likely be achieved more easily by a carefully designed data elicitation instrument than in a sociolinguistic interview. Once detailed descriptions have been obtained, we may consider extending analyses to include a wider range of speech styles. While the analysis of laboratory speech samples in phonetics research has been successfully implemented in the studies reviewed here, and hopefully will continue to be used, it is necessary to emphasize the importance of carefully designing data elicitation instruments through a detailed consideration of participants’ experience with the heritage language. Some controlled tasks, especially those that involve reading, can be particularly problematic for some HSS who have limited experience with Spanish in its written form (Colantoni, Cuza, & Mazzaro 2016), and may result in orthographically influenced pronunciations (e.g., [v] for ). Additionally, since many HSS are not formally educated in Spanish, and thus are not accustomed to exercising their literacy skills in this language, they may find reading tasks particularly alien, which may lead to disfluencies that can impede the analysis of target features. Given the heterogeneity of heritage speaker populations, one of the biggest challenges in linguistic research on HSS concerns how to contend with individual differences. Extralinguistic factors such as age of onset of bilingualism, length of time in contact with the majority language, and self/formal assessment of proficiency and use of the majority and minority languages have been useful in describing some of the observed variation in speech production and perception. 173

Rebecca Ronquest and Rajiv Rao

Grouping HSS based on sociolinguistic generation (i.e., G2, G3), as done by Henriksen (2015) and O’Rourke and Potowski (2016), and conducting longitudinal and/or cross-generational studies of heritage speakers (Nagy & Kochetov 2013), may also prove insightful, allowing for a closer examination of how phonetic categories or mental representations of sounds vary within groups of HSS and how they may change over time. Second language acquisition research has also addressed the impact of cognitive factors such as motivation (see Dörnyei & Skehan 2003), which have, thus far, not been applied to HSS. Collectively, these approaches may permit additional commentary on questions such as incomplete acquisition, attrition, and the role of input, which have not been addressed with specific reference to heritage speakers’ phonetic/phonological systems as widely as they have been for other components of their grammars. Another potential source of individual variation that is particularly relevant to phonetics research concerns the inherent variability of the speech signal. That is, the inter- and intragroup differences observed in pronunciation may stem, in part, from the fact that the segments themselves exist along a continuum of possible realizations. As noted earlier, Henriksen (2015) argues that some of the variation observed in the production of rhotics may not be caused by English influence, but rather by the fact that Spanish rhotics encompass a variety of possible articulations. The same logic could be applied to voiced approximants, which have been shown to be continuous in nature (e.g., Eddington 2011), as well as voiceless stops, which are characterized by a range of possible VOT values (Lisker & Abramson 1964). Examining individual patterns of production, as did Henriksen (2015), will allow for a more thorough assessment of phonetic variability, and could potentially reveal important individual differences that might be obscured by looking at aggregate data alone. Ultimately, such an approach may aid in teasing apart individual variation attributable to extralinguistic factors from that which is due to purely phonetic factors.

Conclusions To revisit the questions posed in our introduction, we have described that while HSS seem to pattern more closely with NSS in some respects (e.g., lenition of /bdg/, style-induced variability in vowel production), they also exhibit characteristics resembling those attested in L2 learner speech (e.g., [v]-like production of , lengthened VOT, unstressed vowel reduction). It is perhaps because of this mix of features that we are faced with two seemingly contradictory observations about heritage Spanish pronunciation: “sounding” like NSS on the one hand but having a “heritage accent” on the other. Although the conjunct of present works does not allow us to provide a definitive answer to these highly complex questions at this time, it is clear that HSS’ production of their heritage language places them on a continuum situated somewhere between L2 learners and monolingual NSS. Combined, the phenomena discussed in this chapter have laid a strong foundation in the area of heritage Spanish phonetics/phonology, offering valuable insight into how HSS both produce and perceive the sounds of their heritage language. They have also raised further questions that researchers will hopefully strive to answer in the upcoming years in an effort to continue advancing our knowledge of heritage Spanish grammars as a whole.

Note 1 Vowel production may also depend on speakers’ location within the U.S. For example, work on English in the South has shown that some Spanish-speaking immigrants and their descendants acquire southern monophthongization (Wolfram, Carter, & Moriello 2004).

174

Heritage Spanish phonetics and phonology

Further reading Carter, P. (2005). Quantifying rhythmic differences between Spanish, English, and Hispanic English. In R. Gess & E. Rubin (Eds.), Theoretical and experimental approaches to Romance linguistics (pp. 63–75). Amsterdam and Philadelphia, PA: John Benjamins. An alternative but related examination of the rhythmic properties overviewed here. Flege, J., & MacKay, I. (2011). What accounts for “age” effects on overall degree of foreign accent? In M. Wrembel, M. Kul, & K. Dziubalska-Kołaczyk (Eds.), Achievements and perspectives in the acquisition of second language speech: New Sounds 2010, Vol. 2, (pp. 65–82). Bern, Switzerland: Peter Lang. An overview of extralinguistic factors and how they influence the L2 sound system. Rohena-Madrazo, M. (2013). Perceptual assimilation of occluded voiced stops by Spanish listeners. In C. Howe, S. Blackwell, & M. Lubbers Quesada (Eds.), Selected proceedings of the 15th Hispanic Linguistics Symposium (pp. 140–156). Somerville, MA: Cascadilla Proceedings Project. A test of the predictions made by the SLM and PAM with respect to the perception of stop voicing in Spanish. Zampini, M. (1998). L2 Spanish spirantization: A prosodic analysis and pedagogical implications. Hispanic Linguistics, 10, 154–188. An innovative analysis accounting for variation in L2 learners’ weakening of /bdg/.

References Alvord, S. (2006). Spanish intonation in contact: the case of Miami Cuban bilinguals. Unpublished doctoral dissertation. University of Minnesota, Minneapolis, MN. Alvord, S. (2010a). Variation in Miami Cuban Spanish interrogative intonation. Hispania, 93(2), 235–255. Alvord, S. (2010b). Miami Cuban Spanish declarative intonation. Studies in Hispanic and Lusophone Linguistics, 3(1), 3–39. Alvord, S., & Rogers, B. (2014). Miami-Cuban Spanish vowels in contact. Sociolinguistic Studies, 8(1), 139–170. Amengual, M. (2012). Interlingual influence in bilingual speech: Cognate status effect in a continuum of bilingualism. Bilingualism: Language and Cognition, 15(3), 517–530. Arvaniti, A. (2009). Rhythm, timing and the timing of rhythm. Phonetica, 66, 46–63. Arvaniti, A. (2012). The usefulness of metrics in the quantification of speech rhythm. Journal of Phonetics, 40, 351–373. Au, T. K., Knightly, L. M., Jun, S.A., & Oh, J. S. (2002). Overhearing a language during childhood. Psychological Science, 13, 238–243. Au, T. K., Oh, J. S., Knightly, L. M., Jun, S.A., & Romo, L. F. (2008). Salvaging a childhood language. Journal of Memory and Language, 58, 998–1011. Benmamoun, E., Montrul, S., & Polinsky, M. (2010). White paper: Prolegomena to heritage linguistics. Harvard University. Retrieved from https://scholar.harvard.edu/files/mpolinsky/files/hl_white_paper_june_ 12.pdf. Best, C. T. (1995). A direct realist view of cross-language speech perception. In W. Strange (Ed.), Speech perception and linguistic experience: Issues in Cross-Language Research (pp. 171–206). Timonium, MD: York Press. Boersma, P., & Weenink, D. (2017). Praat: Doing phonetics by computer [Computer program]. Retrieved from www.praat.org/. Boomershine, A. (2012, October). What we know about the sound system(s) of heritage speakers of Spanish: Results of a production study of Spanish and English bilingual and heritage speakers. Paper presented at the 12th Hispanic Linguistics Symposium, Gainesville, FL. Boomershine, A. (2013). The perception of English vowels by monolingual, bilingual, and heritage speakers of Spanish and English. In C. Howe, S. Blackwell, & M. Lubbers Quesada (Eds.), Selected proceedings of the 15th Hispanic Linguistics Symposium (pp. 103–118). Somerville, MA: Cascadilla Proceedings Project. Boomershine, A. (2014, March). The influence of language experience on speech perception: The case of heritage, monolingual and L2 speakers of Spanish and English. Paper presented at Current Approaches to Spanish and Portuguese Second Language Phonology, Washington, DC.

175

Rebecca Ronquest and Rajiv Rao Bradlow, A. (1995). A comparative acoustic study of English and Spanish vowels. Journal of the Acoustical Society of America, 97(3), 1916–1924. Colantoni, L., Cuza, A., & Mazzaro, N. (2016). Task related effects in the prosody of Spanish heritage speakers. In M. Armstrong, N. Henriksen, & M. Vanrell (Eds.), Interdisciplinary Approaches to Intonational Grammar in Ibero-Romance Intonation (pp. 1–24). Amsterdam: John Benjamins. Delattre, P. (1969). An acoustic and articulatory study of vowel reduction in four languages. International Review of Applied Linguistics in Language Teaching, 7(4), 295–325. Delforge, A. M. (2008). Unstressed vowel reduction in Andean Spanish. In L. Colantoni & J. Steele (Eds.), Selected Proceedings of the 3rd Conference on Laboratory Approaches to Spanish Phonology (pp. 107–124). Somerville, MA: Cascadilla Proceedings Project. Díaz-Campos, M. (2004). Context of learning of Spanish second language phonology. Studies in Second Language Acquisition, 26, 249–273. Dörnyei, Z., & Skehan, P. (2003). Individual differences in second language learning. In C. J. Doughty & M. H. Long (Eds.), The handbook of second language acquisition (pp. 589–630). Oxford: Blackwell. Eddington, D. (2011). What are the contextual variants of [β, ð γ] in colloquial Spanish? Probus, 23, 1–19. Flege, J. (1995). Second language speech learning: theory, findings, and problems. In W. Strange (Ed.), Speech perception and linguistic experience: Issues in cross-language research (pp. 233–277). Timonium, MD: York Press. Flege, J., & Eefting, W. (1988). Imitation of a VOT continuum by native speakers of English and Spanish: Evidence for phonetic category formation. Journal of the Acoustical Society of America, 83, 729–740. Harmegnies, B., & Poch-Olivé, D. (1992). A study of style-induced vowel variability: Laboratory versus spontaneous speech in Spanish. Speech Communication, 11(4–5), 429–437. Henriksen, N. (2012, March). Chicagoland heritage and native Mexican Spanish intonation: Three contact phenomena. Paper presented at Current Approaches to Spanish and Portuguese Second Language Phonology, Columbia, SC. Henriksen, N. (2015). Acoustic analysis of the rhotic contrast in Chicagoland Spanish: An intergenerational study. Linguistic Approaches to Bilingualism, 5(3), 285–321. Hoot, B. (2012). Presentational focus in heritage and monolingual Spanish. Unpublished doctoral dissertation. University of Illinois at Chicago, IL. Hualde, J. I. (2005). The sounds of Spanish. Cambridge, UK: Cambridge University Press. Johnson, K. (2003). Acoustic and auditory phonetics. Malden, MA: Blackwell. Kim, J.-Y. (2011). Discrepancy between perception and production of stop consonants by Spanish heritage speakers in the United States. Unpublished master’s thesis. Korea University, Seoul, South Korea. Kim, J.-Y. (2015). Perception and production of Spanish lexical stress by native Spanish speakers and Spanish heritage speakers. In E. Willis, P. Martín Butragueño, & E. Herrera Zendejas (Eds.), Proceedings from the 6th conference on Laboratory Phonology (pp. 106–128). Somerville, MA: Cascadilla Proceedings Project. Knightly, L., Jun, S.A., Oh, J. S., & Au, T.K. (2003). Production benefits of childhood overhearing. Journal of the Acoustical Society of America, 114, 465–474. Kuhl, P. K. (1993). Early linguistic experience and phonetic perception: Implications for theories of developmental speech perception. Journal of Phonetics, 21, 125–139 Ladd, D. R. (2008). Intonational phonology (2nd ed.). Cambridge, UK: Cambridge University Press. Ladefoged, P. (2005). Vowels and consonants (2nd ed.). Malden, MA: Blackwell. Lipski, J. (1994). Latin American Spanish. London: Longman. Lisker, L., & Abramson, A. (1964). A cross-language study of voicing in initial stops: Acoustical measurements. Word, 20(3), 384–422. Lope Blanch, J. M. (1972). En torno a las vocales caedizas del español mexicano. Estudios sobre el español de México (pp. 53–73). Mexico: Editorial Universidad Nacional Autónoma de México. Maddieson, I. (1984). Patterns of sounds. Cambridge, UK: Cambridge University Press. Martínez Celdrán, E. (1985). Cantidad e intensidad en los sonidos obstruyentes del castellano: hacia una caracterización de los sonidos aproximantes. Estudios de Fonética Experimental, I, 71–130. Martínez Celdrán, E. (1995). En torno a las vocales del español. En torno a las vocales del español: análisis y reconocimiento. Estudios de Fonética Experimental, VII, 195–218. Mazzaro, N., Cuza, A., & Colantoni, L. (2016). Age effects and the discrimination of consonantal and vocalic contrasts in heritage and native Spanish. In C. Tortora, M. den Dikken, I. Montoya, & T. O’Neill (Eds.), Romance linguistics 2013: Selected papers from the 43rd Linguistic Symposium on Romance Languages (LSRL), New York, April 17–19, 2013 (pp. 277–300). Amsterdam: John Benjamins.

176

Heritage Spanish phonetics and phonology Menke, M., & Face, T. (2010). Second language Spanish vowel production: An acoustic analysis. Studies in Hispanic and Lusophone Linguistics, 3(1), 181–214. Nagy, N., & Kochetov, A. (2013). VOT across the generations: A cross-linguistic study of contact-induced change. In P. Siemund, I. Gogolin, M. Schulz, & J. Davydova (Eds.), Multilingualism and language diversity in urban areas (pp. 19–38). Amsterdam: John Benjamins. O’Rourke, E., & Potowski, K. (2016). Phonetic accommodation in a situation of Spanish dialect contact: Coda /s/ and /r̄ / in Chicago. Journal of Hispanic and Lusophone Linguistics, 9(2), 355–399 Pascual y Cabo, D., & Rothman, J. (2012). The (il)logical problem of heritage speaker bilingualism and incomplete acquisition. Applied Linguistics, 33(4), 450–455. Pierrehumbert, J. (1980). The phonology and phonetics of English intonation. Unpublished doctoral dissertation. MIT, Cambridge, MA. Pike, K. L. (1945). The intonation of American English. Ann Arbor, MI: University of Michigan Press. Poch, D., Harmegnies, B., & Martín Butragueño, P. (2008). Influencia del estilo de habla sobre las características de las realizaciones vocálicas en el español de la ciudad de México. In Actas del XV Congreso Internacional ALFAL, CD 6–7. Montevideo, Uruguay. Prieto, P., & Roseano, P. (Eds.). (2010). Transcription of intonation of the Spanish language. Munich: LINCOM Publishers. Quilis, A. (1993). Tratado de fonología y fonética españolas. Madrid: Gredos. Quilis, A., & Esgueva, M. (1983). Realización de los fonemas vocálicos españoles en posición fonética normal. In M. Esgueva & M. Cantarero (Eds.), Estudios de fonética (pp. 159–251). Madrid: Consejo Superior de Investigaciones Científicas. Rao, R. (2014). On the status of the phoneme /b/ in heritage speakers of Spanish. Sintagma, 26, 37–54. Rao, R. (2015). Manifestations of /b, d, g/ in heritage speakers of Spanish. Heritage Language Journal, 12(1), 48–74. Rao, R. (2016). On the intonational phonology of heritage speakers of Spanish. In D. Pascual y Cabo (Ed.), Advances in Spanish as a heritage language (pp. 51–80). Amsterdam: John Benjamins. Robles-Puente, S. (2014). Prosody in contact: Spanish in Los Angeles. Unpublished doctoral dissertation. The University of Southern California, Los Angeles, CA. Ronquest, R. (2012). An acoustic analysis of heritage Spanish vowels. Unpublished doctoral dissertation. Indiana University, Bloomington, IN. Ronquest, R. (2013). An acoustic examination of unstressed vowel reduction in heritage Spanish. In C. Howe, S. Blackwell, & M. Lubbers Quesada (Eds.), Selected proceedings of the 15th Hispanic Linguistics Symposium (pp. 151–171). Somerville, MA: Cascadilla Proceedings Project. Willis, E. (2005). An initial examination of Southwest Spanish vowels. Southwest Journal of Linguistics, 24, 185–198. Wolfram, W., Carter, P., & Moriello, B. (2004). Emerging Hispanic English: New dialect formation in the American South. Journal of Sociolinguistics, 8(3), 339–358. Xu, Y. (2010). In defense of lab speech. Journal of Phonetics, 38, 329–336. Zampini, M. (1994). The role of native language transfer and task formality in the acquisition of Spanish spirantization. Hispania, 77, 470–481.

177

12 THE LEXICON OF SPANISH HERITAGE LANGUAGE SPEAKERS Marta Fairclough university of houston, usa

Anel Garza san jacinto college, usa

Introduction “Words are the basic building blocks of language, the units of meaning from which larger structures such as sentences, paragraphs and whole texts are formed” (Read 2000: 1). But the body of words that a person knows, i.e., their vocabulary knowledge or lexicon, is a complex construct that includes several dimensions. Nation (2013) distinguishes between receptive and productive vocabulary, recognizing that productive skills require higher levels of knowledge than receptive skills. Two other dimensions that have been traditionally studied as part of lexical competence are vocabulary size or breadth and vocabulary depth. Some internal (e.g., age of acquisition) and external factors (e.g., linguistic contact) can have great impact on an individual’s vocabulary. In addition to studies measuring vocabulary from those different perspectives, a number of researchers have looked at correlations between measures of vocabulary knowledge with other language skills such as reading and listening comprehension, and even with general language proficiency (Schmitt 2010). However, most of the studies have focused on the English language, while here the focus is on Spanish. Despite the complexity of lexical knowledge and the importance of vocabulary in language learning, research in this area, especially from learners whose target language is Spanish, has been neglected in favor of acquisition of grammar (Lafford et al. 2003). Even less is known about the vocabulary knowledge of heritage Spanish speakers. After a brief explanation of several aspects of vocabulary and key issues regarding lexical knowledge, use, and acquisition, the chapter will present an overview of research carried out to date on the lexicon of SHL speakers. The chapter will then offer some recommendations for practice and suggest a future research agenda.

178

Spanish heritage language speakers

The notion of vocabulary and its dimensions: some key issues Defining vocabulary Read and Chapelle (2001: 1–2) summarize several perspectives on “vocabulary” as a construct. From a pragmatic point of view, some propose that vocabulary should always be in context (Singleton 1999) – i.e., it can only be truly examined within a specific situation or discourse – while others perceive it as simply memorized lexical units (Skehan 1998). Many think of vocabulary as consisting of individual word forms (Laufer 1998; Laufer and Nation 1999). However, to the question of “What is a ‘word’?” Read (2000: 17–19) explains that a “lemma” is a base plus its inflected forms and a “word family,” a set of words that share a common meaning. The author distinguishes between function or grammatical words (e.g., prepositions, articles) and content or lexical words (e.g., nouns, adverbs). Schmitt and McCarthy (1997) use as the main unit the lexeme (or lexical item), which is “an item which functions as a single meaning unit, regardless of the number of orthographical words it contains. Fly, pain-induced, and put your nose to the grindstone are all lexical items” (1997: 329). Using this type of formulaic sequence as examples, also known as “collocations,” “chunks,” and “multi-word units” (Wray 2002), Schmitt (2010) illustrates that “Vocabulary typically behaves not as single words which are held together by syntax, but rather has a strong tendency to occur in multiple word phraseological units” (2010: 117).

Learning a word in L1 and L2: types of lexical knowledge In the broadest terms, researchers distinguish between size or breadth of lexical knowledge (i.e., the number of known words, even if that knowledge is limited to some aspects of its meaning) and depth of knowledge (i.e., quality of vocabulary knowledge). We now describe each one in detail. Regarding size, the vocabulary of an educated native speaker is considered to be around 15,000 to 20,000 word families (Nation and Waring 1997). The number of words needed to read authentic texts in a second language appears to fluctuate between 3,000 to 5,000 word families. However, most of this data comes from English, and “typical vocabulary size” appears to be a flexible concept that varies from one language to another (Meara 2003: 271). In Spanish, an educated native speaker’s lexicon is estimated to be comprised of about 30,000 words, while the average number of words shared by members of a community seems to range from a surprisingly lower 3,000 to 5,000 items (Alvar Ezquerra 2004: 21). Davies (2005) posits that “a frequency list with about 4,000 words would cover about 90% of all words that would be heard in a ‘typical’ conversation, that 7,000 words would achieve 90% coverage in fiction writing, and 8,000 words for non-fiction” (2005: 110). Learning a word appears to be a process that moves from receptive to productive knowledge. The difference between the two is defined by Nation (2013) in this way: Essentially, receptive vocabulary use involves perceiving the form of a word while listening or reading its meaning. Productive vocabulary use involves wanting to express a meaning through speaking or writing and retrieving and producing the appropriate spoken or written word form. 2013: 47

179

Marta Fairclough and Anel Garza

Therefore, receptive vocabulary size is estimated to be larger than the active lexicon, although it is not clear how much larger it is (Eyckmans 2004: 12). Regarding vocabulary breadth, as Cameron (2002) posits, “clearly, there is much more to knowing a word than just recognizing it . . . Deep word knowledge includes spelling, word associations, grammatical information and meaning” (2002: 149–150). To this list, when dealing with an L2, Mochida and Harrington (2006) add that “functional knowledge of a word can include the ability to recognize the form, translate it into the L1, recognize it in context, or accurately use it in an informative context” (2006: 95). A further aspect of vocabulary development that needs to form part of any theoretical model is how vocabulary knowledge is used productively in communication. Lafford et al. (2003), making reference to Nation (1990) and Nation and Waring (1997), indicate that “knowledge of a word requires an understanding of its spoken and written form, frequency, grammatical patterns and collocations, semantic, pragmatic, stylistic and register constraints, sociolinguistic aspects, and connotations as well as its associations with other related words” (2003: 134).

Vocabulary research with heritage Spanish speakers Several factors may have an impact on a heritage speaker’s knowledge of vocabulary. First, we will briefly discuss dialectal variation in the lexicon of US Hispanics. Then, the following sub-sections will describe some key studies regarding vocabulary among this population as it correlates with: (a) age of acquisition, (b) receptive knowledge, and (c) productive knowledge.

Dialectal lexical variation As noted in the introduction to this volume, the 54+ million Hispanics in the United States come from all Spanish-speaking countries. In 2016 the majority (64%) of US Latinos were of Mexican background, 9.4% Puerto Rican, 3.8% Salvadorans, and 3.7% Cubans, and the remaining included individuals from other Spanish-speaking nations. While all Spanish varieties share the vast bulk of lexicon, each of these dialectal groups has vocabulary that differs from region to region – typically (but not always) words that are of relatively lower frequency. Lipski (2000) underscored the difficulties that emerge when regional vocabulary from one country is used with speakers from other regions/countries, causing alienation or lack of understanding to speakers of other varieties, especially children in transitional bilingual programs, in which educational materials contained primarily Mexican (or occasionally Caribbean) vocabulary. The author explained that “Mexican’s ándale (let’s go), papalote (kite), güero (blond, fairskinned), chamaco and huerco (child), lana (money), and popote (soda straw), are all unknown to Salvadorans” (2000: 198). Lexical regional variety also confuses speakers from different backgrounds when different Spanish words are used for one same referent. For instance, the word chucho, meaning ‘light switch’ to Cubans, means ‘dog’ among Nicaraguans and Guatemalans, or chele is used as ‘blond, fair skinned individuals’ among the former, while canche is the term used by the latter (Lipski 2000). Regarding the Mexican American variety, traditional sociolinguistic research (e.g., Sánchez 1982; Valdés 2000) has revealed the presence of some archaic forms (e.g., asina, pos) which are also typical of working-class or rural speakers in many parts of the Spanish-speaking world and are often stigmatized (Bernal-Enríquez and Hernández-Chávez 2005; Parodi 2008). According to Zentella (1990), the fact that all these varieties of Spanish coexist in the United States may either lead to maintenance of each group’s lexicon; assimilation to the largest or most prestigious variety; or lexical leveling the production of a new lexicon shared by the majority. She also notes that in many cases, Anglicisms play the role of neutralizer 180

Spanish heritage language speakers

between competing dialectal variants, acting as the lingua franca that resolves the conflict without favoring one group at the expense of another (1990: 1101). A common example is cake, which in Spanish can be torta, tarta, pastel, or bizcocho, and as a result many U.S. Spanish-speakers simply use the Anglicism queic.

Age of acquisition Age of acquisition (AoA) is a construct that refers to the chronological age at which certain information (e.g., a lexical item) is acquired by an individual. Cameirão and Vicente (2010) describe it as an important psycholinguistic variable because it affects the speed and accuracy of lexical processing in tasks such as word naming, picture naming, and lexical decision (2010: 474). Studies on brain activity in monolingual subjects have shown that words learned in early childhood stimulate brain areas related to the processing of speech sounds (auditory processing), while words learned in late childhood are associated with brain activity in areas related to the effortful processing of meaning (Fiebach et al. 2003 in Hernández 2013). Even as adults, words are accessed in the same way as they were learned, with individuals “sounding out” early-learned words but making connections to words with similar meanings in the case of latelearned words (Hernández 2013). It has also been shown that words learned earlier in life are accessed more accurately and at a faster rate than those acquired later, and that this AoA effect is independent of word frequency (Bonin et al. 2004; Carrol and White 1973a, 1973b; Ellis and Morrison 1998). Ellis and Lambon Ralph’s (2000) explanation for this word AoA effect has to do with the fact that words that are learned early enter the lexical system first and form stronger links (phonological, semantic, orthographic) than those words that enter later because there is less flexibility for late-learned words to influence the system (Hernández 2013: 20). AoA affects processing of not only L1 lexicon, but also L2 lexicon. Izura and Ellis (2002) found AoA effects in both the L1 and the L2 in object naming and lexical decision tasks of Spanish L1 speakers who started learning English as L2 in Spain after age 10. For example, they were able to recognize and produce early-acquired English words faster than late-acquired English words. Their follow-up study with visual translation and lexical decision tasks (2004) confirmed the previous findings. They concluded that early-acquired words in Spanish L1 and English L2 were processed faster and more accurately than words acquired later in Spanish L1 and English L2, respectively. In the case of heritage learners, a common occurrence is L1 attrition (i.e. heritage language erosion), and lexical access is one area that has been shown to be vulnerable to attrition (de Bot 1998; Weltens and Grendell 1993). A study of three generations of Dutch immigrants in New Zealand performing picture-naming and picture-word matching tasks in both Dutch and English (Hulsen 2000) showed that production is affected more than comprehension in the heritage language, and that second- and third-generation speakers are more affected than first-generation speakers. In addition, children and immigrants arriving in a new country before puberty lose their L1 vocabulary at a faster rate than adult immigrants. This coincides with Polinsky’s (2005) findings for Russian heritage speakers. In lexical recognition and translation tasks of nouns, verbs, and adjectives, she also found that Russian native speakers showed balanced control of the three word classes, while heritage speakers of low proficiency showed lower retention of nouns and adjectives than of verbs. Montrul and Foote (2014) studied the role of global AoA of Spanish and the AoA of individual words on the speed and accuracy of lexical access in SHL learners and late L2 learners of Spanish with similar proficiency levels and whose dominant language is English. Both groups performed a lexical decision task in Spanish and an English-Spanish translation decision task. Both groups tended to be faster and more 181

Marta Fairclough and Anel Garza

accurate on nouns than on verbs and adjectives, although the difference was not statistically significant. In terms of global AoA of L1 and L2, the results showed no differences in accuracy, but the L2 learners were faster in their responses than the heritage learners. In terms of word AoA, there were differences within the groups: heritage learners were faster and more accurate than L2 learners with words learned early in L1 Spanish, while the L2 learners were faster and more accurate than heritage learners with words learned early in L2 Spanish. The results suggest that visual word recognition is not affected by ‘maturational effects’ (i.e., no effects of age for acquisition of lexical items). The effect of AoA of words shows that lexical access depends mostly on the language-learning experience and the context and modality of acquisition. While SHL speakers are exposed to Spanish at home since early childhood and learn the HL primarily via aural input, L2 learners acquire Spanish usually around puberty and mainly in the classroom, where they learn words through both aural and written input. That is, earlier learned lexicon was responded to more quickly and more accurately by both heritage learners and second language learners.

Receptive lexical knowledge Several studies about SHL learners’ receptive vocabulary knowledge were conducted using a lexical decision (or “lexical recognition”) task, a two-choice procedure in which a string of letters stimulus is presented and participants have to decide if it is a real word or not (Fairclough 2011). Fairclough and Ramírez’ (2009) pilot study found a high correlation between general language proficiency and a 200-word test that included high frequency Spanish words (Davies 2006) and pseudowords to control for guessing. If vocabulary levels do reflect language development more generally, and if existing tests can be shown to be valid and reliable, then vocabulary testing might offer a relatively quick and easy way for researchers and schools to monitor progress in language development. Cameron 2002: 151 Fairclough & Ramirez’ study was replicated by Fairclough (2011) with a larger group of participants (N = 330). The purpose of the study was to assess the effectiveness of the lexical recognition task as a placement tool that distinguishes among levels of Spanish L2 and HL learners. University students from four different levels (in addition to a control group of bilingual graduate students) completed the test, which correlated well with other measures of language proficiency. However, while the lexical test based on the 5,000 most frequent words in Spanish worked well with Spanish L2 students and lower level HL learners, a lower frequency lexicon would be needed to avoid the ceiling effect with higher proficiency students. By measuring and comparing the number of words SHL learners know, parameters could be determined to differentiate among different levels of proficiency in the heritage language. Results of a later study (Fairclough 2013a) showed that the receptive SHL students enrolled in first year Spanish courses (i.e., “Receptive” refers to students lacking in productive abilities, but able to comprehend basic oral and written material) who participated in the research had a lexicon of up to 3,000 words, two-thirds of them from the highest frequency bands, while higher proficiency SHL learners recognized at least 90% of the 5,000 most frequent words in Spanish. Velásquez (2015) looked at the correlation between lexical knowledge and reading comprehension in a study with 240 beginner, intermediate, and advanced SHL university students. Participants completed two lexical recognition tests and two reading comprehension tests. The results showed that 98% vocabulary coverage was needed to show “adequate” (70%) 182

Spanish heritage language speakers

comprehension of the text. Velásquez concluded that the relationship between lexical competence and reading comprehension is not linear (i.e., the relationship between vocabulary knowledge and reading comprehension does not increase proportionally) as suggested by Schmitt et al. (2011), with other factors such as grammatical knowledge, ability to infer meaning, text difficulty, and length contributing to the reading comprehension process.

Productive lexical knowledge “Available lexicon in defined as (1) the sum of words that speakers have in their mental systems and (2) whose use is conditioned by a particular topic” (Moreno Fernández 2007: 41). Given a semantic field (e.g., transportation) individuals will produce, in a pre-determined amount of time, all the words that come to mind related to that field. The lexical availability of native speakers of Spanish has been widely researched, both in monolingual contexts (Ávila Muñoz and Villena Ponsoda 2010; Hernández Muñoz 2006, 2015; Hernández Muñoz et al. 2006, among many others) and in language contact situations (see Samper Padilla et al. 2003 for an overview, as well as Nielson Parada, this volume, about Spanish lexicon among Chileans in Sweden). However, relatively less research on lexical availability has been conducted among Spanish-English bilinguals in the United States. We will review three of these studies here. Cooper (1971) used a word naming task with Puerto Ricans in New Jersey to examine specific social domains: family, neighborhood, religion, education, and work. The task consisted of producing within one minute as many words as they could that were associated to the specific context. Results showed that participants produced more words in the language they normally used in each domain. That is, if they used mostly Spanish at church, they produced more words in Spanish in the “religion” domain, and if they used mostly English at work, they produced more words in English for that domain. Moreno-Fernández (2007) looked at the presence of Spanish and English in the available lexicon of first- and second-generation Hispanic adolescents in Chicago. The author elicited vocabulary associated with 22 different semantic fields including parts of the body, clothing, parts of the house, school, and transportation. Word frequency and “availability index” were calculated. Results showed that overall only 15% of produced words were Anglicisms, and among the 20 most commonly listed words in each semantic field, only 6.5% were Anglicisms. This suggests that the Spanish lexicon of young American Hispanics is less influenced by English than may have been suspected. The author also found an inverse correlation between Spanish proficiency and English production: the higher the level of Spanish in which students were enrolled, the lower the number of English lexical items they produced. Similarly, Garza (2013) found that as the level of Spanish proficiency increases, English transfer phenomena decrease. Following traditional methodology designed for lexical availability studies (López Meirama 2008), Fairclough (2010) conducted a study with 150 heritage language learners enrolled in first-, second-, third-, and fourth-year Spanish courses. Participants completed a timed vocabulary task in which they were asked to generate words based on 12 different semantic fields (e.g., parts of the body, nature, animals, education). The frequency and availability indexes were calculated and compared to those from studies in monolingual and other bilingual contexts. The results showed that the longer SHL students studied Spanish, the closer the results matched those of participants from the monolingual studies, which suggests a positive role of heritage language study in lexical development. Other studies of productive lexical knowledge include academic vocabulary, lexical richness, and code-mixing. 183

Marta Fairclough and Anel Garza

Garza’s (2013) vocabulary types in the analysis of written and oral academic texts include formal, technical, colloquial, and stigmatized forms. The vocabulary in the written samples shows a gradual decrease in colloquial items and stigmatized forms as the educational level increases (Group 1 to Group 4), and a gradual increase in words classified as formal. Technical vocabulary is also lower in educational group 1 compared to Groups 2, 3, and 4, although in this category, Group 3 displays the highest percentage of all groups. The oral samples also evidence a gradual decrease in colloquial items and stigmatized forms. Formal vocabulary is lowest in Group 1 and highest in 4, although the pattern of increase is not continuous (Group 2 has a higher percentage than 3). Technical vocabulary also increases, but Group 3 has the highest percentages in all four groups. Overall, these results show a positive influence of increased exposition to Spanish in the heritage courses, as there is a gradual increase in formal and technical vocabulary, and a gradual decrease in colloquial, transfer, and stigmatized/archaic forms in both the written and oral academic registers of these students. The results of these studies are not surprising since it is a well-known fact that heritage language speakers acquire the minority language in a natural context (usually orally and within the home environment), with limited exposure to formal registers, which are learned in academic settings. The feature of lexical richness in the investigation by Fairclough and Belpoliti (2015) encompasses lexical density, lexical frequency, and lexical variation. The lexical density index in this study is calculated as the number of Spanish lexical words divided by the total running words. Proper names were subtracted from the total number of words. The 46.4% lexical index situates the written texts of these students approximately midway along Halliday’s (1989) written-oral continuum and reveals a large proportion of use of non-lexical items, which is indicative of an oral register. However, as Laufer and Nation (1995) point out, this measure by itself is not enough to measure vocabulary, and other measures are needed to clearly represent knowledge and use of words. Therefore, lexical frequency and variation are also included in the analysis. Lexical frequency is measured by classifying all Spanish words into six levels based on Davies’ (2006) frequency dictionary and the CREA database. In terms of word frequency, the writing of these students shows a reduced vocabulary; with level 1K (the 1,000 most frequent words in Spanish) accounting for 92.9% of the vocabulary used. 2K level (from 1,000 to 2,000) words occur 3% of the time; 5K+ (over 5,000) words, 2.7% of the time, and 3K, 4K, and 5K words have frequencies of 0.8%, 0.2%, and 0.4%, respectively. Lexical variation, measured using both the Type/Token Ratio (TTR) and Guiraud’s index (Root TTR), reveals a high level of repetition of words. Overall, the analysis of lexical richness demonstrates that these learners still lack the lexical resources to develop their ideas. Garza’s calculation of lexical density was also measured as the ratio of the number of content words to the total number of words. The results of both written and oral texts show a gradual increase in lexical density from Group 1 to Group 4 (G1 to G4), and the percentages in the written samples are higher than those in the oral samples, in accordance with Halliday’s (1989) observed differences between spoken and written language. These findings demonstrate advancement toward a more academic register as heritage students take more Spanish courses. The researcher’s study of word frequency was also based on Davies’ frequency dictionary. The written samples revealed a gradual decrease (G1 to G4) in the 1K category (the 1–1,000 most common words) and a gradual increase in the +5K category (less frequent words, i.e., the 5,000 + category). Words in the other frequency bands did not show a definite pattern, but what became evident was that G1 used less variety of words within each frequency band. In addition, the variety of vocabulary employed increased as the level of the students went up, with G4 showing greater vocabulary variety than the other groups. Unlike the written sample, which was obtained without any prior preparation on the students’ part, the oral presentations allowed 184

Spanish heritage language speakers

for students to research their topic and prepare in advance. This was probably a contributing factor to the increased variety of words used by all groups in comparison to the written samples, as well as to the overall similarities in terms of percentages, within the frequency bands. Fairclough (2006) described the different types of language mixing phenomena (singleword switches, borrowings, and calques as well as multiple-word switches and calques) in a context where participants would be expected to be in a “monolingual mode” (Grosjean 1997). The study was based on 150 samples written as part of exams completed by heritage learners of Spanish who placed into 5 different levels upon entering the Spanish program at the University of Houston. Results showed low percentages of English transfer in the analyzed samples (2.32%, or in terms of means, 7.15 words per average sample of 308 words). As expected, the influence of English appeared to decrease as competence in Spanish increased. Single- and multiple-word switches to English as well as loans and single-word calques were very sporadic. Multiple-word calques were the most frequent across the levels, yet the means were still very low when compared to the total number of words (total number of words produced: 46,244; total number of CS: 1,073; total number of multiple-word calques: 751). A follow-up study (Fairclough and Belpoliti 2015) also centered on the writing of beginninglevel heritage learners (receptive bilinguals) in essays produced as part of a placement/credit exam. They found a relatively low percentage of lexical transfer (4.95%), with lexical creations emerging as the most frequent (2.50%). The low transfer from English in these receptive bilingual learners indicates that they seem to be in monolingual Spanish mode during the exam. Sánchez-Muñoz (2010) studied the oral production of heritage learners across registers, focusing on vocabulary features produced in the following three contexts ranging from most formal to least formal register: (1) class presentations, (2) interviews, and (3) conversations. Two of the three linguistic elements considered by Sánchez-Muñoz were lexical transfer of English words into Spanish (i.e., deadline), including fillers (i.e., you know), and lexical creations in Spanish based on English words (i.e., endurar “endure”). Confirming what one would predict – that there would be greater English use in more informal contexts – the highest percentage of switches to English occurred in conversations, followed by the interview, and finally in the presentations. The variation of these features across registers suggests that the Spanish spoken by these participants is not a monostylistic variety, but rather that they are able to differentiate among discursive situations (i.e., academic vs non-academic) and adapt their language accordingly, using the resources available to them. The purpose of Fairclough (2013b) was to verify that 12 different high frequency calques were in the process of lexicalization in U.S. Spanish even though more widely used equivalent forms were previously taught in the classroom. The study compared 101 English-Spanish translations of three brief paragraphs completed by university students of Hispanic background from different proficiency levels. The results suggest that a number of calques are undergoing a process of lexicalization in that they were used by a considerable number of the participants in the sample. Some examples of the most frequent included “solicitud / aplicación” (application), “inscribirse / registrarse” (register), “universidad / colegio” (college), “tomar una decisión / hacer una decisión” (make a decision).

