The Handbook of Bilingualism and Multilingualism [2 ed.] 1444334905, 9781444334906

Citation preview

The Handbook of Bilingualism and Multilingualism

Blackwell Handbooks in Linguistics This outstanding multi-volume series covers all the major subdisciplines within linguistics today and, when complete, will offer a comprehensive survey of linguistics as a whole. Already published: The Handbook of Child Language Edited by Paul Fletcher and Brian MacWhinney

The Handbook of English Linguistics Edited by Bas Aarts and April McMahon

The Handbook of Phonological Theory, Second Edition Edited by John A. Goldsmith, Jason Riggle, and Alan C. L. Yu

The Handbook of World Englishes Edited by Braj B. Kachru; Yamuna Kachru, and Cecil L. Nelson

The Handbook of Contemporary Semantic Theory Edited by Shalom Lappin The Handbook of Sociolinguistics Edited by Florian Coulmas The Handbook of Phonetic Sciences, Second Edition Edited by William J. Hardcastle and John Laver The Handbook of Morphology Edited by Andrew Spencer and Arnold Zwicky The Handbook of Japanese Linguistics Edited by Natsuko Tsujimura The Handbook of Linguistics Edited by Mark Aronoff and Janie Rees-Miller The Handbook of Contemporary Syntactic Theory Edited by Mark Baltin and Chris Collins The Handbook of Discourse Analysis Edited by Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton The Handbook of Language Variation and Change Edited by J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes The Handbook of Historical Linguistics Edited by Brian D. Joseph and Richard D. Janda The Handbook of Language and Gender Edited by Janet Holmes and Miriam Meyerhoff The Handbook of Second Language Acquisition Edited by Catherine J. Doughty and Michael H. Long The Handbook of Bilingualism and Multilingualism, Second Edition Edited by Tej K. Bhatia and William C. Ritchie The Handbook of Pragmatics Edited by Laurence R. Horn and Gregory Ward The Handbook of Applied Linguistics Edited by Alan Davies and Catherine Elder The Handbook of Speech Perception Edited by David B. Pisoni and Robert E. Remez The Handbook of the History of English Edited by Ans van Kemenade and Bettelou Los

The Handbook of Educational Linguistics Edited by Bernard Spolsky and Francis M. Hult The Handbook of Clinical Linguistics Edited by Martin J. Ball, Michael R. Perkins, Nicole Müller, and Sara Howard The Handbook of Pidgin and Creole Studies Edited by Silvia Kouwenberg and John Victor Singler The Handbook of Language Teaching Edited by Michael H. Long and Catherine J. Doughty The Handbook of Language Contact Edited by Raymond Hickey The Handbook of Language and Speech Disorders Edited by Jack S. Damico, Nicole Müller, Martin J. Ball The Handbook of Computational Linguistics and Natural Language Processing Edited by Alexander Clark, Chris Fox, and Shalom Lappin The Handbook of Language and Globalization Edited by Nikolas Coupland The Handbook of Hispanic Linguistics Edited by Manuel Díaz-Campos The Handbook of Language Socialization Edited by Alessandro Duranti, Elinor Ochs, and Bambi B. Schieffelin The Handbook of Intercultural Discourse and Communication Edited by Christina Bratt Paulston, Scott F. Kiesling, and Elizabeth S. Rangel The Handbook of Historical Sociolinguistics Edited by Juan Manuel Hernández-Campoy and Juan Camilo Conde-Silvestre The Handbook of Hispanic Linguistics Edited by José Ignacio Hualde, Antxon Olarrea, and Erin O’Rourke The Handbook of Conversation Analysis Edited by Jack Sidnell and Tanya Stivers

The Handbook of Bilingualism and Multilingualism Second Edition Edited by

Tej K. Bhatia and William C. Ritchie

A John Wiley & Sons, Ltd., Publication

This second edition first published 2013 © 2013 Blackwell Publishing, Ltd Edition History: Blackwell Publishing, Ltd (1e, 2004) Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Registered Office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Offices 350 Main Street, Malden, MA 02148-5020, USA 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, for customer services, and for information about how to apply for permission to reuse the copyright material in this book please see our website at www. wiley.com/wiley-blackwell. The right of Tej K. Bhatia and William C. Ritchie to be identified as the authors of the editorial material in this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data The handbook of bilingualism and multilingualism / edited by Tej K. Bhatia and William C. Ritchie. – Second edition.     pages cm   Extensive revision of: The handbook of bilingualism. 2004.   Includes bibliographical references and index.   ISBN 978-1-4443-3490-6   1.  Bilingualism.  2.  Multilingualism.  I.  Bhatia, Tej K.  II.  Ritchie, William C.  III.  Handbook of bilingualism.   P115.H365  2012   404'.2–dc23 2012010708 A catalogue record for this book is available from the British Library. Jacket image: © Artelia / Dreamstime.com Jacket design by Workhaus. Set in 10/12 pt Palatino by Toppan Best-set Premedia Limited 1  2013

To the memory of the thirty-five Syracuse University students who lost their lives in the terror bombing of Pan American Airways flight 103 over Lockerbie, Scotland, on December 21, 1988, and to all other innocent victims of terrorism.

Contents

Acknowledgments Contributors Introduction Tej K. Bhatia and William C. Ritchie

xi xii xxi

Part I  Overview and Foundations

1



3

Introduction Tej K. Bhatia   1 Bilingualism and Multilingualism: Some Central Concepts John Edwards   2 Conceptual and Methodological Issues in Bilingualism and Multilingualism Research Li Wei

5 26

Part II  Neurological and Psychological Aspects of Bilingualism and Multilingualism

53



Introduction William C. Ritchie

55

The Neurology of Bilingualism and Multilingualism

59

  3 Bilingual Aphasia: Theoretical and Clinical Considerations Elizabeth Ijalba, Loraine K. Obler, and Shyamala Chengappa Approaches to Bilingualism, Multilingualism, and Second-Language Acquisition   4 The Bilingual Child Ludovica Serratrice   5 Bilingualism/Multilingualism and Second-Language Acquisition Yuko Goto Butler   6 Multilingualism: New Perspectives on Syntactic Development Éva Berkes and Suzanne Flynn

61

85 87 109 137

viii  Contents   7 Bilingualism and the Heritage Language Speaker Silvina Montrul Bilingual and Multilingual Language Use: Knowledge, Comprehension, and Production   8 Two Linguistic Systems in Contact: Grammar, Phonology, and Lexicon Pieter Muysken   9 The Comprehension of Words and Sentences in Two Languages Judith F. Kroll and Paola E. Dussias 10 An Appraisal of the Bilingual Language Production System: Quantitatively or Qualitatively Different from Monolinguals? Elin Runnqvist, Ian FitzPatrick, Kristof Strijkers, and Albert Costa Bilingualism and Multilingualism: Memory, Cognition, and Emotion 11 Bilingual Memory Roberto R. Heredia and Jeffrey M. Brown 12 Bilingualism and Emotion: Implications for Mental Health Ines Martinovic and Jeanette Altarriba The Bilingual’s and Multilingual’s Repertoire: Code-Mixing, Code-Switching, and Communication Accommodation

168

191 193 216 244

267 269 292

321

13 Code-Switching and Grammatical Theory Jeff MacSwan 14 Sign Language–Spoken Language Bilingualism and the Derivation of Bimodally Mixed Sentences Gerald P. Berent 15 Social and Psychological Factors in Language Mixing William C. Ritchie and Tej K. Bhatia 16 Accommodating Multilinguality Itesh Sachdev, Howard Giles, and Anne Pauwels 17 Bilingualism and Gesture Marianne Gullberg