Recommendations for practice and future directions A number of studies, mostly in English as an L2, have corroborated correlations between lexical knowledge with other linguistic measures including global language proficiency (e.g., Albrechtsen et al. 2008; Alderson 2005; Laufer 1992; Laufer and Goldstein 2004; Meara 1996; Mecartty 2000; Read 1997). With heritage learners, Polinsky (2006) and Polinsky and Kagan 185

Marta Fairclough and Anel Garza

(2007) observed a strong correlation between lexical and grammatical knowledge in heritage speakers of Russian in the United States. According to the authors, similar correlations were evidenced in studies of other heritage languages. The research presented in this chapter has shown correlations between vocabulary knowledge and general language proficiency, and more specifically with reading comprehension, writing, and speaking abilities. Given such correlations, it is imperative to underscore the importance of the lexicon in language learning and make it a priority in instruction. Fairclough and Mrak’s (2003) error analysis of the speech of SHL speakers, for example, shows a very low percentage of morphosyntactic errors (about 2% of the total number of words produced), leading the authors to recommend a focus on vocabulary teaching in HL courses to expand learners’ lexical range and reduce their linguistic insecurity. Although Fairclough (2011, 2013b) suggests that HL learners are usually at an advantage over L2 learners in receptive knowledge of basic Spanish vocabulary (defined as the 5,000 most frequent words), Swender et al. (2014) underscore HL learners’ limited vocabulary knowledge, especially when dealing with abstract topics and those beyond everyday experiences. Furthermore, Zyzik (2016) recommends that: [i]n addition to simply learning more words (i.e., breadth of knowledge), vocabulary instruction tailored to HL learners should focus on improving depth of vocabulary knowledge. The ability to use a word productively involves many layers of knowledge beyond the primary form-meaning link. Schmitt (2010: 26–28) lists a range of factors that facilitate vocabulary acquisition. According to the author, some of the following aspects should be focused on: frequency of exposure, attention focused on lexical items, intention to learn lexical item, manipulation of lexical item and its properties, a requirement to learn lexical item (by teacher, test, syllabus), and a need to learn and use the lexical item (for task or for a personal goal). The author states that practically “anything that leads to more exposure, attention, manipulation, or time spent on lexical items adds to their learning . . . even the process of being tested on lexical items appears to facilitate better retention” (Schmitt 2010: 28). A detailed explanation of all the factors that contribute to the acquisition of the lexicon is out of the scope of this chapter. For further information, please see Schmitt (2008, 2010), as well as other publications on vocabulary learning. Finally, given the limited amount of vocabulary research, more studies from all the different perspectives and dimensions mentioned in this chapter are needed. Zyzik (2016) indicates that since vocabulary knowledge appears to be explicit for all language learners (including HLLs), L2 vocabulary research, methods, and materials are largely transferrable. Future studies on the lexicon of SHL speakers should be guided by L2 research and it could include studies about multiword lexical items, vocabulary richness, comparison of the lexicon in different types of registers, depth of knowledge (e.g., word associations), among many other areas. A better understanding of SHL learners’ lexical knowledge and acquisition processes can help meet students’ learning needs and serve as a guide to professionals in the field.

Further reading Nation, I. S. P. (2013). Learning Vocabulary in another Language. (2nd ed.). Cambridge, UK: Cambridge University Press. A manual for understanding vocabulary teaching and learning. Reed, J. (2000). Assessing Vocabulary. Cambridge, UK: Cambridge University Press. Comprehensive overview of vocabulary testing. Schmitt, N. (2010). Researching Vocabulary: A Vocabulary Research Manual. London: Palgrave Macmillan. A current guide to lexical research.

186

Spanish heritage language speakers

References Albrechtsen, D., Haastrup, K., and Henriksen, B. (2008). Vocabulary and Writing in a First and Second Language: Process and Development. Basingstoke, UK: Palgrave Macmillan. Alderson, J. C. (2005). Diagnosing Foreign Language Proficiency. London: Continuum. Alvar Ezquerra, M. (2004). La frecuencia léxica y su utilidad en la enseñanza del español como lengua extranjera. Actas del XV Congreso Internacional de ÁSELE, 19–39. Retrieved from http://cvc.cervantes. es/ensenanza/biblioteca_ele/asele/pdf/15/15_0017.pdf/2015/03/02. Ávila Muñoz, A. M., and Villena Ponsoda, J. A. (eds). (2010). Variación social del léxico disponible en la ciudad de Málaga. Diccionario y análisis. Málaga, Spain: Sarriá. Bernal-Enríquez, Y., and Hernández-Chávez, E. (2005). La enseñanza del español en Nuevo México. ¿Revitalización o erradicación de la variedad chicana? In A. Roca and M. C. Colombi (eds.), Mi lengua: Spanish as a Heritage Language in the United States (pp. 96–119). Washington, DC: Georgetown University Press. Bonin, P., Barry, C., and Chalard, M. (2004). The influence of age of acquisition in word reading and other tasks: A never ending story. Journal of Memory and Language 50: 456–476. de Bot, K. (1998). The psycholinguistics of language loss. In G. Extra and L. Verhoeven (eds.), Bilingualism and Migration (pp. 345–361). Berlin: Mouton de Gruyter. Cameirão, M. L., and Vicente, S. G. (2010). Age-of-acquisition norms for a set of 1,749 Portuguese words. Behavior Research Methods 42(2): 474–480. Cameron, L. (2002). Measuring vocabulary size in English as an additional language. Language Teaching Research 6(2): 145–173. Carrol, J., and White, M. (1973a). Age of acquisition norms for 220 picturable nouns. Journal of Verbal Learning and Verbal Behavior 12: 563–76. Carrol, J., and White, M. (1973b). Word frequency and age-of-acquisition as determiners of picturenaming latency. Quarterly Journal of Experimental Psychology 25: 85–95. Cooper, R. L. (1971). Degree of bilingualism. In J. A. Fishman, R. L. Cooper, and R, Ma (eds.), Bilingualism in the Barrio (pp. 273–309). Bloomington, IN: Research Center for the Language Sciences, Indiana University. Davies, M. (2005). Vocabulary range and text coverage: Insights from the forthcoming Routledge Frequency dictionary of Spanish. In D. Eddington (ed.), Selected Proceedings of the 7th Hispanic Linguistics Symposium (pp. 106–115). Somerville, MA: Cascadilla Proceedings. Davies, M. (2006). A Frequency Dictionary of Spanish: Core Vocabulary for Learners. London: Routledge. Ellis, A., and Lambon Ralph, M. (2000). Age of acquisition effects in adult lexical processing reflect loss of plasticity in maturing systems: Insights from connectionist networks. Journal of Experimental Psychology: Learning, Memory and Cognition 26: 1103–1123. Ellis, A., and Morrison, C. (1998). Real age of acquisition effects in lexical retrieval. Journal of Experimental Psychology: Learning, Memory and Cognition 24: 515–523. Eyckmans, J. (2004). Measuring receptive vocabulary size: Reliability and validity of the yes / no vocabulary test for French-speaking learners of Dutch. Unpublished doctoral dissertation. Utrecht, the Netherlands: LOT, Landelijke Onderzoekschool Taalkunde. Fairclough, M. (2006). Spanish / English interaction in U.S. Hispanic heritage learners’ writing. In C. Mar-Molinero and M. Steward (eds.), Globalization and Language in the Spanish-speaking World: Macro and Micro Perspectives (pp. 76–93). London: Palgrave Macmillan. Fairclough, M. (2010). Lexical availability in bilingual contexts: A preliminary study of Spanish lexicon in Houston. Paper presented at the 5th International Conference of Hispanic Linguistics, University of Limerick, Ireland. Fairclough, M. (2011). Testing the lexical recognition task with Spanish/English bilinguals in the United States. Language Testing 28(2): 273–297. Fairclough, M. (2013a). El ‘(re)conocimiento’ del léxico español del estudiante hispano bilingüe en los Estados Unidos. In D. Dumitrescu (ed.), El español en Estados Unidos: e pluribus unum? Enfoques multidisciplinarios (pp. 167–184). Academia Norteamericana de la Lengua Española. Madrid: Santillana. Fairclough, M. (2013b). Muestras de lexicalización en estudiantes hispanos bilingües en los Estados Unidos: un estudio exploratorio. Revista Internacional d’Humanitats 16: 105–114. Fairclough, M., and Belpoliti, F. (2015). Emerging literacy in Spanish among Hispanic heritage language university students in the United States: A pilot study. International Journal of Bilingual Education and Bilingualism 19(2): 185–201.

187

Marta Fairclough and Anel Garza Fairclough, M., and Mrak, A. (2003). La enseñanza del español a los hispanohablantes bilingües y su efecto en la producción oral. In A. Roca and M.C. Colombi. (eds.), Mi Lengua: Spanish as a Heritage in the United States, Research and Practice (pp. 198–212). Washington, DC: Georgetown University Press. Fairclough, M., and Ramírez, C. (2009). La prueba de decisión léxica como herramienta para ubicar al estudiante bilingüe en los programas universitarios de español. Íkala, Revista de Lenguaje y Cultura 14(99): 84–99. Fiebach, C. J., Friederici, A. D., Muller, K., von Cramon, D. Y., and Hernandez, A. R. (2003). Distinct brain representations for early and late learned words. NeuroImage 19: 1617–1637. Garza, A. (2013). El registro académico en estudiantes hispanohablantes de herencia. Unpublished doctoral dissertation. University of Houston, Houston, TX. Grosjean, F. (1997). Processing mixed language: Issues, findings, and models. In A. M. de Groot and J. F. Kroll (eds.), Tutorials in Bilingualism: Psycholinguistic Perspectives (pp. 225–254). Mahwah, NJ: Erlbaum. Halliday, M. A. K. (1989). Spoken and Written Language, 2nd ed. Oxford, UK: Oxford University Press. Hernández, A. (2013). The Bilingual Brain. New York: Oxford University Press. Hernández Muñoz, N. (2006). Hacia una teoría cognitiva integrada de la disponibilidad léxica. El léxico disponible de los estudiantes castellano-manchegos. Salamanca, Spain: ediciones de la Universidad de Salamanca. Hernández Muñoz, N. (2015). La evaluación de la competencia léxica adulta: una aproximación a través de la disponibilidad léxica y la especialización académica en preuniversitarios. Revista de Filología 33: 79–99. Hernández Muñoz, N., Izura C., and Ellis, A. (2006). Cognitive aspects of lexical availability. European Journal of Cognitive Psychology 18(5): 734–755. Hulsen, M. (2000). Language loss and language processing. Three generations of Dutch migrants in New Zealand. Unpublished doctoral dissertation) University of Nijmegen, the Netherlands. Izura, C., and Ellis, A. (2002). Age of acquisition effects in word recognition and production in first and second languages. Psicológica 23: 245–81. Lafford, B. A., Collentine, J. G., and Karp, A. S. (2003). The acquisition of lexical meaning by second language learners: An analysis of general research trends with evidence from Spanish. In B. A. Lafford and R. Salaberry (eds.), Spanish Second Language Acquisition: State of the Science (pp. 130–159). Washington, DC: Georgetown UP. Laufer, B. (1992). How much lexis is necessary for reading comprehension? In P. J. L. Arnaud and H. Béjoint (eds.), Vocabulary and Applied Linguistics (pp. 126–132). London: Macmillan. Laufer, B. (1998). The development of passive and active vocabulary in a second language: Same or different? Applied Linguistics 19(2): 255–271. Laufer, B., and Goldstein, Z. (2004). Testing vocabulary knowledge: Size, strength, and computer adaptiveness. Language Learning 54(3): 399–436. Laufer, B., and Nation, P. (1995). Vocabulary size and use: Lexical richness in L2 written production. Applied Linguistics 16(3): 307–322. Laufer, B., and Nation, P. (1999). A vocabulary-size test of controlled productive vocabulary. Language Testing 16(1): 307–322. Lipski, J. (2000). The linguistic situation of Central Americans. In S. L. McKay and S. C. Wong (eds.), New Immigrants in the United States (pp. 189–215). Cambridge UK: Cambridge University Press. López Meirama, B. (2008). Léxico disponible en el español de Galicia. Santiago de Compostela, Spain: Universidad de Santiago de Compostela. Meara, P. M. (1996). The dimensions of lexical competence. In G. Brown, K. Malmkjaer and J. Williams (eds.), Performance and Competence in Second Language Acquisition (pp. 35–53). Cambridge, UK: Cambridge University Press. Meara, P. M. (2003). Designing vocabulary tests for English, Spanish and other languages. In C. S. Butler et al. (eds.), Dynamics of Language Use: Functional and Contrastive Perspectives, (pp. 271–285). Philadelphia, PA: John Benjamins Publishing Company. Mecartty, F. (2000). Lexical and grammatical knowledge in reading and listening comprehension by foreign language learners of Spanish. Applied Language Learning 11(2): 323–348. Mochida, A., and Harrington, M. (2006). The yes/no test as a measure of receptive vocabulary knowledge. Language Testing 23(1): 73–98. Montrul, S., and Foote, R. (2014). Age of acquisition interactions in bilingual lexical access: A study of the weaker language in L2 learners and heritage speakers. The International Journal of Bilingualism 18(3): 274–303.

188

Spanish heritage language speakers Moreno-Fernández, F. (2007). Anglicismos en el léxico disponible de los adolescentes hispanos de Chicago. In K. Potowski and R. Cameron (eds.), Spanish in Contact: Policy, Social and Linguistics Inquiries (pp. 41–58). Amsterdam: John Benjamins. Nation, I. S. P. (1990). Teaching and Learning Vocabulary. Rowley, MA: Newbury House. Nation, I. S. P. (2013). Learning Vocabulary in another Language. (2nd ed.). Cambridge, UK: Cambridge University Press. Nation, I. S. P., and Waring, R. (1997). Vocabulary size, text coverage and word lists. In N. Schmitt and M. McCarthy (eds.), Vocabulary: Description, Acquisition and Pedagogy (pp. 6–19). Cambridge: Cambridge University Press. Parodi, C. (2008). Stigmatized Spanish inside the classroom and out: A model of language teaching to heritage speakers. In D. M. Brinton, O. Kagan, and S. Bauckus (eds.), Heritage Language Education: A New Field Emerging (pp. 199–214). New York: Routledge. Polinsky, M. (2005). Word class distinction in an incomplete grammar. In D. Ravid and H. Bat-Zeev Shyldkrot (eds.), Perspectives on Language and Language Development (pp. 419–436). Dordrecht, the Netherlands: Kluwer. Polinsky, M. (2006). Incomplete acquisition: American Russian. Journal of Slavic Linguistics 14: 191–262. Polinsky, M., and Kagan, O. (2007). Heritage languages: In the ‘wild’ and in the classroom. Language and Linguistics Compass 1(5): 368–395. Read, J. (1997). Vocabulary and testing. In N. Schmitt and M. McCarthy (eds.), Vocabulary: Description, Acquisition and Pedagogy (pp. 303–320). Cambridge, UK: Cambridge University Press. Read, J. (2000). Assessing Vocabulary. Cambridge, UK: Cambridge University Press. Read, J., and Chapelle, C. A. (2001). A framework for second language vocabulary assessment. Language Testing 18(1): 1–32. Samper Padilla, J. A., Bellón Fernández, J. J., and Samper Hernández, M. (2003). El proyecto de estudio de la disponibilidad léxica en español. In R. Ávila, J.A. Samper, U. Hiroto, et al. (eds), Pautas y pistas en el análisis del léxico hispano(americano) (pp. 27–140). Madrid: Iberoamericana. Sánchez, R. (1982). Our linguistic and social context. In Amastae, J. and Elías-Olivares, L. (Eds.), Spanish in the United States: Sociolinguistic Aspects, (pp. 9–45). New York: Cambridge University Press. Sánchez-Muñoz, A. (2010). Different words for different contexts: Intra-speaker variation in Spanish as a heritage language. In S. V. Rivera-Mills and D. J. Villa (Eds.), Spanish of the U.S. Southwest: A Language in Transition (pp. 337–352). Frankfurt, Germany and Madrid: Vervuert/ Iberoamericana. Schmitt, N. (2008). Instructed second language vocabulary testing. Language Teaching Research 12(3): 329–363. Schmitt, N. (2010). Researching Vocabulary: A Vocabulary Research Manual. London: Palgrave Macmillan. Schmitt, N., Jiang, X., and Grabe, W. (2011). The percentage of words known in text and reading comprehension. Modern Language Journal 95(1): 26–43. Schmitt, N., and McCarthy, M. (1997). Glossary. In N. Schmitt and M. McCarthy (eds.), Vocabulary: Description, Acquisition and Pedagogy (pp. 327–31). Cambridge, UK: Cambridge University Press. Singleton, D. (1999). Exploring the Second Language Mental Lexicon. Cambridge, UK: Cambridge University Press. Skehan, P. (1998). A Cognitive Approach to Language Learning. Oxford, UK: Oxford University Press. Swender, E., Martin, C., Rivera-Martínez, M., and Kagan, O. (2014). Exploring oral proficiency profiles of heritage speakers of Russian and Spanish. Foreign Language Annals 47(3): 423–446. Valdés, G. (2000). Bilingualism and language use among Mexican Americans. In S. L. McKay and S. C. Wong (eds.), New Immigrants in the United States (pp. 99–136). Cambridge UK: Cambridge University Press. Velásquez, E. V. (2015). Competencia léxica y comprensión de lectura: un estudio con hablantes de español como lengua de herencia. Unpublished doctoral dissertation. University of Houston, Houston, TX. Weltens, B., and Grendell, M. (1993). Attrition of vocabulary knowledge. In R. Schreuder and B. Weltens (eds.), The Bilingual Lexicon (pp. 135–56). Amsterdam: John Benjamins. Wray, A. (2002). Formulaic Language and the Lexicon. Cambridge, UK: Cambridge University Press. Zentella, A. C. (1990). Lexical leveling in four New York City Spanish dialects: Linguistic and social factors. Hispania 73(4): 1094–1105. Zyzik, E. (2016). Towards a prototype model of the heritage language learner: Understanding strengths and needs. In M. Fairclough and S. M. Beaudrie (eds.), Innovative Approaches in Heritage Language Teaching: From Research to Practice. Washington, DC: Georgetown University Press.

189

13 HERITAGE SPANISH PRAGMATICS Derrin Pinto university of saint thomas, usa

Introduction In the realm of pragmatics and discourse-related phenomena (henceforth, pragmatics only), research on Spanish as a heritage language is still in the developing stage, and, as a whole, remains somewhat unfocused and inconsistent. While certain topics such as speech acts and discourse markers have surfaced in multiple studies, other areas continue to be under investigated. For the most part, the examination of the speech behavior of bilingual and multilingual speech communities from a pragmatic perspective has not developed as rapidly as the interdisciplinary fields of Interlanguage Pragmatics, involving the study of second-language learners, or CrossCultural Pragmatics, the comparison of two or more speech communities. Although neither field has necessarily excluded bilingual groups as a subject of study, their inclusion has not been the norm. Nevertheless, if one considers bilingual and multilingual contexts involving different languages, there has still been a steady trickle of studies over the past few decades dealing with topics such as speech acts, discourse markers, and politeness (Barnes, 2001; Blum-Kulka, 1990; Blum-Kulka & Sheffer, 1993; Clyne, 1979; Clyne, Ball, & Neil, 1991; Hlavac, 2006; Farghal & Haggan, 2006; Marti, 2006). More recently, with the increase of attention in linguistics dedicated to bilingualism and multilingualism, it is likely that pragmatic research focusing on languages in contact will continue to develop. This chapter will attempt to synthesize a relatively disparate body of literature in a systematic fashion by applying a distinction that has been widely utilized in pragmatics. Pragmatic features are often categorized according to two different perspectives designated as pragmalinguistic and sociopragmatic (Leech, 1983). Essentially these can be considered two sides of a continuum, thus the distinction is not necessarily categorical and there are times when the two concepts intersect or overlap. Nevertheless, the pragmalinguistic end of the spectrum refers to matters of a more structural or linguistic nature, while the sociopragmatic side includes phenomena related to social norms or appropriateness. Dating back to Thomas (1983), this distinction has been used for discerning different types of errors or pragmatic failure in second-language learning. For Thomas, while pragmalinguistic failure for language learners reflects a linguistic problem, sociopragmatic failure “stems from cross-culturally different perceptions of what constitutes appropriate linguistic behaviour” (p. 99). In spite of the convenient categorical distinction,

190

Heritage Spanish pragmatics

the overlapping nature of pragmalinguistic and sociopragmatic perspectives is unavoidable. For example, if a speaker uses tú commands when usted would be more acceptable, this could be due to unfamiliarity with the appropriate social norms, which is more sociopragmatic, but it could also be attributed to confusion between formal and familiar commands, which is more of a structural concern involving pragmalinguistics. Nevertheless, for the present discussion on SHL, these two categories provide a general framework for examining a wide range of features. From the outset, it is worth pointing out that one of the underlying assumptions of crosscultural pragmatics may be at odds with the stance that some researchers adopt regarding SHL. As the name indicates, there is a basic understanding that different speech communities are comparable and that such comparisons can potentially yield interesting results, especially since “speech communities tend to develop culturally distinct interactional styles” (Blum-Kulka, House, & Kasper, 1989, p. 7). Ideally when making comparisons, one would hope that the two groups being studied represent similar backgrounds, but in reality, this criterion may be difficult to meet, particularly with heritage speakers who represent a unique set of experiences. For example, it has been pointed out that heritage speakers of Spanish use Spanish to carry out particular communicative needs in specific contexts (largely those that are informal and non-academic), have little to no contact with normative monolingual varieties of Spanish, and possess limited reading and writing skills but strong speaking and listening skills (Carreira & Kagan, 2011; Pascual y Cabo & Rothman, 2012; Valdés, 2005). Given their distinct profile, one might argue that from the outset heritage speakers will never be fully comparable to monolingual Spanish speakers. From this perspective, comparing SHL to other monolingual varieties of Spanish should be considered an analytical exercise that does not necessarily imply any of the following: 1) that heritage speakers of Spanish strive, or should strive, to speak like monolingual speakers; 2) that the monolingual variety is superior; 3) that any differences reported for SHL represent deficiencies that need to be corrected; and 4) that formal registers are more valid than informal registers, although one or the other may be more appropriate in a particular context. Nevertheless, in the real world, it is still likely that speakers of SHL, especially in some professional or academic contexts, will be expected to perform in accordance with the monolingual speech behavior or norms that are associated with a formal register. As Achugar (2003) emphasizes, creating an awareness among teachers and learners of how language and power are related, and of how different registers are considered more or less valuable, is essential for successful language socialization and effective pedagogical practices.

Pragmalinguistic issues The topic of discourse markers (DM) could serve as a prototypical case of what would be considered more pragmalinguistic in nature, further toward the linguistic side of the spectrum, although an analysis of their usage can also take into account sociopragmatic factors. DMs are single or multi-word units that fulfill one or more language functions in a given context, at least one of which is more procedural in nature, meaning that it conveys information about how an utterance is to be interpreted (Fraser, 1999; Fraser & Malamud-Makowski, 1996). Among the discourse markers investigated in SHL are tú sabes (‘you know’), entonces (‘so’), bueno (‘well’) and pues (‘well’). Starting with children’s speech, one of the functions highlighted in the play discourse of Spanish-English bilinguals is that of marking register. In a study of children between the ages of 6 and 10 residing in southern California, Brizuela et al. (1999) found that, especially in stacks

191

Derrin Pinto

or clusters (e.g. Bueno, pues . . . ), these items were used by the children to signal the discourse associated with people of higher social status, mainly parents and teachers. Similar to the idea of associating certain DMs with a particular register, their level of mastery could be, according to the authors, at least partially determined by social class. Andersen et al. (1999), in their data collected from bilingual children in southern California, report that while the monolingual English subjects used up to three DMs together in English, the bilingual group never used more than two in Spanish. As a possible explanation for the sparser use of DMs by these Spanish-speaking children, the authors speculate that two potential factors might be their bilingualism and the fact that they pertain to a lower social class than the monolingual group. However, no significant evidence is provided to substantiate these claims other than a reference to some preliminary data suggesting that “middle-class, monolingual Argentinian children” exhibit greater use of DMs than their Chicano participants (p. 1349). These observations about the stacking of DMs signal a promising avenue of future research, within the context of speech produced by bilingual children or adults. Given the prominence of English DMs in the Spanish of Spanish-English bilinguals, some of the research in this area overlaps with studies on code-switching or mixing. The use of English DMs such as ‘so,’ ‘I mean,’ and ‘y’know’ was analyzed in the bilingual speech of Puerto Rican adults in New York and these lexical items were found to enter originally as instances of code-switching before becoming more established over time as borrowings (Torres, 2002). When comparing ‘so’ to entonces in New Mexican Spanish, Aaron (2004) found them to be functionally similar, although ‘so’ triggered more code-switches and was used more frequently in the data corpus. Torres and Potowski (2008, forthcoming) also observed functional similarities between ‘so’ and entonces in Mexican, Puerto Rican, and MexiRican Spanish in Chicago. In the more recent study, when looking at the discourse marker production of 31 Mexican and 31 Puerto Rican individuals, they found that Puerto Ricans used significantly more ‘so’ than Mexicans, but there was no significant difference in the two groups’ use of entonces. They also reported that use of ‘so’ spiked among G2 speakers, being significantly higher than the use among the G1 and G3 speakers. In the speech of bilingual university students in Florida, Said-Mohand (2006) discovered that the three most common functions of entonces are narrative progression, cause, and conclusion. In the case of narrative progression, the author points out that many instances of this usage are equivalent to ‘and’ when linking ideas, a divergence from its most basic semantic meaning. With regard to tú sabes, Said-Mohand (2006, 2007) reported that this marker tended to perform the functions of narrative progression, conclusion, and reformulation, such as repeating or correcting an utterance (e.g. porque por ejemplo en high school, tú sabes, bueno en high school siempre existe como . . . [‘because for example in high school, you know, well in high school there always exists like’] (2007, p. 80)). None of these functions are necessarily unique to SHL, although one of the same author’s findings for como does appear to signal a non-canonical usage; namely, using it as a connector on the sentence level and to indicate a direct quote (e.g. él estaba como ‘¿qué estás haciendo?’ [‘he was like, what are you doing?’), similar to ‘like’ in English (Said-Mohand, 2008, p. 84). In this latter case, the author is careful to emphasize that there is not conclusive evidence indicating that this represents a transfer of ‘like’ from English since similar cases are also observed in the Spanish of monolinguals. In addition to the research on DMs, the pragmalinguistic aspects highlighted in the literature on speech acts involve different types of cross-linguistic variation, or lack thereof, in the choice of strategies, formulas, or mitigators, as well as non-standard grammatical structures that researchers attribute to an underdeveloped proficiency in Spanish. In the directives produced by two young bilingual children of first-generation Mexican families residing in Los Angeles, García and Leone (1984) found that by the age of 6, the subjects demonstrated a range of 192

Heritage Spanish pragmatics

linguistic structures in both English and Spanish. However, the data showed some grammatical errors and misuse of formal and familiar commands, such as using píntelo (‘paint it’) to a sibling rather than píntalo (‘paint it’). Bhimji (2002) applies an ethnographic, discourse analytic framework to the directives used by caregivers to children in three working class Latino families in Los Angeles. She argues against the view that in families of lower-economic backgrounds, parents reduce the complexity of their language when talking to children. The directives analyzed show a wide gamut of structures, from direct imperatives to declaratives and interrogatives, often accompanied with various mitigating devices and justifications. Furthermore, the children were shown to be active participants in the interactions, responding to directives with clarification questions, verbal displays of compliance and noncompliance, and justifications when they opted not to comply. This study is relevant to the discussion of Spanish heritage speakers due to the fact that these bilingual children, who grow up to become adult heritage speakers, were immersed in a learning environment in Spanish that was found to be pragmatically rich. Valdés (1981) and Valdés and Pino (1981), among the findings of a pragmalinguistic nature, reported a prevalence of code-switching when examining requests and compliment responses respectively in adult bilingual speech. In some cases, these switches appeared to fulfill a mitigating function. This occurred in the Valdés and Pino data, for example, when one speaker employed the politeness formula cuando quieras (‘whenever you want’) while responding to a compliment in English. In this same study the Mexican monolingual group used more politeness formula than the bilingual group, including the fixed expression muy a la orden (‘at your service’), but apart from the frequent use of code-switching and a higher incidence of tú over usted, the bilingual group exhibited similar compliment response patterns. Dumitrescu (2005) observed the influence of English grammatical structures in the SHL spoken by university students, including calques (e.g. Gracias por un buen tiempo. ‘Thanks for a good time’) and instances of transfer (e.g. Espero hacer esto de nuevo. ‘I hope to do this again,’ when thanking a friend for dinner). Pinto and Raschio (2007, 2008) reported similar cases of noncanonical structures for requests and complaints due to English interference. To ask to borrow notes from a classmate, one heritage speaker used estaria bien si me las prestarias? [sic] (‘would it be okay if you lent them to me?’), while another asked a friend to pay up a debt with estaba pensando si me podrías pagar porque tuve una emergencia y necesito el dinero [sic] (‘I was wondering if you could pay me because I had an emergency and I need money’). Similarly, in both their complaints and requests there was some evidence of English interference in their choice of mitigators, including quisiera saber si (‘I wanted to know if’) and quisiera ver si (‘I wanted to see if’), combinations that demonstrate possible English transfer and the use of formal-sounding structures in informal contexts. For compliments and compliment responses, Yáñez (1990) found that in her relatively small data corpus collected from Chicana women, the utterances were highly formulaic in nature and syntactically similar to the structures reported for monolingual English speakers (Pomerantz, 1978; Wolfson, 1983). In conjunction with this finding, the author observed a narrow range of adjectives for complimenting, including a high incidence of bonito (e.g. ¡Qué bonito se te ve el pelo! ‘How pretty your hair looks!’). Beyond the speech act, some authors have examined oral discourse in the context of longer narrations. With the objective of studying spontaneous oral speech, Leone and Cisneros (1985) looked at how bilingual children retell stories in Spanish, and they observed different communicative strategies that the children utilized for their narratives. For example, they used English as metalinguistic comments (i.e. ‘I don’t know how to say it in Spanish’), which served as a type of ‘disclaimer’ showing a heightened awareness of their own language ability and of the expectations of the test administrators. Other strategies consisted of restart-switches (e.g. ‘one day, he . . . un día’), indicating a process of self-monitoring and an awareness of the 193

Derrin Pinto

appropriate resources for the task at hand, and requests for confirmation, such as ¿el muchacho? (‘the boy?’), which served to trigger a response from the administrator regarding the correctness or appropriateness of the lexical item. The authors concluded that these strategies indicate not only a linguistic competency under development, but also an awareness of this process by the children. Another study by the same authors (Cisneros & Leone, 1993) explore the unplanned oral discourse of eight Spanish speakers living in San Antonio, Texas, although the subjects here are adult Spanish speakers learning English, hence the results are not particularly representative of SHL. The same could be said for the participants in Elías-Olivares (1995) and García Pastor (1999), where the focus is on adult Spanish-speaking immigrants. Regardless of the limited English proficiency of the speakers in Elías-Olivares (1995), who originated from a Mexican community in Chicago, they still show evidence that is characteristic of language contact, such as loanwords and calques. While the topic of academic writing is beyond the scope of this chapter, the use of oral academic registers is relevant to the overarching topic of pragmatics. More specifically, the challenges that heritage learners face regarding academic registers can primarily be broken down in pragmalinguistic terms. Valdés and Geoffrion-Vinci (1998) found that, while their bilingual subjects showed a sensitivity to register differences and an effort to adjust their speech accordingly, when compared to a monolingual group from Mexico, they struggled more to express themselves in an academic register. Their speech was characterized by a more limited vocabulary, the use of invented words (e.g. ojalamente ‘hopefully,’ arruinamiento ‘ruin,’ and mayoridad ‘majority’), and fewer strategies for organizing their discourse (e.g. por otra parte ‘on the other hand’; no obstante ‘nevertheless’). In a comparison of the oral presentations of two graduate students, one Spanish dominant and the other English dominant, Achugar (2003) documented how the English-dominant speaker relied on English to clarify and expand on the material presented in Spanish and, in general, displayed a scarcity of discursive features required to enhance and elaborate on the subject matter. In contrast, the Spanish-dominant student employed a wide range of discursive strategies to expand on the information, make connections, and adopt a more critical stance. While heritage speakers understandably struggle with both written and oral formal registers, this is an aspect of their proficiency that can be improved through consciousness-raising tasks and practice (Acevedo, 2003). The pragmalinguistic features mentioned in this summary could essentially be reduced to lexical items, grammatical structures, or a combination of both that, for one reason or another, differ slightly from what one might observe in non-contact varieties of Spanish. In some instances, these dissimilarities merely entail varying frequencies of usage, primarily due to a lower proficiency in Spanish and only partial access to a comparable repertoire of words, expressions, and structures that a typical monolingual Spanish speaker possesses. As Lapidus and Otheguy (2005) specify with regard to grammar, most instances of contact-induced change in the Spanish of immigrants and their children tends to be restricted to alterations in the frequencies of linguistic forms. This appears to hold true for many cases within the pragmalinguistic domain as well. However, English can play a more obtrusive role in certain circumstance involving varying degrees of interference, either through calques or unconventional collocations (e.g. quisiera ver si). To conclude this section, it may be worth mentioning that there are also a series of studies focusing on morphological or syntactic phenomena that entail pragmatics indirectly. In a number of these investigations, certain contextual factors are believed to motivate grammatical choices. This is often the case, for example, when the topics under consideration involve the use of the subjunctive or subject pronouns (e.g. Ávila-Shah, 2000; Bayley & Pease-Alvarez, 1997; Coles, 2012; Flores-Ferrán, 2002; Lapidus & Otheguy, 2005; Martínez Mira, 2009; Torres Cacoullos & Travis, 2010). 194