323

Part III  Societal Bilingualism/Multilingualism and its Effects

439



Introduction Tej K. Bhatia

441

Language Contact, Maintenance, and Endangerment

443

18 The Bilingual and Multilingual Community Suzanne Romaine

351 375 391 417

445

Contents  ix 19 Language Maintenance, Language Shift, and Reversing Language Shift Joshua A. Fishman 20 Linguistic Imperialism and Endangered Languages Robert Phillipson and Tove Skutnabb-Kangas 21 Multilingualism, Indigenization, and Creolization Jeff Siegel 22 Multilingualism and Family Welfare Xiao-Lei Wang Bilingualism and Multilingualism: The Media, Education, Literacy, and the Law 23 Bilingualism and Multilingualism in the Global Media and Advertising Tej K. Bhatia and William C. Ritchie 24 Bilingual Education Wayne E. Wright 25 The Impact of Bilingualism on Language and Literacy Development Ellen Bialystok 26 Bilingualism and Writing Systems Benedetta Bassetti 27 Multilingualism and Forensic Linguistics Tej K. Bhatia and William C. Ritchie

466 495 517 542

563 565 598 624 649 671

Part IV  Global Perspectives and Challenges: Case Studies

701



703

28 29 30 31 32 33 34 35

Introduction William C. Ritchie Bilingualism and Multilingualism in North America William F. Mackey Bilingualism in Latin America Anna María Escobar Bilingualism in Europe Andrée Tabouret-Keller Turkish as an Immigrant Language in Europe Ad Backus Multilingualism in Southern Africa Nkonko M. Kamwangamalu Multilingualism in Greater China and the Chinese Language Diaspora Sherman Lee and David C. S. Li Bilingualism and Multilingualism in South Asia Tej K. Bhatia and William C. Ritchie Multilingualism and Language Renewal in Ex-Soviet Central Asia Birgit N. Schlyter

707 725 745 770 791 813 843 871

x  Contents 36 Bilingualism/Multilingualism in the Middle East and North Africa: A Focus on Cross-National and Diglossic Bilingualism/Multilingualism Judith Rosenhouse Index

899 920

Acknowledgments

We are grateful first and foremost to the contributors, whose cooperation in putting this work together was nothing short of remarkable. Our heartfelt thanks are also due to a number of people at Blackwell including particularly Danielle Descoteaux, the Acquisitions Editor in Linguistics, Julia Kirk, our Project Editor, and Tessa Hanford, Project Manager, for their unstinting patience and encouragement. We are also grateful to our teachers and colleagues, Yamuna and Braj Kachru, Barbara Lust, James Gair, K. Machida, Rajeshwari Pandharipande, Hans Hock, Meena and S. N. Sridhar who have deeply influenced our work. The volume also benefited immeasurably from the advice and counsel of a number of valued colleagues: Virginia Valian, Jyotsna Vaid, Loraine Obler, Judith Kroll, François Grosjean, and John Edwards. For Bhatia: The D.A.V school system (India) is particularly worthy of his deepest appreciation for imparting a commendable academic training and providing much-needed global and multicultural vision during his formative years in India. Our families – Shobha, Kevin, Kanika, and Ankit; Laurie, Jane, Peter, and Marnie – have supported us immensely with their love and affection; no words can express our deepest appreciation to them. Finally, we are grateful to Dr. Ben Ware, Vice President of Research and Computing at Syracuse University (for assistance with the first edition of this work) and the College of Arts and Science at Syracuse, George M. Langford, Dean, for their support of this project, particularly in the form of a Research Leave for Ritchie, so he could devote full time to the project. In addition, we wish to thank Senior Associate Dean Gerald Greenberg for his support. This work has also benefited from the support of the Syracuse University Humanities Center in the form of a grant to finance a symposium held in the fall of 2009 at Syracuse entitled ‘The Bilingual Mind.’ We express our deep gratitude to Gregg Lambert, Director of the Center and PI of the Andrew W. Mellon Foundation Grant for the Mellon Central New York Humanities Corridor for this support.

Contributors

Jeanette Altarriba is a Professor of Psychology at the University at Albany, State University of New York where she directs the Cognition and Language Laboratory. Her research interests include bilingualism, second-language acquisition, emotion, and the interaction between language, perception, and memory. She has published numerous scientific research articles and books on these topics and is currently co-editing a volume on bilingual memory and language. Currently, she also serves as Chair of the Department of Communication, at the University at Albany. Ad Backus studied Linguistics at Tilburg University in Holland, where he received his Ph.D. in 1996. He has held research fellowships from the Netherlands Science Foundation and the Royal Netherlands Academy of Sciences and is now an Associate Professor at Tilburg University. His work is mostly on Turkish, language contact, and cognitive linguistics. Benedetta Bassetti is a Lecturer at the University of York. She is interested in bilingualism and biliteracy, including second-language reading and the effects of orthography on metalinguistic awareness and phonology in second-language learners and bilinguals. She coedited the volume Second Language Writing Systems (with Vivian Cook, 2005) and is founding co-editor of the journal Writing Systems Research. Gerald P. Berent is a Professor at the National Technical Institute for the Deaf, Rochester Institute of Technology, New York. He conducts second-language research on deaf learners’ acquisition of spoken-language grammatical, lexical, and semantic-pragmatic knowledge. He also conducts research on acquisitional parallels between deaf learners and hearing second-language learners of English, on bimodal bilingualism, and on English language teaching methodologies. Éva Berkes studied theoretical linguistics and language acquisition at the Ortega y Gasset Graduate Institute (Madrid) and received her Ph.D. from the Complutense University of Madrid in 2002. She is currently a Lecturer at the University of Applied Sciences Burgenland, Austria. Her research interests concern multilingualism, syntactic development, and representation focusing on Hungarian, English, and German. Tej K. Bhatia is Professor of Linguistics and Director of South Asian Languages at Syracuse University. He is a recipient of the Chancellor’s Citation Award for Excellence in Research.

Contributors  xiii He has published a number of books, articles and book chapters in the areas of bilingualism, multiculturalism, media discourse, socio- and psycholinguistics, and the structure and typology of English and South Asian languages. Publications include Colloquial Urdu: The Complete Course for Beginners (2000), Colloquial Hindi: The Complete Course for Beginners, 2nd Edition (2007), Advertising and Marketing in Rural India, 2nd Edition (2007), and The New Handbook of Second Language Acquisition (co-edited with William C. Ritchie, 2009). Ellen Bialystok is Distinguished Research Professor of Psychology at York University. Her research investigates language and cognitive development in children and cognitive change across the lifespan focusing on the effect of bilingualism on these developments. Her recent work has extended to patient studies and neuroimaging methods to understand how bilingualism modifies cognitive behavior. She has published numerous research articles, book chapters, and books, and is the recipient of several major awards, including the 2010 Killam Prize for Outstanding Contribution to the Social Sciences. Jeffrey M. Brown received his Ph.D. in Psychology from Texas A&M International University (TAMIU) in 1996. He is a tenured Associate Professor of Psychology and currently serves as the Dean of Graduate Studies and Research at TAMIU. He currently oversees over 25 different graduate programs in four colleges, with a total graduate enrolment of over 1,000 students. He also oversees the Lamar Bruni Vergara Education Fund, which annually distributes over $1,000,000 in scholarship and assistantship funds to support graduate education at TAMIU. Additionally, he supervises the operation of the Office Research and Sponsored Projects, a support unit that assists administrators, faculty, and staff. He has also published in the areas of eyewitness memory, bilingual language processing, and the tip-of-the-tongue phenomena, to name a few. Yuko Goto Butler is an Associate Professor of Educational Linguistics in the Graduate School of Education at the University of Pennsylvania, where she is also Director of the Teaching English to Speakers of Other Languages (TESOL) Program. Her research interests include language assessment and second and bilingual language learning among children. Shyamala Chengappa is a Professor of Language Pathology and Head of the Department of Speech-Language Pathology at the All India Institute of Speech and Hearing, Mysore. Her areas of interest are child and adult language disorders and bi-/multilingualism with larger work on autism spectrum disorders and aphasia in multilinguals. Her books include Simultaneous Acquisition of Two Languages (1988, with M.S. Thirumalai), Language Disorders in Children (2010), Bilingual Aphasia (editor, 2010) and An Introductory Handbook on Autism (2011). Albert Costa attended the Ph.D. program ‘Cognitive Science and Language’ at the Universitat de Barcelona where he received his Ph.D. in Psychology in 1997. From 1998 to 1999 he was a postdoctoral fellow at the Brain and Cognitive Sciences department, MIT. From 1999 to 2000, he was a postdoctoral researcher at the Cognitive Neuropsychology Laboratory, Harvard University. In 2001, he moved to the Cognitive Neuroscience department at the International School for Advanced Studies in Trieste. From 2002 to 2005 he was a research fellow at the University of Barcelona where he became an Associate Professor in the Department of Psychology in 2006. Since 2008 he has been a Research Professor at ICREA-Universitat Pompeu Fabra in Barcelona. His research is focused on the cognitive

xiv  Contributors mechanisms and representations involved in the production of speech, both in monolingual and bilingual contexts. Paola E. Dussias is Associate Professor of Spanish, Linguistics, and Psychology and Associate Director of the Center for Language Science at Pennsylvania State University. Her research takes a cross-disciplinary approach to bilingual sentence processing. Using a variety of behavioral methods, ranging from off-line questionnaires to eye-tracking methods during reading and spoken language comprehension, she examines the way in which second-language readers and listeners negotiate the presence of two languages in a single mind. Her work, which has been supported by grants from the National Science Foundation and the National Institutes of Health, has appeared in journals such as Bilingualism: Language and Cognition, the International Journal of Bilingualism, Studies in Second Language Acquisition, Second Language Research, as well as in a number of edited volumes. Together with Janet van Hell, Ping Li, and Judith Kroll, Dussias is a co-PI on a PIRE grant (Partnerships for International Research and Education) from the National Science Foundation to develop an international research network and program of training to enable language scientists at all levels (undergraduate, graduate, and postdoctoral) and early career faculty to pursue research abroad on the science of bilingualism. John Edwards received a Ph.D. (in Psychology) from McGill University in 1974. After working as a Research Fellow at the Educational Research Centre, St Patrick’s College, Dublin (now affiliated with Dublin City University), he moved to Nova Scotia, where he is Professor of Psychology at St. Francis Xavier University. His research interests are in language, identity, and the many ramifications of their relationship. He is on the editorial boards of a dozen international language journals, and is editor of the Journal of Multilingual and Multicultural Development and the Multilingual Matters book series, which now comprises about 150 titles. He has authored nine books and a tenth, Sociolinguistics (Oxford University Press), is forthcoming. He is a member of several psychological and linguistic societies, as well as scholarly organizations for the study of ethnicity and nationalism. He is a fellow of the British Psychological Society, the Canadian Psychological Association, and the Royal Society of Canada. Anna María Escobar works at the University of Illinois at Urbana-Champaign. Her areas of research, publication, and teaching are on contact linguistics and bilingualism, with a focus on Spanish–Quechua contact and Andean sociolinguistics, and on language variation and change, with a focus on grammaticalization and semantic change in Spanish. Joshua A. Fishman is Distinguished University Research Professor of Social Sciences, Emeritus at Yeshiva University, Albert Einstein College of Medicine. He is credited with being the founder of the field of sociology of language or macrosociolinguistics, and has made significant contributions to the fields of language and ethnicity, language planning, bilingualism and bilingual education, Yiddish, and medical anthropology. He is a prolific author with over 1,000 publications that include over 85 books and over 900 articles and chapters in books. Among his seminal contributions are: Handbook of Language and Ethnic Identity (Vols. I and II); Do Not Leave Your Language Alone (2006); Can Threatened Languages be Saved? (2001); and Bilingual Education: An International Sociological Perspective (1976). He has been honored many times around the world and in 2004 received the Linguapax Prize.

Contributors  xv Ian FitzPatrick conducted his Ph.D. research on lexical interactions in nonnative speech comprehension at the Max Planck Institute for Psycholinguistics and Donders Institute for Brain, Cognition and Behaviour. From January 2010 to June 2011 FitzPatrick worked as a postdoctoral researcher at the Department of Technology and Information of the University Pompeu Fabra in Barcelona. As of July 2011, FitzPatrick is working as a postdoctoral researcher at the Heinrich Heine University in Düsseldorf in collaboration with the Donders Institute. Suzanne Flynn is a Professor of Linguistics and Language Acquisition at MIT. Her research focuses on the acquisition of syntax by both children and adults in bilingual, second-, and third-language acquisition contexts. More recently, her work has focused on the neural representation of the multilingual brain. She also conducts research on language impairment and early onset of Alzheimer’s disease. Howard Giles is Professor of Communication at the University of California, Santa Barbara. He is founding editor and co-editor of the Journal of Language and Social Psychology and the Journal of Asian Pacific Communication. He was past President of the International Communication Association and the International Association of Language and Social Psychology. His research interests encompass intergroup communication in intergenerational, police– civilian, and other settings. Marianne Gullberg is Professor of Psycholinguistics and Director of the Humanities Laboratory at Lund University. Her research targets adult second-language acquisition and bilingualism, and the production and comprehension of gestures. She led a research group on multilingual and multimodal language processing at the Max Planck Institute for Psycholinguistics, the Netherlands, 2002–9, applying linguistic, psycholinguistic, and neurocognitive perspectives, and also founded the Nijmegen Gesture Centre. She has published extensively on second-language acquisition, bilingualism and gestures, and has also edited a series of special issues and volumes on these topics including a volume of the Encyclopedia of Applied Linguistics (2012). She also serves as an editor for three international journals – as associate editor for both Language Learning and Language, Interaction, and Acquisition, and as information editor for Gesture – and has served as Vice-President of the European Association of Second Language Research (EUROSLA). Roberto R. Heredia is Professor of Psychology in the Department of Psychology and Communication at Texas A&M International University (TAMIU), and founder of the Cognitive Science Laboratory, and Cognitive Science Research Group. Roberto’s major interests are in bilingual lexical access at the sentence level and figurative language processing. He served as Chair of the Department of Behavioral Sciences for two years. He is currently Director and Principal Investigator of the Graduate Retention Enhancement at the TAMIU (GREAT) Program, funded by a multi-million dollar grant from the U.S. Department of Education Title V Promoting Postbaccalaureate Opportunities for Hispanic Americans. He is currently collaborating on three projects: Understanding Bilingual Memory: Theory and Applications (Springer); Bilingual Figurative Language Processing (Cambridge University Press); and Methods in Bilingual Reading Comprehension Research (Springer). Elizabeth Ijalba is an Assistant Professor in Linguistics and Communication Disorders at Queens College, CUNY. She teaches courses on language acquisition with a focus on bilingualism. Her research interests are on reading difficulties and their relationship to