Heritage Spanish pragmatics

Sociopragmatic issues Starting with speech acts, in addition to the pragmalinguistic aspects summarized earlier, research on this topic has also emphasized sociopragmatic phenomena including levels of directness or formality and different types of sociolinguistic variation. Walter (1981) carried out a study on the requests of Spanish-English bilingual children, primarily from Puerto Rican origin, with a focus on the effect of setting, topic, and gender. The data were collected by using puppets to elicit requests from children between the ages of 7 and 11. Two significant findings were that the factors of setting and sex of the addressee triggered varying levels of deference in the requests. For example, in a scenario that involved selling cookies, the bilingual children were less deferential when directing their requests to males than to females. While Walter (1981) worked with elicited data, García and Leone (1984) used naturalistic data to investigate the directives produced by two bilingual children (ages 4.5 to 6) in southern California. The authors show that these two children used a variety of directives in their dominant language (Spanish for one, English for the other), with a preference for direct imperatives and no politeness markers with peers and siblings, and for indirect requests and inferred strategies (e.g. ‘Teacher, where do I get mine?’) with adults and other children when they seek special favors. Furthermore, the children’s performance was influenced by contextual factors such as the setting and the status of the addressee. For instance, the girl participant was more verbally aggressive with her siblings at home than with her classmates at school. The fact that both Walter (1981) and García and Leone (1984) found notable differences due to setting and the addressee’s status or sex underscores the sensitivity that bilingual children develop from a young age. In a comprehensive study of eighth-grade students in a dual immersion school in Chicago, Potowski (2007) includes sociolinguistic appropriateness as one of many testing measures. Subjects were assessed though written and oral tasks, targeting both formal and informal requests. When compared to a group of L2 Spanish students and another of Spanish-dominant recent arrivals, the heritage speakers were more similar to the latter group than the former in that they received high ‘difference scores’ in the language used in formal situations versus that used for informal situations. In other words, unlike the L2 students, in both their written and oral requests the heritage speakers demonstrated a perceptible sensitivity to the difference between formal and informal registers. Cashman (2005, 2006) delves into some questions regarding the potentially conflictive side of verbal interaction; namely, impoliteness and disagreement. In her 2005 study, she applies a conversation analytic approach to explore how three bilingual children manage disagreement, how they perform stereotypical male or female talk, and whether or not code-switching plays a role in managing their interaction. For the first question, Cashman observed the tendency in the children to aggravate their disagreement moves. For example, one exchange contains a threat plus an increase in volume in the move ‘I’m gonna DESTROY YOU,’ while another sequence includes the insults puto and menso. Regarding the second research question, Cashman found that stereotypical male and female strategies were employed, but not necessarily by the corresponding gender. Such an example occurred when one of the girls used the aforementioned insults in an aggravated opposition move. Overall the contributions produced by girls were not always of a collaborative nature, going against the female stereotype. For the last issues, codeswitching played a role in helping the children manage their disagreement, such as allowing them to change affiliation between interlocutors or move from disagreeing to justifying. Based primarily on Culpeper’s (1996) model of impoliteness and Spencer-Oatey’s (2002) concept of rapport management, Cashman (2006) investigates the impoliteness strategies, and reactions to them, in the spontaneous verbal interaction of 22 second-grade bilingual children. The children employed strategies such as name-calling and ridicule to attack their interlocutor’s face. 195

Derrin Pinto

The children also threatened their peers’ equity rights, the desire to be treated equally by others, by producing directives to impose on their interlocutor (e.g. ya te callas tú también, no te metas ‘be quiet already you too, don’t butt in’), as well as threats (e.g. o te voy- doy una cachetada ‘or I’ll- I’ll slap you’). The participants also threatened their peers’ association rights, the belief that one is entitled to interact with others in accordance with the given relationship, by using silence and other verbal resources to disassociate themselves from the group. Regarding their reactions to impolite behavior, which contribute to co-constructing impoliteness, they ranged from not responding to countering both defensively and offensively. Moving on to adult speakers, Pinto and Raschio (2007, 2008) highlight the hybrid nature of the requests and complaints in their SHL data. Participants employed request strategies in Spanish that were less direct than the monolingual Spanish speakers from Mexico, more in line with the Anglo-speaking preference for indirectness. For complaints, the bilingual participants demonstrated some similarities to the native English-speaking group, including a propensity for supplying their complaints with multiple forms of mitigation. Nevertheless, these same heritage learners also employed a percentage of complaints with zero downgraders that was comparable to the Mexican monolingual group. Overall the authors found that the heritage learners in these two studies represented the inbetweenness of SHL, especially when considering both pragmalinguistic and sociopragmatic aspects of speech act production. Given this quality of indeterminacy that is prevalent in SHL, researchers may choose to focus their attention on one or more of the three possibilities mentioned in Pinto and Raschio (2008): (1) how and when SHL resembles monolingual Spanish; (2) how and when SHL resembles monolingual English; and (3) how and when SHL distinguishes itself from either monolingual Spanish, monolingual English, or both. In a partial replication of Pinto and Raschio’s (2007, 2008) work, Finestradt-Martínez and Potowski (2016) administered discourse completion tasks to a total of 85 speakers and incorporated four additions: (1) third-generation speakers; (2) scenarios involving refusals; (3) scenarios in both Spanish and English (for the bilingual participants); and (4) scenarios with varying degrees of power, distance, and imposition between the imagined interlocutors. The authors reasoned that the most pragmatically indirect and mitigated responses would be used when social distance and imposition were high, but familiarity was low, and the opposite would happen when interlocutors were separated by low social distance and the degree of imposition was low. Results showed that all groups of speakers preferred indirect strategies (indirectness and hints) over direct strategies, but there were significant differences between the four groups. In particular, the monolingual Mexicans made the least use of indirect strategies, while the monolingual English speakers used the most. The bilinguals used increasingly more indirect strategies with each increasing generational group. In addition, bilinguals were more indirect when responding in English than when responding in Spanish. For bilinguals, then, indirectness increased in this order: G2 Spanish < G2 English < G3 Spanish < G3 English. Gutiérrez-Rivas (2008, 2011) looks at the requests produced by three generations of Cubans and Cuban Americans living in Miami. In her 2008 article, the author emphasizes the preference of first-generation speakers to use positive politeness strategies, (e.g. me hace falta que me des una mano en esto ‘I need you to give me a hand with this’), with friends and strangers. In contrast, among the findings reported in the second publication is the tendency for third-generation speakers, especially women, to approximate English-speaking norms of using more indirectness when requesting, including asking questions rather than affirming, apologizing, and employing more verbose strategies in general. This study supports the idea that pragmatic acculturation takes place over time and that politeness norms of the predominant culture begin to seep in. In another study dealing with performing complaints in English and Spanish, Elias (2013) reported that a group of second-generation, Mexican American bilinguals displayed a slightly 196

Heritage Spanish pragmatics

higher preference in Spanish for establishing solidarity with their interlocutor. In English, the participants were more focused on showing respect for the other. The author concludes that the subjects seem to demonstrate “two separate pragmatic systems in dealing with the act of complaining” (p. 64), although these systems clearly entail extensive overlap. Dumitrescu (2005) described cases of sociopragmatic infelicities with regard to heritage speakers violating the norms of politeness. The author attributes such infractions to the unawareness of, or perhaps insensitivity to, the use of registers. This occurred, for instance, when subjects employed informal expressions (e.g. Ahí te wacho / Simón) in the context of a service encounter, which in some contexts might suggest a higher degree of familiarity than the situation would normally involve (p. 394). Arellano (2000) utilized a multiple-choice questionnaire to investigate the request preferences of a mixed group of monolingual and bilingual farmworkers in California. The results signal an expected tendency toward higher levels of indirectness and mitigation when directing requests to an authority figure. A predictable relationship was also found for requests involving a higher degree of imposition, which, relatively speaking, are more likely to be performed with less direct means. The extended speech event of leave-taking in the context of a Mexican American family in Los Angeles is the subject of García (1981). The author identified four phases of interaction, initiation, preparation, joking interchanges, and final leave-taking, which she describes as uniquely Mexican American given the dynamic nature of the talk exchange (e.g. verbal overlapping, simultaneous interchanges, and loud voices), the use of Spanish and English with nonstandard features (e.g. ¿Ónde está Juan ‘Where is Juan?’), as well as switching between Spanish and English names (e.g. Elena/Helen, Consuelo/Connie, etc.). It should be pointed out that the lengthy and dynamic nature of leave-taking is certainly not specific to Mexican American communities, hence the degree of uniqueness here is relative and surfaces in Garcia’s study due to the crosscultural comparison to Anglo norms. The existing research on singular second-person subject pronouns, tú, usted, and vos, covers different geographical areas within the U.S. Among the issues investigated are usage patterns, situational variation, and generational differences in how these pronouns are used. Among the studies carried out in the Southwestern region of the U.S., Brown (1975) utilized questionnaires to examine which pronoun, tú or usted, was preferred by Mexican American students at the University of Arizona when addressing their parents and found that twice as many of the subjects reported using usted. However, since the 1970s when Brown’s study was implemented, second-person pronoun usage has continued to evolve. Jaramillo (1990) carried out a study with Spanish-speaking adults in rural New Mexico and identified two general domain clusters: tú informal, including nuclear family and friends, and usted formal, encompassing the realm of ceremonial family, employment, and low and high status professionals. The author concludes that usage patterns in this particular community, while similar to other Spanishspeaking populations, trend toward the conservative. This tendency surfaces in compadrazgo interactions between the godparents and parents of a child, with a high incidence of reciprocal usted (88%), and when subjects address their godparents, which usually prompts non-reciprocal usted (96%). In another study in the same geographical area, Jaramillo (1995) found that two groups, adults over 51 and those with less than eight years of education, reported a higher rate of interactions with reciprocal usted. Jaramillo’s observation that tú is gradually encroaching on contexts traditionally reserved for usted is consistent with the trend observed in other parts of the Spanish-speaking world. Using a data corpus comprised of the speech of Salvadorians in Houston, Texas, Schreffler (1994) considered the fate of the three pronouns tú, vos (‘you’ informal), and usted among second-generation speakers in contact with English and other varieties of Spanish. As might be 197

Derrin Pinto

expected, the use of voseo has diminished as tuteo has become the preferred option for informal address, the consequence of extended contact with Mexican Spanish and the higher prestige associated with tú. Sigüenza-Ortiz (1996) collected data in eastern Los Angeles and observed that for English-dominant speakers, tú appears to have virtually displaced usted, leaving it as the exclusive second-person singular pronoun, even for addressing people in first encounters, a context in which the Mexican-born participants in the study prefer usted. Another finding in this study reveals the importance of conversational setting given that all participants reported using a higher incidence of usted at church than at home. Gutiérrez-Rivas (2010) examines the use of tú and usted in the context of requests produced by second- and third-generation Cuban Americans in Miami, Florida. Among the notable findings of this study are the fact that some participants combined both tú and usted, whether it be pronouns or verb forms, in a single utterance (e.g. disculpe y perdóname ‘sorry and excuse me’). In some cases, the author attributes this inconsistency to a lack of language proficiency, while in other cases she finds this to be a strategic approach to show both respect (with usted) and solidarity (with tú). The speakers in this study, especially those from the third generation, do not shy away from using tú in contexts involving distance between the interlocutors. One of the general conclusions regarding the study of second-person subject pronoun usage in SHL is the encroachment of tú on territory once reserved for usted. This probably reflects a combination of factors, including contact with English and the ongoing universal trend, both in Spanish-speaking communities and in other languages, that favors informality over formality (Carrasco Santana, 1998; Fairclough, 2001; Marín, 1972). Once again, direct contact with English may be accelerating tendencies that are also occurring in other varieties of Spanish.

Concluding remarks From a historical perspective, pragmatic changes to the Spanish of the U.S. are not necessarily a recent occurrence. Evidence of early pragmatic change is documented in a study of letters written in California and New Mexico during the 19th century (Balestra, 2008). During this period, the forms of address employed in letters underwent modifications due to contact with both the English language and Anglo culture, evolving from more elaborate and deferential openings (e.g. Mui señor mío y venerado amigo ‘Distinguished sir and venerated friend’) to less verbose alternatives that are more in sync with current day uses (e.g. Querida Julia ‘Dear Julia’). While pragmatic changes themselves are not new, the attention dedicated to investigating them in present-day SHL is relatively recent and is still in the early stages of development. As noted earlier, besides obvious cases of English intrusion through code-switching and/or lexical borrowing, contact with English often acts on developmental tendencies or changes that have already manifested themselves in the Spanish language (Silva-Corvalán, 1994, 2001). For this reason, distinguishing between externally and internally induced change is often impossible (Montrul, 2012). It should also be emphasized that on a global level, subtle English influence on Spanish can be more indirect and ubiquitous, acting through different channels of multimedia such as film, television and social media. Within the realm of pragmatics, the result can be a situation in which Anglicized expressions (e.g. ¡Correcto!; ¡Dame un respiro!; ¡Olvídalo!) are preferred in translations over those that would be more conventionalized alternatives in Spanish (Gómez Capuz, 2001). From this perspective, direct contact with English in Spanish-speaking communities in the U.S. can be considered an intensification of the contact that occurs on a more global level. As such, the difference between contact and non-contact varieties of Spanish is not necessarily a categorical distinction. 198

Heritage Spanish pragmatics

The drawbacks that one can attribute to the body of literature on pragmatic phenomena in SHL, beyond its relatively limited lifespan and erratic development, have been described previously (Pinto, 2012). Among these are the shortage of studies that incorporate comparative and/or truly comparable data in monolingual Spanish and L1 English, the lack of follow-up studies, the need to combine both quantitative and qualitative analyses, and the predominant focus on speech production over reception/interpretation and perceptions. This latter topic requiring the elicitation of heritage speakers’ perceptions of different aspects of their speech behavior can provide enlightening evidence about their own learning and interactions (e.g. Marijuan, 2015). This is one promising avenue of research that will surely be explored more thoroughly in the future. It is also worth pointing out that the wide-ranging nature of pragmatics as a field of study inevitably leads to multiple strands of research developing in multiple directions. Especially when compared to the investigation of SHL in formal linguistic or psycholinguistic studies, where there have been some underlying theoretical assumptions about the nature of language acquisition, pragmatics has lacked an analogous set of guiding principles. One could identify, perhaps, the general aim of looking into the nature of language contact, and more specifically, the influence of English in SHL and the consequences of this contact when speakers possess a less developed proficiency in Spanish. However, this exploratory goal does little to narrow the scope of research. The pragmatic domain of language, as Blum-Kulka and Sheffer (1993) note, may be the first area to be affected in contexts of intense contact with a second language. The permeability of pragmatics in contact situations may be due to a variety of overlapping causes. One likely cause is the fact that the pragmatic component of a given language largely comprises the appropriate use of lexical items and therefore does not involve deep structural issues. For instance, while SHL in the U.S. tends to show English lexical influence either through code-switching, borrowing, false cognates, or calques, these lexical substitutions are relatively superficial and can be incorporated with ease. Another reason that pragmatic aspects are easily affected in contact environments is that pragmatics encompasses variable and loosely defined prescriptive rules or norms. That is, at least when compared to morphosyntax or phonology, sociopragmatic norms are not as easily accessible or as easy to articulate due to an unlimited potential for contextual fluctuation. In spite of dialectical variation, one can still consult prescriptive rules for the subjunctive, verb conjugations, or pronunciation, but establishing the prescriptive norms for something like requesting or complaining is less straightforward. Even the choice between tú and usted involves competing variables such as age, distance, power, status, and setting, and there is no concrete formula that yields the correct option for any given interaction. One might even say that no two linguistic encounters are ever exactly identical, thus speakers are constantly faced with the need to evaluate each situation independently. Within a particular dialect, this vagueness surrounding the prescriptive standards for pragmatic-related phenomena in the collective conscious of the corresponding speech community makes the realm of pragmatics more susceptible to language interference.

References Aaron, J. E. (2004). So respetamos una tradición del uno al otro. Spanish in Context, 1(2), 161–179. Acevedo, R. (2003). Navegando a través del registro formal: Curso para hispanohablantes bilingües. In A. Roca & M. C. Colombi (Eds.), Mi lengua: Spanish as a Heritage Language in the United States, Research and Practice (pp. 257–268). Washington, DC: Georgetown University Press. Achugar, M. (2003). Academic registers in Spanish in the U.S.: a study of oral texts produced by bilingual speakers in a university graduate program. In A. Roca & M. C. Colombi (Eds.), Mi lengua: Spanish as a

199

Derrin Pinto Heritage Language in the United States, Research and Practice (pp. 213–234). Washington, DC: Georgetown University Press. Andersen, E., Brizuela, M., DuPuy, B., & Gonnerman, L. (1999). Cross-linguistic evidence for the early acquisition of discourse markers as register variables. Journal of Pragmatics, 31, 1339–1351. Arellano, S. (2000). A hierarchy of requests in California Spanish: Are indirectness and mitigation polite? In A. Roca (Ed.), Research on Spanish in the United States (pp. 319–332). Somerville, MA: Cascadilla Press. Ávila-Shah, B. (2000). Discourse connectedness in Caribbean Spanish. In A. Roca (Ed.), Research on Spanish in the United States (pp. 238–251). Somerville, MA: Cascadilla Press. Balestra, A. (2008). Formas de tratamiento en correspondencia en español: California y Nuevo México, 1800–1900. In A. Balestra, G. Martinez & I. Moyna (Eds.), Recovering the U.S. Hispanic Linguistic Heritage: Sociohistorical Approaches to Spanish in the United States (pp. 75–88). Houston, TX: Arte Público Press. Barnes, J. (2001). Politeness in English, Basque and Spanish: Evidence from a trilingual child. Jakingarriak, 45, 40–45. Bayley, R., & Pease-Alvarez, L. (1997). Null pronoun variation in Mexican-descent children’s narrative discourse. Language Variation and Change, 9, 349–371. Bhimji, F. (2002). ‘Dile family’: Socializing language skills with directives in three Mexican families in south central Los Angeles. Unpublished doctoral dissertation, University of California, Los Angeles, CA. Blum-Kulka, S. (1990). Don’t touch your lettuce with your fingers. Journal of Pragmatics, 14, 259–288. Blum-Kulka, S., House J., & Kasper G. (Eds.). (1989). Requests and Apologies. Norwood, NJ: Ablex Publishing. Blum-Kulka, S., & Sheffer, H. (1993). The metapragmatic discourse of American-Israeli families at dinner. In G. Kasper & S. Blum-Kulka (Eds.), Interlanguage Pragmatics (pp. 196–224). New York: Oxford University Press. Brizuela, M., Andersen E., & Stallings, L. (1999). Discourse markers as indicators of register. Hispania, 82(1), 128–141. Brown, D. (1975). The use of ‘tú’ and ‘usted’ with parents by some Mexican American students. Hispania, 58(1), 126–127. Carrasco Santana, A. (1998). Tendencias relacionales de los españoles en las interacciones verbales, I: La distancia social en la conversación. Cuadernos de Lazarillo, 15, 55–61. Carreira, M., & Kagan, O. (2011). The results of the national heritage language survey: Implications for teaching curriculum design and professional development. Foreign Language Annals, 44(1), 40–64. Cashman, H. R. (2005). Aggravation and disagreement: A case study of a bilingual, cross-sex dispute. Southwest Journal of Linguistics, 24(1), 31–51. Cashman, H. R. (2006). Impoliteness in children’s interactions in a Spanish/English bilingual community of practice. Journal of Politeness Research, 2, 217–246. Cisneros, R., & Leone, E. A. (1993). Literacy stories: Features of unplanned oral discourse. In A. Roca & J. M. Lipski (Eds.), Spanish in the United States (pp. 103–120). Berlin: Mouton de Gruyter. Clyne, M. (1979). Communicative competences in contact. ITL Review of Applied Linguistics, 43, 17–37. Clyne, M., Ball, M., & Neil, D. (1991). Intercultural communication at work in Australia: Complaints and apologies in turns. Multilingua, 10(3), 251–273. Coles, F. (2012). Stance and the subjunctive in Isleño Spanish. Hispania, 95(2), 285–298. Culpeper, J. (1996). Towards an anatomy of impoliteness. Journal of Pragmatics, 25, 349–367. Dumitrescu, D. (2005). Agradecer en una interlengua. In J. M. Medrano (Ed.), ACTAS, II Coloquio Internacional del Programa EDICE (pp. 375–406). Costa Rica: Programa EDICE. Elias, M. (2013). ‘Tengo bien harto esperando en la línea’: Complaint strategies by second-generation Mexican-American bilinguals. Unpublished master’s thesis, Arizona State University, Tucson, AZ. Elías-Olivares, L. (1995). Discourse strategies of Mexican American Spanish. In C. Silva-Corvalán (Ed.), Spanish in Four Continents. Studies in Language Contact and Bilingualism (pp. 227–240). Washington, DC: Georgetown University Press. Fairclough, N. (2001). Language and Power. London: Longman. Farghal, M., & Haggan, M. (2006). Compliment behaviour in bilingual Kuwaiti college students. International Journal of Bilingual Education and Bilingualism, 9, 94–118. Finestradt-Martínez, I., & Potowski, K. (2016). Requests and refusals among bilingual Mexican-Americans. Paper presented at the 8th Workshop on Spanish Sociolinguistics, April 14–16, San Juan, Puerto Rico. Flores-Ferrán, N. (2002). Subject Personal Pronouns in Spanish Narratives of Puerto Ricans in New York City: A Sociolinguistic Perspective. Munich: Lincom Europa.

200

Heritage Spanish pragmatics Fraser, B. (1999). What are discourse markers? Journal of Pragmatics, 31, 931–952. Fraser, B., & Malamud-Makowski, M. (1996). English and Spanish contrastive discourse markers. Language Sciences, 18, 863–881. García, M. E. (1981). Preparing to leave: Interaction at a Mexican-American family gathering. In R. P. Durán (Ed.), Latino Language and Communicative Behavior (pp. 195–215). Norwood, NJ: Ablex. García, M. E., & Leone, E. (1984). The use of directives by two Hispanic children: an exploration of communicative competence. National Center of Bilingual Research Report Series, 25. Los Alamitos, CA: National Center for Bilingual Research. García Pastor, M. D. (1999). Directive use and performance in a U.S. Hispanic community: A cultural approach. Studies in English Language and Linguistics, 1, 147–170. Gómez Capuz, J. (2001). La interferencia pragmática del inglés sobre el español en doblajes, telecomedias y lenguaje coloquial: una aportación al estudio del cambio lingüístico en curso. Revista Tonos Digital, 2. Retrieved from www.um.es/tonosdigital/znum2/estudios/indicestudios.htm. Gutiérrez-Rivas, C. (2008). Actos de habla mixtos: Reflexiones sobre la pragmática del español en referencia a la teoría y métodos actuales de análisis. Núcleo, 20(25), 149–171. Gutiérrez-Rivas, C. (2010). Los usos de ‘tú’ y ‘Ud.’ en los actos de habla: una aproximación a la pragmática del bilingüe. ALPHA, 31, 85–102 Gutiérrez-Rivas, C. (2011). El efecto del género en el discurso bilingüe. Un estudio sobre peticiones. Estudios de Lingüística Aplicada, 29, 37–59. Hlavac, J. (2006). Bilingual discourse markers: Evidence from Croation-English code-switching. Journal of Pragmatics, 38, 1870–1900. Jaramillo, J. (1990). Domain constraints on the use of tú and usted. In J. J. Bergen (Ed.), Spanish in the United States: Sociolinguistic Issues (pp. 14–22). Washington, DC: Georgetown University Press. Jaramillo, J. (1995). Social variation in personal address etiquette. Hispanic Linguistics Journal, 6–7, 191–224. Lapidus, N., & Otheguy, R. (2005). Overt nonspecific ellos in the Spanish of New York. Spanish in Context, 2, 157–176. Leech, G. (1983). Principles of Pragmatics. London: Longman. Leone, E. A., & Cisneros, R. (1985). Children’s storytelling strategies. In L. Elfas-Olivares, E. A. Leone, R. Cisneros, & J. Gutiérrez (Eds.), Spanish Language Use and Public Life in the USA (pp. 89–112). Berlin: Mouton Publishers. Marijuan, S. (2015). Spanish Vernacular Use in College L2 Language Classrooms Amongst Spanish Heritage Speakers in Washington DC. Paper delivered at the Annual Convention of the Modern Language Association. Vancouver, Canada, January 7–11. Marín, D. (1972). El uso de ‘tu’ y ‘usted’ en el español actual. Hispania, 55, 904–908. Marti, L. (2006). Indirectness and politeness in Turkish-German bilingual and Turkish monolingual requests. Journal of Pragmatics, 38, 1836–1869. Martínez, M. (2009). Spanish heritage speakers in the Southwest: Factors contributing to the maintenance of the subjunctive in concessive clauses. Spanish in Context, 6(1), 105–126. Montrul, S. (2012). The grammatical competence of Spanish heritage speakers. In S. Beaudrie & M. Fairclough (Eds.), Spanish as a Heritage Language in the United States (pp. 101–120). Washington, DC: Georgetown University Press. Pascual y Cabo, D., & Rothman, J. (2012). The (il)logical problem of heritage speaker bilingualism and incomplete acquisition. Applied Linguistics, 33(4), 450–455. Pinto, D. (2012). Pragmatics and discourse: Doing things with words in Spanish as a heritage language. In S. Beaudrie & M. Fairclough (Eds.), Spanish as a Heritage Language in the United States (pp. 121–138). Washington, DC: Georgetown University Press. Pinto, D., & Raschio, R. (2007). A comparative study of requests in heritage speaker Spanish, L1 Spanish, and L1 English. International Journal of Bilingualism, 11(2), 135–155. Pinto, D., & Raschio, R. (2008). Oye, ¿qué onda con mi dinero? An analysis of heritage speaker complaints. Sociolinguistic Studies, 2(2), 221–249. Pomerantz, A. (1978). Compliment responses. In J. Schenkein (Ed.), Studies in the Organization of Conversational Interaction (pp. 79–111). New York/San Francisco/London: Academic Press. Potowski, K. (2007). Language and Identity in a Dual Immersion School. Clevedon, UK: Multilingual Matters. Said-Mohand, A. (2006). Estudio sociolingüístico de los marcadores como, entonces y tú sabes en el habla bilingüe estadounidense. Unpublished doctoral dissertation, University of Florida. Said-Mohand, A. (2007). A Sociolinguistic approach to the discourse marker tú sabes ‘you know’ in the speech of young US bilinguals. Southwest Journal of Linguistics, 26(2), 67–93.

201

Derrin Pinto Said-Mohand, A. (2008). Aproximación sociolingüística al uso de entonces en el habla de jóvenes bilingües estadounidenses. Sociolinguistic Studies, 2(1), 97–130. Schreffler, S. B. (1994). Second-person singular pronoun options in the speech of Salvadorans in Houston, TX. Southwest Journal of Linguistics, 13(1–2), 101–119. Sigüenza-Ortiz, C. (1996). Social deixis in a Los Angeles Spanish-English bilingual community: Tú and usted patterns of address. Unpublished doctoral dissertation, University of Southern California. Silva-Corvalán, C. (1994). The gradual loss of mood distinctions in Los Angeles Spanish. Language Variation and Change, 6, 255–272. Silva-Corvalán, C. (2001). Sociolingüística y pragmática del español. Washington, DC: Georgetown University Press. Spencer-Oatey, H. (2002). Managing rapport in talk: Using rapport sensitive incidents to explore the motivational concerns underlying the management of relations. Journal of Pragmatics, 34, 529–545. Thomas, J. (1983). Cross-cultural pragmatic failure. Applied Linguistics, 4(2), 91–112. Torres, L. (2002). Bilingual discourse markers in Puerto Rican Spanish. Language in Society, 31(1), 65–83. Torres, L., & Potowski, K. (2008). A comparative study of bilingual discourse markers in Chicago Mexican, Puerto Rican, and MexiRican Spanish. International Journal of Bilingualism, 12(4), 263–279. Torres, L., & Potowski, K. (forthcoming). Manuscript in preparation. Torres Cacoullos, R., & Travis, C. E. (2010). Variable yo expression in New Mexico: English influence? In S. Rivera-Mills & D. J. Villa (Eds.), Spanish of the U.S. Southwest: A Language in Transition (pp. 185–206). Madrid: Iberoamericana. Valdés, G. (1981). Codeswitching as deliberate verbal strategy: A microanalysis of direct and indirect requests among bilingual Chicano speakers. In R. P. Durán (Ed.), Latino Language and Communicative Behavior (pp. 95–108). Norwood, NJ: Ablex. Valdés, G. (2005). Bilingualism, heritage language learners, and SLA research: Opportunities lost or seized? The Modern Language Journal, 89(3), 410–426. Valdés, G., & Geoffrion-Vinci, M. (1998). Chicano Spanish: The problem of the ‘underdeveloped’ code in bilingual repertoires. The Modern Language Journal, 82(4), 473–501. Valdés, G., & Pino, C. (1981). Muy a tus órdenes: compliment responses among Mexican-American bilinguals. Language in Society, 10, 53–72. Walter, J. (1981). Variation in the requesting behavior of children. International Journal of the Sociology of Language, 27, 77–92. Wolfson, N. (1983). An empirically based analysis of complimenting in American English. In N. Wolfson & E. Judd (Ed.), Sociolinguistics and Language Acquisition (pp. 82–95). Rowley, MA: Newbury House. Yáñez, R. H. (1990). The complimenting speech act among Chicano women. In J. Bergen (Ed.), Spanish in the United States: Sociolinguistic Issues (pp. 79–85). Washington, DC: Georgetown University Press.

202

14 NEUROLINGUISTIC APPROACHES TO SPANISH AS A HERITAGE LANGUAGE Harriet Wood Bowden and Bernard Issa university of tennessee knoxville, usa

Introduction The field of neurolinguistics––i.e., the study of how language is processed and represented in the human brain––finds itself at the intersection of the fields of neuroscience, linguistics and cognitive science more generally. The history of neurolinguistics can be traced back as far as the 1860s with Paul Broca’s seminal observations and descriptions of patients with speech impairments caused by brain lesions to the posterior inferior frontal gyrus. This region later became known as Broca’s area, one of the most studied language-related brain regions (Dronkers, Plaisant, IbaZizen & Cabanis 2007). While the field has continued to gain insight into how the brain processes language throughout its history, it has been revolutionized in the last forty years as a result of developments in neuroimaging techniques (Ingram 2007). The vast majority of this recent work has examined language processing in monolingual speakers (see, for example, reviews by Poeppel & Hickok 2004 and Ross 2010), and has resulted in a rich body of literature which includes the investigation of a number of issues, including the time-course of both semantic and syntactic language processing (Friederici 2002), the neural correlates underlying semantic processing (Binder, Desai, Graves & Conant 2009), as well as how words are recognized (Lee & Federmeier 2012), comprehended (Kaan & Swaab 2002) and produced (Indefrey 2011). More recently, neurolinguistic research has expanded to focus on language processing in bilingual individuals, examining how and where a bilingual’s two languages are represented in the brain, and how a bilingual manages these two systems during real-time language processing (see Kroll & Sunderman 2003 and Kroll, Bobb & Hoshino 2014 for relevant reviews). Heritage speakers (HSs) constitute a unique type of bilingual with regard to their profile of (a) age of exposure and (b) experience and proficiency with their two languages. In particular, HSs generally have an early age of acquisition of both their home language (their L1) and the majority language (their second language or L2), with increasing exposure to the L2 upon entering schooling, and possibly reduced exposure to and concomitant proficiency in their L1 at this time. Thus, the investigation of HSs can shed light on how a number of age- and experiencerelated factors may influence language processing. In fact, multiple researchers (e.g., Montrul 2011, 2012; O’Grady, Kwak, Lee & Lee 2011) have pointed out that this population is uniquely poised in terms of their potential to answer theoretical questions about language acquisition. 203

Harriet Wood Bowden and Bernard Issa

Research has just begun to explore language processing in the brains of heritage speakers using neurophysiological measures, which can provide insights into real-time processing and neural representation of language. In this chapter, we will focus on three neurolinguistic methods that have been used to investigate Spanish as a heritage language either during language recognition or production tasks: functional magnetic resonance imaging (fMRI), functional near infrared spectroscopy (fNIRS) and event-related potentials (ERP). These techniques measure changes in blood-flow and electrical activity in the brain and shed light on the neural substrates and/or time course of language processing. These methods may thus be particularly informative as they provide insight into the mind-brain in ways that traditional behavioral linguistic assessments (e.g., acceptability judgment tasks, production tasks, recognition tasks) and psycholinguistic assessments (e.g., reaction time data) may not allow. The remainder of this chapter will be organized as follows. First, we will review current neurolinguistic research on Spanish as a heritage language. Within this review, we will briefly describe each of the neurolinguistic research methods that have been used to investigate this topic. Second, we will explore how neurolinguistic models of language acquisition and representation may account for HSs’ language processing based on the evidence from current research. Finally, we will discuss open questions for future research and discuss how neurolinguistic methodologies can be applied to important theoretical questions surrounding language acquisition generally and heritage language in particular.

Review of neurolinguistic research on Spanish as a heritage language Our initial literature search identified very few neurolinguistic studies that reported to investigate heritage language. We thus performed literature searches in LLBA and PsycINFO databases for “heritage” as well as “bilingual”-focused studies, together with one of several neurophysiological methods: fMRI, fNIRS, ERP, Positron Emission Tomography (PET), Magnetoencephalography (MEG) and Diffusion Tensor Imaging/Diffusion Magnetic Resonance Imaging (DTI/DMRI). We found that, when examining the descriptions of participants, several of the “bilingual” studies in fact appeared to investigate HSs of Spanish. We followed the characterization of Bolger and Zapata (2011: 1) in operationalizing HSs of Spanish as those who “are exposed to Spanish or both English and Spanish since birth or before the age of 5, but receive most of or all their schooling in English.” In cases that were unclear, we contacted the authors to determine whether the participants could be classified as HSs of Spanish. Thus, we were able to identify seven published studies that have investigated the processing of Spanish as a heritage language with neurolinguistic methods, and in addition we include an as yet unpublished study of our own. These studies used only three different neurophysiological methods: fMRI, fNIRS and ERP.

Functional Magnetic Resonance Imaging (fMRI) fMRI is a neuroimaging technique that reflects “changes in blood oxygenation and flow that occur in response to very localized changes in neural activity” (Rodden & Stemmer 2008: 62). Specifically, magnetic fields are used to measure hemodynamic changes known as blood oxygenation level dependent effects (the BOLD signal).1 fMRI has excellent spatial resolution (i.e., activation can be localized to regions of the brain on the order of millimeters), whereas its temporal resolution is on the order of seconds due to the time-course of the hemodynamic response to neural activity. Thus, the chief advantage of fMRI in language study is its ability to identify 204

Neurolinguistic approaches

regions that are involved in language processing. Typically, fMRI language studies use blocked designs, where tasks (or processing of particular types of stimuli) are performed over blocks of time, and brain activity over the entire block is measured and compared to activation either during rest or during a baseline task. Some fMRI research has also used event-related designs, in which it is possible to intermix stimuli or tasks of different types, and statistically separate and measure the activation related to each one. We were able to identify five fMRI studies that investigated the processing of Spanish in heritage speakers. Three of these examined language switching, one investigated singleword reading, and one studied sentence-level processing. In the first of the language switching studies, Hernandez, Martinez and Kohnert (2000) used fMRI to investigate language switching during a naming task in bilingual speakers who were in fact HSs of Spanish. Participants (4 female, 2 male) were right-handed bilinguals with a mean age of 23.5 (SD = 3.21)2 who learned both Spanish and English prior to age 5. They had on average 16 years of formal study of English, compared to 3 of Spanish. Participants’ vocabulary was assessed with the Boston Naming Test, which consists of 60 line drawings graded in difficulty, which participants are asked to name. They scored higher in English (mean = 54) than Spanish (mean = 40), and thus were deemed to be stronger in English than Spanish. In the study, participants performed a cued picture-naming task, in which a visual cue (the written word “say” in English or “diga” in Spanish) indicated which language participants should use to respond. The two conditions were blocked (cues all in one language) or mixed (alternating languages). Data were collected from pre-defined regions of the brain spanning the dorsolateral prefrontal cortex and the supramarginal gyrus. Analyses revealed no differences in areas activated or in intensity of activation in the blocked conditions (English vs. Spanish). The mixed condition, however, evidenced increased activation in the dorsolateral prefrontal cortex as compared to the blocked conditions. The authors interpreted this activation as being tied to language switching, which they viewed as part of a general executive attentional system.3 Hernandez (2009) also examined language switching during a naming task in a group of HSs of Spanish via fMRI. Participants in this study (5 female, 7 male) were right-handed and had a mean age of 21.4 (SD = 1.75). They had acquired Spanish as their first language and were exposed to English “upon entering school at age 5” (Hernandez 2009: 135). They scored higher on the Boston Naming Test in English (mean = 47) vs. Spanish (mean = 32), and thus were considered to be dominant in English. In this study, picture-naming was covert (i.e., silent), and analyses were not limited to pre-defined regions of interest, but rather a random effects analysis was employed across all scanned brain regions, which included nearly the entire brain (i.e., a much broader analysis as compared to the above-mentioned study). Results were taken to replicate and extend the previous findings through the examination of additional brain regions. In particular, there was increased activity in right dorsolateral prefrontal cortex during language switching, but this activation was found to extend farther (to the right precentral gyrus, the supplementary motor area, and the right superior parietal lobule). This finding was taken to indicate that not only the prefrontal cortex (implicated in executive function) but a larger network is in fact recruited for language switching. The authors suggested that these areas are involved in executive function as well as in articulatory and motor planning, and that they could be recruited in the switching task either to inhibit the non-target language or to activate the target language response, or both. Results also revealed differences in areas of activation between languages in blocked conditions, whereas in the prior study, no such differences were found (for English vs. Spanish, there was increased activity in right post-central gyrus, superior temporal gyrus, hippocampus, insula, and pre-supplementary motor areas; for Spanish vs. English, in left hippocampus, thalamus, amygdala, inferior frontal gyrus, and right anterior insula). 205

Harriet Wood Bowden and Bernard Issa

Note that increased activity for the L1 (Spanish) was largely left-lateralized whereas increased activity for the L2 (English) was right-lateralized. It was not entirely clear how to interpret these findings, but the authors suggested that there could be differences in sensorimotor learning based on age of acquisition, and that the L1 may be more closely linked to word knowledge and emotional content.4 A third fMRI study of language switching during a naming task was conducted by Weissberger, Gollan, Bondi, Clark and Wierenga (2015). Participants (15 females, 5 males) were right-handed and had a mean age of 20.45 (SD = 1.9). They were exposed to Spanish very early (mean age = 0.3 years, SD = 0.5) and English later (mean = 5.1 years, SD = 3.5). Based on reaction times and error rates for a picture naming task, 90% were classified as English dominant. By self-report, 70% considered themselves English dominant, 15% balanced, and 15% Spanish dominant. In this study, participants performed (a) a cued language-switching task in which they had to covertly name numbers (1–9) in Spanish or English (cued by a Mexican or American flag), and (b) a cued color-shape-switching task in which they had to name either the color or shape of a form (cued by a rainbow-colored rectangle or a series of shapes, respectively; all participants chose to carry out this task in English). The fMRI design was hybrid (both blocked and event-related). Analyses examined neural responses to trials of three types as compared to rest: (a) single trials (a trial in the blocked condition, where the task did not alternate), (b) stay trials (a trial of the same type as the preceding trial during a mixed block), and (c) switch trials (a trial of the opposite type as the preceding trial during a mixed block). Analyses also compared brain responses during the language task as compared to the color-shape task for these three trial types. Results revealed few differences in activation between responses to the language vs. color-shape task on switch and single trials, but large differences on stay trials, with the activation being more widespread for the color-shape than the language task. This was taken to suggest that (a) switching mechanisms are shared across linguistic and non-linguistic domains, and (b) bilinguals are more efficient at sustaining the inhibition of the non-target language (and thus recruit fewer neural resources) in the stay trials of the language task as compared to the stay trials in the color-shape task. The authors suggest that this efficiency might come from bilinguals’ need to suppress a non-target language when not switching languages.5 An fMRI study of single-word reading in HSs of Spanish was carried out by Meschyan and Hernandez (2006). Participants (7 females, 5 males) were right-handed, with an average age of 22.3 years (SD = 1.35) and had learned Spanish early (mean = 0.25 years, SD = 0.45) and English later (mean = 4.33 years, SD = 1.16). They had on average 15.6 years of formal study of English compared to 3.4 years in Spanish. Their average score on the Boston Naming Test was also higher in English (mean = 46.83, SD = 4.59) than Spanish (mean = 35.42, SD = 3.42). Self-assessments of language ability were also higher in English than Spanish, leading the authors to conclude that participants appeared more proficient in English than Spanish. In the experimental task, participants silently read blocks of words in Spanish or English while undergoing scanning. Results showed greater activation in Spanish (vs. English) of the articulatory motor areas in the right hemisphere (supplementary motor area/cingulate, insula, and putamen) and in phonological processing regions (left superior temporal gyrus). In English (vs. Spanish), there was greater activation in visual processing and word recoding regions, i.e., areas involved in converting letters to sounds (occipito-parietal border and inferior parietal lobe). The greater right hemisphere activation in Spanish was taken to reflect more effortful processes given that Spanish was participants’ weaker language. The differences in phonological and visual processing activation between languages was attributed to the orthographic transparency of Spanish (leading to more phonological processing) vs. the opacity of English (leading to more visual processing). Together these results suggest that there are differences in processing an HS’s two 206

Neurolinguistic approaches

languages, which may be related to differences in proficiency between the languages as well as to differences in the orthographic transparency of the languages.6 Kovelman, Baker and Petitto (2008) investigated processing of sentences with fMRI during a syntactic “sentence judgment task.” Participants, who were heritage speakers (the “bilingual” group; 7 females, 4 males), were right-handed and had a mean age of 19 years (range = 18–22 years). All were exposed to Spanish from birth, and to English between birth and age 5. All learned to read in English at school and in Spanish at home. Five participants had some formal bilingual education prior to university, and all had at least one formal Spanish class. Participants considered themselves to be equally fluent in both languages and equally active in both communities, and achieved at least 80% accuracy in both languages on a language screening task called the Language Competence/Expressive Proficiency test. In the study, participants read sentences in a blocked design (English in one block and Spanish in the other) and gave plausibility judgments after each sentence. The authors exploited specific differences in the languages, in particular differential reliance on word order vs. morphosyntax, to explore whether there were differences in neural responses to the two languages.7 Participants were told in which language each block would be presented (Spanish or English) in order to induce a monolingual processing mode, rather than a switching or bilingual mode. Stimuli included both plausible and implausible sentences with either object- or subject-relative clauses (referred to as OS and SO, respectively). It was predicted that if bilinguals have differentiated linguistic systems, they should show a smaller difference in brain activity between the two types of sentences in Spanish (where word order is less informative) than in English (where it is more informative, and where OS sentences are more common and should be easier to process than SO sentences). Results indicated that in English mode, there was indeed greater intensity of activation for SO (more difficult) than OS (easier) sentences. In Spanish mode, however, there were no differences in brain activity between SO and OS sentences. This evidence was taken to support the notion that these early bilinguals “have a differentiated neural pattern of activation for each language” (Kovelman, Baker et al. 2008: 166).8 In sum, the data from fMRI studies suggest that with regard to language switching tasks, HSs may engage executive control-related regions of the brain in order to activate the target language and/or inhibit the non-target language (Hernandez et al. 2000; Hernandez 2009). They may be especially efficient at sustaining inhibition of the non-target language as compared to inhibiting non-linguistic non-target responses (Weissberger et al. 2015), and this may be due to their experience with suppressing the non-target language when operating in a monolingual mode. With regard to brain activation during language processing, the research suggests that while there is apparently a large degree of overlap in areas of activation, HSs’ two languages may be differentiated on a fine-grained neurological level. This is suggested by the fact that HSs show differential intensity of activation for object- vs. subject-relative clauses in English but not in Spanish (Kovelman, Baker et al. 2008), and that they show differential activation between languages in articulatory motor regions, phonological processing regions and word recoding regions, which may be driven by either proficiency, orthographic transparency or both (Meschyan & Hernandez 2006).

Functional Near Infrared Spectroscopy (fNIRS) fNIRS (also referred to as optical imaging or optical topography) is a non-invasive technique that measures hemodynamic changes in the brain via sensors at the scalp. These sensors detect near infrared light transmitted through neural tissue from sources also placed at the scalp; the scattering and absorption of this light depends on hemodynamic changes that reflect brain 207

Harriet Wood Bowden and Bernard Issa

activity. The temporal resolution of fNIRS is more fine-grained than that of fMRI but less than that of ERPs (see later), whereas its spatial resolution is less precise than fMRI but more precise than that of ERPs (on the order of centimeters). Thus it can be considered a middle-of-the road technique with regard to both temporal and spatial resolution. Note, however, that fNIRS research usually focuses only on selected regions of the scalp near where the sources and sensors are placed. It is also noteworthy that fNIRS equipment is portable and (unlike either fMRI or ERP) amenable to movement by participants during the completion of experimental tasks (see Quaresima, Bisconti & Ferrari 2012 for a review of fNIRS used with language paradigms). We were able to identify only one study of Spanish as a heritage language employing this relatively new method (Kovelman, Shalinsky, Berens & Petitto 2008). Participants (6 females, 4 males) were right-handed and averaged 20 years of age. They had been consistently exposed to both English and Spanish by the age of 5, with half exposed to both Spanish and English at home from birth, and half exposed to only Spanish at home from birth and to English in daycare or kindergarten starting at 3–5 years of age. They learned to read in English from ages 5–7 and in Spanish from ages 5–12. All participants were screened for proficiency in English and Spanish on standardized grammaticality judgment tasks in each language, and all were deemed to have “high, monolingual-like, language proficiency in each of their two languages” (Kovelman, Shalinsky et al. 2008: 1459), having scored at least 80% on each. In the study, these bilingual participants were tested (as were a control group of monolingual English speakers) on semantic processing with a version of the Pyramids and Palm Trees Task. In this task, each trial presented a sequence of two words, followed by a picture that corresponded to either the first or second word; participants responded via button-press to indicate which word matched the picture.9 There were four blocked conditions: monolingual English, monolingual Spanish, bilingual with language integration (one word in Spanish and the other word in English within each trial) and bilingual with language alternation (languages alternated randomly between trials but only one language within each trial). Results indicated that in monolingual mode, HSs (as compared to monolingual English speakers) showed greater intensity of activation in areas tied to verbal working memory and attention (dorsolateral prefrontal cortex and inferior frontal cortex). Moreover, within the HS group, in bilingual mode (as compared to monolingual mode), results indicated that there was greater recruitment of these same structures in the right hemisphere. These results were taken to support a view in which bilinguals recruit neural resources differentially from monolingual speakers, and in particular recruit regions that have previously been tied to language switching. Interestingly, this increased activation was observed in HSs not just in bilingual mode as compared to monolingual mode, but also in monolingual mode as compared to monolingual speakers.

Event-Related Potentials (ERP) During human cognition, there are electrical signals embedded within the electroencephalogram (i.e., recordings of electrical activity at the scalp) that are affected by language processing. When these signals are time-locked to specific stimuli and averaged over numerous similar trials, the brain’s electrophysiological response (i.e., changes in the brain’s electrical activity) to a specific cognitive event can be observed. These time-locked signals are known as event-related potentials (ERPs), and the characteristic positive and negative potentials produced are known as components (see Handy 2004 and Luck 2014 for overviews of the ERP technique). ERPs are particularly well-suited for examining language processing as this method has very precise temporal resolution (on the order of milliseconds). Moreover, various components have been identified in L1 ERP research as being reliably elicited by specific types of language processing, 208

Neurolinguistic approaches

though these components are not unique to language (see Kutas & Federmeier 2011; Steinhauer & Connolly 2008 for reviews on monolingual language processing).10 The processing of lexical/semantic information is usually indexed by a negative potential that occurs approximately 400 ms after the onset of a stimulus (such as a semantic anomaly) and is typically maximal at central and parietal regions of the scalp (the N400; Kutas & Federmeier 2011; Kutas & Hillyard 1980). The brain’s most commonly reported electrophysiological response to the processing of grammatical information is a positive potential occurring approximately 600 ms after the onset of a grammatical anomaly and is typically largest at central and posterior regions of the scalp (the P600; Osterhout & Holcomb 1992). This component may or may not be preceded by an earlier left and/or anterior negativity (the LAN) and may additionally be accompanied by a later anterior negativity (Neville, Nicol, Barss, Forster & Garrett 1991). Generally, research with bilinguals has shown that the higher the level of proficiency and the more experience with a language, the more L1-like their ERP signatures for grammatical conditions are, whereas ERP signatures in response to semantic errors are L1-like from much earlier on in the learning process (McLaughlin et al. 2010; Steinhauer, White & Drury 2009). We identified one published ERP study of Spanish as a heritage language in our search, and in this chapter, we also report a recent study of our own. Moreno and Kutas (2005) recorded ERPs to examine semantic processing in the dominant and non-dominant languages of two groups of Spanish-English bilinguals. One of these groups, the English-dominant group as established via the Boston Naming Test, was in fact a group of HSs of Spanish (personal communication). These participants (15 females, 5 males; right-handed except for one ambidextrous participant) had a mean age of 23 years (SD = 3). They had relatively early exposure to both Spanish and English, with 17 reporting exposure to Spanish at birth and all exposed to English by age 6. All underwent the majority of their schooling in English (personal communication). Dominance was operationalized as having a difference in scores between the two languages of at least 8 on the Boston Naming Test. Participants in this group scored on average 29 (SD = 8) in Spanish and 51 (SD = 6) in English (mean difference = 22; range of difference scores = 11–42) thus all participants in this group were deemed to be English-dominant. Participants also reported continued use of Spanish approximately 44% of the time with family and friends. In this study, participants completed a blocked sentence-reading task, in which they read a series of semantically congruent and incongruent sentences for comprehension (first in one language and then the other). Analyses found that the N400 peak latency (i.e., the amount of time from stimulus onset to the peak amplitude of this component) was delayed in the bilinguals’ less dominant language (Spanish) as compared to their more dominant language (English). Surprisingly, these results suggest that despite having exposure to Spanish from birth, HSs’ time course of semantic processing was unlike that of monolinguals, as reflected by the delay in N400 peak amplitude. This delay is typically found in L2 learners who are exposed to their L2 relatively later in life. The authors take this as evidence that age of acquisition alone cannot explain differences in processing shown in this group. Moreover, correlations evidenced between vocabulary proficiency and N400 peak latency suggested that proficiency may also play a crucial role in the latency of the integration of semantic information. In a recent ERP study (Bowden, Issa & Morgan-Short 2015), we examined processing of both lexical and grammatical structures in the Spanish of HSs. Participants (11 females, 8 males) were enrolled in a university-level heritage language class. All were right-handed and their mean age was 21.1 years (SD = 3.1). They reported having acquired Spanish very early (mean = 0.6 years, SD = 0.8) and English a bit later (mean = 4.1 years, SD = 2.8). Most (n = 15) considered themselves dominant in English, and all had undergone the majority of their schooling in English. On average they rated themselves higher in speaking, understanding 209

Harriet Wood Bowden and Bernard Issa

and reading English (mean = 8.8 out of 10) as compared to Spanish (mean = 7.2). In the experimental task, participants read Spanish sentences (presented one word at a time) and made acceptability judgments. The stimuli consisted of intermixed correct and violation sentences of various types, including the following conditions: phrase structure and semantic conditions (taken from Bowden, Steinhauer, Sanz & Ullman 2013) and gender and number agreement conditions (taken from Bowden 2007). All sentences were constructed in a balanced paradigm following Steinhauer et al. 2009. These conditions are exemplified in Table 14.1. We included 40 sentences each for the phrase structure and semantic conditions, and 32 each for the gender and number agreement conditions (equally divided among masculine, feminine, singular and plural critical items). EEGs were recorded during the sentence-reading task, across 32 electrodes and at a 512 Hz sampling rate, with impedances kept below 5 kΩ. ERPs were time-locked to the onset of the target word (underlined in Table 14.1) and a −200 ms baseline correction was applied.11 Time windows for analyses were determined by visual inspection of grand average ERPs. We performed standard analyses (ANOVAs with step-down analyses) across the whole group (grand averages) and additionally explored individual differences based on response dominance (following Tanner, Mclaughlin, Herschensohn & Osterhout 2013) in early and late time windows. Behavioral responses indicated that as a group, all conditions were above chance level (d’s of 1.87 for semantic, 2.33 for phrase structure, 1.06 for gender and 0.87 for number). Regarding the ERP signatures, the semantic condition elicited an N400 (300–500 ms; see Figure 14.1). This result is consistent with what was found by Bowden et al. (2013) for L1 Spanish as well as for L2 speakers at low and advanced levels of experience and proficiency. For phrase structure, there was an N400 (300–600 ms) followed by a small P600 (900–1100 ms) across the group; however, the P600 was significant at only one electrode. No significant ERP effects were found for gender, while for number, there was only a small negativity (400– 600 ms) in the centro-anterior region. Note, these responses to grammatical violations do not correspond to typical L1- or L2-like responses. Subsequently, following Tanner et al. (2013), we examined individual variability in ERP responses. We characterized participants in terms of their effect magnitude (i.e., size of the effect, whether negative or positive) and response dominance (i.e., more negative or positive) over a specified centro-parietal region of interest in earlier and later time windows (see Table 14.2).12 Table 14.1  Sample stimuli sentences Condition

Violation

Correct

Semantic

La profesora espera ir en autobús a la #semana. ‘The professor hopes to go by bus to the #week.’

Phrase structure

Tengo que ∗millas muchas correr esta semana. ‘I have to ∗miles many run this week.’

Gender

El lago es ∗tranquila por la mañana. ‘The lakems is tranquilfs in the morning.’

Number

El lago es ∗tranquilos por la mañana. ‘The lakems is tranquilmp in the morning.’

210

Tengo muchas millas que correr esta semana. ‘I have many miles to run this week.’

El lago es tranquilo por la mañana. ‘The lakems is tranquilms in the morning.’

Figure 14.1  E  RPs for semantic condition (grand mean across all participants). Waveforms show ERPs to correct (solid line) and violation (dashed line) critical words recorded at electrode Pz, with negative plotted up. The arrow points to significant differences between responses to correct and violation critical words. Scalp maps show potentials (violation-correct) across early and late time windows. Table 14.2  Time windows for individual ERP analyses Condition

Early

Late

Phrase structure Gender Number

300–600 ms 400–600 ms 400–600 ms

900–1100 ms 900–1100 ms 900–1100 ms

Figure 14.2  S catterplots showing the distribution of N400 and P600 effect magnitudes across participants, averaged over a centro-parietal region of interest, for each condition (semantic, phrase structure, gender, number). Each dot represents a participant, with P600 magnitude (later time window) on the x-axis and N400 magnitude (earlier time window) on the y-axis. The best-fit line from a regression analysis is represented by a solid line, whereas the dashed line represents equal responses. Participants to the right of the dashed line are positive-dominant and to the left, negative-dominant.

Harriet Wood Bowden and Bernard Issa

Figure 14.3  N  egative- and positive-dominant individuals by grammatical condition (phrase structure, gender, number). Waveforms show ERPs to correct (solid line) and violation (dashed line) critical words recorded at electrode Cz, with negative plotted up. Arrows point to significant differences between responses to correct and violation critical words. Scalp maps represent potentials (violation-correct) across early and late time windows.

As can be seen in Figure 14.2, in the semantic condition, the N400 effect magnitude (earlier time window) did not correlate with the P600 effect magnitude (later time window). In the three grammatical conditions, however, there was a negative correlation—that is, the larger a person’s N400 effect in the early time window, the smaller their P600 effect in the later time window tended to be, or vice-versa.13 In addition, it can be observed that the dashed line (the dividing line between a negative or positive response dominance), divides the participants roughly in half for all conditions except the semantic condition, in which all but two participants have predominantly negative responses. This indicates that there was significant variability in response dominance in the three grammatical conditions, but very little in the semantic condition. Thus we separately explored the responses of the positive- and negative-dominant participants for each grammatical condition.14 In doing so, a very different picture emerged (see Figure 14.3). In brief, within each of the grammatical conditions (phrase structure, gender, number), the negative-dominant participants

212

Neurolinguistic approaches

showed a significant N400 and the positive-dominant participants a significant P600. The N400 extended into the later time window in the gender and number conditions. Additionally, in the gender condition, the positivity started in the early time window as well. This pattern of results suggests that except for the semantic condition, there were underlying individual differences that were obscured by group averages. Recent research by Tanner and Van Hell (2014) has revealed that even within L1, morphosyntactic processing may show similar variability. Our results extend this finding to the heritage language context. From the two ERP studies presented above, we can draw some preliminary conclusions regarding heritage language processing of Spanish at the sentence level. First, there is evidence from both studies that lexical/semantic processes in heritage language rely on the same neurocognitive systems as monolinguals and L2 learners (as seen in findings from previous studies), given that processing semantically incongruent stimuli elicits N400s in both studies. There is also evidence that lexical/semantic processing in real time may be delayed in the heritage language as compared to HSs’ more dominant L2 (Moreno & Kutas 2005). In addition, there is evidence of variability in processing among HSs for grammar (Bowden et al. 2015). It remains to be seen to what extent this variability and the factors that may account for it (e.g., proficiency, experience, cognitive resources) differ from or correspond to variability in L1 and L2. In sum, it appears that from the fMRI and fNIRS research reviewed earlier that HSs exhibit a high degree of overlap in neural substrates that underlie their two languages, even while they show some differential activation of brain regions when using one language as compared to the other. Moreover, they show increased activation of particular regions of the brain apparently related in part to executive function when switching between their two languages. In fact, HSs may be particularly adept at inhibiting a non-target language as compared to non-linguistic inhibition. These findings echo previous research on simultaneous bilinguals, indicating that bilingualism has positive effects on cognitive function, specifically with regard to inhibition (Kroll & Sunderman 2003). In the ERP studies reviewed earlier, HSs showed neural signatures in Spanish that have been reported previously for both L1 and L2 processing, although the time-course of their lexical processing may be somewhat delayed. Together, these results suggest that HSs of Spanish resemble native monolinguals in some ways but also differ from them in fine-grained ways in the processing of their heritage language. At least some of these differences appear to be related to management of two languages in the brain.

Theoretical models Although the studies we have summarized are all related in that they investigated heritage language processing through neurolinguistic methods, they differ in their broader conceptualization of how language is represented and processed in the brain and the type of research questions they aim to answer. Examining the theoretical models that motivated these studies can provide insights into how to better understand and contextualize the results reported, as well as yield questions to be asked in future research. At least in broad strokes, the fMRI (Hernandez 2009; Hernandez et al. 2000; Kovelman, Baker et al. 2008; Meschyan & Hernandez 2006; Weissberger et al. 2015) and fNIRS (Kovelman, Shalinsky et al. 2008) studies reported earlier appear to provide evidence consistent with the cognitive control model (Abutalebi 2008; Abutalebi & Green 2007, 2008). This model suggests that while bilinguals largely engage similar neural substrates in L1 and L2 processing, bilinguals may engage more extended neural areas in L2 processing, especially at lower levels of proficiency, because increased cognitive control is necessary for processing a language that is weaker. Thus,

213

Harriet Wood Bowden and Bernard Issa

this interpretation holds that larger areas of activation observed for L2 processing (especially in the left prefrontal cortex) are not due to the need to engage more areas for linguistic processing per se, but rather due to increased activation of brain regions involved in controlling the language networks when using the weaker language. This control seems to include not only activation of the weaker language but inhibition of the stronger language, which would be especially plausible in contexts in which both languages are active, such as in language-switching tasks. A more recent version of the cognitive control model, called the adaptive control hypothesis (Green & Abutalebi 2013), makes additional predictions about how control mechanisms are activated based on the interactional context that a speaker most often encounters. This model distinguishes between three interactional contexts: (a) the single-language context, in which a speaker uses one language in one environment (e.g., at work) and their other language in a different environment (e.g., at home); (b) the dual-language context, where “both languages are used but typically with different speakers” (Green & Abutalebi 2013: 518); and (c) the dense code-switching context, “in which speakers routinely interleave their languages in the course of a single utterance,” (Green & Abutalebi 2013: 518). HS data may be able to directly test predictions made by this hypothesis as HSs may fall into any one of the above-mentioned interactional contexts depending on their linguistic profile and the linguistic community to which they belong. In addition, this model could potentially make predictions that would differ depending on other task and speaker characteristics. For example, on tasks that HSs are more adept at performing (for example oral vs. written or less vs. more metalinguistic), the model would likely expect less engagement of these control networks as compared to tasks on which they are less proficient. These are questions that future research on heritage language could address by employing neurophysiological methods. Another prominent theoretical model that may lend itself to interpreting the results of some of the studies summarized earlier, as well as to making novel predictions with regard to heritage language, is the Declarative/Procedural model (see Ullman 2001, 2005, 2015a, 2015b for more details about the model itself as well as for reviews of evidence supporting the model). This neurobiologically informed model of language generally holds that two memory systems, declarative and procedural memory, underlie much of language learning, processing and use. In particular, the model posits that idiosyncratic knowledge of language (including lexical knowledge and knowledge of irregular forms) must rely on declarative memory, where learning depends largely on the hippocampus, but once consolidated, knowledge may come to rely on related cortical areas. Grammatical knowledge (especially knowledge of rules and sequences), on the other hand, may rely on procedural memory, which is based in frontal and basal ganglia circuits in the brain and is taken to be a compositional system. Because learning in declarative memory is relatively fast, learning of both lexical and grammatical forms may initially depend largely on declarative memory, for both children and adults. However, with increasing experience and proficiency with a language, procedural knowledge of the grammar gradually builds up and grammatical processing may come to rely on the procedural system. Importantly, learning in procedural memory is better in children than adults. Thus, it should be easier to acquire grammar procedurally in early-learned L1 than later-learned L2, and later-learned languages may tend to rely more on declarative memory even for knowledge that in L1 may rely on procedural memory. The Moreno and Kutas (2000) and Bowden et al. (2015) studies are both consistent with predictions that this model would make, although the model has not been specifically extended to heritage language and it is not yet entirely clear to what degree ERP components directly reflect declarative or procedural memory systems. In both studies, semantic conditions reveal N400 responses, which have been argued to depend on declarative memory (Ullman 2004). 214

Neurolinguistic approaches

Furthermore, the results from Bowden et al. (2015) highlight the variability among HSs in their neural signatures for the processing of grammar, which may reflect, at least in part, variability in reliance on declarative and procedural memory systems for this type of processing, which is within the scope of the Declarative/Procedural model. Whether this variability in ERP responses in fact reflects differential reliance on declarative and procedural memory systems remains to be investigated. Moreover, the model could make predictions with regard to factors that should explain the variability in responses, in particular, age, amount and type of exposure to the target language, individual differences in declarative and procedural memory, and sex, among others. Bolger and Zapata (2011) have explored how the Declarative/Procedural model could be extended to the case of heritage language. Two key issues in heritage language that they consider are those of attrition and incomplete acquisition of the often less-used L1. As they point out, L1 attrition could be explained by this model especially in the case of declarative knowledge, since it is more susceptible to changes in frequency in the input (e.g., reduced exposure to L1 for HS speakers). Moreover, incomplete acquisition of grammar could also be predicted by this model, if, for example, not all aspects of the heritage grammar have been proceduralized when the amount of input is drastically reduced, which might happen when heritage learners of Spanish begin attending school in English and spend less time immersed in Spanish. Thus, for example, as Bolger and Zapata suggest, perhaps only “core” aspects of grammar have been acquired at this point, but not “peripheral” aspects (which might lead to incomplete acquisition, another key issue in heritage language). Specific predictions would depend on when particular aspects of grammar are typically learned as well as when exposure to English begins, and whether this reduces exposure to Spanish and inhibits L1 learning. Moreover, if, as they suggest, procedural knowledge is less susceptible to environmental changes than declarative knowledge, then grammar that is acquired and proceduralized early should be less susceptible to attrition than either grammar that is acquired later and not proceduralized, or declarative knowledge. Thus, we might expect that early-acquired grammar would be L1-like in its neural representation and signatures, whereas neural responses to later-acquired grammar (i.e., that which might show incomplete acquisition) might be less like that of monolinguals and might show increased reliance on declarative memory. In addition, if peripheral aspects of grammar depend more on semantics (i.e., grammar governed by the syntax-semantics interface) and thus rely more on declarative memory, then for this reason in addition, peripheral aspects of grammar might be more susceptible to erosion (and thus might not be native-like in terms of neural processing). Thus, this model may predict both lexical and grammatical attrition and/or incomplete acquisition, with lexical knowledge generally being more susceptible to such effects than grammatical knowledge, but with later-acquired grammar being more susceptible than early-acquired grammar. These predictions could be tested through the use of neurophysiological research methods.

Future directions for research In addition to testing predictions of the above models as extended to the heritage language context, there are many fruitful areas of neurolinguistic research that would allow us not only to better understand this unique population but also to better understand language acquisition and processing more generally. Here we provide some suggestions for such future neurolinguistic research. To begin with, heritage languages can be more fully investigated in their own right in order to understand how they are acquired, represented and processed. Neurolinguistic methods such as ERP, fMRI and fNIRS could also be used to explore whether HSs process language similarly to or different from both L1 and L2 speakers of different types. Additionally, they 215

Harriet Wood Bowden and Bernard Issa

could provide a window for investigating theoretical questions about what aspects of language have undergone attrition vs. incomplete acquisition, questions that have so far been examined exclusively with behavioral and psycholinguistic techniques (see Montrul 2008). These questions could be examined by testing HSs’ neural signatures of processing at different points in their lifespan (thus probing what has been acquired and may be subject to attrition) as well as by comparing HSs to matched monolingual Spanish speakers and balanced bilinguals. In addition, HSs may contribute to our understanding of the roles of age of acquisition and proficiency in language processing. In the field of second language acquisition, it is widely accepted that age of acquisition has an important effect on L2 processing and performance. However, the case has been made that other factors such as L2 use, L2 proficiency or L1–L2 pairing may in fact be more important than age of acquisition in affecting whether or not L2 learners demonstrate native-like language processing or performance (e.g., Birdsong & Molis 2001). That said, there are limitations to investigating adult L2 learners alone in probing these questions. By definition, L2 learners have a later age of acquisition, thus limiting researchers’ ability to investigate the age factor. Furthermore, L2 speakers with native-like proficiency are not abundantly available for research in foreign-language learning contexts. HSs, on the other hand, have early experience with both languages, and often high levels of proficiency in their heritage language, though it may be lower than that of monolingually raised native speakers. Thus the study of HSs using neurophysiological methods would elucidate how such factors as dominance, use and proficiency affect language processing while holding constant an early age of acquisition. Another avenue for future research to pursue is to employ neurophysiological methods to investigate how speaker and task characteristics together influence processing. In the case of HSs, research could explore why HSs tend to show advantages on oral and more implicit tasks as compared to written and more metalinguistic tasks by investigating neurocognition during such tasks. Moreover, these methods could shed light on questions related to cognitive control. For example, some previous ERP studies with simultaneous bilinguals (but not HSs to our knowledge) have identified modulations in specific ERP components that appear to be related to language switching, and thus might be tied to cognitive control (e.g., Jackson, Swainson, Cunnington & Jackson 2001, and Proverbio, Leoni & Zani 2004). Future research could investigate this issue in HSs of Spanish. Besides using the above-mentioned neurophysiological methods in new ways to investigate heritage language processing and language acquisition questions more generally, there are a number of additional neurolinguistic methods that, to our knowledge, have not been used to study the processing and neural correlates of Spanish as a heritage language, such as MEG, DTI or PET (see Rodden & Stemmer 2008 for a review of these methods). While each neurophysiological method has unique strengths and weaknesses, the main drawback of many of them is cost. In many cases, interdisciplinary research collaborations can provide language researchers with access to needed technologies and expertise to analyze heritage language processing. Ultimately, neurolinguistic data on heritage language processing should provide complementary and converging evidence to that of behavioral and psycholinguistic methods to gain a more holistic understanding of heritage language in all its complexity.

Notes 1 For discussion of the interpretation of the BOLD signal, see Attwell and Iadecola (2002), Ekstrom (2010) and Raichle (1998). 2 Note that the amount and precision of statistical information presented in this chapter reflects what was reported in each study, for example whether or not standard deviations were included, and the number of decimal places reported.

216

Neurolinguistic approaches 3 In a behavioral version of the task, performed prior to fMRI and with auditory cues (‘say’ or ‘diga’), reaction times (RTs) were measured. Participants responded faster in English (in both blocked and mixed conditions), whereas overall, they were faster on blocked than mixed conditions. Thus it was faster for HSs to access English (and/or cost them more time to suppress English while naming in Spanish), and language switching was also costly in terms of RT. 4 In this study, the RT results differed from the previous study; these HSs showed faster (and more accurate) naming in Spanish than English. Also, though the RT difference between mixed and blocked conditions was in the expected direction, it was not statistically significant. It is not clear what these differences should be attributed to. There might be differences in the dominance and/or proficiency of the participants between the two studies, or there may have been methodological differences in the naming task and/or stimuli. 5 RT analyses of an overt task performed prior to the fMRI task showed a cost for both mixed and switch trials, with a slightly greater cost for switching in the language task than the color-shape task. 6 Participants were slower at naming in Spanish on a behavioral version performed prior to fMRI. 7 They also investigated differences between monolinguals’ and bilinguals’ processing of English, which is not discussed here. 8 Analyses of RTs to the judgment task showed that HS were slower in Spanish than English overall, and were slower on OS sentences in Spanish than in English. In addition, they were faster (like English monolinguals) on plausible than implausible sentences. 9 The task was done both in auditory and visual modes, which were combined in analyses. 10 Note, however, that ERPs are only measured at the scalp. Although components have characteristic scalp distributions and functional interpretations have been made for many components based on the processes that elicit them, there is no straightforward way to localize the underlying source(s). For more information, see especially chapter 2 of Nunez and Srinivasan (2006). 11 An IIR Butterworth filter with a highpass frequency of .1 Hz and a low-pass frequency of 20 Hz was applied to the data offline, and trials contaminated with artifacts were excluded from analysis. 12 In addition, we performed all analyses (both grand mean and for response dominance) for each condition with a consistent early time window (300–600 ms) for the sake of comparability across conditions. The pattern of results was almost identical to that reported above. In the grand mean analyses with this early time window, the same components were evidenced, except that under the new analyses, the small early negativity in response to number violations was not significant. 13 Results from the correlation analysis examining the relationship between the N400 and the P600 effect magnitudes with the consistent early time window also produced an identical pattern of results, with the semantic condition yielding a non-significant correlation and the grammatical conditions resulting in significant negative correlations. 14 With regard to response dominance, analyses with the consistent early time window classified participants into the same groups as previously, except that one additional participant was classified as positive-dominant for semantic violations. Thus we again performed analyses by response dominance groups for all conditions except semantic, and again, the same components were evidenced, with one exception: the positivity for the number condition extended into the earlier time window.