xvi  Contributors second-language learning. Her current research focuses on training parents who speak a language other than English to develop early literacy in the home-language with their children with language impairment. She has presented widely at national and international venues on topics associated with bilingualism, dyslexia, and aphasia. Nkonko M. Kamwangamalu is Professor of Linguistics in the Department of English at Howard University, Washington, DC. He is polity editor for the series ‘Current Issues in Language Planning’; author of the monograph The Language Planning Situation in South Africa; and of numerous refereed articles on topics in language planning, code-switching, multilingualism, World Englishes, and African linguistics. Judith F. Kroll is Distinguished Professor of Psychology, Linguistics, and Women’s Studies and Director of the Center for Language Science at Pennsylvania State University. Together with Annette de Groot, she co-edited Tutorials in Bilingualism: Psycholinguistic Perspectives (1997) and the Handbook of Bilingualism: Psycholinguistic Approaches (2005). The research that she and her students conduct concerns the acquisition, comprehension, and production of two languages during second-language learning and in proficient bilingual performance. Their work, using behavioral and neurocognitive methods, is supported by grants from the National Science Foundation and the National Institutes of Health. She was one of the founding organizers of Women in Cognitive Science, a group developed to promote the advancement of women in the cognitive sciences and supported by the National Science Foundation. Together with Janet van Hell, Ping Li, and Paola Dussias, Kroll is a co-PI on a PIRE grant (Partnerships for International Research and Education) from the National Science Foundation to develop an international research network and program of training to enable language scientists at all levels (undergraduate, graduate, and postdoctoral) and early career faculty to pursue research abroad on the science of bilingualism. Sherman Lee is an Assistant Professor at Hong Kong Shue Yan University. She obtained her Ph.D. in Hong Kong for research examining language shift among the Hakka population, and her B.A. (Manchester) and M.Phil. (Cambridge) from the UK in the fields of linguistics and computer speech. Her research and teaching interests include sociolinguistics, bilingualism, discourse analysis, English for academic purposes (EAP), and research methods. David C. S. Li obtained his B.A. in Hong Kong, M.A. in France, and Ph.D. in Germany. Being a native speaker of Cantonese, he developed a keen interest in foreign language learning from a very young age. His research interests are mainly related to social aspects of language use in multilingual settings. He has published in World Englishes on ‘Hong Kong English’, code-switching in Hong Kong and Taiwan, and EFL learners’ learning difficulties and error correction strategies. William F. Mackey is the author of 20 books and some 200 articles on bilingualism, language education, geolinguistics, and language policy. He was Senior Lecturer (1948–51) at the University of London Institute of Education and the Founding Director of the International Centre for Research on Bilingualism in Laval University (Canada), where he is currently Emeritus Professor. Jeff MacSwan is a Professor of Education and Linguistics (by courtesy) at the University of Maryland. His research focuses on the linguistic study of bilingualism and code-switch-

Contributors  xvii ing, on the role of language in theories of academic achievement differences among language minority students, and on education policy related to English language learners in U.S. schools. He is editor of the International Multilingual Research Journal, and currently serves on several editorial boards. He is the author of over 50 publications. Examples of his work appear in Bilingualism: Language and Cognition, Lingua, Hispanic Journal of Behavioral Sciences, Teachers College Record, Education Policy Analysis Archives, and in edited collections. He has served as a visiting scholar at UCLA, MIT, UC Santa Barbara, and the University of Hamburg. Ines Martinovic received her Bachelor of Science degree in Physics in 2009 from Rensselaer Polytechnic Institute. She is currently pursuing her master’s and doctoral degrees in Cognitive Psychology at the University at Albany, State University of New York. She is interested in emotion, bilingualism, and memory and is currently investigating how survival-based processing affects memory in bilingual populations. Silvina Montrul is Professor and Head of the Department of Spanish, Italian and Portuguese and Professor of Linguistics at the University of Illinois at Urbana-Champaign, where she is also Director of the University Language Academy for Children and of the Second Language Acquisition and Bilingualism Laboratory. She is co-editor of the journal Second Language Research. Her research focuses on linguistic and psycholinguistic approaches to adult second-language acquisition and bilingualism, in particular syntax, semantics, and morphology. She has expertise in language loss and retention in minority language-speaking bilinguals, or heritage speakers. She is author of The Acquisition of Spanish (2004), Incomplete Acquisition in Bilingualism (2008), and El Bilingüismo en el mundo hispanohablante [Bilingualism in the Spanish-Speaking World] (2012). Pieter Muysken is Academy Professor of Linguistics at Radboud University Nijmegen (the Netherlands), having previously taught at Amsterdam and Leiden. He has carried out research and fieldwork in the Andes, Curacao, and the Netherlands. Recent books include Bilingual Speech: A Typology of Code-mixing (2000), The Languages of the Andes (with Wilhelm Adelaar, 2004), Functional Categories (2008), and Lenguas de Bolivia I–IV (with E. I. Crevels, 2009–11). His current research is concentrated in the Languages in Contact Group in the Centre for Language Studies, Radboud University Nijmegen, where he is studying the effect of language contact at four time scales. He is also collaborating with Marianne Gullberg on the interaction of linguistic and processing models for code-switching. Loraine K. Obler is a Distinguished Professor in the Ph.D. Programs in Speech-LanguageHearing Sciences and Linguistics at the City University of New York Graduate Center. Her research has focused on bilingual aphasia, cross-language study of agrammatism, language changes in healthy aging, and talented and challenged second-language learning. Her books include The Bilingual Brain: Neuropsychological and Neurolinguistic Aspects of Bilingualism (with Martin Albert) and Language and the Brain (with Kris Gjerlow). Anne Pauwels is Professor of Sociolinguistics and Dean of the Faculty of Languages and Cultures at the School of Oriental and African Studies, University of London. Her research deals with the social and sociolinguistic aspects of language and communication, with particular attention to multilingual and transnational settings. Her research foci include multilingualism, language maintenance/shift, language policy in relation to language learning in schools and universities, as well as various aspects of the relationship between

xviii  Contributors gender and language. Her most recent book publications include Language and Communication: Diversity and Change (2007), Maintaining Minority Languages in a Transnational Context (2007) and Boys and Language Learning (2008). Robert Phillipson is British, with degrees from Cambridge, Leeds, and Amsterdam. He worked in four countries for the British Council before emigrating to Denmark in 1973. His current research interests are the use of English worldwide, and multilingualism at university level. He is a Professor Emeritus at Copenhagen Business School. William C. Ritchie is Associate Professor of Linguistics at Syracuse University, where he teaches courses in general linguistics and second-language acquisition. His publications include an edited volume entitled Second Language Acquisition Research: Issues and Implications (1978). He has co-edited four handbooks with Tej K. Bhatia, including the Handbook of Child Language Acquisition (1999). Suzanne Romaine has been Merton Professor of English Language at the University of Oxford since 1984. She has published numerous books and articles on linguistic diversity, multilingualism, language death, language revitalization, language change and contact. She was a member of the UNESCO Expert Group that produced its position paper on Education in a Multilingual World (2003), and also wrote the background paper on Languages and Cultural Identities for UNESCO’s report Investing in Cultural Diversity and Intercultural Dialogue (2009). Judith Rosenhouse specializes in Arabic dialectology but has also worked in many other linguistic fields (e.g., phonetics). She has published over 100 scientific articles on Arabic, Hebrew, and other languages. Her books include The Bedouin Arabic Dialects: General Problems and a Close Analysis of North Israel Bedouin Dialects (1984), and Medical Communication in Colloquial Arabic (1989), which won the New Israel Fund Prize. She was awarded the Svend Smith Award for Applied Phonetics by ISPhS (2004). After retirement (2005) from the Technion-Israel Institute of Technology as Head of Department, she has been pursuing her research interests and working with SWANTECH (Sound Waves Analysis and Technologies Ltd.) in linguistic consultancy. Elin Runnqvist is currently finishing her Ph.D. research on bilingual language control and bilingual speech production disadvantages within the program ‘Cognitive Science and Language’ at the Universitat de Barcelona. Itesh Sachdev was born and brought up in Kenya, completed secondary and undergraduate education in the UK (Psychology, University of Bristol), and doctoral training in Social Psychology in Canada (McMaster University, Ontario). He then taught in Applied Linguistics at Birkbeck College, University of London, and is now Professor of Language and Communication at the School of Oriental and African Studies (SOAS, University of London), where he has also been Director of the SOAS-UCL Centre for Excellence in Teaching and Learning ′Languages of the Wider World′. He has served as President of the British Association for Canadian Studies, and is the current President of the International Association for Language and Social Psychology. He has published widely in the social psychology of language and intergroup relations, having conducted research with various ethnolinguistic groups including those in/from Bolivia, Canada, France, Hong Kong, India, Japan, Taiwan, Thailand, Tunisia, and the UK.

Contributors  xix Birgit N. Schlyter is Associate Professor and Head of Forum for Central Asian Studies at the University of Stockholm. She is a Lecturer in Turkish, Uzbek, Central Asian Linguistics, and Central Asian Cultural History. She has published works on modern Turkish and is at present conducting research on language development and language policies in the Central Asian region. Her works include Case Marking Semantics in Turkish (Ph.D. dissertation), Stockholm, 1985, ‘Turkish semantics revisited’, in H. Boeschoten and L. Verhoeven (eds.), Turkish Linguistics Today (1991); Return to the Silk Routes: Current Scandinavian Research on Central Asia (co-edited, 1999); and an article on the language situation in Turkey and Turkish language research in An International Handbook of the Science of Language and Society (2006). Ludovica Serratrice’s research interests are in the acquisition of syntax, pragmatics, and semantics in monolingual and bilingual children. Her work investigates children’s understanding and use of referential expressions in pre-school and school-aged children and the role of priming in the acquisition of subordinate clauses. Jeff Siegel is Adjunct Professor in Linguistics at the University of New England in Australia. His main areas of research concern the processes involved in the development of contact languages, and the use of such languages in formal education. He has worked specifically on Melanesian Pidgin, Hawai‘i Creole, Pidgin Fijian, and Pidgin Hindustani. His most recent books are The Emergence of Pidgin and Creole Languages (2008) and Second Dialect Acquisition (2010). Tove Skutnabb-Kangas, bilingual from birth, is Finnish, with doctoral degrees from Helsinki and Roskilde. Her research interests are multilingual mother-tongue-based education, genocide and crimes against humanity in education, linguistic human rights, and the relationship between biodiversity and linguistic and cultural diversity. Kristof Strijkers conducted his Ph.D. research on the time course of lexical selection in object naming within the program ‘Cognitive Science and Language’ at the Universitat de Barcelona. Currently he is a Marie Curie postdoctoral fellow at the CNRS-Laboratoire de Psychologie Cognitive, Université de Aix-Marseille. At the time of the original publication of her Chapter 30 in 2004/2006, Andrée TabouretKeller was Professor Emeritus, University Louis Pasteur, Strasbourg II, Editor of Education et societies plurilingues/Educazione e societa plurilingue (Aosta, Italy), and President of Centre d’information sur l’education bilingue et plurilingue (Aosta, Italy). In addition, she was responsible for the Seminare ‘Comment lier epistemologie et politique’, Maison des Sciences de l’Homme, Paris and a member of Groupe d’etude du plurilinguisme en Europe (University Marc Bloch, Strasbourg). Her interests at that time were contemporary language contact phenomena, language and psychoanalysis. Xiao-Lei Wang, Ph.D., is a Professor in the School of Education at Pace University. Her research covers a wide range of topics such as cultural parenting styles, nonverbal communication, multilingual acquisition and development, and moral development. Her recent books Growing up with Three Languages and Learning to Read and Write in the Multilingual Family focus on the challenges and strategies of raising multilingual children. Li Wei is Professor of Applied Linguistics, Pro-Vice-Master for Research of Birkbeck College, University of London, and Director of the Birkbeck Graduate Research School,

xx  Contributors University of London. He is Principal Editor of the International Journal of Bilingualism and author and editor of many publications including the Blackwell Guide to Research Methods in Bilingualism and Multilingualism (with Melissa Moyer, 2008), which won the 2009 British Association for Applied Linguistics Book Prize. Wayne E. Wright is an Associate Professor in the Department of Bicultural-Bilingual Studies at the University of Texas at San Antonio. He is editor of the Journal of Southeast Asian American Education and Advancement and is the book review editor for the International Multilingual Research Journal. His research related to language and educational policies and programs for language minority students has been published widely in books and leading academic journals. He is the author of Foundations for Teaching English Language Learners: Theory, Research, Policy, Practice (2010). He has many years of experience teaching in bilingual (Khmer–English) and ESL classrooms with students from kindergarten to adults, and was a Fulbright Scholar in Cambodia at the Royal University of Phnom Penh.

Introduction TEJ K. BHATIA AND WILLIAM C. RITCHIE

In addressing issues in the study of bilingualism and multilingualism, one immediately encounters a terminological issue. The terms bilingualism and multilingualism have come to be used, respectively, to refer to the knowledge and use of two languages and the knowledge and use of three or more languages. Hence, a term is needed to refer to the full range of phenomena including both bilingualism and multilingualism in these senses. Rather than repeat the awkward ‘bi-/multilingualism’ in this introduction and the other introductions in this volume, we will use the term plurilingualism to refer to both bilingualism and multilingualism, as Mackey (chapter 28, this volume) and others have proposed. Whatever the terminology, there is no doubt that plurilingualism constitutes a major fact of life in the world today. Plurilingualism is not such a rare phenomenon; there are, in fact, more bilingual/multilingual speakers in the world than there are monolinguals. The Ethnologue (2009) estimates more than 7,000 languages (7,358) are spoken in the 194 countries of the world, or approximately 38 languages per country. According to the Ethnologue, 94% of the world’s population employs approximately 5% of the world’s languages. Furthermore, many languages such as Hindi, Chinese, Arabic, Bengali, Punjabi, Spanish, Portuguese, and, of course, English are spoken in many countries around the globe. Such a linguistic situation necessitates that many people live with plurilingualism. In fact, David Crystal (2003: 69) estimates that two-thirds of the world’s children grow up in a plurilingual environment. Considering only bilingualism involving English, the statistics that Crystal has gathered indicate that, of the approximately 750 million people worldwide who speak English, over 41% or 235 million are plurilingual in English and some other language(s). The processes of globalization now in progress can only increase the extent and character of plurilingualism, as people the world over continue to recognize the advantage of adding a world language to their verbal repertoires. One must conclude that, far from being exceptional, as many laymen believe, plurilingualism (which, of course, goes hand-in-hand with multiculturalism in many cases) is currently the rule throughout the world and will become increasingly so in the future.

xxii  Introduction Perhaps not surprisingly, research on plurilingualism, whether theory-driven or practically-oriented, has grown dramatically in quantity, quality, and breadth in recent years – in particular, since the publication of the first edition of the current work. These developments have resulted in the founding of three major journals for the publication of basic research in the field since 1990s – The International Journal of Bilingualism (1997–), Bilingualism: Language and Cognition (1998–), and the International Journal of Bilingual Education and Bilingualism (1998–). Two notable journals prior to 1990s were the: Bilingual Research Journal (1975–) and Journal of Multilingual and Multicultural Development (1980–). The International Symposium on Bilingualism (ISB) held its first meeting in 1997 and has met biennially since then; the eighth meeting, held in Norway at the University of Oslo in summer 2011, attracted over 600 abstracts (Elizabeth Lanza, personal communication). In addition to international conferences like the ISB, many more national and regional conferences on the topic of plurilingualism are held on every continent. This Handbook – the second edition of a work published by the same publisher and co-editors in 2004 (hardcover) and 2006 (paperback) under the title The Handbook of Bilingualism – is intended to contribute to these vital trends with 36 chapters presenting state-of-the-art reviews of developments in both theoretical and practical research areas ranging from the study of the bi-/multilingual brain to bilingual education and literacy to the state of plurilingualism in a number of critical regions of the world. The addition of ‘Multilingualism’ to the title reflects the dramatic development over the last eight years of research on the knowledge and use of three or more languages – research that was just getting under way at the time of the compilation of the first edition. These developments have been recognized generally in a number of ways. To cite just one example, beginning in 2013 the international symposium referred to above will be called the ‘International Symposium on Bilingualism and Multilingualism.’ As was true of the first edition, the contributors to this volume – each one a top, internationally known scholar in his or her field – have been given full rein to develop their chapters in the way that seems most fitting to their areas of expertise. The volume is fully revised and updated from the first edition. There are new chapters by new authors, new chapters by many of the authors who contributed to the first edition, and updated and revised chapters from others. In fact, this edition is sufficiently different from the first edition that it should properly be thought of as a companion to the first edition rather than a replacement of it. Since the contributors represent fields as diverse as linguistic theory, neurolinguistics, psycholinguistics, sociolinguistics, speech and hearing, experimental psychology, developmental psychology, clinical psychology, social psychology, sociology, anthropology, language education, deaf education, and area studies, the Handbook is not only an exhaustive area-by-area treatment of the field, but a mosaic of different approaches to the study of plurilingualism as well. The new Handbook is divided into four Parts, each addressing one aspect of plurilingualism, with each Part introduced by one of the editors. Part I provides a general orientation in the study of plurilingualism; Part II addresses questions

Introduction  xxiii concerning the plurilingual individual; Part III surveys the role of plurilingualism in society in general; and Part IV consists of case studies of plurilingualism in eight different areas of the world. In bringing the work of the field together in one place for the advanced student and the researcher in plurilingualism as well as those who apply such research, we hope to have contributed to the deepening and broadening of our understanding not only of the many facets of plurilingualism, but also of the human mind/ brain in general.