Further reading Jegerski, J., & VanPatten, B. (2013). Research Methods in Second Language Psycholinguistics. London and New York: Routledge. Handbook that introduces researchers to different psycho- and neurolinguistic research methods. See chapter 6 (Morgan-Short and Tanner on ERP research design and methods) and chapter 7 (Newman on fMRI research design and methods) for relevant neurolinguistic methods. Luck, S. J. (2014). An Introduction to the Event-Related Potential Technique. Cambridge, MA: MIT press. Introduction to the fundamentals of ERP research. Stemmer, B., & Whitaker, H. A. (Eds.). (2008). Handbook of the Neuroscience of Language. Waltham, MA: Academic Press. Handbook with accessible introductions and reviews of various neurophysiological methods used to examine language processing.

217

Harriet Wood Bowden and Bernard Issa VanPatten, B., & Williams, J. (Eds.). (2014). Theories in Second Language Acquisition, 2nd ed. New York: Routledge. Includes chapter on Declarative/Procedural model of Ullman as well as other theories of second language acquisition that may be relevant to the heritage context.

References Abutalebi, J. (2008). Neural aspects of second language representation and language control. Acta Psychologica, 128, 466–478. Abutalebi, J., & Green, D. (2007). Bilingual language production: The neurocognition of language representation and control. Journal of Neurolinguistics, 20, 242–275. Abutalebi, J., & Green, D. (2008). Control mechanisms in bilingual language production: Neural evidence from language switching studies. Language and Cognitive Processes, 23(4), 557–582. Attwell, D., & Iadecola, C. (2002). The neural basis of functional brain imaging signals. Trends in Neurosciences, 25(12), 621–625. Binder, J. R., Desai, R. H., Graves, W. W., & Conant, L. L. (2009). Where is the semantic system? A critical review and meta-analysis of 120 functional neuroimaging studies. Cerebral Cortex, 19(12), 2767–2796. Birdsong, D., & Molis, M. (2001). On the evidence for maturational constraints in second-language acquisition. Journal of Memory and Language, 44(2), 235–249. Bolger, P. A., & Zapata, G. C. (2011). Psycholinguistic approaches to language processing in heritage speakers. Heritage Language Journal, 8(1), 1–29. Bowden, H. W. (2007). Proficiency and second-language neurocognition: A study of Spanish as a first and second language. Unpublished doctoral dissertation, Georgetown University, Washington, DC. Bowden, H. W., Issa, B., & Morgan-Short, K. (2015, March). Brain processing of heritage language: Group and individual profiles as revealed by ERPs. Paper presented at the meeting of the American Association for Applied Linguistics (AAAL), Toronto, Canada. Bowden, H. W., Steinhauer, K., Sanz, C., & Ullman, M. T. (2013). Native-like brain processing of syntax can be attained by university foreign language learners. Neuropsychologia, 51(13), 2492–2511. Dronkers, N. F., Plaisant, O., Iba-Zizen, M. T., & Cabanis, E. A. (2007). Paul Broca’s historic cases: High resolution MR imaging of the brains of Leborgne and Lelong. Brain, 130(5), 1432–1441. Ekstrom, A. (2010). How and when the fMRI BOLD signal relates to underlying neural activity: The danger in dissociation. Brain Research Reviews, 62(2), 233–244. Friederici, A. D. (2002). Towards a neural basis of auditory sentence processing. Trends in Cognitive Sciences, 6(2), 78–84. Handy, T. C. (2004). Event-Related Potentials: A Methods Handbook. Cambridge, MA: MIT Press. Hernandez, A. E. (2009). Language switching in the bilingual brain: What’s next? Brain & Language, 109, 133–140. Hernandez, A. E., Martinez, A., & Kohnert, K. (2000). In search of the language switch: An fMRI study of picture naming in Spanish–English bilinguals. Brain and language, 73(3), 421–431. Indefrey, P. (2011). The spatial and temporal signatures of word production components: A critical update. Frontiers in Psychology, 2, Article 225. Ingram, J. C. (2007). Neurolinguistics: An Introduction to Spoken Language Processing and its Disorders. Cambridge, UK: Cambridge University Press. Jackson, G. M., Swainson, R., Cunnington, R., & Jackson, S. R. (2001). ERP correlates of executive control during repeated language switching. Bilingualism: Language and Cognition, 4(2), 169–178. Kaan, E., & Swaab, T. Y. (2002). The brain circuitry of syntactic comprehension. Trends in Cognitive Sciences, 6(8), 350–356. Kovelman, I. Baker, S. A., & Petitto, L. (2008). Bilingual and monolingual brains compared: A functional magnetic resonance imaging investigation of syntactic processing and a possible ‘neural signature’ of bilingualism. Journal of Cognitive Neuroscience, 20(1), 153–169. Kovelman, I., Shalinsky, M. H., Berens, M. S., & Petitto, L. A. (2008). Shining new light on the brain’s ‘bilingual signature’: A functional near infrared spectroscopy investigation of semantic processing. NeuroImage, 39(3), 1457–1471.

218

Neurolinguistic approaches Kroll, J. F., Bobb, S. C., & Hoshino, N. (2014). Two languages in mind bilingualism as a tool to investigate language, cognition, and the brain. Current Directions in Psychological Science, 23(3), 159–163. Kroll, J. F., & Sunderman, G. (2003). Cognitive processes in second language learners and bilinguals: The development of lexical and conceptual representations. In C. J. Doughty & M. H. Long (Eds.), The Handbook of Second Language Acquisition, 104–129. Malden, MA: Blackwell. Kutas, M., & Federmeier, K. D. (2011). Thirty years and counting: Finding meaning in the N400 component of the event related brain potential (ERP). Annual Review of Psychology, 62, 621–647. Kutas, M., & Hillyard, S. A. (1980). Event-related brain potentials to semantically inappropriate and surprisingly large words. Biological Psychology, 11(2), 99–116. Lee, C. L., & Federmeier, K. D. (2012). In a word: ERPs reveal important lexical variables for visual word processing. In Faust, M. (Ed.), Blackwell Handbooks of Behavioral Neuroscience: Handbook of the Neuropsychology of Language (184–208). Hoboken, NJ: Wiley-Blackwell. Luck, S. J. (2014). An Introduction to the Event-Related Potential Technique. Cambridge, MA: MIT press. McLaughlin, J., Tanner, D., Pitkanen, I., Frenck-Mestre, C., Inoue, K., Valentine, G., et al. (2010). Brain potentials reveal discrete stages of L2 grammatical learning. Language Learning, 60(1), 123–150. Meschyan, G., & Hernandez, A. E. (2006). Impact of language proficiency and orthographic transparency on bilingual word reading: An fMRI investigation. NeuroImage, 29, 1135 – 1140. Montrul, S. A. (2008). Incomplete Acquisition in Bilingualism: Re-examining the Age Factor (Vol. 39). Amsterdam: John Benjamins Publishing. Montrul, S. A. (2011). The linguistic competence of heritage speakers. Studies in Second Language Acquisition, 33, 155–161. Montrul, S. A. (2012). Is the heritage language like a second language? EUROSLA Yearbook, 12, 1–29. Moreno, E. M., & Kutas, M. (2000). Processing semantic anomalies in two languages: An electrophysiological exploration in both languages of Spanish-English bilinguals. Cognitive Brain Research, 22, 205–220. Neville, H., Nicol, J. L., Barss, A., Forster, K. I., & Garrett, M. F. (1991). Syntactically based sentence processing classes: Evidence from event-related brain potentials. Journal of Cognitive Neuroscience, 3(2), 151–165. Nunez, P. L., & Srinivasan, R. (2006). Electric Fields of the Brain: The Neurophysics of EEG. Oxford, UK: Oxford University Press. O’Grady, W., Kwak, H. Y., Lee, O. S., & Lee, M. (2011). An emergentist perspective on heritage language acquisition. Studies in Second Language Acquisition, 33(2), 223–245. Osterhout, L., & Holcomb, P. J. (1992). Event-related brain potentials elicited by syntactic anomaly. Journal of Memory and Language, 31(6), 785–806. Poeppel, D., & Hickok, G. (2004). Towards a new functional anatomy of language. Cognition, 92(1), 1–12. Proverbio, A. M., Leoni, G., & Zani, A. (2004). Language switching mechanisms in simultaneous interpreters: an ERP study. Neuropsychologia, 42(12), 1636–1656. Quaresima, V., Bisconti, S., & Ferrari, M. (2012). A brief review on the use of functional near-infrared spectroscopy (fNIRS) for language imaging studies in human newborns and adults. Brain and Language, 121(2), 79–89. Raichle, M. E. (1998). Behind the scenes of functional brain imaging: A historical and physiological perspective. Proceedings of the National Academy of Sciences USA, 95, 765–772. Rodden, F. A., & Stemmer, B. (2008). A brief introduction to common neuroimaging techniques. In B. Stemmer & H. Whitaker (Eds.), Handbook of the Neuroscience of Language (pp. 57–65). Oxford, UK: Elsevier. Ross, E. D. (2010). Cerebral localization of functions and the neurology of language: Fact versus fiction or is it something else? The Neuroscientist, 16, 222–243. Steinhauer, K., & Connolly, J. F. (2008). Event-related potentials in the study of language. In B. Stemmer & H. Whitaker (Eds.), Handbook of the Neuroscience of Language (pp. 91–104). Oxford, UK: Elsevier. Steinhauer, K., White, E. J., & Drury, J. E. (2009). Temporal dynamics of late second language acquisition: Evidence from event-related brain potentials. Second Language Research, 25(1), 13–41. Tanner, D., Mclaughlin, J., Herschensohn, J., & Osterhout, L. (2013). Individual differences reveal stages of L2 grammatical acquisition: ERP evidence. Bilingualism: Language and Cognition, 16, 367–382. Tanner, D., & Van Hell, J. G. (2014). ERPs reveal individual differences in morphosyntactic processing. Neuropsychologia, 56, 289–301.

219

Harriet Wood Bowden and Bernard Issa Ullman, M. T. (2001). The neural basis of lexicon and grammar in first and second language: The declarative/procedural model. Bilingualism: Language and Cognition, 4, 105–122. Ullman, M. T. (2004). Contributions of memory circuits to language: The declarative/procedural model. Cognition, 92(1–2), 231–270. Ullman, M. T. (2005). A cognitive neuroscience perspective on second language acquisition: The declarative/procedural model. In Sanz, C. (Ed.), Mind and Context in Adult Second Language Acquisition: Methods, Theory and Practice (pp. 141–178). Washington, DC: Georgetown University Press. Ullman, M. T. (2015a). The declarative/procedural model: A neurobiologically motivated theory of first and second language. In B. Van Patten and J. Williams (Eds.), Theories in Second Language Acquisition, 2nd ed. New York: Routledge. Ullman, M. T. (2015b). The declarative/procedural model: A neurobiological model of language learning, knowledge and use. In G. Hickok & S. A. Small (Eds.), The Neurobiology of Language. Oxford, UK: Elsevier. Weissberger, G. H., Gollan, T. H., Bondi, M. W., Clark, L. R., & Wierenga, C. E. (2015). Language and task switching in the bilingual brain: Bilinguals are staying, not switching, experts. Neuropsychologia, 66, 193–203.

220

15 PSYCHOLINGUISTIC PERSPECTIVES ON HERITAGE SPANISH Jill Jegerski university of illinois at urbana-champaign, usa

Introduction Psycholinguistics In the most conventional sense, psycholinguistics is a sub-field of psychology dedicated to the study of language behavior and cognition (i.e., mental processes) as observed primarily in the processing of written and spoken language in comprehension and language processing in speech production, and also in first language acquisition and language disorders. Psycholinguistics is closely related to linguistics and the two fields of study overlap to an increasing degree, but there are still subtle differences. Whereas the more humanistic field of linguistics approaches language as a natural phenomenon in its own right and has traditionally relied mostly on introspection and contemplation as methods of research, psycholinguistics sees language more as an instance of cognition and employs experimental methods for its research. Broadly speaking, psycholinguistic data collection methods have included a wide variety ranging from basic pencil-and-paper questionnaires to advanced imaging via fMRI (functional magnetic resonance imaging), but in the study of bilingualism and second language acquisition, the term psycholinguistics has come to imply the more technologically advanced methods that are common in psycholinguistics research, such as computerized reaction time measures and eye tracking (and also brain-based methods like event-related brain potentials, which are covered by Wood Bowden and Issa, Chapter 14, this volume). Psycholinguistics can thus be conceived of as both a field of study with unique theoretical perspectives and as a set of research tools. The present chapter focuses on the latter, reviewing existing research on online language processing among heritage Spanish bilinguals and placing particular emphasis on the specific methodologies used in psycholinguistic research. This first section of the chapter introduces the fundamental concepts of language processing and the classification of some research methods as online, and also provides basic descriptions of four specific research methods that have been or will likely be employed in the study of Spanish as a heritage language (SHL).

221

Jill Jegerski

Language processing The fundamental tasks of recognizing individual words and discerning the complex relationships between multiple words in a series are what enable us to select an intended meaning from among an infinite number of potential meanings that can be communicated using human language. This is language processing: the unconscious and effortless steps that occur in a language user’s mind in order to comprehend words and sentences. Language processing at the word level involves the perception of different letters, sounds, and morphological sub-components that enable the reader or listener to access the appropriate word entry in a mental dictionary known as the lexicon. This is called lexical access. Language processing at the sentence level, known as sentence processing, requires more than just basic lexical access for each subsequent word that occurs as the sentence is read or heard. It also requires the incremental processing of syntactic, semantic, and discourse relationships between those words, in order to arrive at the intended meaning, thus achieving successful sentence comprehension. Of course, a similar sequence of complex processes is required to encode a speaker’s intended meaning during speech production, but the present chapter focuses on language processing for the purpose of comprehension.

Online methods With regard to experimental research methods that are employed to study language processing, these can be either offline or online, with greater emphasis in psycholinguistics on the latter. Offline methods are those that measure or test the end result of language comprehension, as in the meaning of a word or a sentence that was ultimately understood by the language user. Online methods, on the other hand, examine the moment-by-moment cognitive processes that occur along the way in order to arrive at that meaning, and especially the timing and order of different steps. To illustrate with an example using the visual world eyetracking paradigm (Jegerski & Sekerina, 2017, discussed in greater detail below in the section on ‘Research methods’), in which participants listen to auditory stimuli while tracking records their eye movements across an array of pictures related to the audio, an offline measure might be a mouse click on one of the pictures in response to a question. After listening to a brief narrative, the participant hears a stimulus question like Who saved the goat in the hole? and clicks on a picture of a rabbit, then the experimenter compares the accuracy of the mouse click responses that occurred with different types of auditory stimuli, with different types of pictures, or between different participant groups, for an offline measure of sentence comprehension. To measure online processing, on the other hand, the experimenter synchronizes the timing of the auditory stimulus with the participants’ eye movements while they were listening and analyzes the amount of time spent looking at the different pictures while each word of the auditory stimulus was playing. In this way, one can tell, for instance, if one participant group started to look at the picture of the rabbit sooner than the other group did, even if both groups were equally accurate in the mouse click responses to the question. This would suggest a difference in online processing as it occurred in real time, as opposed to a difference in offline sentence comprehension. The online/offline distinction is generally of greater importance in sentence processing research than with lexical or word processing, but lexical processing research often includes basic response time measures and sometimes even eyetracking or ERPs, so online measures can still be relevant (see e.g., CansecoGonzález, Brehm, Brick, Brown-Schmidt, Fischer, & Wagner, 2010, discussed in greater detail below in the section on ‘Recommendations for practice’). The present chapter will cover SHL research using online methods, leaving coverage of offline methods to other chapters in the volume (e.g., Fairclough and Garza, Chapter 12: Lexicon; Montrul, Chapter 10: Morphosyntax). 222

Psycholinguistic perspectives

Four common online research methods The most basic online psycholinguistic methods record the amount of time (in milliseconds) that participants take to react or respond to stimuli, usually individual words, and can be carried out on a regular computer using specialized software. For instance, in a lexical decision task, the participant reads or hears a word and indicates whether or not it is indeed a word. In self-paced reading, the participant reads sentences one word at a time, using a button to move on to each subsequent word, which gives the researcher a separate reading time for each individual word in a sentence. Eyetracking, on the other hand, requires highly specialized equipment that is usually also quite expensive, in addition to specialized software. In eyetracking with text, the amount of time a participant spends reading each word in a sentence is measured via their eye movements, tracked by one or more cameras that record infrared light reflected off the eyes. With visual world eyetracking (i.e., looking-while-listening), the participant listens to the stimulus sentence while their pattern of eye movements across an array of pictures related to the stimulus is recorded. With all four of these methods, online data are analyzed to draw inferences regarding the moment-by-moment cognitive processes that comprise language comprehension.

Critical issues and topics The existing body of research on SHL that has adopted psycholinguistic methods is comprised of only a handful of investigations, but at least two key issues have emerged thus far. First, the identification of appropriate research methods for the empirical study of SHL has long been of interest, and some recent work suggests that online psycholinguistic measures may be particularly useful in this regard. Second, the question of cross-linguistic influence from English during the processing of Spanish by heritage speakers is of fundamental theoretical significance and has begun to be examined empirically as well.

Research methods for the study of heritage Spanish Within the linguistic study of heritage languages in general, an important issue is the appropriateness of research methods, which are often adopted from related fields of study with longer research traditions. One specific concern is whether the grammaticality judgments that have been commonly used in linguistics can accurately gauge the implicit knowledge that characterizes heritage speakers (e.g., Benmamoun, Montrul, & Polinsky, 2013), assuming that this is the investigative goal. Grammaticality judgments, in which speakers of a language indicate whether a given phrase or sentence is possible or not, often rely on metalinguistic knowledge about the language, as opposed to the ability to use the language to communicate meaning. A person can speak a language fluently and yet have very little accuracy when it comes to recognizing specific grammar errors in the language, especially if they have not had extensive formal education in the language, as is the case with SHS. It has thus been argued that metalinguistic (i.e., explicit) tasks or tests are generally not the most appropriate for the study of SHL—except where the research goal is to examine explicit knowledge, as may be the case with some studies of instruction in the heritage language—and this claim has begun to be explored empirically. For instance, researchers and educators alike are continually searching for methods of assessment that are appropriate for use with heritage languages. Proficiency tests are traditionally metalinguistic, offline measures like cloze tests, but there is evidence from Russian that suggests that an alternative measure like speech rate in words per minute may be a more accurate test of general heritage language proficiency (Polinsky, 2008). Speech rate is an example of a psycholinguistic method that is online, because the data reflect how language processing unfolds 223

Jill Jegerski

in real time, moment by moment. Online techniques can also be useful in research as well because real-time language tasks are probably less likely to rely on metalinguistic knowledge than those that do not require an immediate response. Although the two factors are not always easy to tease apart, the classification of a particular research methodology as metalinguistic is independent from its classification as online. Thus, both are important methodological issues in the experimental study of SHL.

Cross-linguistic influence from English Beyond the potential contribution with regard to the identification of appropriate research methods, of the three main factors affecting SHL—incomplete acquisition, attrition, and dominant language transfer (Benmamoun, Montrul, & Polinsky, 2013)—psycholinguistic research stands to make a particularly strong contribution in the area of dominant language transfer from English. From a psycholinguistic perspective, constructs related to the notion of linguistic transfer include parallel or non-selective activation and competition within the bilingual mind. The most fundamental question driving the psycholinguistic study of bilingual language processing is regarding selectivity, that is, whether language comprehension involves the selective activation of only the relevant language or the non-selective activation of both languages in parallel, regardless of relevance. More specifically, in the case of heritage Spanish bilinguals, at issue is whether translation equivalents and other related English words are activated as words are accessed in the lexicon in Spanish, and whether sentence processing strategies for English are similarly activated as sentences are processed in Spanish (and vice versa, for both words and sentences). Such nonselective or parallel activation opens the door for competition, which can take the form of either interference or facilitation. Cross-linguistic competition among bilinguals is in general a robust finding in psycholinguistics (see Kroll, Dussias, Bice, & Perrotti, 2015, for a recent overview), particularly with word-level processing or lexical access, probably because online methods can reveal subtle effects that would not be evident in offline measures because such cross-linguistic influence often does not affect the end product of language comprehension.

Current contributions and research This section provides an overview of existing work that has examined Spanish as a heritage language using psycholinguistic research methods. Although processing-oriented research on SHL has been quite limited thus far, two general trends will be discussed. The first part of this section will describe several investigations that have either intentionally or incidentally explored issues in research methodology and will conclude that psycholinguistic methods that are online and meaning-oriented (i.e., not metalinguistic or invoking explicit grammar rules) appear to have the greatest potential to contribute to our existing knowledge of heritage Spanish. The second part of this section will discuss research that has just begun to explore cross-linguistic influence from English during online language processing of Spanish and has shown that such influence is not always evident.

Research methods There are at least five psycholinguistic studies of SHL that are relevant to the discussion of research methods. In the first of these, Montrul, Davidson, de la Fuente, and Foote (2014) used a series of three different tasks for their descriptive study of gender agreement among heritage and non-native speakers of Spanish. Interestingly, how the two groups compared to each other 224

Psycholinguistic perspectives

and to a reference group of monolingually raised native Spanish speakers varied according to both the degree of explicitness of the task and, within each task, whether the data were from a timed or untimed measure. The stimuli for all three tasks were three-word noun phrases comprised of a determiner, an adjective, and a noun, as in la quinta calle “the fifth street,” which varied with regard to the grammaticality of the gender agreement between the determiner and the noun. For the first task, an aural monitoring task that was very explicit in its focus on gender agreement, participants listened to the noun phrases and indicated whether the grammatical gender of each noun was masculine or feminine. All three participant groups were affected by the grammaticality of the gender agreement in the stimuli, with higher accuracy scores and faster response times for grammatical stimuli than for ungrammatical stimuli, but accuracy scores among the monolingually raised native speakers were affected the least by grammaticality, followed by the second language learners, followed by the heritage bilinguals. For the second task, a grammaticality judgment that was also clearly metalinguistic, participants again listened to three-word noun phrases and this time indicated whether each was grammatical or ungrammatical. The accuracy scores were highest for the monolingually raised native speakers and lower for the heritage speakers and second language learners, whose scores were statistically similar to each other. The online response time data, however, revealed a difference between the two experimental groups, with the heritage speakers responding faster overall than the nonnative speakers. For the third and final task, a word repetition task that was less explicit than the other two tasks in its focus on gender agreement, participants once again listened to three-word noun phrases, but this time were asked to merely repeat the last word of the phrase, the noun, as quickly and accurately as possible. The accuracy scores were unrevealing because of a ceiling effect, in which all three groups correctly repeated the noun 100% of the time. The online response time data did reveal differences, with the monolingually raised native and heritage speakers repeating nouns more quickly when they appeared in grammatical versus ungrammatical phrases and the second language learners showing no overall response time difference for grammatical versus ungrammatical noun phrases. Overall, across the three experimental tasks in this study, the offline data suggest that the heritage speakers were more similar to second language learners than to monolingually raised native speakers in terms of their knowledge of gender agreement, while the online data suggest that they were more similar to the monolingually raised native speakers in this regard, consistently showing greater sensitivity to agreement than the second language learners. As concerns the explicitness of the tasks, the heritage speakers appeared most similar to the monolingually raised native speakers on the word repetition task, the least explicit of the three, followed by the grammaticality judgment and then the gender monitoring task. The authors concluded that explicit experimental tasks can place heritage speakers at a disadvantage, particularly when they are being compared to second language learners, who typically have more experience with Spanish in the formal instructional settings that build metalinguistic knowledge and familiarity with explicit grammar than do heritage speakers. Thus, the outcome of this three-part investigation highlights the importance of methodology in experimental research on SHL and suggests that measures that are more implicit and online may be more appropriate for most descriptive studies because explicit tasks and offline data can sometimes undershoot heritage language knowledge and ability. In the second study, similar patterns in online versus offline data were apparent in the interpretation of null and overt pronouns, and heritage bilinguals were more similar to monolinguals when the two were compared on the basis of online data. Keating, Jegerski, and VanPatten (2016) investigated the interpretation of Spanish null and overt pronouns by heritage speakers using a self-paced reading time measure. In contrast with English, Spanish subject pronouns are 225

Jill Jegerski

optional; when the pronoun is expressed it is said to be overt and when it is omitted but understood it is said to be null, as illustrated in (1) and (2). While most contexts allow for either a null or an overt pronoun to appear, the referential tendencies of the two types of pronoun differ. For instance, both the null pronoun in (1) and the overt pronoun in (2) are ambiguous in that they could refer to either el sospechoso “the suspect” or el policía “the policeman,” but in most varieties of Spanish the null pronoun usually refers to el sospechoso because it is the subject of the first clause and the overt pronoun more often refers to el policía because it is not the subject. (1)

Null pronoun Después de que el sospechoso habló con el policía, hizo tres llamadas. “After the suspect spoke with the policeman, Ø made three phone calls.”

(2)

Overt pronoun Después de que el sospechoso habló con el policía, él hizo tres llamadas. “After the suspect spoke with the policeman, he made three phone calls.”

A prior study (Keating, VanPatten, & Jegerski, 2011) with heritage bilinguals had employed an offline interpretation task, in which participants were presented with ambiguous stimuli like those in (1) and (2) and asked to indicate whether they thought it was el sospechoso “the suspect” or el policía “the policeman” that made the phone call. A group of heritage speakers most often interpreted both null and overt pronouns as referring to subjects, while a monolingual group interpreted the two types of pronoun differently, with null pronouns usually referring to subjects and overt pronouns usually referring to objects. In short, the heritage bilinguals and monolinguals appeared quite different from each other when tested with an offline questionnaire. In the subsequent study with a similar group of heritage bilinguals drawn from the same population (Keating, Jegerski, & VanPatten, 2016), the researchers observed that heritage speakers did interpret the two types of pronoun differently, but this preference was only evident during an online self-paced reading task, and was again absent when the task was offline interpretation. The heritage bilingual group was not identical to the comparison group, with the two differing from each other even online in the case of overt pronouns, but the overall similarity between the groups was greater when pronoun interpretation preferences were observed during sentence processing rather than after (i.e., online versus offline). Thus, the outcome of this study provides further evidence that the choice of research method can be crucial in the study of SHL, and that offline methods can sometimes underestimate heritage speakers’ knowledge and ability in Spanish. A third investigation that has demonstrated the potential importance of including online psycholinguistic research methods in studies of SHL is that of Foote (2011), which targeted the online processing of subject-verb number agreement and adjective-noun gender agreement during meaning-oriented self-paced reading. Additional aspects of the study will be discussed below in the section on cross-linguistic influence from English, but here the focus is on those details that may be informative with regard to the use of online research methods. While not part of the statistical analysis, the reading time data revealed that the heritage speakers read stimulus sentences more slowly in general than did a group of second language learners, who in turn read at a speed that was similar to or slightly slower than a comparison group of monolingually raised native speakers. On the other hand, all three participant groups were similar with regard to online sensitivity to both types of agreement, which suggests that the difference in global reading speed did not affect this aspect of sentence processing. Still, it is an interesting result that may 226

Psycholinguistic perspectives

reflect the fact that heritage speakers tend to have less experience with Spanish in formal educational contexts and with printed text in Spanish than do second language learners and native speakers raised in Spanish speaking countries. In addition, the results from an offline written test of grammatical agreement that was administered as part of the study suggested that the heritage speakers may have been at a disadvantage there as well. Even though both groups showed similar effects of agreement during the online self-paced reading task, the second language learners scored higher than the heritage speakers on the offline written task, although both groups did score quite high and the significance of the difference (98.7% L2 versus 93.2% heritage) was not confirmed with statistical testing because it was not relevant to the objectives of the study. In the fourth study, Jegerski (2017) examined the online processing of the a marker on direct and indirect objects in SHL. The outcome of this investigation illustrates how online research methods can make important additions to our knowledge of SHL, even when they produce results that are broadly consistent with previous research using offline methods. The particle a, commonly referred to as “personal a,” marks all indirect objects in Spanish and a select subset of direct objects, primarily based on the animacy of the direct object, in a system referred to as differential object marking, which is illustrated in (3) and (5). Previous research using offline methods had found that monolingual Spanish speakers have strong preferences to mark direct objects that are animate with a, as in (3), but this preference is less robust among heritage speakers, who sometimes leave animate direct objects unmarked, as in (4). With direct objects that are inanimate, on the other hand, both groups have clear preferences to leave them unmarked, as in (5). One factor that has been proposed to partially account for these observations is the non-saliency of the a marker, which is just a single letter that could easily be overlooked during reading or not heard during listening and thus contribute to the incomplete acquisition of differential object marking among heritage bilinguals (Montrul, 2014). (3)

Animate direct object with a La estudiante visita a la maestra todos los meses. “The student visits [a] the teacher every month.”

(4)

Animate direct object without a La estudiante visita la maestra todos los meses. “The student visits the teacher every month.”

(5)

Inanimate direct object without a La estudiante visita el museo todos los meses. “The student visits the museum every month.”

The experimental measures for Jegerski’s (2017) study were an offline scaled acceptability judgment task and a meaning-oriented, online, self-paced reading task. The two tasks were completed by a group of heritage speakers and a comparison group of late Spanish-English bilinguals. Both participant groups demonstrated robust preferences online and offline for marking with a for indirect objects and for not marking with a for inanimate direct objects. With animate direct objects, both groups also showed the expected preference for marking with a offline, but neither showed a significant preference online during meaning-oriented self-paced reading. This outcome was consistent with previous research in that the most dynamic area of object marking in heritage Spanish was with animate direct objects. Nonetheless, three new contributions came specifically from the self-paced reading results: (1) heritage Spanish bilinguals do 227

Jill Jegerski

reliably notice the marker a in some contexts during real-time language processing—albeit not with direct objects—even though it is only one letter; (2) other bilingual populations in the U.S. can also show attenuated preferences for marking animate direct objects; and (3) offline metalinguistic preferences for object marking are not necessarily evident during online processing for meaningful comprehension. Thus, even though there had already been multiple research studies targeting differential object marking in Spanish as a heritage language using a variety of research methods, and the outcome of this study was broadly consistent with prior work, the psycholinguistic perspective of the study made it uniquely informative to our understanding of why differential object marking is so interesting in heritage Spanish. We now know that it is not just because the single letter a is easily missed during reading or listening, we know that the tendency towards indifference to marking with animate direct objects goes beyond heritage Spanish and may be present at the community level (see also Montrul, 2014; Montrul & Sánchez-Walker, 2013), and we know that the type of research method (i.e., online or offline) used to study object marking among heritage speakers can affect the results. In a related investigation, Jegerski and Sekerina (2017) used the visual world eyetracking method to examine the online processing of the particle a, but this time in questions starting with quién “who” or a quién “whom,” where the inclusion of non-canonical word order (i.e., both SVO and OVS vs. just SVO) and the sentence-initial position of the a case marker might make it more salient and likely to be noticed. Sentence processing was also examined during listening rather than reading, as in previous studies. The participants were a group of SHS and a comparison group of late L1 Spanish – L2 English bilinguals. For the visual world eyetracking procedure, participants listened to a series of brief stories while looking at four pictures of the characters and objects from each story while their eye movements were recorded. After each story, they heard a question about how two of the characters were affected by the main action that occurred in the story (e.g., ¿A quién salvó el conejo en el hoyo? “Who(m) did the rabbit save in the hole?”) and responded by clicking on one of the four pictures and saying the correct answer aloud. The verbal responses of the comparison group were more accurate than those of the heritage speakers, both in terms of the a case marker and the picture referent. But the online comprehension measures, meaning the speed and accuracy of mouse clicks and the timing of eye movements across the four pictures, suggested that the two groups processed the stimulus questions similarly. In addition, fine-grained analysis of the online eye movement data revealed that both groups started to look more at the correct picture 200 milliseconds sooner with the object-first than the subject-first questions (i.e., “¿A quién . . .? vs. ¿Quién . . .?), which means that the a marker was at least noticed during processing, something that was not evident in the other measures. Thus, once again there was a difference between the online data (eye movements) and some of the offline data (verbal responses) from the same experiment. To summarize the five studies in this section, all provide interesting evidence that online methods may be particularly useful in the study of SHL (or, from a psycholinguistic perspective, heritage Spanish bilingualism), even though most were not intentionally designed to explore the issue. A related outcome was that experimental tasks that are meaning-oriented, as opposed to metalinguistic, can also be especially informative.

Cross-linguistic influence from English The question of cross-linguistic influence in the processing of SHL is not as well represented in the literature, but so far there are at least two published research studies that have explored the issue. A recent investigation by Jegerski, Keating, and VanPatten (2016) targeted a type of 228

Psycholinguistic perspectives

syntactic ambiguity that has long been of interest in the empirical study of sentence processing in main stream psycholinguistics. In sentences such as (6) below, the relative clause who had a very long moustache could refer to either the boss or the clerk. Despite the overall ambiguity, some languages tend towards the first noun and others tend towards the second, preferences that are known as NP1 and NP2 attachment, respectively (NP stands for “noun phrase”). Monolingual Spanish users typically interpret the relative clause who had a very long moustache as referring to the boss, whereas monolingual English users usually interpret it as referring to the clerk, so relative clause attachment preferences are ideal for observing cross-linguistic influence among Spanish-English bilinguals. (6)

Ambiguous relative clause Miguel argued with the boss of the clerk who had a very long moustache.

(7)

Disambiguated relative clauses a b

Miguel discutió con el jefe de la vendedora que tenía un bigote muy largo. NP1 Miguel discutió con la jefa del vendedor que tenía un bigote muy largo.   NP2

“Miguel argued with the bossM/F of the clerkF/M that had a very long moustache.” Previous research had found that this aspect of bilingual sentence processing was subject to strong cross-linguistic influence. Specifically, late Spanish-English bilinguals, monolingually raised Spanish speakers who had acquired English as adults and experienced long-term immersion in English, were shown to have similar relative clause attachment preferences in both of their languages (Dussias, 2003). These reflected an English-like preference for NP2 attachment, even while reading in their native Spanish (Dussias & Sagarra, 2007). The NP2 attachment preference was evident in longer reading times on NP1 versus NP2 sentences, as illustrated in (7) above. It was therefore proposed that relative clause attachment preferences in either of a bilingual’s languages reflected the preference of the language of greatest exposure, which predicts that heritage Spanish bilinguals would adopt English-like attachment preference in Spanish and that they would have a similar preference in both languages. Interestingly, Jegerski et al. (2016) observed the opposite trend, in which heritage speakers of Spanish demonstrated a monolingual-like preference for NP1 attachment while reading relative clauses in Spanish. In a related study using an offline measure, Jegerski, VanPatten, and Keating (2016) also found that heritage speakers of Spanish had distinct attachment preferences in Spanish and English, each of which was similar to what would be expected of monolingual native speakers of the language. The authors tentatively concluded that heritage Spanish was less susceptible to this type of cross-linguistic influence than was the Spanish of other Spanish-English bilinguals, because heritage bilinguals are early bilinguals, having acquired both languages in childhood, whereas the participants in the previous investigations were late bilinguals. Further research is needed to corroborate both the outcome of these studies and the account proposed by the authors, but this one study demonstrates the potential of psycholinguistically oriented research to be uniquely informative with regard to the effects of cross-linguistic influence from English on SHL. A second psycholinguistic investigation of the effects of English on heritage Spanish also found no evidence of cross-linguistic influence. Foote (2011; also discussed in the above section on research methods) examined the online processing of two types of agreement among heritage bilinguals (and also second language learners of Spanish), subject-verb number agreement and noun-adjective gender agreement. The two types of agreement were selected because subjectverb number agreement also occurs in English, but gender agreement does not, so the latter 229

Jill Jegerski

might be more likely to be subject to variability in SHL. Nevertheless, the results of a self-paced reading task suggested that this was not the case, as a group of heritage bilingual participants demonstrated online sensitivity to both types of agreement that was similar to that of a monolingually raised comparison group, and so did a group of very advanced second language learners. In addition, the tendency to slow down upon encountering agreement errors while reading diminished as the distance between the two words in the sentence that agreed increased, as illustrated in (8) and (9) below, a pattern that was evident among all three participant groups. Thus, similarity with English did not seem to affect this aspect of sentence processing in SHL because both types of agreement were the same, regardless of their similarity with English. (8)

Gender agreement between adjacent words a b

Dicen que el libro blanco está en esa mesa.   Grammatical Dicen que el libro blanca está en esa mesa.   Ungrammatical

“They say that the white book is on that table.” (9)

Gender agreement between non-adjacent words a El pollo del taco está rico pero picante.     Grammatical b El pollo del taco está rica pero picante.     Ungrammatical “The chicken in the taco is tasty but spicy.”