REFERENCES Crystal, David (2003). English as a Global Language, 2nd edn. Cambridge: Cambridge University Press.

Ethnologue (2009). www.ethnologue.com/ show_country.asp?name=lb. 16th edn (accessed 1/5/12).

Part I  Overview and Foundations

Introduction TEJ K. BHATIA

The investigation of plurilingualism is a broad and complex field, which includes the study of the nature of the individual plurilingual’s knowledge and use of two (or more) languages as well as the broader social and cultural consequences of the widespread use of more than one language in a given society. The two chapters that make up Part I provide a general orientation to this complex field. In Chapter 1 – ‘Bilingualism and Multilingualism: Some Central Concepts’ – John Edwards provides a bird’s-eye view of the field by examining a wide range of issues that are addressed in greater depth in later, more specialized chapters of the book. Edwards opens his chapter with a description of ways in which individual and societal plurilingualism differ and how societal plurilingualism arises. He then proposes a framework for classifying instances of societal plurilingualism and discusses a number of ways in which these cases of plurilingualism have been dealt with in these societies. Edwards then addresses issues in the study of individual plurilingualism, including popular questions about the capacity to acquire more than one language, ways in which degree of bilingualism has been measured, the relationship of bilingualism to general intelligence and cognitive development, and the process of becoming bilingual (that is, second-language acquisition). Finally, Edwards discusses the relationship between language and social identity in instances of plurilingualism. Li Wei’s Chapter 2, ‘Conceptual and Methodological Issues in Bilingualism and Multilingualism Research,’ opens with the presentation of three presuppositions required for the coexistence of more than one language in a given society: that there are different languages in the world at large; that speakers of different languages (and the languages themselves) come into contact in a particular sense; and that the usual consequence of this contact is plurilingualism both in The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

4  Overview and Foundations individual speakers and in the society in which they interact. He then moves on to discussions of three different approaches to the study of plurilingualism and their accompanying methods – the psycholinguistic, linguistic, and sociolinguistic approaches – and calls for the development of a combined transdisciplinary approach to the study of plurilingualism in the future. (It should be noted that this chapter is intended to complement – rather than replace – François Grosjean’s Chapter 2 in the first edition.)

1  Bilingualism and Multilingualism: Some Central Concepts JOHN EDWARDS

Introduction Bilingualism and multilingualism have both de facto existences and important places in the psychological, political, and social debates that define social and ethnic groups, communities, and regions. Very widespread phenomena, they arise for a number of well-understood reasons; in the main, however, they are also quite unremarkable phenomena, fuelled by necessity up to, but rarely beyond, appropriately useful levels of competence. They imply both heightened and lessened opportunities for interpersonal and intercultural exchange: multilingual capacities at an individual level can obviously broaden possibilities, but a world of many languages is also one in which communicative problems exist. In such a world, lingua francas and translation are required.1 While almost everyone knows at least a few words in other languages, we generally require a little more competence than that before we are willing to acknowledge bilingual or multilingual ability. Where, however, to draw the line? Where does bilingualism ‘start’? And how are we to accommodate different levels of fluency? Still, there are those who we confidently put in the monolingual category. And, at the other end of this linguistic spectrum, there are those who have virtually maternal multilingual capabilities. After rigorous self-examination – of which language emerges spontaneously at times of emergency or elevated emotion, which variety is dreamt in, which is associated with the earliest memories – George Steiner (1992), for example, claimed equal fluency in English, French, and German. As in other social arenas, however, it is the grayer areas between extremes that are at once more common and more interesting.

The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

6  Overview and Foundations As noted, competence in more than one language can be approached from social as well as individual perspectives, and these need not be as neatly connected as might first be thought. While it is true that a country (or any other recognizably bordered region) full of multilingual people is itself multilingual in a broad sense, it may nevertheless officially sanction only one or two varieties and thus, in another sense, be something less than multilingual. Conversely, a country may be officially bilingual or multilingual and yet most of its citizens may fall into the monolingual basket. Many states in Africa have two official languages (usually a strong indigenous variety and a widely used European one) for highly heterogeneous and multilingual populations. In Canada, too, there is official recognition of two languages, but the situation on the ground hardly resembles the linguistically rich, varied, and – above all – interpenetrating settings common in Africa. Both personal and social manifestations of bilingualism are of course important, but it should be noted that the emphases are quite different: a thoroughgoing discussion of individual bilingualism involves, for example, linguistic and psycholinguistic dimensions which figure much less prominently, if at all, at the social level where other dimensions – historical, educational, political, and so on – arise for consideration. There are important differences between individual bilingualism and collective or social bilingualism, regardless of whether or not the latter is officially endorsed. In many settings, ancient and modern, the latter is an enduring quantity. Individual bilingualism, on the other hand, may be less permanent, often reflecting a generational way-station on the road between two monolingualisms. Thus, the classic pattern for immigrants to the United States has been bilingualism (mother tongue and English) by the second generation and English monolingualism by the third. A more enduring collective bilingualism rests upon continuing necessities which become absent among most immigrant populations. Where these involve different functions and domains of use for each language, the situation is often referred to as diglossia. Of course, even stability is relative. The French– English diglossia that prevailed in England after the Norman conquest eventually broke down, for example. Arrangements of societal bilingualism vary across contexts. The passage of the Official Languages Act (1969), which legally underpins French and English in Canada, was preceded by an official investigatory commission. Paying special attention to the linguistic situations in Belgium, Finland, Switzerland, and South Africa, the commissioners closely examined the so-called ‘personality’ and ‘territorial’ principles relating to bilingualism. In South Africa, language rights are seen to inhere in individuals, wherever they may live within a state; the Belgian territorial principle, however, accords official status by region – Flemish in the north, French in the south. The commission opted for the application of the personality principle in Canada, even though official-language minorities were (and still are) small in all provinces except Quebec and New Brunswick, a fact that would seem to point towards a territorial disposition. The sensitive and sometimes volatile relationship between the two ‘founding peoples,’ however, was a prime consideration. The recommendation, therefore, was for federal bilingualism

Bilingualism and Multilingualism: Some Central Concepts  7 and the provision of bilingual services at the provincial level – but only Quebec, Ontario, and New Brunswick were to become ‘officially’ bilingual. Today, however, only New Brunswick has proclaimed itself bilingual. Indeed, broader hopes for Canada-wide bilingualism have faded, with the country steadily moving towards a Belgian-like territoriality: French in Quebec and English elsewhere, with a bilingual ‘belt’ in those regions of Ontario and, especially, New Brunswick that abut Quebec. The assimilation of francophones outside Quebec, and that province’s rejection of bilingualism, reflect the importance of the political and social frameworks within which stable bilingualism occurs. Policies of social engineering in democratic states must ultimately, it seems, be reconciled with widespread popular perceptions of social reality and self-interest. When perceptions differ among powerful ethnic groups – in Canada, the anglophones, francophones, aboriginals, and ‘allophones’ are all central players, though no group is itself monolithic – then centrally inspired conceptions of bilingualism and diglossia are seen to be quite delicate (see Edwards 1995, 2009).