To summarize the two studies reviewed in this section, we cannot assume that cross-linguistic influence from English will always be evident during the processing of Spanish as a heritage language, despite the strong theoretical motivation to make such a prediction. On the contrary, neither of the two studies discussed found evidence of such influence, showing instead that when it comes to online processing, the Spanish of heritage bilinguals can be indistinguishable from the Spanish of those that acquired it abroad as a majority language, at least in some specific contexts. Of course, this is but an interesting observation made on the basis of a few initial investigations, both of which examined language processing at the sentence level. Extensive further research is needed before any consistent trend can emerge.

Recommendations for practice For the purposes of this chapter, practice will be used to refer to research with heritage Spanish bilinguals that employs online psycholinguistic methods. As discussed in the first part of the previous section surveying existing research, the choice of a research method that relies on online data collected as language is being processed in real time may in and of itself be an important step in advancing our knowledge of Spanish as a heritage language, and particularly our knowledge of what distinguishes a heritage language from a second language. Another important concern for researchers that was discussed is that experimental measures that are less explicit can eliminate a potential source of bias, in which heritage speakers are at a disadvantage as compared to native speakers raised abroad and especially as compared to second language learners. From the perspective of the researcher, the implication of this observation from previous studies is that experimental tasks should be meaning-based to the greatest extent possible. Metalinguistic tasks that ask participants to identify grammar errors can be biased in favor of formally instructed second language learners because they introduce esoteric skills and types of knowledge that are learned primarily in formal educational contexts, in which heritage speakers typically have 230

Psycholinguistic perspectives

had less experience with Spanish. Tasks with responses based on meaningful comprehension (e.g., “Where does the student go?” or “Click on the picture of the beetle”) may serve to minimize or eliminate such biases and thereby create a more level playing field on which to compare data from different participant populations. Looking beyond the advantages of research methods that are online and meaning-oriented, one additional factor that researchers may want to consider when conducting psycholinguistic studies of heritage Spanish bilingualism, but that has not yet been investigated among this particular bilingual population and therefore was not discussed in the previous section of this chapter, is the role of bilingual language mode. Grosjean’s (1985) notion of language mode proposes that the relative level of mental activation of a bilingual’s two languages varies according to the communicative context in which they find themselves at a given point in time. Activation level is continuous, ranging from monolingual mode, with only one language or the other activated, to completely bilingual mode, with both languages fully activated, and is determined by a number of contextual characteristics such as the language proficiency and use habits of both interlocutors, the physical location and presence of others, the content of the ideas being communicated and the register, and the purpose of the language act. Grosjean (1998) has suggested that language mode is an important concern for researchers, as it can influence the results of psycholinguistic experiments that do not take it into consideration. This may be especially true with online methods, because they rely on reaction times and other types of time-sensitive measures that might be affected by degree of activation. In theory, an experimental procedure determines the language mode of bilingual participants and thus can affect how quickly they can access one or the other of their languages and the degree of crosslinguistic influence from one language to the other. Thus far, the results of the few studies that have tested the proposal empirically have been inconclusive and none of this research has explicitly targeted heritage bilinguals. Despite this gap in existing research, there are two studies of language mode that appear to have included heritage Spanish bilinguals. Results from the two have been conflicting. First, Dunn and Fox Tree (2014) examined the role of language mode in word recognition among fluent Spanish-English bilinguals, some of whom appeared to be heritage Spanish speakers— age of acquisition was not reported, but there were 75 bilingual participants from Southern California who self-identified as fluent bilinguals dominant in either English or Spanish. For this carefully designed experiment, monolingual and bilingual participants were recruited through a university subject pool database on the basis of existing language background information so that they were not aware that the research targeted bilingualism. For the first part of the experiment, all participants completed a timed lexical decision task in English, in which 100 words and nonwords appeared individually on a computer screen and participants pressed a button to indicate whether each item was a word or not. The second step in the experiment was to watch a five-minute silent cartoon and provide an oral retelling of the story. Half of the bilingual participants completed this second step in Spanish, while the other half of the bilinguals and the monolinguals completed it in English. Finally, the third part of the experiment was another lexical decision task in English. The language of the story re-telling affected the results, as all three groups performed equally on the first lexical decision task, and those who completed the entire experiment in English also took the same amount of time as the monolinguals to recognize words and nonwords in the second lexical decision task, but those who did the story re-telling in Spanish in the interim took longer to reject English nonwords in the second lexical decision task than the other two groups. In addition, this apparent effect of competition from Spanish while in bilingual mode was greatest among those participants who were Spanish dominant, that is, whose proficiency in Spanish outweighed their proficiency in English, and least among 231

Jill Jegerski

those who were English dominant. The authors concluded that when bilinguals switch from monolingual to bilingual mode, cognitive inhibition of words in the non-relevant language is lowered and cross-linguistic competition occurs. The second of the two existing studies of language mode that have included heritage Spanish bilinguals was an eyetracking study that targeted cross-linguistic competition in English from words that sound similar in Spanish (Canseco-González et al., 2010). Participants saw three pictures on a computer screen and heard instructions to click on one of the three pictures while their eye movements were recorded. In one example, the instructions were click on the beetle and the three pictures on the screen were of a beetle, a moustache (the Spanish bigote starts with sounds that are very similar to those at the beginning of beetle), and a jail cell. Cross-linguistic competition from Spanish was evident when participants were more distracted by the competitor picture of the moustache (bigote) than the neutral picture of the jail cell (cárcel) before finally clicking on the target picture of the beetle. This type of distraction is presumably unconscious and is very short-lived, occurring over a span of around a third of a second (i.e., 300–400 milliseconds), but it can be observed in eye movements because participants look around very quickly at the different pictures before focusing in on one and clicking on it, and these eye movements are believed to reflect the cognitive processes involved in word recognition. The heritage bilingual participant group in this study did show evidence of subtle competition from words in Spanish while listening to and identifying words in English, but less so than a comparison group of Spanish dominant bilinguals raised abroad. Crucially, the manipulation of language mode did not appear to affect the level of cross-linguistic competition with either participant group. Thus, the two existing studies that can speak to the role of bilingual language mode in research on heritage Spanish bilinguals have contradicted each other in terms of outcome. Nevertheless, because neither of these studies was explicitly designed to target heritage bilinguals, they have some important limitations that warrant further empirical investigation of this issue. First, both studies examined the role of language mode on cross-linguistic competition from Spanish while the bilinguals were processing English, but competition would likely be weaker in this direction among heritage speakers than in the opposite direction, from English during the processing of Spanish, because proficiency and bilingual language dominance affect the degree of cross-linguistic influence. This may explain why the competition effects in general were not robust in either study. Second, the construct of language mode might be more relevant to the Spanish of heritage bilinguals than in other bilingual contexts because, as adults, they are most often in communicative contexts that call for either monolingual English mode or bilingual mode and not nearly as often in monolingual Spanish mode. In fact, given the prevalence and pervasiveness of English in the U.S., it may not be even possible to create a context that is conducive to monolingual Spanish mode. Still, language mode is a continuum, so even if the procedure for an experiment conducted in the U.S. cannot put a bilingual participant in completely monolingual Spanish mode, it still may be possible to create relative degrees of activation of their two languages. This also highlights the importance of research on SHL in other communicative contexts, such as in heritage language immersion abroad.

Future directions So far, there has been very little research on SHL that has adopted psycholinguistic methods, so there are numerous potential avenues for future research. Two areas that seem promising are those discussed in this chapter, the issue of online psycholinguistic research methods in the study of SHL and that of cross-linguistic influence from English during online processing. Compiled below are these and other broad, unanswered questions that might serve as inspiration for future investigations. 232

Psycholinguistic perspectives

  I How do experimental outcomes vary according to whether a measure is online or offline? If timed tasks rely less on explicit knowledge and heritage Spanish is characterized by primarily implicit knowledge, then future research will continue to support the proposal that online methods are especially appropriate for use with heritage bilinguals. II How does online processing vary according to whether an experimental task is metalinguistic or meaning-oriented (e.g., whether sentential stimuli are followed by a grammaticality judgment or a comprehension question)? If heritage bilinguals typically have less explicit knowledge than L2 learners and native speakers with schooling in Spanishmajority contexts, then inclusion of any metalinguistic component in an experimental task introduces a potential source of bias. III How does online processing vary according to whether linguistic stimuli are presented in written or auditory form? Given that heritage bilinguals typically have much more experience with oral language than with written text, then text-based measures might undershoot their Spanish ability. IV Is there cross-linguistic influence or competition from English during online processing among heritage Spanish bilinguals? The exposure based Tuning Hypothesis (Cuetos, Mitchell, & Corley, 1996), for instance, claims that sentence processing behavior reflects past exposure to input, so extensive exposure to English would lead to pronounced crosslinguistic influence on sentence processing in heritage Spanish. V Can language mode affect online processing among heritage Spanish bilinguals? If heritage speakers are most often in bilingual mode or monolingual English mode, including while participating in research studies, this could be a mechanism of cross-linguistic influence from English.

Further reading Jegerski, J., & VanPatten, B. (Eds.). (2014). Research methods in second language psycholinguistics. New York: Routledge. Each of the chapters in this volume gives a detailed and practical overview of one psycholinguistic or neurolinguistic research method, including self-paced reading, eye tracking, event-related brain potentials (ERPs) and others. Keating, G. D., & Jegerski, J. (2015). Experimental designs in sentence processing research: A methodological review and user’s guide. Studies in Second Language Acquisition, 37(1), 1–32. This article is a practical guide for researchers on how to carry out experiments on sentence processing using self-paced reading and eyetracking, with a particular focus on experimental design, the creation of stimulus sentences, statistical analysis and common pitfalls to avoid. Kroll, J. F., Dussias, P. E., Bice, K., & Perrotti, L. (2015). Bilingualism, mind, and brain. Annual Review of Linguistics, 1, 377–394. This review article surveys existing psycholinguistic research on bilingualism, including studies of both early and late bilinguals, and argues that there have been three major developments in this area since the late 1990s: that both languages are always active in the bilingual mind and thus influence each other, that a second language can also influence a first language, and that there are cognitive consequences of bilingualism that go beyond those areas of the mind that are dedicated specifically to language. Sekerina, I. A., Fernández, E. M., and Clahsen, H. (Eds.) (2008). Developmental psycholinguistics: Online methods in children’s language processing. Amsterdam: John Benjamins. The chapters in this volume cover a variety of issues pertaining to the experimental study of language processing in children and provide an introduction to the four online research methods that are most commonly used in psycholinguistic and neurolinguistic research on children: reaction time techniques, visual world eyetracking, free viewing eyetracking and event-related brain potentials (ERPs).

233

Jill Jegerski

References Benmamoun, E., Montrul, S., & Polinsky, M. (2013). Heritage languages and their speakers: Opportunities and challenges for linguistics. Theoretical Linguistics, 46, 129–181. Canseco-González, E., Brehm, L., Brick, C. A., Brown-Schmidt, S., Fischer, K., & Wagner, K. (2010). Carpet or cárcel: The effect of age of acquisition and language mode on bilingual lexical access. Language and Cognitive Processes, 25(5), 669–705. Cuetos, F., Mitchell, D. C., & Corley, M. M. B. (1996). Parsing in different languages. In M. Carreiras, J. E. García-Albea, & N. Sebastián-Gallés (Eds.), Language Processing in Spanish (pp. 145–187). Mahwah, NJ: Lawrence Erlbaum Associates. Dunn, L. D., & Fox Tree, J. E. (2014). More on language mode. International Journal of Bilingualism, 18(6), 605–613. Dussias, P. E. (2003). Syntactic ambiguity resolution in second language learners: Some effects of bilingualism on L1 and L2 processing strategies. Studies in Second Language Acquisition, 25, 529–557. Dussias, P. E., & Sagarra, N. (2007). The effect of exposure on parsing in Spanish-English bilinguals. Bilingualism: Language and Cognition, 10, 101–116. Foote, R. (2011). Integrated knowledge of agreement in early and late English–Spanish bilinguals. Applied Psycholinguistics, 32, 187–220. Grosjean, F. (1985). The bilingual as a competent but specific speaker-hearer. Journal of Multilingual and Multicultural Development, 6, 467–477. Grosjean, F. (1998). Studying bilinguals: Methodological and conceptual issues. Bilingualism: Language and Cognition, 1, 131–149. Jegerski, J. (2017). The processing of the object marker a by heritage Spanish speakers. International Journal of Bilingualism. Advance online publication. http://journals.sagepub.com/doi/10.1177/136700 6916681083. Jegerski, J., Keating, G. D., & VanPatten, B. (2016). Online relative clause attachment strategy in heritage speakers of Spanish. International Journal of Bilingualism, 20(3), 254–268. Jegerski, J., & Sekerina, I. (2017, April). The processing of Spanish wh- questions with case marking among heritage bilinguals. Paper presented at the 26th Conference on Spanish in the United States, Brigham Young University, Provo, UT. Jegerski, J., & VanPatten, B. (Eds.). (2014). Research Methods in Second Language Psycholinguistics. New York: Routledge. Jegerski, J., VanPatten, B., & Keating, G. (2016). Relative clause attachment preferences in early and late bilinguals. In D. Pascual y Cabo (Ed.), Advances in Spanish as a Heritage Language (pp. 81–98). Amsterdam: John Benjamins. Keating, G. D., & Jegerski, J. (2015). Experimental designs in sentence processing research: A methodological review and user’s guide. Studies in Second Language Acquisition, 37(1), 1–32. Keating, G. D., Jegerski, J., & VanPatten, B. (2016). Online processing of subject pronouns in monolingual and heritage bilingual speakers of Mexican Spanish. Bilingualism: Language and Cognition, 19(1), 36–49. Keating, G. D., VanPatten, B., & Jegerski, J. (2011). Who was walking on the beach? Anaphora resolution in Spanish heritage speakers and adult second language learners. Studies in Second Language Acquisition, 33, 193–221. Kroll, J. F., Dussias, P. E., Bice, K., & Perrotti, L. (2015). Bilingualism, mind, and brain. Annual Review of Linguistics, 1, 377–394. Montrul, S., (2014). Structural changes in Spanish in the United States: Differential object marking in Spanish heritage speakers across generations. Lingua, 151, 177–196. Montrul, S., Davidson, J., de la Fuente, I., & Foote, R. (2014). Early language experience facilitates gender agreement processing in Spanish heritage speakers. Bilingualism: Language and Cognition, 17(1), 118–138. Montrul, S., & Sánchez-Walker, N. (2013). Incomplete acquisition of differential object marking in child and adult Spanish heritage speakers. Language Acquisition, 20, 1–31. Polinsky, M. (2008). Russian gender under incomplete acquisition. Heritage Language Journal 6(1), 40–71. Sekerina, I. A., Fernández, E. M., & Clahsen, H. (Eds.) (2008). Developmental Psycholinguistics: Online Methods in Children’s Language Processing. Amsterdam: John Benjamins.

234

16 CHILD HERITAGE SPEAKERS’ MORPHOSYNTAX Rate of acquisition and crosslinguistic influence Naomi Shin university of new mexico

Introduction This chapter reviews research on bilingual children’s Spanish morphosyntax. The review focuses primarily on Latin@ children in the United States who are exposed to Spanish at home and are often English-dominant bilinguals (Merino, 1983; Zentella, 1997). Given that these children are learning Spanish as a minority language, they are often considered ‘heritage speakers’ (e.g., Montrul, this volume; Valdés, 2005). As such, and in keeping with the theme of this volume, I will use the term ‘heritage speakers’ when referring to children who are acquiring Spanish in the U.S. I will also sometimes refer to these heritage speakers as English-Spanish bilinguals, especially when highlighting the role that English plays in the children’s acquisition of Spanish. The primary aim of the chapter is to extract broad generalizations from the literature, and to interpret those generalizations within current theories of language acquisition and bilingualism. In particular, two questions are explored: 1 2

Do child heritage speakers and monolingual children acquire Spanish morphosyntax at the same rate? Do child heritage speakers’ morphosyntactic patterns reflect influence from English?

The literature reviewed suggests that reduced exposure to Spanish results in slower acquisition of Spanish morphosyntax. With respect to the second question, it appears that English influences Spanish morphosyntax during heritage language development; however, with age, heritage speakers become increasingly adept at suppressing features that do not correspond to communicative expectations, which in turn reduces the likelihood of structural convergence. The chapter also provides discussion that illuminates unanswered questions, and thus suggests topics that are ripe for future investigation.

235

Naomi Shin

Rate of acquisition To explore whether child heritage speakers acquire Spanish morphosyntax at the same rate as monolingual children, I focus on morphosyntactic phenomena that do not have counterparts in English. These phenomena are more likely to provide evidence of protracted development than grammatical structures that are present in both Spanish and English because the presence of a grammatical category in one language can bolster acquisition of that same grammatical category in another language (Schwartz et al., 2015). In particular, this section reviews and discusses research on gender assignment and agreement, differential object marking, and verb morphology.

Gender assignment and agreement In order to examine acquisition of morphological gender, it is helpful to distinguish between gender assignment and gender agreement. Gender assignment is the process by which Spanish nouns are classified as either masculine or feminine. Masculine nouns that end in –o and feminine nouns that end in –a are considered ‘canonical’ or prototypical of the Spanish gender system. Nouns that end in e, i, u, or a consonant are considered ‘noncanonical.’ Canonical noun gender is more predictable than noncanonical noun gender for several reasons. One reason is type frequency: there are more nouns that fit the canonical profile than the noncanonical one. For example, Teschner and Russell (1984, pp. 116–118) found that 67% of 41,879 nouns in a Spanish dictionary followed the masculine-o/feminine-a pattern. The remaining 33% were noncanonical nouns, a category that consists of 22 types, including common ones such as those ending in –d or –e, as well as rare types, such as those ending in –u, or –j. A second reason canonical noun gender is especially predictable is that nouns ending in –o or –a are highly regular, meaning there are few exceptions to this pattern. In contrast, noncanonical nouns are less consistent. In Teschner and Russell’s (1984, pp. 116–118) study, 99.9% of all nouns that ended in –o were masculine, and 96.3% of nouns that ended in –a were feminine. In contrast, 89% of nouns that ended in –e were masculine, and 62% of nouns that ended in –z were feminine. Nevertheless, there is evidence that Spanish-speaking adults track the likelihood that noncanonical nouns will be masculine or feminine; for example, they tend to assign masculine gender to unfamiliar words that end in –e and feminine gender to those that end in –z (Eddington, 2002, p. 70). Let us now consider gender agreement. Articles, adjectives, and object pronouns match the gender of the noun they modify or refer to (Ambadiang, 1999, p. 4882). For example, in (1) esa ‘that’ and redonda ‘round’ end in –a because they refer to mesa.   (1) Quiero comprar esa mesa redonda. ‘I want to buy that round table.’ Gender agreement is often postulated to be an automatic syntactic function whereby the gender of the noun is copied onto the words that agree with it (Franck et al., 2008). Indeed, monolingual Spanish-speaking adults’ gender agreement is generally flawless (Montrul, Davidson, de la Fuente, & Foote, 2014, p. 119). Yet, mismatches sometimes occur, as in su rendimiento en esta prueba no está asociada con . . . (Igoa, García-Albea, & Sánchez-Casas, 1999, p. 178). Such mismatches are more common when there is distance between the noun and its modifier. Also, mismatches are more easily elicited with noncanonical than with canonical nouns (Franck et al., 2008). 236

Child heritage speakers’ morphosyntax

When do children learn gender assignment and agreement? Monolingual children’s earliest combinations of nouns with articles or adjectives are lexically restricted chunks in which the gender of the article or adjective correctly matches the noun, as in la cuna ‘the crib,’ la llave ‘the key,’ otro bus ‘other bus,’ este coche ‘this car,’ otra sartén ‘other frying pan’ (Hernández Pina, 1984, p. 235). Shortly thereafter, children begin to treat the article and the noun as separate units and produce gender mismatches, such as la globo ‘the balloon’ and un llave ‘a key’ (Hernández Pina, 1984, p. 236; Montrul, 2004, pp. 46, 52). By around age 2;6, monolingual children’s article + noun combinations become systematically matched in gender (Hernández Pina, 1984, p. 233; Mariscal, 2008). Noun + adjective agreement is also systematic by around age 3, and researchers report few gender mismatches beyond this age (Hernández Pina, 1984, p. 236; Mariscal, 2008; Montrul, 2004, p. 53; Pérez Pereira, 1991). Thus, children’s acquisition of grammatical gender follows a U-shaped trajectory that proceeds from target-like utterances like la llave, to mismatches like un llave, and finally reverts back to gender-matched utterances. That is, performance is initially high, then dips, and then recovers. Such U-shaped development characterizes the acquisition of other morphological categories, such as past tense verbs in English (Marcus et al., 1992; Plunkett & Marchman, 1991), and suggests that the earliest ‘correct’ utterances do not represent knowledge of abstract gender assignment or agreement ‘rules’ or schemas. The age at which bilingual children detect and reproduce gender patterns very likely depends on the amount of input they receive (Granfeldt, 2016). Silva-Corvalán’s (2014, pp. 77–78) longitudinal study of her two English-dominant bilingual grandsons, Nico and Brennan, whom she recorded between ages 0 to 6, provides some initial information about gender assignment and agreement in the early years. In a random selection of the boys’ data, there were only three gender mismatches (e.g., Está muy bonito-MASC la tapa-FEM ‘It is very pretty, the lid’). As with monolingual children, the very earliest combinations are surely chunks and thus cannot provide evidence of generalized knowledge of gender patterns. Clearer evidence comes from Brennan’s overgeneralization of the masculine-o pattern at age 4;4 in (2), which suggests productive knowledge of the masculine-o gender assignment pattern. Also, by age 2;10, Nico produces predicate adjectives that agree with nouns, even when the noun and adjective are not adjacent (3), suggesting productive knowledge of agreement.   (2) Pero tengo mis manos-FEM limpios-MASC (Silva-Corvalán, 2014, p. 77) ‘But I have my hands clean’   (3) Bibi, mis sopas-FEM van a poner muy riquísimas-FEM (Silva-Corvalán, 2014, p. 78) ‘Bibi, my soups are going to turn out very good.’ Whereas Silva-Corvalán found only three mismatches, studies of school-age heritage speakers have uncovered many more. Mismatches appear to be most frequent between ages 6 and 8 and then decline between ages 9 and 11 (Anderson, 1999; Cuza & Pérez-Tattam, 2015, tables 4 and 5; Gathercole, 2002, 2007; Montrul & Potowski, 2007). Thus, the research suggests that child heritage speakers experience a second U-shaped pattern of development whereby the initial period of nearly flawless gender assignment and agreement is followed by a period of increased gender mismatches, and then finally an increase in target-like assignment and agreement. Do child heritage speakers acquire gender assignment and agreement more slowly than monolingual children do? It is possible that restricted Spanish input prolongs the earliest stage during which abstract assignment or agreement rules are not productive, but further research is needed to test this hypothesis. Montrul and Potowski (2007, p. 322) propose that child heritage 237

Naomi Shin

speakers’ gender mismatches may indicate that they never fully acquire gender agreement and “continue to make gender errors at the rate of a two-year-old monolingual child.” Indeed, child heritage speakers produce high rates of gender mismatches in experimental settings (Cuza & Pérez-Tattam, 2015; Montrul & Potowski, 2007, pp. 317–318). Yet, in their naturalistic production data, school-age English-dominant heritage speakers produce target-like agreement more often than not (Montrul & Potowski, 2007, pp. 315–316). A possible explanation is that heritage speakers do indeed acquire abstract gender agreement rules or patterns by school age, but the patterns are not as strongly entrenched for these bilinguals as they are for monolinguals. Such an explanation could account for sporadic deviations. With respect to gender assignment patterns in particular, child heritage speakers who experience reduced exposure to Spanish may take an especially long time to generalize over lexical items, especially when the pattern they must detect is weak. Protracted development is especially likely for noun types whose gender is relatively unpredictable, like those ending in –z. Due to low type frequency and high irregularity, learners need numerous exemplars to notice that nouns ending in –z are likelier to be feminine than masculine. On the other hand, high token frequency should override obstacles presented by low type frequency and pattern irregularity. For example, if a learner frequently hears la nariz ‘the-fem. nose,’ she will categorize nariz as feminine and will be unlikely to produce mismatches with this noun. But accuracy with particular words does not indicate grasp of a generalized pattern. A learner might produce nariz with feminine modifiers 100% of the time, but then vacillate between masculine and feminine for voz ‘voice.’ The question of whether child heritage speakers take longer to converge on the various patterns of gender assignment will best be answered by studies that examine both canonical and noncanonical noun gender assignment patterns and the fate of particular lexemes. Another question is whether heritage speakers experience attrition of gender assignment patterns. Surely all speakers may learn and then forget the gender of an infrequent noncanonical noun (Montrul, this volume; Montrul et al., 2014, p. 135), but whether or not learners may also acquire and then forget generalized patterns is a question that would be best answered by a longitudinal study. In summary, the literature thus far is inconclusive regarding rate of development of Spanish gender assignment and agreement. Children acquiring Spanish as a minority language may experience protracted development of gender patterns, but further research examining when abstract generalizations emerge is needed to test this hypothesis. At the same time, the research provides abundant evidence that school-age heritage speakers produce more gender mismatches than monolingual children do. Abundant mismatches may indicate slower acquisition of gender patterns or attrition of previously learned patterns. A third possibility is that child heritage speakers acquire and retain generalized gender patterns, but the patterns are not as deeply entrenched and thus are more susceptible to intermittent divergences.

Differential object marking (DOM) In Spanish, direct object marker a precedes some direct objects (DOs), as in (4a) but not (4b). (4a) Veo a la mujer. ‘I see the woman.’ (4b) Veo la mesa. ‘I see the table.’ Among adult Spanish speakers, direct object marking with a is probabilistically predicted by several factors including animacy and specificity of the referent, and whether the DO is a proper noun (Tippets, 2011; Torrego, 1999). For example, a precedes la mujer, which is animate and 238

Child heritage speakers’ morphosyntax

specific (4a), but not la mesa, which is inanimate (4b). Also, verbs that usually occur with an animate DO, such as matar ‘kill’ and saludar ‘greet’ are more likely to be followed by DOM a than verbs that occur with both animate and inanimate direct objects, such as tocar ‘touch,’ ver ‘see,’ and visitar ‘visit’ (Montrul & Sánchez-Walker, 2013, pp. 117–118; von Heusinger & Kaiser, 2011). When do children acquire differential object marking patterns? In a study of four monolingual children’s natural production data, Rodríguez-Mondoñedo (2008) found that 82% of 55 specific, animate DOs occurred with DOM marker a. Two children, whose data spanned ages 1;7 to 3;0, produced 42 of the 55 animate, specific DO tokens; the other two children, ages 0;9 to 2;11, produced only 13 tokens. While this study demonstrates that children begin to produce DOM between the ages of 2;0 and 3;0, several interesting questions remain. At what age do monolingual children produce DOM consistently? Do children produce a before any animate, specific DO and with any transitive verb, or do they first produce chunked sequences that are common in their input? Further research is needed to determine when monolingual children generalize over items and construct a grammar that probabilistically generates DOM based on factors such as animacy, specificity, and verbal semantics. Studies comparing monolingual and bilingual children suggest that the former produce higher rates of DOM than the latter, which in turn suggests protracted development among bilingual children. Ticio (2015) examined DOM produced by seven bilingual children, ages 1;1 to 3;6, who were simultaneously acquiring Spanish and English (four in contexts where English was the dominant language, two were in Spain, and one was in Cataluña). She found that a preceded 25% of the children’s 67 animate, specific DOs, i.e., a much lower rate than the 82% found by Rodríguez-Mondoñedo for monolingual children. Research on school-age children also suggests protracted development of DOM among bilingual children. Montrul and SánchezWalker (2013) compared 39 child heritage speakers in Chicago to 20 monolingual children in Mexico. The children, ages 6;0 to 17;0, narrated a story and also described pictures designed to elicit sentences with one participant doing something to a person or an object. The monolingual children produced a before animate, specific DOs significantly more often than the child heritage speakers did, especially during the picture description task (p. 117). Although the research thus far indicates that child heritage speakers acquire DOM patterns more slowly than monolingual children, it is not yet entirely clear why that is so. Perhaps it takes longer for child heritage speakers to generalize over lexical items due to being exposed to fewer exemplars that illustrate the relevant pattern. Further research investigating lexical restrictions on DOM during the early stages of language acquisition is needed to test this hypothesis.

Verb morphology Spanish verb endings encode information regarding person, number, and tense/mood/aspect. For example, in the verb tomábamos ‘we used to drink,’ the verb ending –bamos is comprised of the imperfective indicative past tense marker –ba-, and the first person plural marker –mos. How and when do monolingual children learn to produce these verbal morphemes? Research indicates that initially children produce only one verb form per verb (AguadoOrea & Pine, 2015; Fernández Martínez, 1994; Gathercole, Sebastián, & Soto, 1999). In other words, during this early stage, if a child produces quiere ‘want-3sg,’ she will not produce quiero ‘want-present-1sg’ or quise ‘want-pret.-1sg’ or any other form of querer ‘to want.’ Thus, early utterances show no person, tense, aspect, or mood contrasts in verb morphology. Person contrasts emerge at around 2 years of age, but at first the contrast is only between the first and third person and only evident in the present tense (e.g., quiero ~ quiere) (Gathercole et al., 1999, p. 145). Tense and aspect contrasts, which emerge later than person contrasts 239

Naomi Shin

(Pratt & Grinstead, 2007), also unfold in a piecemeal fashion and depend on the frequency of the form in the input (Jackson-Maldonado & Maldonado, 2001; Silva-Corvalán, 2014, pp. 290–295, 300, 344). For example, among Mexican monolingual children the first tense/aspect contrast is the present simple/preterit (Jackson-Maldonado, 2012, p. 162), whereas among monolingual children in Spain it is the present simple/present perfect (Gathercole et al., 1999, p. 145; Grinstead, 2000, p. 128). This difference in the order in which tense/aspect contrasts are acquired is indicative of distributional differences in the input. The present perfect is used much more frequently in Spain than in Mexico. Once a morphological paradigm becomes productive and not lexically restricted, we can assume the child has constructed an abstract grammatical representation of that paradigm. Does this process of abstraction happen more slowly for bilingual children than for monolingual children? Gathercole (2007) has argued that bilingual children follow the same pathways of development, but do so at a slower pace. Research on English-Spanish bilingual children supports this view. Consider Silva-Corvalán’s (2014, pp. 325–326) grandsons’ person morphology. Both boys substituted third person preterit forms for first person forms, as in yo mató ‘I killed-3sg,’ but Brennan, whose exposure to Spanish was more restricted, produced more person substitution errors than his brother Nico. Furthermore, Brennan’s substitution errors persisted throughout his second year of life, whereas Nico’s decreased substantially between 2.5 and 2.11. Brennan’s extended period of substitutions suggests that reduced Spanish input results in slower acquisition of person morphology. There is some evidence that child heritage speakers also experience protracted development of tense/mood/aspect (TMA) morphology. Jackson-Maldonado (2012, pp. 162–163) examined monolingual and bilingual children’s production of verbs by TMA category, at ages 1.8 and 2.4. The bilingual children in this study were in California and Spanish was their dominant language. The monolingual and bilingual children produced similar verb token frequencies in each TMA category. For example, in the simple present tense, monolinguals and bilinguals produced 85 and 92 verb tokens, respectively; in the periphrastic future, they produced 27 and 25 verb tokens, respectively. On the other hand, the monolinguals produced a greater variety of verb lexemes (Jackson-Maldonado & Maldonado 2001, pp. 186–187). Given that the ability to produce inflections with a larger range of word types is the hallmark of productivity, JacksonMaldonado’s (2012) study provides evidence of protracted development of productive TMA morphology among bilingual children, including those whose dominant language is Spanish. This conclusion is bolstered by Silva-Corvalán’s (2014, p. 332) study of Nico’s and Brennan’s verbs. Nico produced a wide variety of tenses by age six, including the present and imperfect subjunctive. In contrast, Brennan relied on frequent tenses, such as the indicative simple present, preterit, imperfect, and periphrastic future. The effects of reduced exposure to Spanish are also evident in bilingual children’s preterit and imperfect past tenses. In adult Spanish, stative and atelic predicates occur more frequently in the imperfect (era, estaba, trabajaba, etc.) than in the preterit (fue, estuvo, trabajó, etc.), whereas telic predicates occur more frequently in the preterit (e.g., murió, nació) than in the imperfect (e.g., moría, nacía) (Clements, 2009, pp. 15–17; Delgado-Díaz, 2014, p. 28). Scholars have argued that child heritage speakers struggle with contexts that contradict the typical pairings (Cuza & Miller, 2015; Potowski, 2005; Silva-Corvalán, 2014, p. 342; Zentella, 1997, p. 187). For instance, Silva-Corvalán (2014, p. 341) notes Nico’s use of imperfect era where preterit fue is expected in examples like Ed le destruyó algo a Tony y era un accidente ‘Ed tore something that belonged to Tony and it was an accident.’ Similarly, in her study of eighth graders in Chicago, Potowski (2005, p. 131) found that children who had lived in the U.S. since before they started school demonstrated an overreliance on the imperfect with stative verbs, whereas the 240

Child heritage speakers’ morphosyntax

eighth graders who had recently arrived in the U.S. from Latin America did not. Child heritage speakers also rely heavily on the pairing of the preterit with accomplishment and achievement predicates (Cuza & Miller, 2015). In summary, the literature indicates that the ability to pair the preterit with stative predicates and the imperfect with accomplishment verbs may take an especially long time to develop.

Summary: rate of acquisition Overall, the literature reviewed supports the generalization that child heritage speakers who experience restricted Spanish input acquire Spanish morphosyntax more slowly than monolingual children do. This difference in rate of acquisition is consistent with several theories of language acquisition. Usage-based approaches posit that children need to encounter a critical mass of exemplars before they generalize over items and extract a linguistic pattern (Gathercole, 2007; Tomasello, 2003). Current generative approaches also place a heavy burden on the amount of input needed to detect relevant linguistic patterns (Yang, 2003, 2016). Even though cognitive-functional and generative linguists disagree as to whether children construct a grammar or are born with innate knowledge of some abstract parts of grammar, both predict that less frequent exposure to a particular language will slow acquisition of that language’s grammar. The studies discussed provide evidence that acquisition of gender agreement and verb morphology is initially lexically restricted. As such, child heritage speakers’ slower rate of acquisition can be explained by frequency effects: If numerous encounters are needed to extract a generalization, and if child heritage speakers experience fewer encounters in Spanish than monolingual children, then it logically follows that they will take longer to detect generalizations. Slower development of some morphosyntactic patterns in one language must not be confused with overall language delay. There is no evidence suggesting that bilingual children, including those acquiring a minority language, are slower to develop language in a general sense. In fact, bilingual and monolingual children take about the same amount of time to reach developmental milestones. Bilingual and monolingual children alike tend to produce their first word at around 12 months old, and they start combining words anywhere between 18 and 24 months of age (De Houwer, 2009, pp. 37–40). Thus, the finding that English-Spanish bilingual children in the U.S. may experience some delay in their acquisition of Spanish morphosyntax is in no way discouraging, especially since bilingual children can and do learn morphosyntactic generalizations.

Crosslinguistic influence in bilingual children’s Spanish morphosyntax Linguists have proposed that crosslinguistic influence is more likely to affect morphosyntactic structures that show variability in usage (de Prada Pérez, 2015; Silva-Corvalán, 1994; Sorace, 2012). A pattern is considered relatively invariable if it is nearly categorical. For example, in Spanish, articles categorically appear before nouns. Articles in postnominal position would render a sentence ungrammatical, as in ∗Veo libro el. Thus the rule or schema that states that articles precede nouns is invariable or categorical. At the opposite end of the spectrum are highly variable phenomena, like subject pronoun expression. Although Spanish speakers are more likely to omit rather than express subject pronouns, there are very few linguistic contexts that require a pronoun or that prohibit its expression. Instead, linguistic contexts probabilistically constrain usage. For example, switching reference increases the likelihood that speakers will express a subject pronoun. According to Silva-Corvalán’s (2014) parallel structure hypothesis, crosslinguistic influence will only affect morphosyntax if the two languages have a structure in common, and if that structure 241

Naomi Shin

is used variably in the non-dominant language. Consider subject placement in declarative sentences. Spanish and English have a parallel structure: subject-verb word order. But in Spanish verb-subject order is also frequent, whereas in English it is not. The same reasoning applies to subject pronoun expression. The parallel structure is the expressed subject pronoun. But in Spanish subject omission is also frequent, whereas in English omission is rare. Does crosslinguistic influence affect morphosyntactic patterns that are variable more so than patterns that are invariable? The research to date suggests that invariable Spanish structures show clear influence from English in early childhood, but as children grow older their production of these English-like structures wanes. With respect to highly variable phenomena, there is evidence of English influence throughout the lifespan.

Crosslinguistic influence: low variability phenomena This section reviews research on preposition stranding, pre-nominal adjectives, and subject-verb word order in wh- questions, and considers whether child heritage speakers’ patterns of use reflect influence from English. From this point forward, glosses appear with examples if they are necessary for understanding the structure in question. Preposition stranding. In both Spanish and English, the object of a preposition typically appears after the preposition (5). Spanish and English differ, however, with respect to prepositional phrases in wh- questions (6) and relative clauses (7). In Spanish, the object of the preposition appears after the preposition; in English, the object appears at the front of the clause, resulting in a phenomenon called ‘preposition stranding.’1   (5) Estoy jugando con María. ‘I am playing with Maria.’  (6) ¿Con quién estás jugando? With who are-2SG play-GER ‘Who are you playing with?’   (7) María es la  niña con quien estoy jugando. Maria is the girl with whom am-1SG play-GER ‘Maria is the girl who I’m playing with.’ Between the ages of 3;0 and 5;0 child heritage speakers in the U.S. produce stranded prepositions in Spanish, whereas monolingual children do not (Silva-Corvalán, 2014, pp. 70–71; Vásquez Carranza, 2009). Examples from Silva-Corvalán’s grandson, Nico, are provided in (8) and (9).   (8) Bibi, ¿qué es esto para? ‘Bibi, what is this for?’   (9) Esa es la  cosa   que lo abriste    con That is the thing  that it opened-2SG  with ‘That is the thing that you opened it with.’ 242

Child heritage speakers’ morphosyntax

Given that monolingual children do not produce these constructions, preposition stranding in child heritage speakers’ Spanish is attributable to crosslinguistic influence from English. After age 5, preposition stranding in Spanish is rare in heritage speakers’ natural production data (Silva-Corvalán, 2014, p. 356; Vásquez Carranza, 2009, p. 95). Nevertheless, in experimental settings, even adult heritage speakers sometimes accept and produce it (Pascual y Cabo & Gómez Soler, 2015). Pre-nominal adjectives. In Spanish, restrictive adjectives typically appear after the noun they modify (Demonte, 1999), as in (10). (10) Quiero  ese lápiz  azul. Want-1SG that pencil blue ‘I want that blue pencil.’ Monolingual children produce restrictive adjectives in post-nominal position (Hernández Pina, 1984, p. 224; Montrul, 2004, pp. 38, 85, footnote 6). In contrast, child heritage speakers sometimes place restrictive adjectives in prenominal position, as in (11), produced by Silva-Corvalán’s (2014, p. 76) grandson Nico at age 3.6. (11) ¿Esta es chilena leche, Bibi? ‘This is Chilean milk, Bibi?’ Given that monolingual children do not produce examples like (11) – neither in naturalistic production data nor in experimental settings – child heritage speakers’ pre-nominal restrictive adjectives can be interpreted as the result of crosslinguistic influence from English. At the same time, these constructions are rare even among child heritage speakers, and after age 4;0, they become even scarcer (Silva-Corvalán, 2014, p. 75). Nevertheless, they can be elicited from older English-Spanish bilinguals in experimental settings (Cuza & Pérez-Tattam, 2015; Montrul & Potowski, 2007, p. 316), especially when prompts are given in English (Hsin, Legendre, & Omaki, 2013). Subject-verb word order in wh- questions. In non-Caribbean varieties of Spanish, subjects generally appear after the verb in wh-questions. Notice that in (12) the subject appears after the main verb in Spanish, whereas it appears before the main verb in English. (12) ¿Qué comió Carla? What eat-3SG Carla ‘What did Carla eat?’ Verb-subject (VS) wh- questions constitute a low variability phenomenon because VS order is typically considered obligatory in argument wh- questions in non-Caribbean Spanish (Austin, Blume, & Sánchez, 2013, p. 543, but see Goodall, 2004 and Raña Risso, 2013, p. 124). VS order is not obligatory in adjunct wh- questions, where SV order sometimes occurs, especially in por qué ‘why’ questions (Goodall, 2004; Torrego, 1984). Studies have found that monolingual children acquiring non-Caribbean varieties of Spanish categorically produce VS word order in both argument and adjunct wh- questions, as in (13) and (14), respectively, taken from Pérez-Leroux and Dalious (1998, p. 102) (also see Hernández Pina, 1984; Montrul, 2004, pp. 269–270). 243

Naomi Shin

(13) ¿Qué es esto? (Alfonso 2.6) ‘What is this?’ (14) ¿Dónde está la vaca? (Juan 2.3) ‘Where is the cow?’ In contrast, child heritage speakers sometimes produce wh- questions with subject-verb (SV) word order. Silva-Corvalán’s (2014, p. 66) grandsons, who were exposed to a non-Caribbean variety of Spanish, produced examples like (15) and (16) between ages 3 and 4. (15) ¿Cuándo tú   vas      a  vengar   a nuestra casa? When you going-2SG to come-INF to our house ‘When are you going to come to our house?’ (16) ¿Por qué  ella no    los  llevó? Why  she NEG  them take-3SG ‘Why didn’t she take them?’ Silva-Corvalán (2014, p. 66) notes that SV word order commonly occurred with por qué, but never with qué ‘what.’ Given that examples like (15) and (16) have not been found among monolingual children acquiring non-Caribbean varieties of Spanish, we might attribute them to influence from English. An English-contact interpretation is supported by the boys’ English wh- questions at this stage since they omit auxiliary do and thus produce questions that are unambiguously SV. For example, at age 2;8 Brennan said “Mommy, why Bibi – why Bibi have to change, change my clothes?” On the other hand, as discussed earlier, preverbal subjects do occur with adjunct wh- questions in adult Spanish. Perhaps child heritage speakers’ SV whquestions reflect structures found in the input, which are then reinforced by English word order. Experimental studies support the interpretation that child heritage speakers transfer whquestion word order from English into Spanish. Austin et al. (2013) conducted an elicited imitation experiment with 13 children in the U.S. who were ages 5;0 to 6;0 at the beginning of the study and 7;0 to 9;0 at the end. Children were asked to repeat wh- questions with postverbal subjects, as in (17) (17) ¿Qué comen tus papás? What eat-3PL your parents? ‘What do your parents eat?’ Results showed that the children almost always repeated VS word order, but there were four instances of SV rather than VS (e.g., ¿Qué tus papás comen?). Moreover, two of the four SV tokens occurred during the last session of the study, at which point the children were older and had had more exposure to English (Austin et al., 2013, p. 551). The conclusion that English wh- question word order influences child heritage speakers’ Spanish is supported by Cuza’s (2016) experimental study of 27 child heritage speakers born and raised in the United States, ages 5;0 to 13;3, and 18 monolingual children in Mexico, ages 6;6 to 12;4. Children were presented with short stories and prompts designed to elicit wh- questions 244

Child heritage speakers’ morphosyntax

in matrix and embedded clauses. The child heritage speakers produced SV wh- questions at high rates, whereas the monolingual children did not. Most of the SV wh- questions occurred with cómo, cuándo, and dónde ‘how,’ ‘when,’ and ‘where,’ rather than qué ‘what’ and quién ‘who’ questions. Thus, the child heritage speakers’ subject-verb wh- questions were concentrated in the very contexts that show variability in adult Spanish. Even so, the study uncovers a stark difference between school-age heritage speakers and monolinguals, especially for embedded wh- questions. Furthermore, the older child heritage speakers, who tended to be Englishdominant, produced more SV wh- questions than the younger ones, many of whom were Spanish-dominant. Overall, the research suggests that child heritage speakers produce more SV wh- questions than monolingual children and their amount of SV wh- questions increases with English dominance. A reasonable interpretation of these studies is that English wh- question word order influences these children’s Spanish. At the same time, child heritage speakers show the same asymmetry found among monolingual adults: they produce SV word order in adjunct whquestions, but rarely in argument wh- questions. Perhaps, then, child heritage speakers’ SV wh- questions reflect structures found in the input and these patterns are further reinforced by English word order. Summary: less variable morphosyntactic patterns. Very young heritage speakers produce English-like structures such as preposition stranding and pre-nominal restrictive adjectives in their Spanish. In contrast, adult heritage speakers rarely produce such structures in natural settings. These findings indicate that the impact of English word order on Spanish becomes less common as heritage speakers grow older. At the same time, these English-influenced structures can be elicited from adult and older child heritage speakers in experimental settings, suggesting that the structures remain in the speakers’ linguistic competence, but are suppressed as Spanish language patterns get reinforced with increased exposure. With respect to preverbal subjects in wh- questions, the discussion uncovered clear differences between monolingual children and child heritage speakers. The latter produce SV wh- questions more than the former do. On the other hand, child heritage speakers’ SV wh- questions are primarily adjunct questions, which are known to be variable in adult Spanish. Thus, their higher rate of SV wh- questions could be considered an example of a variable phenomenon where the variant that has a parallel structure in English is reinforced among bilinguals.

Crosslinguistic influence: variable phenomena This section reviews research on Spanish subject pronoun expression, subject-verb word order in declarative sentences, and variable clitic placement, and considers whether child heritage speakers’ patterns of use reflect crosslinguistic influence from English. Subject-verb word order in declarative sentences. Subjects of finite verbs in declarative sentences tend to appear before verbs (e.g., Gabriel está durmiendo ‘Gabriel is sleeping’), but sometimes appear after verbs (Está durmiendo Gabriel). The variation between pre- and post-verbal subject placement in adult Spanish is constrained by several linguistic factors. Post-verbal position is likelier when (i) the subject is a lexical noun phrase or clause rather than a pronoun, (ii) the subject referent is not human, (iii) the subject referent is new in the discourse, and (iv) the verb type is unaccusative or stative (e.g., Bentivoglio, 1988, 1989; Erker et al., 2017 and references therein). There is some evidence that, like adults, monolingual children produce VS word order more often with unaccusative verbs than other types (Bel, 2003, 2005), when the subject is new in the discourse, and when it is inanimate (Díaz-Campos, 2001). 245

Naomi Shin

If English influences Spanish subject placement, we should find in the Spanish spoken by English-Spanish bilinguals a higher rate of pre-verbal subjects and possibly some desensitization to the factors that constrain subject placement. Indeed, adult bilinguals raised in New York City produce higher rates of preverbal subjects than more recent arrivals do. Yet, the factors that guide speakers’ choice to place a subject in pre- or post-verbal position remain mostly intact (Erker et al., 2017). Research on child heritage speakers’ variable word order thus far indicates they too show elevated rates of preverbal subject placement. For example, SilvaCorvalán (2014, p. 206) finds that Nico’s and Brennan’s rates of preverbal subjects in Spanish were higher than those of monolingual Spanish-speaking children. Furthermore, their rates of preverbal subjects increased with age. For example, Brennan’s rate of preverbal subjects was 75% when he was 3;0–3;11 and 85% when he was 4;0–5;11. In contrast, three Chilean monolingual children, ages 4;6 to 4;11, produced preverbal subjects at a rate of 66%. At the same time, SilvaCorvalán’s (2014, p. 215) analysis of subject placement with particular verb lexemes and with animate versus inanimate subjects suggests that, for the most part, her grandsons were sensitive to semantic and discourse-pragmatic factors that constrain subject placement. Subject pronoun expression has frequently been used to test the hypothesis that English influences Spanish spoken by U.S. bilinguals. Spanish subject pronouns tend to be omitted more often than they are expressed (Otheguy & Zentella, 2012), whereas in English they are almost always expressed (Travis & Lindstrom, 2016). Many studies of adult bilinguals find that, compared to monolinguals, U.S. bilinguals express Spanish subject pronouns at higher rates, and show some changes in factors that guide pronoun expression (Otheguy & Zentella, 2012; Shin, 2014; Shin & Montes-Alcalá, 2014; Shin & Otheguy, 2009; but also see Torres Cacoullos & Travis, 2011). Studies of U.S. child heritage speakers also find increased subject pronoun use. SilvaCorvalán’s (2014, p. 153) grandson Brennan expressed Spanish subject pronouns at a rate of approximately 70% between ages 3;0 and 5;11. Such a high rate of subject pronoun use is not attested in studies of adult bilinguals in the U.S. (Shin & Otheguy, 2013). Also, Brennan’s pronoun use appeared to be unaffected by ‘switch-reference,’ which refers to whether a speaker maintains or switches reference across two consecutive grammatical subjects. All adult Spanish speakers and monolingual children express pronouns more often when reference is switched than when it is maintained (Carvalho, Orozco, & Shin, 2015). But Brennan expressed pronouns at a rate of 68% in same-reference contexts and 69% in switch-reference contexts (Silva-Corvalán, 2014, p. 157). An example of Brennan’s pronouns in same-reference contexts is illustrated by the boldface tokens of él ‘he’ in (18). (18) Sabes que cuando yo dijo: “para atrás” y no es parte de mi familia, él, dijo él “para atrás” y él empuja para atrás, tan, tan atrás porque él tiene esos [antenas]. ‘ . . . when I said “back” and is not part of my family, he, he said “back” and he pushes back, so, so much back because he has those [antennas].’ Brennan, age 4.19 (Silva-Corvalán 2014, pp. 158–159). Similar to Silva-Corvalán (2014), Montrul and Sánchez-Walker (2015) find elevated pronoun rates and a weaker impact of switch-reference among child heritage speakers in the U.S. as compared to monolingual children in Mexico. To summarize, compared to monolingual Spanish-speaking children, bilingual children – particularly child heritage speakers whose exposure to Spanish is relatively infrequent – express high rates of subject pronouns, and their usage is less constrained by switch-reference. Do these findings provide evidence of crosslinguistic influence? Some scholars have argued that it is 246

Child heritage speakers’ morphosyntax

not the language pair per se, but rather bilingualism itself, that results in an increase in subject pronouns (Michnowicz, 2015; de Prada Pérez, 2015; Sorace, 2012). Nevertheless, Brennan’s overall pronoun rate and distribution of pronouns in same- and switch-reference are unattested in other studies. Thus, it may be that English morphosyntax leaves its footprints on bilingual children’s Spanish subject pronoun expression, but only among those children who experience very restricted exposure to Spanish (Shin & Van Buren, 2016). Variable clitic placement. Another syntactic phenomenon that fits Silva-Corvalán’s (2014) criterion of variability is object clitic placement in [finite + nonfinite verb] constructions. An example of a clitic in preverbal position (‘proclisis’) is in (19), and an example in postverbal position (‘enclisis’) is in (20). (19) La      voy a  ve-r  mañana. her.acc.f3sg go.prs.1sg to see-inf tomorrow ‘I’m going to see her tomorrow.’ (20) Quiero ve-r= la want.prs.1sg see-inf=her.acc.f3sg ‘I want to see her.’ Given that direct object pronouns follow the verb in English, we might expect EnglishSpanish bilinguals to produce higher rates of enclisis than monolingual Spanish speakers. Pérez-Leroux, Cuza, and Thomas (2011) argue that this is what they find among 20 bilingual children in Canada, ages 3 to 8 years old. In their experiment children repeated sentences with proclisis, e.g., La princesa Jazmín lo puede ver esta noche ‘Princess Jasmine can see him tonight,’ or enclisis, e.g., Por la tarde Aladín quiere darme un caramelo ‘In the afternoon, Aladdin wants to give me a candy.’ For 25% of the prompts with proclisis, the bilingual children responded with enclisis. In a similar study, monolingual children produced enclisis when prompted with proclisis only 6% of the time (Eisenchlas, 2003, p. 203). Thus the bilinguals’ higher rate of repositioning to enclisis supports Pérez-Leroux et al.’s (2011, p. 230) conclusion that “word order preferences are affected by English in bilingual children.” Shin, Requena and Kemp (2017) also hypothesized that bilingual children would show an increase in enclisis in their natural production data. They examined proclisis/enclisis in sociolinguistic interviews conducted with 17 child heritage speakers in the U.S., ages 6;0–11;9. They compared these children to 21 adults from their same community and 43 monolingual children in Mexico, ages 6;3–11;9. They found no increase in enclisis and thus no evidence for impact from English. Instead, the children in Mexico and the children in the U.S. appeared to match the clitic placement patterns found among adults in their same communities. Nevertheless, the bilingual children studied by Shin et al. (2017) lived in communities with many monolingual Spanish speakers. Future studies should examine child heritage speakers who experience limited input in Spanish to further investigate whether contact with English results in increased enclisis.

Summary: crosslinguistic influence The results from studies examining crosslinguistic influence paint a complex picture where age and the variability of the linguistic phenomena are relevant. Broad generalizations extracted from the literature are presented in Table 16.1. 247

Naomi Shin Table 16.1  Crosslinguistic influence on child heritage speakers’ Spanish morphosyntax Less variability < age ~5

> age ~5

Preposition stranding

✔

Pre-nominal restrictive adjectives

✔

SV wh- questions

?

Experiments ✔ Natural production data ✗ Experiments ✔ Natural production data ✗ Experiments? Natural production data?

More variability SV in declaratives Subject pronoun expression Variable clitic placement

Natural production data ✔ Natural production data ✔ Experiment ✔ Natural production data ✗?

With respect to morphosyntactic phenomena that are relatively invariable, the studies indicate that before age 5 English impacts child heritage speakers’ Spanish. After age 5, however, Englishlike structures are absent in natural Spanish production data, but can be elicited during experiments. ‘SV wh- questions’ appears with questions marks in Table 16.1, illustrating the need to further examine this topic. Most of the child heritage speakers’ SV wh- questions were adjunct questions (por qué, cuándo, etc.). It remains unclear how often these children produce SV word order in wh- argument questions and whether such constructions occur in natural production data at all. With respect to morphosyntactic phenomena that are highly variable, there is evidence of crosslinguistic influence both before and after age 5. In particular, child heritage speakers display elevated use of subject pronouns and preverbal subjects. Nevertheless, it is difficult to determine whether these findings reflect impact from English structure or bilingualism in general. Elevated subject pronoun use is also found among bilinguals who speak two languages in which subject omission is common. It is possible that bilingualism accelerates patterns that are already highly frequent in the language (Silva-Corvalán, 1994). And perhaps such acceleration is intensified if the dominant language further reinforces the expanding pattern.

Conclusion This chapter explored bilingual Latin@ children’s Spanish morphosyntax in order to assess (1) whether rate of acquisition is the same for bilinguals and monolinguals, and (2) whether English impacts child heritage speakers’ Spanish. Studies of differential object marking and verbal morphology indicate that child heritage speakers acquire morphosyntactic patterns more slowly than monolingual children. It seems likely that a similar conclusion may be reached for gender assignment and agreement after further investigation. The broad generalization that reduced input slows acquisition of morphosyntax can be explained by current theories of language acquisition, which posit that children must encounter a certain amount of exemplars in order to generalize over items and construct abstract grammatical schema. The chapter also uncovers a complex picture in which linguistic variability is relevant for understanding crosslinguistic influence during bilingual language development. For phenomena that are relatively invariable, English influence is evident before age 5. After age 5, however, 248

Child heritage speakers’ morphosyntax

English influence is found in experimental settings, but not natural production data. The fact that English-like structures, such as preposition stranding in Spanish, can be elicited during experiments suggests that they are not discarded but instead are suppressed. Suppression of English-like structures overtime may be the result of children’s increasing ability to select features that correspond to communicative and social expectations (Otheguy, García, & Reid, 2015). With respect to highly variable linguistic morphosyntactic phenomena, such as Spanish subject pronoun expression and preverbal subject placement, there are differences between monolinguals and bilinguals beyond age 5. Nevertheless, it is difficult to discern whether such differences arise because of the structure of English or bilingualism in general (or both). In either case, the research supports Silva-Corvalán’s (2014) prediction that bilingual and monolingual children will differ in their production of highly variable morphosyntactic patterns.

Future directions The road ahead for researchers of childhood bilingualism and heritage language acquisition in particular is wide open, as many key questions remain unanswered. With respect to rate of acquisition, further research is needed to better understand when children grasp abstract patterns of gender assignment and agreement. We also need to gain a firmer grasp on the extent to which both monolingual and bilingual child language is formulaic. For example, it has been assumed that Spanish-speaking monolingual children acquire differential object marking very early. But we cannot conclude that children’s usage reflects a generalized morphosyntactic pattern until we investigate the possibility that their use of a is lexically restricted. The studies of gender agreement underscore this point. Children begin with formulaic noun phrases and later arrive at generalizations that move beyond particular lexical items. Is the same true for DOM and verbal morphology? If so, child heritage speakers’ slower rate of acquisition reflects prolonged acquisition of particular lexical combinations, especially infrequent ones. Another outstanding issue is that sometimes morphosyntactic patterns are assumed to be categorical, but upon further inspection it turns out they are in fact variable. DOM is a case in point. Rephrasing research questions so that variation is acknowledged would address this issue. For example, rather than asking when children learn to categorically use a before animate, specific direct objects, we can ask when and how children’s use of a becomes probabilistically constrained by several linguistic factors. Variability also appears to be an important factor that shapes crosslinguistic influence. As such, future studies should take into account amount of variability when predicting what parts of grammar will be affected by crosslinguistic influence. Finally, it was noted that experimental settings and corpus studies yielded divergent results. Thus, the use of multiple methods of investigation may help reveal the mental processes that underlie heritage language development and, more generally, the nature of heritage speakers’ mental grammars.

Note 1 Prepositions can also appear before the interrogative or relative pronoun in English, as in ‘With whom are you playing?’ and ‘Maria is the girl with whom I am playing,’ but this pied piping construction is restricted to very formal registers.

Acknowledgments The author thanks Osmer Balam, Jill Morford, Molly Perara-Lunde, Pablo Requena, Barbara Rodríguez, Marty Smith, and two anonymous reviewers for helpful feedback on this chapter. 249

Naomi Shin

Further reading Montrul, S. 2004. The acquisition of Spanish. Morphosyntactic development in monolingual and bilingual L1 acquisition and in adult L2 acquisition. Amsterdam: John Benjamins. Silva-Corvalán, C. 2014. Bilingual language acquisition: Spanish and English in the first six years. Cambridge, UK: Cambridge University Press. Zentella, A. C. 1997. Growing up bilingual: Puerto Rican children in New York. Malden, MA: Blackwell Publishers.

References Aguado-Orea, J., & Pine, J. (2015). Comparing different models of the development of verb inflection in early Child Spanish. PLoS ONE, 10, e0119613. doi:10.1371/journal.pone.0119613. Ambadiang, T. (1999). La flexión nominal. Género y número. In I. Bosque Muñoz & V. Demonte Barreto (Eds.), Gramática descriptiva de la lengua española. Vol. 3 (4843–4914). Madrid: Editorial Espasa Calpe. Anderson, R. (1999). Loss of gender agreement in L1 attrition: Preliminary results. Bilingual Research Journal, 23(4), 389–408. Austin, J., Blume, M., & Sánchez, L. (2013). Syntactic development in the L1 of Spanish-English bilingual children. Hispania, 96(3), 542–561. Bel, A. (2003). The syntax of subjects in the acquisition of Spanish and Catalan. Probus, 15, 1–23. Bel, A. (2005). Aspectos de la adquisición del orden de palabras: La posición del sujeto en Castellano y Catalán. Revista Electrónica de Lingüística Aplicada, 4, 36–48. Bentivoglio, P. (1988). La posición del sujeto en el español de Caracas: un análisis de los factores lingüísticos y extralingüísticos. In R. Hammond & M.C. Resnik (Eds.), Studies in Caribbean Spanish Dialectology (pp. 13–23). Washington, DC: Georgetown University Press. Bentivoglio, P. (1989). Función y significado de la posposición del sujeto nominal en el español hablado. Estudios sobre español de América y Lingüística Afroamericana, 83, 40–58. Carvalho, A. M., Orozco, R., & Shin, N. L. (2015). Subject pronoun expression in Spanish: A cross-dialectal perspective (xiii–xxvi). Washington, DC: Georgetown University Press. Clements, J. C. (2009). The linguistic legacy of Spanish and Portuguese: Colonial Expansion and Language Change. Cambridge, UK: Cambridge University Press. Cuza, A. (2016). The status of interrogative subject-verb inversion in Spanish-English bilingual children. Lingua, 180, 124–138. Cuza, A., & Miller, L. (2015). The protracted acquisition of past tense aspectual values in child heritage Spanish. In R. Klassen, J. Liceras, & E. Valenzuela (Eds.), Issues in Hispanic and Lusophone Linguistics (pp. 211–230). Amsterdam: John Benjamins. Cuza, A., & Pérez-Tattam, R. (2015). Grammatical gender selection and phrasal word order in child heritage Spanish: A feature re-assembly approach. Bilingualism: Language and Cognition, 19(1), 50–68. de Houwer, A. (2009). Bilingual First Language Acquisition. Bristol, UK and Buffalo, NY: Multilingual Matters. de Prada Pérez, A. (2015). First person singular subject pronoun expression in Spanish in contact with Catalan. In Carvalho, A. M., Orozco, R., & Shin, N. L. (Eds.), Subject Pronoun Expression in Spanish: A Cross-Dialectal Perspective (pp. 121–142). Washington, DC: Georgetown University Press. Delgado-Díaz, G. (2014). Teoría versus uso: análisis sobre el pretérito y el imperfecto. Boletín de Filología XLIX, 1, 11–36. Demonte, V. (1999). El adjetivo: Clases y usos. La posición del adjetivo en el sintagma nominal. In I. Bosque Muñoz & V. Demonte Barreto (Eds.), Gramática descriptiva de la lengua española. Vol. 1. (pp. 129–212). Madrid: Editorial Espasa Calpe. Díaz-Campos, M. (2001). La posición de la frase nominal sujeto respeto al verbo: un estudio del habla infantil caraqueña. Caracas, Venezuela: Fondo Editorial de la Facultad de Humanidades y Educación, Universidad Central de Venezuela. Eddington, D. (2002). Spanish gender assignment in an analogical framework. Journal of Quantitative Linguistics, 9(1), 49–75. Eisenchlas, S. (2003). Clitics in child Spanish. First Language, 23(2), 193–211. Erker, D., Ho-Fernández, E. Otheguy, R., & Shin, N. L. (2017). The order and expression of nominal and pronominal subjects among first- and second-generation Cubans in the U.S. In A. Cuza (Ed.), Cuban Spanish Dialectology: Variation, Contact and Change. (pp. 63–82). Washington, DC: Georgetown University Press.

250

Child heritage speakers’ morphosyntax Fernández Martínez, A. (1994). El aprendizaje de los morfemas verbales: Datos de un studio longitudinal. In S. López Ornat, A. Fernández, P. Gallo, & S. Mariscal (Eds.), La adquisición de la lengua española (pp. 29–46). Madrid: Siglo XXI de España Editores. Franck, J., Vigliocco, G., Antón-Méndez, I., Collina, S., & Frauenfelder, U. H. (2008). The interplay of syntax and form in sentence production: A cross-linguistic study of form effects on agreement. Language and Cognitive Processes, 23(3), 329–374. Gathercole, V. (2002). Grammatical gender in bilingual and monolingual children: A Spanish morphosyntactic distinction. In D. K. Oller & R. E. Eilers (Eds.), Language and Literacy in Bilingual Children (pp. 207–219). Clevedon, UK: Multilingual Matters. Gathercole, V. (2007). Miami and North Wales, so far and yet so near: A constructivist account of morphosyntactic development in bilingual children. International Journal of Bilingual Education and Bilingualism, 10(3), 224–247. Gathercole, V., Sebastián, E. & Soto, P. (1999). The early acquisition of Spanish verb morphology: Across-the-board or piecemeal knowledge? International Journal of Bilingualism, 3, 133–182. Goodall, G. (2004). On the syntax and processing of Wh-questions in Spanish. In B. Schmeiser, V. Chand, A. Kelleher, & A. Rodriguez (Eds.), WCCFL 23 Proceedings (pp. 101–114). Somerville, MA: Cascadilla Press. Granfeldt, J. (2016). The development of gender in simultaneous and successive bilingual acquisition of French: Evidence of AoA and input effects. Bilingualism: Language and Cognition. Online. DOI: https:// doi.org/10.1017/S1366728916001140. Grinstead, J. (2000). Case, inflection, and subject licensing in child Catalan and Spanish. Journal of Child Language, 27(1), 119–155. Hernández Pina, F. (1984). Teorías psicosociolingüísticas y su aplicación a la adquisición del español como lengua materna. Madrid: Siglo XXI de España. Hsin, L., Legendre, G. & Omaki, A. (2013). Priming cross-linguistic interference in Spanish-English bilingual children. In S. Baiz, N. Goldman, & R. Hawkes (Eds.), BUCLD 37: Proceedings of the 37th annual Boston University Conference on Language Development (pp. 165–177). Somerville, MA: Cascadilla Press. Igoa, J.-M., García-Albea, J. E. & Sánchez-Casas, R. (1999). Gender-number dissociations in sentence production in Spanish. Rivista di Linguistica, 11(1), 163–196. Jackson-Maldonado, D. (2012). Verb morphology and vocabulary in monolinguals, emerging bilinguals, and monolingual children with primary language impairment. In B. Goldstein (Ed.), Bilingual Language Development and Disorders in Spanish-English Speakers, 2nd edition (pp. 153–173). Baltimore, MD: Brookes. Jackson-Maldonado, D., & Maldonado, R. (2001). Determinaciones semánticas de la flexión verbal en la adquisición temprana de español. In C. Rojas-Nieto & L. De León Pasquel (Eds.), La adquisicioón de la lengua materna: español, lenguas mayas, euskera (pp. 165–200). Mexico City: UNAM-CIESAS. Marcus, G. F., Pinker, S., Ullman, M., Hollander, J. M., Rosen, T. J., & Xu, F. (1992). Overregularization in language acquisition. Monographs of the Society for Research in Child Development, 57(4, Serial No. 228). Mariscal, S. (2008). Early acquisition of gender agreement in the Spanish noun phrase: Starting small. Journal of Child Language, 36(1), 1–28. Merino, B. (1983). Language loss in bilingual Chicano children. Journal of Applied Development Psychology, 10, 477–494. Michnowicz, J. (2015). Subject pronoun expression in Yucatan Spanish. In A. M. Carvalho, R. Orozco, & N. L. Shin (Eds.), Subject Pronoun Expression in Spanish: A Cross-Dialectal Perspective (pp. 101–120). Washington, DC: Georgetown University Press. Montrul, S. (2004). The Acquisition of Spanish: Morphosyntactic Development in Monolingual and Bilingual L1 Acquisition and in Adult L2 Acquisition. Amsterdam: John Benjamins. Montrul, S., Davidson, J., De La Fuente, I., & Foote, R. (2014). Early language experience facilitates the processing of gender agreement in Spanish heritage speakers. Bilingualism: Language and Cognition, 17(1), 118–138. Montrul, S., & Potowski, K. (2007). Command of gender agreement in school-age Spanish-English bilingual children. International Journal of Bilingualism, 11(3), 301–328. Montrul, S., & Sánchez-Walker, N. (2013). Differential object marking in child and adult Spanish heritage speakers. Language Acquisition, 20(2), 109–132. Montrul, S., & Sánchez-Walker, N. (2015). Subject expression in bilingual school-age children in the United States. In A. M. Carvalho, R. Orozco, & N. L. Shin (Eds.), Subject Pronoun Expression in Spanish: A Cross-Dialectal Perspective (pp. 231–247). Washington, DC: Georgetown University Press.

251

Naomi Shin Otheguy, R., & Zentella, AC. (2012). Spanish in New York: Language Contact, Dialectal Leveling, and Structural Continuity. Oxford, UK: Oxford University Press. Otheguy, R., García, O., & Reid, W. (2015). Clarifying translanguaging and deconstructing named languages: A perspective from linguistics. Applied Linguistics Review, 6(3), 281–307. Pascual y Cabo, D., & Gómez Soler, I. (2015). Preposition stranding in Spanish as a heritage language. Heritage Language Journal, 12(2), 186–209. Pérez-Leroux, A., Cuza, A., & Thomas, D. (2011). Clitic placement in Spanish–English bilingual children. Bilingualism: Language and Cognition, 14(2), 221–32. Pérez-Leroux, A., & Dalious, J. (1998). The acquisition of Spanish interrogative inversion. Hispanic Linguistics, 10, 84–114. Pérez Pereira, M. (1991). The acquisition of gender: What Spanish children tell us. Journal of Child Language, 18, 571–590. Plunkett, K., & Marchman, V. (1991). U-shaped learning and frequency effects in a multi-layered perception: Implications for child language acquisition. Cognition, 38, 43–102. Potowski, K. (2005). Tense and aspect in the oral and written narratives of two-way immersion students. In D. Eddington (Ed.), Selected Proceedings of the 6th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages (pp. 123–136). Somerville, MA: Cascadilla Proceedings Project. Pratt, A., & Grinstead, J. (2007). Optional infinitives in child Spanish. In A. Belikova et al. (Eds.), Proceedings of the 2nd Conference on Generative Approaches to Language Acquisition North America (GALANA) (pp. 351–362). Somerville, MA: Cascadilla Proceedings Project. Raña Risso, R. (2013). A corpus-based sociolinguistic study of subject pronoun placement in Spanish in New York. PhD dissertation. CUNY Graduate Center, NY. Rodríguez-Mondoñedo, M. (2008). The acquisition of differential object marking in Spanish. Probus, 20(1), 111–145. Schwartz, M., Minkov, M., Dieser, E. Protassova, E. Moin, V., & Polinsky, M. (2015). Acquisition of Russian gender agreement by monolingual and bilingual children. International Journal of Bilingualism, 19(6), 726–752. Shin, N. L. (2014). Grammatical complexification in Spanish in New York: 3sg pronoun expression and verbal ambiguity. Language Variation and Change, 26(3), 303–330. Shin, N. L., & Montes-Alcalá, C. (2014). El uso contextual del pronombre sujeto como factor predictivo de la influencia del inglés en el español en Nueva York. Sociolinguistic Studies, 8(1), 85–110. Shin, N. L., & Otheguy, R. (2009). Shifting sensitivity to continuity of reference: Subject pronoun use in New York City. In M. Lacorte & J. Leeman (Eds.), Español en Estados Unidos y en otros contextos: Cuestiones sociolingüísticas, políticas, y pedagógicas (pp. 111–136). Madrid/Frankfurt: Iberoamericana/ Vervuert Verlag. 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, 429–452. Shin, N. L., Requena, P., & Kemp, A. (2017). Bilingual and monolingual children’s patterns of syntactic variation: Variable clitic placement in Spanish. In A. Auza and R. Schwartz (Eds.), From Typical Language Development to Language Disorders in Spanish-Speaking Children: Language Processing and Cognitive Functions (pp. 63–88). Cham, Switzerland: Springer. Shin, N. L., & Van Buren, J. (2016). Maintenance of Spanish subject pronoun expression patterns among bilingual children of farmworkers in Washington/Montana. Spanish in Context, 13(2), 173–194. Silva-Corvalán, C. (1994). Language Contact and Change: Spanish in Los Angeles. Oxford, UK: Oxford University Press. Silva-Corvalán, C. (2014). Bilingual Language Acquisition: Spanish and English in the First Six Years. Cambridge, UK: Cambridge University Press. Sorace, A. (2012). Pinning down the concept of interface in bilingual development: A reply to peer commentaries. Linguistic Approaches to Bilingualism, 2(2), 209–217. Ticio, E. (2015). Differential object marking in Spanish-English early bilinguals. Linguistic Approaches to Bilingualism, 5(1), 62–90. Tippets, I. (2011). Differential object marking: Quantitative evidence for underlying hierarchical constraints across Spanish dialects. In L. Ortiz-López (Ed.), Selected Proceedings of the 13th Hispanic Linguistics Symposium (pp. 107–117). Somerville, MA: Cascadilla Proceedings Project. Teschner, R., & Russell, W. (1984). The gender patterns of Spanish nouns: An inverse dictionary-based analysis. Hispanic Linguistics, 1, 115–132.