A Multilingual World Multilingual realities arise in a number of ways. Immigrants, whether settlers or invaders, bring languages into contact and sometimes, as with imperialist and colonial expansion, it is unnecessary for many people to physically move; their language may make its presence felt through military, religious, or economic force requiring relatively small numbers of soldiers, merchants, bureaucrats, and missionaries. Some cultures have had more explicit policies here than have others – compare, for instance, the mission civilisatrice of the French with the more pragmatic attitude of the English (see below for discussion) – but all imperial powers have, directly or indirectly, made their languages attractive and sometimes necessary to conquered or colonized groups. The languages of expansionist regimes often become intertwined with pragmatic advantage and cultural prestige at a local level, and these factors often long outlive the original dominating influence: the continued adherence to European varieties which exists in former colonial areas is an example here. Multilingualism can also arise as a result of political union among different linguistic groups: Switzerland incorporates German, Italian, Romansch, and French populations; Belgium unites (sometimes precariously) French and Flemish speakers; Canada has English and French ‘charter’ groups. In addition to this sort of political association, there are also multilingual federations based upon more arbitrary arrangements, often the result of colonial boundary marking and country creation; modern examples are found in Africa and Asia. Multilingualism is also commonly observed in border areas: two North American examples can be found along the Mexican–American border in the south, and on that between New England and Quebec in the north. These are the primary circumstances underpinning multilingual arrangements, but they are not the only ones. Cultural and educational motivations can also

8  Overview and Foundations expand linguistic repertoires – not only on an individual basis, but in more widespread fashion as well. Also relevant here is the degree to which a language community is open to the use of ‘its’ variety by others. Consider the differences between the English and the French in this regard: the latter have traditionally been much more possessive of their language and, while working hard to bring it to those unfortunate enough not to already speak it (the mission civilisatrice, again), have also been zealous in protecting its ‘purity,’ both at home and abroad. English, on the other hand, has not been treated in the same guarded way; while there are books and journals devoted to the ‘new’ Englishes and to ‘world’ English, there are few similar treatments for French. English is thus becoming ‘internationalized’ in a way that French is not, and an important consequence is that a language once tainted by imperialism is rapidly becoming ‘ours’ in many parts of the world. India provides perhaps the best example of a broadly accepted indigenized variety of English.

Classifying Multilingualism Understanding the dynamics of multilingualism means coming to grips with many complicated linkages between languages and virtually all other areas of social life. While it is no doubt true that every language–contact situation is unique, that uniqueness arises because of the differential weightings and combinations of elements that are, themselves, recurrent across settings – and not because of the presence of elements or factors found nowhere else. This suggests the possibility of frameworks within which many settings might be assessed and compared. Building upon the work of many other researchers, I have constructed a typological framework of language–contact settings, with particular reference to minority linguistic groups. While such a model does not cover all important instances of multilingual contact, it is certainly descriptive of a great many – simply because language contact very often involves varieties of unequal strength. To provide physical context, the framework begins with the adaptation of a geographical scheme first proposed by White (1991). It makes three basic distinctions. The first is among minority languages which are unique to one state (e.g., Breton in France), those which are non-unique but which are still subordinate in all contexts in which they occur (e.g., Basque in Spain and France), and those which are minorities in one setting but majority varieties elsewhere (French in Canada and French in France); thus, we have unique, non-unique, and local-only minorities. The second distinction deals with the type of connection among speakers of the same language in different states; are they adjoining (again, Basque in Spain and France) or non-adjoining (French in Canada and French in France)? Thirdly, what degree of spatial cohesion exists among speakers within a given state? Here, the terms cohesive (Cree in Canada) and non-cohesive (Spanish in the United States) can be used. In addition, the model considers these distinctions as they apply to both immigrant and indigenous groups (French and Cree in Quebec).2

Bilingualism and Multilingualism: Some Central Concepts  9 Such a framework cannot capture all the important nuances, of course, some of which are not immediately apparent. As an example, consider the immigrant– indigenous dimension. Some have argued that only Amerindian languages – which, themselves, arrived via ‘Beringia’ some 12,000 years ago – are indigenous to Canada. French and English, however, have 400-year claims. While less indigenous than the ‘first nations’ might they be considered as more indigenous than later arrivals? How long, in other words, does it take to become indigenous? This is not an idle matter, since scholarly pleas have been made for the differential treatment of indigenous and immigrant groups. There are also problems with the cohesiveness dimension: if a language is spoken sparsely over a wide area, but also possesses a concentrated center, then it could perhaps be described as either cohesive or non-cohesive. Yet another difficulty arises when considering a language that is found in adjoining states; while each group can be classified as cohesive or non-cohesive, the degree of cohesion of its neighbor may also be important. Issues also arise concerning the adjoining/non-adjoining dichotomy itself. For Basques in France and Spain, the adjoining label seems appropriate, but what of language communities groups found in neighboring states but not in their common border areas? Despite the complexities that bedevil even these geographical underpinnings, it is clear that information of many other types is required for fuller understanding of multilingualism. The functions and status of competing language varieties are central here, and have engaged the attention of a number of important scholars – Ferguson (1962, 1966), Stewart (1962, 1968), Haugen (1972), Haarmann (1986), Giles (see, e.g., Giles and Coupland 1991), and others. All of this existing work has been taken into account as my own comprehensive scaffolding of ecolinguistic factors has arisen. The latter builds upon three basic categories of variables: speaker; language; and setting. These are not, of course, watertight and mutually exclusive dimensions, but they may serve as benchmarks. A second and intertwined categorization takes into account different disciplinary perspectives (demography, sociology, linguistics, psychology, and so on). A few examples will illustrate this interweaving. Viewing settings from a demographic perspective could highlight urban–rural distinctions of importance for language maintenance or decline. Sociological attention to speaker variables might consider within- or without-group marriage. Formal linguistic analysis of different varieties could focus upon dialectal variation, and social psychology could investigate language attitudes and beliefs. These questions – and the large number of others that arise when we plot discipline against dimension – may not be specific enough to comprise a completed and applicable typology, but they could formalize and direct our inquiries. Given that many individual case studies, however rich and many–ayered – are essentially one-off exercises, it is clear that further classification could be useful, even if it does no more than rationalize data gathering. As Ferguson (1991: 230) has noted: It is frustrating to read a stimulating case study and find that it lacks information on what the reader regards as some crucial points . . . what I have in mind is not so

10  Overview and Foundations much a well developed theoretical frame of reference as something as simple as a checklist of points to be covered.

The typological approach that I have briefly sketched here has attracted the attention of other scholars. Grenoble and Whaley (1998) draw centrally upon the model, for instance, citing both strengths and weaknesses. Clyne (2003), too, points out the need for more clarity; overall, however, he looks favorably on the model’s contextualization of variables, something that ‘could be considered more in the methodology of future studies’ (244). In their studies of Bashkir, Altai, and Kazakh speakers in the Russian republics of Bashkortostan and Altai, Yag˘ mur and Kroon (2003, 2006) have employed the framework in conjunction with the ethnolinguistic-vitality approach of Giles and his colleagues – a model first introduced by Giles et al. (1977) and extensively referred to since then (for a recent example, see Ladegaard, 2006). Paulston et al. (2007) have cited it in their examination of ‘extrinsic’ linguistic minorities – that is, groups who once belonged to a majority population in a neighboring country: Russians in Latvia are a case in point. Extra and Gorter (2008) discuss the approach in the introduction to their own framework for regional minority languages in Europe. Tsunoda (2006) uses the model as the scaffolding for his chapter on endangered languages, reminding us that typological work is a useful subsection of language ecology. Vail (2006) has employed the scheme in his assessment of Northern Khmer and, while making some critical remarks, styles it ‘the most robust model’ (140) of both macro- and micro-level approaches to the ecology of endangered languages.

Dealing with Multilingual Realities Although the exigencies of language contact give rise to multilingual abilities, there are obviously many occasions when limitations in such abilities necessitate some means of bridging a language gap. There are two main methods here. The first, the use of lingua francas, is either part of the existing multilingual picture, or necessitates an extension of it. ‘Link languages’ fall into three categories: socalled ‘languages of wider communication,’ varieties that have achieved regional or global power; pidgins, creoles, and other restricted linguistic forms whose diminished scope is at once easy to master and sufficient for communicative purposes which are, themselves, quite circumscribed; and constructed or ‘artificial’ languages. There are many historical examples of existing languages becoming important lingua francas – not, of course, because of any intrinsic qualities setting them above other varieties, but because of the power and prestige of their speakers. Greek and Latin are the classical examples, but French, Italian, Arabic, and – today – English have all played bridging roles. In the second category are pidgins, whose limited vocabularies and grammars are sufficient for basic communication. (Pidgins are, then, a rather good example of a much more general tendency, the development of multiple fluencies up to, but not beyond, the requirements of day-to-day use.)