252

Child heritage speakers’ morphosyntax Tomasello, M. (2003). Constructing a Language. A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press. Torrego, E. (1984). On inversion in Spanish and some of its effects. Linguistic Inquiry, 15(1), 103–129. Torrego, E. (1999). El complemento directo preposicional. In I. Bosque Muñoz & V. Demonte Barreto (Eds.), Gramática descriptiva de la lengua española. Vol. 2 (1779–1805). Madrid: Editorial Espasa Calpe. Torres Cacoullos, R., & Travis, C. (2011). Testing convergence via code-switching: Priming and the structure of variable subject expression. International Journal of Bilingualism, 15(3), 241–267. Travis, C., & Lindstrom, A. (2016). Different registers, different grammars? Subject expression in English conversation and narrative. Language Variation and Change, 28(1), 103–128. Valdés, G. 2005. Bilingualism, heritage learners, and SLA research: Opportunities lost or seized? Modern Language Journal, 89, 410–426 Vásquez Carranza, L. M. (2009). Cross-linguistic influence in the syntactic domain in simultaneous language acquisition: Evidence from extraction constructions involving the object of a preposition in the speech of an English-Spanish bilingual child. Káñina, Rev. Artes y Letras, 33(1), 85–105. Von Heusinger, K., & Kaiser, G. A. (2011). Affectedness and differential object marking in Spanish. Morphology, 21(3–4), 593–617. Yang, C. (2003). Knowledge and Learning in Natural Language. Oxford, UK: Oxford University Press. Yang, C. (2016). The price of linguistic productivity: How children learn to break the rules of language. Cambridge, MA: MIT Press. Zentella, A. C. (1997). Growing up bilingual: Puerto Rican children in New York. Malden, MA: Blackwell Publishers.

253

17 SOCIOLINGUISTIC VARIATION IN U.S. SPANISH1 Rena Torres Cacoullos and Grant M. Berry pennsylvania state university, usa

The social context of Spanish in the U.S.: conceptual and methodological challenges The neglect of sociolinguistic patterns of language variation in U.S. Spanish is due at least in part to the abiding preoccupation with contact-induced change and a methodological predilection for acceptability judgments, experimental tasks, or cherry-picked examples. This “Hispanic tradition” of language study as Bills (1975: vi–vii) characterized it nearly half a century ago is hampered by an “interest in the accumulation of speech fragments with little concern for linguistic or sociological context” and “almost exclusive interest in deviations from standard Spanish.” Adherence to the analyst’s idealizations as the benchmark for evaluations leaves working-class varieties of U.S. Spanish in a no-win situation, as pointed out by Ana Celia Zentella; for example, New Mexican Spanish is branded ‘archaic’ “porque se describe en referencia a la norma de otra comunidad [because it is described in reference to the norm of another community],” but “tampoco se vale ser innovador . . . al notar la reacción . . . en contra de . . . lonche . . . y otros préstamos [it isn’t worth being innovative either . . . when one considers the reaction . . . against . . . lonche . . . and other borrowings]” (1990: 157). Methodological issues begin with data collection. As Peñalosa (1981: 7) asserted, appropriate data come from “Labovian-type studies of Chicano speech in a natural setting.” This is because the vernacular—the unreflecting use of language in the absence of the observer, when minimum attention is paid to monitoring speech—is the style that is most regular in structure (Labov 1972: 112). In contrast, when speakers of subordinate varieties are asked direct questions about their language, as is the case with acceptability judgments, their answers shift toward (or away from) the prestige variety in irregular and unforeseeable ways (Labov 1972: 111). Whether data are gathered by an in- or out-group member has also been demonstrated to make a difference, for example in rates of word-final nasal velarization by Salvadorans interviewed by a Mexican in Houston (Hernández 2011: 67). When we turn our attention to community-based samples of vernacular speech, social factors in tandem with linguistic constraints become important for diagnosing stability vs. change in U.S. Spanish varieties; these factors also serve to detect parallels vs. divergences vis-à-vis Spanish varieties spoken outside the U.S. across the Americas (e.g., Otheguy and Zentella 2012). Moreover, even for assessing contact-induced change, it has become 254

Sociolinguistic variation in U.S. Spanish

clear that neglect of the social context of bilingualism is risky, because once speakers are adequately characterized with respect to social factors, phenomena attributed either to majority language influence or to minority language loss may turn out to be conditioned by social class instead. An instructive example concerns use of the subjunctive mood, which undergoes attrition among at least some second- and third-generation speakers of Spanish in the U.S. (e.g., Ocampo 1990; Silva-Corvalán 1994: 86–90). A parallel presumed loss of the French subjunctive in Canada is imputed to contact with English. To test this, Poplack (1997) considered external measures of contact at the individual and community level. If contact with English is playing a role, speakers with higher English proficiency and those living in neighborhoods with a higher proportion of English speakers should show a lower rate of the subjunctive than speakers with lower indices of contact with English. Neither measure of contact correlated with subjunctive rate. Instead, after accounting for a strong lexical effect of the governing verb, systematic quantitative analysis of both internal and external constraints exposed the unsuspected effect of social class, with professionals displaying a proclivity for the subjunctive (Poplack and Levey 2010: 402–404). Moreover, subjunctive use is characterized not by change, but by long-term stable variability, despite centuries of normative injunctions (Poplack, Lealess and Dion 2013). In this chapter, we survey the scant number of reports on social factors in U.S. Spanish, first for stable linguistic variables—those with distribution patterns that persist across time and communities—and then for possible changes in progress—where age distributions are gradient. In the final section, we apply statistical procedures (principal component analysis and regression analysis) to a community-based corpus of New Mexican Spanish to infer and test social factors relevant to conditioning language variation.

Social class and gender in stable variation The “central dogma” of sociolinguistics stated by Labov is that “the community is prior to the individual” (Labov 2006: 5). Individual speaker behavior can be understood only once the community pattern is known, since individual linguistic behavior results from social histories and memberships. For Spanish-speaking communities in the U.S., the social factor most considered, though often in isolation, has been speaker gender. Usually, gender is implicated in claims of changes in progress, with women being seen in some cases as conservative and in others as leaders of linguistic change. Hasty pronouncements of change can stem from equating language change with perceived departures from an idealized norm or even with ordinary variation; that is, failing to differentiate between situations of stable variation and situations of change in progress. Crucially, while all change implies the existence of variability in language, the converse is not true: “not all variability and heterogeneity in language structure involves change” (Weinreich, Labov and Herzog 1968: 188). The “gender paradox” is the pattern of gender differentiation whereby “women conform more closely than men to sociolinguistic norms that are overtly prescribed, but conform less than men when they are not” (Labov 2001: 292–293). This follows from two generalizations for the distinct scenarios of variation: women use stigmatized variants at lower rates than men for stable sociolinguistic variables, but they adopt innovative forms earlier, both for prestige variants and for linguistic changes from below, i.e., changes from within the system which occur below the level of conscious awareness and consequently lack style shifting in initial stages (Labov 2001: 261–293). A stable sociolinguistic variable in Spanish is variation in the realization of the forms of the copula estar. This dates back approximately half a millennium, judging from the 255

Rena Torres Cacoullos and Grant M. Berry

recommendation of Juan de Valdés (1535) that the verb be written without e- to distinguish it from the demonstrative pronoun esta: [m]e ha parecido, por no hacer tropezar al letor, poner la e cuando son pronombres, porque el acento está en ella, y quitarla cuando son verbos, porque, estando el acento en la última, si miráis en ello, la primera e casi no se pronuncia, aunque se escriba [It has seemed a good idea to me, in order not to trip up the reader, to put the letter e before the words that are pronouns, because it is accented there, and not to include it in verbs, because, being that the accent is on the last syllable, if you think about it, the first e is almost not pronounced, even though it is written]. Drawing on 32 interviews recorded in San Antonio with Mexican Americans raised in South Texas, Garcia and Tallon (2000) examined three variants of estar: está, ‘stá (with apheresis of the vowel) and ‘tá (N = 1,025) in multivariate analysis. They find phonological conditioning by preceding segment. In addition, and in accordance with the generalization for stable variation, female speakers favored “canonical” está, while males favored the ‘tá variant, leading the authors to suggest that the latter may be “the least formal variant” and “a marker of male speech” (Garcia and Tallon 2000: 356–357). Another variable showing social stratification is the alternation between para and pa’. This is conditioned by morphosyntactic and phonological, but also social factors. A study based on data extracted from recordings with 171 speakers in San Antonio (token numbers not reported) indicates that males use the abbreviated form of the preposition at nearly double the rate of females (42% vs. 23%) (Lantolf 1982). What is more, gender interacts with education and occupation: although males display a higher rate of pa’ across all occupation and education levels, the gender difference is as high as 48 percentage points for blue-collar workers and as low as 5 for professionals, with a 20-point difference within white-collar workers (the largest group sampled) (Lantolf 1982: 172). Considering internal factors, the reduced form was favored in directional (locative) uses (Lantolf 1982: 167). The social and linguistic conditioning of para ~ pa’ in San Antonio parallels that found outside the U.S. Based on two analyses in Venezuelan corpora (48 speakers and 1,599 tokens in one, 72 speakers and 2,144 tokens in the other), Bentivoglio and Sedano (2011: 169–171) report that expressions of directionality are favorable contexts for pa’ (me fui pa Nueva York ‘I went to New York’), while para is preferred for purposives (para terminar ‘in order to finish’), and furthermore that following consonants promote the abbreviated form (as in pa’ comprar ‘in order to buy’), but that the strongest effect is that of socioeconomic level: the “low level” showed the highest rates of the reduced variant pa’. A class-based account is also proposed for a higher Spanish subject pronoun rate among Colombians and possibly Cubans who have lived in New York City for more than five years as compared with newcomers from those same countries. Shin and Otheguy (2013: 442–443) point to the high affluence rankings of these Latino national-origin groups in Census data, offering the conjecture that affluent Latinos are susceptible to influence from English due to looser social networks and more interaction with speakers of English. However, among the Colombians and Cubans sampled (N = 45), no effect is found for social class or education. Nevertheless, a “woman effect” is reported, which is most pronounced among those who were Latin-American born but had lived in NYC more than five years (Shin and Otheguy 2013: 439). While the gender effect among Colombians and Cubans may be because, as suggested by Shin and Otheguy (2013: 446), women have more contact with U.S.-born children or friends than men do, it also may be the case that women have higher rates of pronominal subject expression than men in Colombia to begin with (Orozco 2015: 30; see also Martín Butragueño and Lastra 2015: 50). 256

Sociolinguistic variation in U.S. Spanish

Understanding variation in Spanish among Latino New Yorkers necessitates knowledge of their social context (Otheguy and Zentella 2012: 149–150), in the same way that it is imperative to distinguish language contact settings due to immigration, as a result of conquest, or across national boundaries (Guadalupe Valdés 1982; cf. Poplack and Levey 2010: 396–397). The hypothesis of susceptibility to English or other-dialect influence would necessarily be tested by measures of degree of contact. Such measures have relied on self-reports, for example, speakers have been divided into in-group vs. out-group orientation groups based on reported frequency of interactions with speakers from other dialect regions (Otheguy and Zentella 2012: 109–112). Metrics for degree of contact with English may be derived from demographic data; in particular, the proportion of Spanish vs. English speakers in neighborhoods of residence (see Poplack and Levey 2010: 399, 402). A parallel measure could be applied to gauge participants’ level of interaction with speakers of other dialects, as has been done for contact between Salvadorans and Mexicans in Houston (Hernández 2009: 598–600). Direct measures of contact are best developed from sociolinguistic profiles culled from content analysis of recorded conversations constituting a corpus; for example, concerning time and location of acquisition of English, preferred or “most comfortable” language, language choice according to interlocutor, and general affect toward the bilingual situation (Poplack, Walker and Malcolmson 2006: 196–207; Torres Cacoullos and Travis 2018: 62–71). Poplack’s (1979) dissertation with Puerto Ricans living in Philadelphia remains a model study that has yet to be repeated in a U.S. Spanish community. Based on 24 sociolinguistic interviews collected from a neighborhood block in Philadelphia over a period of one year (Poplack 1979: 28–37), it was one of the first to apply rigorous statistical analysis (logistic regression and principal component analysis) to data on linguistic variation and show the effects of social predictors. Of the 24 participants, 15 were female, and most were working class or unemployed and had limited formal education and social mobility (Poplack 1979: 38–43). Poplack tested gender, age, education, language proficiency, and geographic origin (1979: 48–50) as they conditioned lenition of coda /s/ (N = 19,284), /n/ (N = 8,648), and /r/ (N = 7,142) (1979: 64, 108, 143).2 While there was very limited social conditioning of /n/ lenition (cf. Poplack 1979: 123, 127) for monomorphemic /r/ she found an increased lenition rate for males (Poplack 1979: 165), while for infinitival /r/ she found a slight effect of education (Poplack 1979: 172). With respect to coda /s/ lenition, social factors were particularly important. For plural /s/, only geographic origin and language proficiency were selected by the model (Poplack 1979: 86). For monomorphemic and verbal coda /s/, however, Poplack found that each of the five social factors tested (age, speech style, education, geographic origin, and language proficiency) conditioned lenition in word-final position (1979: 75, 96). It is important to note that the speakers studied here were from an immigrant population with a relatively short history (less than 50 years) in Philadelphia, which raises the question of the role of social factors given varying degrees of community stability and geographic permanence.

Changes in progress Linguistic change in Spanish in the U.S. is often proclaimed, though not as often demonstrated. Making a reasonable case for language change requires, above all, a robust quantitative pattern, which is verified in the speech of a community-based sample of speakers selected in a principled manner (Poplack et al. 2012). Changes in progress can be detected synchronically in apparent time—the distribution of variant forms across age cohorts (Labov 1994: 43–72). For example, in her pioneering sociolinguistic study of Panama City in 1969–71, Henrietta Cedergren (1973) observed a process of deaffrication from [tʃ] to the fricative [ʃ] in apparent 257

Rena Torres Cacoullos and Grant M. Berry

time, with an inverse relationship between lenition and age. The lenited variant increased as age decreased, peaking in the second youngest age group of 27- to 32-year-olds and slightly declining in the 15–26 group. Cedergren obtained data using the same sampling procedure in 1983 to see whether real-time evidence would suggest a genuine change in progress or age-grading—that is, change with age that repeats in each generation and results in stable community behavior in aggregate. In fact, her comparison revealed age-grading—the same pattern was followed across apparent time at each point in real time—but with lenition incrementally higher for all but the two youngest groups; this is interpreted to mean that [tʃ] lenition in this community had peaked (Labov 1994: 94–97). Deaffrication of [tʃ] showed a correlation with age in Tomé, in the Río Abajo region of New Mexico, just south of Albuquerque. Excluding postnasal and postlateral cases (planchar ‘to iron,’ el chile ‘the chili pepper’), which are categorically realized as affricates, Jaramillo and Bills (1982) give an apparent time interpretation to the distribution of the variants across age groups in a sample of 36 speakers (N = 1,029). They find a shift from the fricative variant to the affricate, as the rate of [ʃ] is nearly halved in the youngest (17–30) age group compared to approximately 80% in the older groups. The interpretation of a shift toward the more standard pronunciation is supported by considering the effect of education, operationalized as years of formal instruction. Since age and educational attainment partially overlap (a greater proportion of younger than older people had college education), Jaramillo and Bills (1982: 161) cross-tabulated age and education and found an independent effect for education. In fact, within the young group, eight speakers with a college education had a lenition rate approximately four times lower than that of the other four young speakers. Speakers with more than two years of formal study of Spanish also tended to lenite less often. The researchers conclude that the “perceived change” away from the “long-established” fricative variant “appears to simply reflect a sociological change related to education” whereby some residents are “expanding their command of different varieties of Spanish” (Jaramillo and Bills 1982: 163–164). Change in progress may be inferred from comparison of variation patterns in communities of origin. In Salvadoran communities, sequences of front vowels in hiatus with other vowels (as in vea ‘he/she/you(formal) sees (Subj)’) alternate with a hiatus-breaking [ʝ] variant (veya) (Lipski 1994: 258). Hernández (2015) compared rates and conditioning of the hiatus-breaking [ʝ] variant for the immigrants in Houston to comparable data from San Sebastián, El Salvador, the municipality of origin for most families. He reports that the rate of hiatus-breaking [ʝ] in Houston (6%, N = 737) is less than a third of that in San Sebastián (20%, N = 811), receding to 2% (N = 288) in the second immigrant generation. While in San Sebastián the hiatus-breaking variant was favored by older speakers and disfavored by women and those with a secondary school education, in Houston, with the now overwhelming preference for the hiatus variant, none of the social factors investigated—education level, gender, and age of the speaker—make a statistically significant contribution. Contrariwise, speaker gender does appear to contribute to linguistic variation in Kennett Square, Pennsylvania, though in a diminished way. Matus-Mendoza (2004) analyzed variable assibilation of word-final /r/ to a voiceless retroflex sibilant (/ɾ/→ [ř] \ __#; e.g., deci[ř] ‘to say, tell’) in a corpus of 83 sociolinguistic interviews with speakers in Moroleón, Guanajuato and Kennett Square, where many mushroom industry workers are from Moroleón. The linguistic conditioning of assibilated /r/ was the same in Moroleón (N = 2,796; Matus-Mendoza 2004: 21) and Kennett Square (unknown N). As for extralinguistic factors, rates in Moroleón differed across locales, with more frequent assibilation in urban than in rural areas, and across genders, with women assibilating more than men (Matus-Mendoza 2004: 20–22). Differences according to occupation and education level also indicate that assibilation is a prestige variant 258

Sociolinguistic variation in U.S. Spanish

in Mexico (Matus-Mendoza 2004: 26–27). In Kennett Square, the rates of assibilation increase with more schooling and among women, but the percentages are “extremely low . . . compared to . . . Moroleón” (6% among women in Kennett Square vs. 24% in Moroleón), suggesting an “equalizing situation” in the shared working environment (Matus-Mendoza 2004: 27). Contraction of a phonetic variant has also occurred in Houston, where Salvadorans live and work alongside Mexicans. Composition of neighborhoods of residence provides one measure to approximate degree of dialect contact. In Houston’s Segundo Barrio, Hispanics make up 90% of the population and the ratio of Salvadorans to Mexicans is on the order of one-to-ten, while in Holly Spring, where Hispanics constitute just 12% of the population, it is closer to one-to-two (Hernández 2011: 55). Hernández (2011) capitalizes on this difference to compare variable word-final nasal velarization (/n/→[ŋ] \ __#, as in los pueblos fueron [ˈfwe.ɾoŋ] los que sufrieron [su.ˈfɾje.ɾoŋ] más ‘the towns were the ones that suffered the most’) in three Salvadoran communities. The rate of nasal velarization declines in Houston compared with San Sebastián, El Salvador, the community of origin (23%, N = 430), but more so in Segundo Barrio (3%, N = 476) than in Holly Spring (14%, N = 981) (Hernández 2011: 66). On this basis, Hernández proposes that differences between the two Houston communities are explained by amount of exposure to speakers of Mexican Spanish (cf. Trudgill 1986: 39). One scenario of possible change in progress in a U.S. Spanish immigrant community, then, is dropping an alternation that constitutes a linguistic variable in the community of origin (see Weinreich [1968: 18–19], and Erker this volume, on Spanish dialectal contact in the U.S.). A contrary development may be the spread of a new linguistic variable. This appears to be the case with intervocalic ⟨ll⟩ ([ʝ]) deletion (e.g., iba a ir el bos por eØa a Brownsville ‘the bus was going to go to Brownsville for her’ vs. estudiar a Matamoros con ella ‘to study in Matamoros with her’) in the Segundo Barrio and Holly Spring neighborhoods in Houston. Hernández (2015) compared speech data from sociolinguistic interviews conducted among Salvadoran and Mexican immigrants. The participants were first generation, second generation or (in the case of the Mexican speakers) third generation, and most were from families from San Sebastián, El Salvador, or Matamoros, Mexico. The rate of intervocalic ⟨ll⟩ deletion in Houston is twice as high in the second and third than in the first generation among Mexicans (N = 383) and three times as high in the second generation than in the first among Salvadorans (N = 622) (Hernández 2015). This means that second-generation Mexican and Salvadorans show a closer elision rate (31% and 23%, respectively) than do their first-generation counterparts (17% and 5%, respectively). Though not significant in the multivariate analysis, there appears to be a tendency for higher elision rates for men than women, in both national origin groups, for this expanding phonological variable. As indicated by the studies surveyed in this chapter, linguistic realizations and social categories are linked, yet social factors remain understudied—particularly socio-economic status. Common belief holds that linguistic patterns in U.S. Spanish are unaffected by speakers’ socioeconomic status (Bills and Vigil 2008: 250). Some researchers even assert that speakers’ occupation or education should not be expected to correlate with minority language patterns since Spanish is not instrumental for success in the employment market (e.g., Garcia and Tallon 2000: 358, n.1). Contributing to the lack of studies of social factors is the problem of grouping speakers according to sociological characteristics. This is at least no less exacting in minoritylanguage situations than elsewhere, as the appropriateness of the criteria must be independently established for each community. For example, a solution for immigrant communities is offered by Orozco (2007: 105), who classified NYC Colombians into three groups by taking into consideration their occupations both in NYC and in Colombia: those who retained white-collar jobs, blue-collar workers before and after immigration, and blue-collar workers in NYC who held white-collar positions in Colombia. 259

Rena Torres Cacoullos and Grant M. Berry

But a remaining problem in general is that social categories, unlike linguistic categories, have no standard or agreed-upon methods of demarcation. An additional obstacle is that social groupings often correlate with one another, and as such it is disadvantageous to include them in omnibus in a statistical model. We now illustrate an alternative approach which can circumvent this problem by grouping speakers based on their linguistic behavior to infer social grouping.

Predicting social variation with linguistic behavior: clustering and stratification in New Mexican Spanish Hints of the social conditioning of variable usage can already be discerned in the earliest linguistic study of Spanish in the U.S. Over a century ago, Espinosa (1911: 10) in The Spanish Language in New Mexico and Southern Colorado suggests a social evaluation of the aspirated variant of /s/, qualifying it as “widespread among the rural uneducated classes.” Dating back to 16th- to 17th-century settlement from New Spain (today, Mexico), Northern New Mexico is home to (Traditional) New Mexican Spanish. As Lipski (2000: 2–4) has noted, New Mexican Spanish was deemed by Espinosa and contemporary linguists in Latin America and Spain to be nothing less than another national variety of the language. In New Mexico, it is the speakers of English, not Spanish, who are (descendants of) immigrants. In 1850, the area became a U.S. territory, and in 1878, the railroad arrived along with accelerating Anglo-American immigration. In 1912, New Mexico was admitted to the Union as the 47th state and English increasingly displaced Spanish in schools—even in northern, longstanding Spanish-speaking communities—by the 1940s. Today Spanish is taught as a foreign language and, while Hispanics represent as much as 80% of the population in some northern counties, there is a continued shift toward English (Bills and Vigil 2008, inter alia). The remaining speakers of New Mexican Spanish provide an invaluable window into Spanish language use in a native community. The New Mexico Spanish-English Bilingual (NMSEB) corpus consists of spontaneous speech collected by in-group community members and thoroughly transcribed in prosodic units. Participants were selected to cover a range of demographic backgrounds to permit the assessment of extra-linguistic constraints on linguistic variation (Travis and Torres Cacoullos 2013; Torres Cacoullos and Travis 2018: 24–33). To identify those social factors that may contribute a consistent effect on linguistic variation in the data, we cast a wide net by looking at the problem in reverse (Horvath and Sankoff 1987; Poplack 1979: 190–223). We will use the linguistic behavior of speakers in NMSEB to cluster them via a principal component analysis (PCA), and then interpret the resultant configurations in terms of our extralinguistic knowledge of the speakers to identify the social characteristics that individuals within those clusters have in common. PCA is a data optimization method used to partition a multidimensional space into several orthogonal components that reduce the dimensionality of that space; the dimensions that contribute toward partitioning the variance of that space are called the principal components. A PCA works best when there is a high amount of variance in the space; typically, this is found when each row vector (in this case, speaker) has more than ten numeric variables, or dimensions (in this case, linguistic features) (cf. Horvath and Sankoff 1987: 186). We focus on four phonetic variables, which yield 14 such dimensions: onset (syllable-initial) /s/ lenition, coda (syllable-final) /s/ lenition, intervocalic /d/ elision, and intervocalic ⟨ll⟩ lenition. Each of these has been studied and implicated as either characteristic of New Mexican Spanish or as a stable, socially stratified variable in other dialects of Spanish (e.g., Espinosa 1909: 260

Sociolinguistic variation in U.S. Spanish

72, 75; Gutiérrez 1981; Lapesa 1968: 354, 356; Lipski 2011: 75–83; Samper-Padilla 2011: 105–114). The dimensions for the PCA were based on the linguistic constraints for each of these four phonetic variables. Onset /s/ lenition: Favored and other contexts. Since onset /s/ lenition was most strongly favored by preceding non-high vowels—as in ese ‘that one’ or la señora ‘the woman’ (cf. Brown 2005a)—counts of onset /s/ in favorable phonetic contexts (preceding non-high vowels) were separated from counts of onset /s/ in other phonetic environments. Additionally, since complete elision of onset /s/ was rare, tokens were divided into full ([s]) and lenited variants and counts were included separately. Doing so produced counts of four variants: full onset /s/ preceded by a non-high vowel, lenited onset /s/ preceded by a non-high vowel, full onset /s/ in other contexts, and lenited onset /s/ in other contexts. Coda /s/: Favored and other contexts. Lenition was most strongly favored when the following phone was a voiced consonant, as in desde ‘since/from’ or los viejitos ‘the old people’ (cf. Brown 2005b). Unlike onset /s/, coda /s/ showed a mix of full ([s]), aspirated ([h]), and elided (Ø) variants, so we considered counts of each separately. This produced six additional variants per speaker (each of the three variants followed by a voiced consonant and in other environments). Intervocalic /d/ elision: We take counts of intervocalic approximants against the number of elided intervocalic tokens (in which there was no perceptible frication as well as audible vowel coarticulation, e.g., casado [kasau]; casada [kasa=ː] ‘married’). ⟨ll⟩ lenition: In words like ellos ‘they/them,’ reduced and completely elided forms were grouped together. In all, there were 14 variants across the 4 phonetic variables. Since onset and coda /s/ were further subdivided by phonetic environment, this produced six categories: onset /s/ in a favorable environment for lenition, onset /s/ in other environments, coda /s/ in a favorable environment for lenition, coda /s/ in other environments, intervocalic /d/, and intervocalic ⟨ll⟩. If a participant had fewer than 20 tokens in total for any of these categories (summing up all variants within those categories), their counts for all variants in that category were zeroed out to keep low token counts from warping the PCA output. The principal components resulting from the PCA3 were then plotted based on the amount of variance each principal component accounted for. Three principal components accounted for 78% of the total variance in the dataset. We then examined the associations of each of the 14 variants with each of these three principal components. Many of the variants showed moderate associations, or loadings (with magnitude greater than 0.3; |PCx| >0.3) (cf. Horvath and Sankoff 1987: 194), indicated by bolded text and cell shading in Table 17.1; variants with weaker associations (0.25 ≤ |PCx| < 0.3) are listed in bold without shading. We interpret the loadings as follows. Principal Component 1 (PC1), which accounts for 46% of the variance, appears to represent lenition in general. That is, speakers who have (negative) associations with this component are more likely than other speakers to aspirate onset /s/ and will also tend to lenite (aspirate) coda /s/ and to use lenited intervocalic /ʝ/ as well. Principal Component 2 (PC2), accounting for another 21% of the variance, is largely the complement of PC1. Here retention of full variants is (positively) associated with PC2.4 PC 3, which accounts for 11% of the total variance, is more complicated. Both onset /s/ retention in other than preceding non-high-vowel environments (i.e., in disfavorable contexts for /s/ aspiration) and coda /s/ lenition, especially ø, pattern in the same direction (negatively), and these are contrasted with intervocalic ⟨ll⟩ lenition and, though its association is marginal, onset /s/ aspiration in preceding non-high-vowel contexts. This component, then, groups more standard and general Spanish linguistic patterns, namely onset /s/ retention and coda /s/ lenition, in opposition to traditional New Mexican variants, that is, intervocalic ⟨ll⟩ lenition and onset /s/ aspiration. 261

Rena Torres Cacoullos and Grant M. Berry Table 17.1  Loadings of 14 consonantal variants in New Mexico (NMSEB) on principal components Variable

Variant (Dimension)

PC1

PC2

PC3

Onset /s/

[s].Preceding NonHighV [h]/ø.Preceding NonHighV [s].Other environments [h]/ø.Other environments [s].Following Voiced C [h].Following Voiced C ø.Following Voiced C [s].Other environments [h].Other environments ø.Other environments Approximant intervocalic /d/ Elided intervocalic /d/ (ø) Full intervocalic /ʝ/ Lenited intervocalic /ʝ/ Variance accounted for: Interpretation:

0.24 −0.33 0.13 −0.31 0.20 −0.30 −0.28 0.28 −0.34 −0.28 −0.15 −0.25 −0.23 −0.30 46% Lenition (general)

0.40 −0.11 0.38 −0.03 0.34 0.13 0.09 0.30 0.05 −0.09 0.43 0.31 0.34 0.21 21% Retention (general), except /d/

−0.11 −0.01 −0.45 0.21 −0.15 −0.33 −0.40 −0.04 −0.26 −0.36 0.26 0.21 0.22 0.32 11% NM Spanish vs. other dialects

Coda /s/

Intervocalic /d/ Intervocalic ⟨ll⟩

Effectively, the PCA has taken a 14-dimensional space representing each variant of our 4 variables and reduced it to a three-dimensional space where highly correlated items pattern together. This permits a spatial representation of the data, which can elucidate similarities in speaker behavior, but it also, crucially, illustrates the associations of the phonetic variants to one another. Through such an analysis, we apprehend that participant groupings are strongly determined by patterns of lenition. With a general linguistic interpretation of the principal components in mind, we then ask how individual speakers associate with each of the principal components. In Figure 17.1, by plotting each participant according to their loading on the first two Principal Components, and letting the shading indicate the third Principal Component, we capture the results of the PCA visually and use those results to cluster participants. In doing so, we observe that the participants naturally fall into three main groups, primarily delineated by PC1 (indicated by shape in Figure 17.1). Using these speaker clusters based on linguistic behavior, we compared sociodemographic characteristics of the speakers to assess what was shared among most members. Group 1 mostly consists of miners, factory workers, or ranchers who are men with a middle or high school education. Group 2 is mainly constituted by middle- or high-school-educated men and women, some in production (e.g., factory workers), and some in service (e.g., in dry cleaning) occupations. Group 3 is a more urban, predominately female group, in which we find most of the participants with (some) college education and/or professional occupations (e.g., teachers). Based on these clusters, then, it appears that socioeconomic status (occupation and education), gender, and rural vs. urban locale should be considered as candidates for conditioning linguistic variation in NM Spanish. For a composite socioeconomic index based on occupation and education, we grouped speakers into ‘production workers’ (N = 14), ‘service employees’ (N = 15), and ‘professionals’ (N = 9). There were 22 women and 16 men. As to locale, ‘urban’ were those participants from cities with 10,000 or more residents (Albuquerque, Española, Las Vegas, Los Lunas, Santa Fe) (N = 11). 262

Sociolinguistic variation in U.S. Spanish

Figure 17.1  Grouping of NMSEB speakers by linguistic behavior (from PCA)

To obtain additional evidence that these social factors may be predictors of variation, we determined whether they were distributed unevenly among the three speaker clusters, via Fisher’s exact tests. According to these, gender (p < 0.05) was disproportionately distributed among groups, with males being more common in Group 1 and females in Group 3. Subsequently, Fisher’s exact tests conducted pairwise indicated that occupation-education was also differentially distributed across Groups 1 and 3 (p < 0.05), suggesting that both social class and gender may be useful categories for conditioning linguistic behavior. Additionally, there seems to be a slight bias toward rural speakers in Group 1 (7 of 8 are rural), though this does not reach statistical significance. Thus, we include Rural vs. Urban locale as a social predictor, understanding that this characteristic may not be as robust as gender or socioeconomic status in distinguishing participants’ linguistic behavior. These three social predictors were considered together with linguistic factors in generalized linear mixed models (conducted using the lme4 package [Bates et al. 2014] in R [R Core Team 2015]).5 A separate model was fit for each of the four phonetic variables, and this model was compared to models with only linguistic and only social predictors (via likelihood ratio tests). While both onset and coda /s/ lenition were primarily determined by linguistic factors (as also reported by Brown 2005a, 2005b), model fits for intervocalic /d/ elision and ⟨ll⟩ lenition improved with the inclusion of a combination of linguistic and social predictors. Tables 17.2 and 17.3 show the results of generalized linear mixed models for intervocalic /d/ elision and ⟨ll⟩ lenition, respectively. The Intercept refers to the estimated log-likelihood of a dependent variable at a given reference level (reference levels are listed below each table). Levels of each predictor are assigned a weighting (β-coefficient), or Estimate, with positive values indicating an increased likelihood and negative values indicating a decreased likelihood for a given level (factor). For example, in the case of intervocalic /d/ elision, the positive Estimate for a preceding non-high vowel suggests that this phonetic context increases the likelihood of elision.6 Also indicated in the model outputs is significance (determined by estimated p-values computed via a Wald test). Intervocalic /d/ deletion is strongly affected by social class. In agreement with reports on Latin American varieties of Spanish including Panamanian (Cedergren 1973) and Venezuelan 263

Table 17.2  Social and linguistic factors conditioning intervocalic /d/ elision in NMSEB (N = 3,447)∗ Factor Intercept Preceding non-high V (Preceding high V) Participle (Not a participle) Production vs. other occupations Service vs. professional (Production occupation) (Service occupation) (Professional occupation) Rural locale (Urban locale) Men (Women) Preceding non-high V: participle (Prec. non-high V, non-participle) (Prec. high V, participle) (Prec. high V, non-participle)

Estimate

Std. error

Sig.

−3.74 3.08

0.54 0.57

∗∗∗ ∗∗∗

1.65

0.59

∗∗

−1.28 −0.12

0.42 0.42

∗∗

−0.09

0.29

0.60

0.29

∗

1.52

0.78

∗

N

% Elision

2,788 659 471 2,976

16% 3% 34% 10%

1,753 1,132 444 2,376 1,065 1,633 1,808 320 753 150 288

18% 10% 9% 14% 12% 17% 11% 48% 14% 3% 1%

∗Generalized linear mixed model, lme4 package (Bates et al. 2014) in R (R Core Team 2015) Random effects (SD): Speaker (0.55); Word (2.01) Reference level: /d/ present, prec high V, non-participle, urban, female |∗∗∗ p