Bilingualism and Multilingualism: Some Central Concepts  11 If the social contacts that give rise to pidgins remain, themselves, of a rudimentary nature, then those pidgins may have considerable longevity; if contact situations persist, however, and become more complicated, pidgins may evolve into creoles. The developments here stem from the need for richer and more expressive forms, and these, in turn, often arise because nobody’s mother tongue (a pidgin) is in the process of becoming somebody’s mother tongue (a creole). Constructed languages provide a third potential bridge across linguistic divides. Although there have been a great many schemes over the centuries, the best known and the most successful example is Ludwig Zamenhof’s Esperanto, first published in 1887. In most cases, the initial desire of the ‘language makers’ has been to produce an easily learned auxiliary medium that possesses a sort of neutrality unavailable to powerful ‘natural’ languages, tied as they are to important groups and burdened by specific histories. But Zamenhof and others like him have also believed that constructed link languages, untainted by imperialist pasts, might contribute to some ‘trans-national’ identity, and thus to global harmony. Although there are several important reasons for the lack of acceptance of constructed languages, a basic problem is this: Esperanto (for example) might be more appealing if there were a significant community of speakers one could join. Without such a community, however, motivating potential learners to take the plunge is difficult – yet how else will a speech community come into being? (Fuller details on constructed languages like Esperanto may be found in Edwards 2010a.) The other great bridging method is translation. Apart from an almost useless word-for-word exercise, every act of translation involves interpretation and judgment. For this reason, it has sometimes been supposed that ‘true’ translation is impossible; however, although a perfect version which captures every nuance and allusion is rather unlikely – and becomes more so as the material to be translated becomes less prosaic – we have nonetheless translated, for practical purposes, throughout history. (Seeing translation as interpretation also links, incidentally, cross-language exercises with communications within the same language. That is, even the simplest of conversations between two speakers of the same language involves interpretation, and is analogous to ‘reading between the lines’ in written language.) If, as is generally the case, the aim is to adhere to what the well-known classical translator, Émile Rieu, once called the ‘law of equivalent effect’ (see Tancock 1954: 15), then the greatest threats to accurate translation appear at opposite ends of the literary continuum. At the ‘lower’ end, the non-standard language of the streets, heavy with ever-changing slang and obscenity, poses real problems; at the ‘higher’, poetic or philosophical productions lay traps in their use of metaphor, allusion, or dense, abstract reasoning.

The Definition and Measurement of Personal Fluencies Early in his partnership with Watson, Sherlock Holmes explained his ignorance of many things by saying that the brain was like an attic, that one should fill it

12  Overview and Foundations wisely according to one’s needs, and that ‘it is a mistake to think that that little room has elastic walls and can distend to any extent’ (Conan Doyle 1966: 12). But while there are limits, Holmes grievously underestimated them; he could easily have remedied his ignorance of literature and astronomy without displacing his knowledge of poisons or the many varieties of cigarette ash. I mention this because the notion of some finite or ‘containerized’ language competence bedeviled the language literature for some time. At its simplest, the suggestion was that what you gain on the swings of one language you lose on the roundabouts of the other. All that we know about brain function and capacity, however, completely undercuts this idea – but it still has currency beyond the academic cloisters. Many parents interested in immersion education for their children, for example, need reassurance on the point. As may be imagined, it is easy to find definitions of bilingualism that reflect widely divergent responses to the question of degree with which I opened this chapter. Some views acknowledged bilingualism only where two well-developed and roughly equal fluencies were found; Bloomfield (1933) provided an example here. Some, like Weinreich (1953) were content to remain vague in matters of degree. And still others (e.g. Haugen 1953) have suggested that linguistic repertoire expansion begins with the ability to produce complete and meaningful utterances in a second language. Roughly speaking, earlier definitions were more restrictive, while later ones have made room for greater variation in competence. But this more liberal view proves to be as unsatisfactory as its more narrowly conceived predecessors. This is simply because any attempt to come to grips with bilingual competences – whether for purposes of educational best practice, or for attempts to relate those competences to other aspects of personal or social life, or simply to more fully open another window on language and cognition – must obviously start from definable levels or degrees. Several specific complications can be briefly touched upon here. Bilingualism and multilingualism, for example, clearly represent extensions along more than one language dimension. Given the four basic skills (listening, speaking, reading, and writing), and given further subdivisions so as to take into account divergent possibilities under headings like vocabulary, grammar, and pronunciation, we soon arrive at a considerable number of important elements. They will all figure in the assessment of language competence, but it does not follow that strength in one means strength in another: a pupil may be able to understand spoken English and Welsh, speak English fluently but Welsh only haltingly, read in Welsh with a reading age of six and in English with a reading age of eight, write poorly in English and not at all in Welsh. Is that pupil bilingual? (Baker 1988: 2)

Surely the broad answer here is yes – but equally surely, any attempt to depart from such a generality will prove problematic. Many tests have been used to measure bilingualism, including rating scales and tests of fluency, flexibility, and dominance. The first of these can involve inter-

Bilingualism and Multilingualism: Some Central Concepts  13 views, language-usage measures, and self-assessment procedures, and these personal reports have a lot to recommend them. Their strengths, however, rest upon the ability (and willingness) to self-report accurately, and a roughly equivalent sense across informants of what competence means. Such matters cannot be taken for granted, and the inaccuracies of census language information provide an illustrative example of the difficulties in relying upon self-assessment. Apparently more objective tests of fluency and flexibility are also far from perfect predictors of either performance or competence, and can be quite inadequate in assessing patterns of relative dominance. Factors such as attitude, age, gender, intelligence, memory, inter-linguistic distance, and context of testing are all potentially confounding. Even if we were able to gauge bilingual or multilingual capacities with some accuracy, there would remain problems of adequate labeling, for it is hardly to be expected that measured individuals would neatly fall into a small number of categories of ability. There even remains confusion as to what term ought to be applied to those much sought-after individuals whose bilingual capacities are great: they have been described as balanced bilinguals, ambilinguals, and equilinguals, among other terms. The first has become the most common, and its use need not imply solely Steiner-like perfection; even a rougher equivalence of fluencies, however, still implies a category in which most bilingual or multilingual individuals cannot be placed. The capabilities of most of those who may reasonably be styled as ‘bilingual’ fall well below any line of equivalence. There are other important matters to be considered. A useful distinction, for instance, can be drawn between receptive (or passive) and productive (or active) bilingualism; the difference here is between those who understand a language – either spoken or written – but cannot produce it themselves, and those who can do both. Additive or subtractive tendencies must also be reckoned with: does learning a new language represent a repertoire expansion or a replacement of the earlier variety? Outcomes here tend to reflect different social pressures and needs. Additive bilingualism generally occurs where both languages continue to be useful and valued; the subtractive variety, on the other hand, typically reflects a setting in which one language is more dominant, where one is on the ascendant and the other is waning. Yet another common distinction is that between primary and secondary bilingualism, between a dual competence acquired naturally through contextual demands, and one where systematic and formal instruction has occurred. These are not watertight compartments, of course. One might, for example, develop a fluent conversational grasp of a language in a relatively informal way, and only later feel the need to add some formal literacy skills. This would, incidentally, recapture the process by which a mother tongue is developed, and it is noteworthy that more enlightened school curricula have tried to reflect this in their second-language programs. Still, it is not difficult to appreciate that there are some important and socially relevant differences between those who become bilingual informally (often, to add to the point, relatively early in life) and those whose second competence is more self-consciously acquired. Compare those English–Gaelic bilinguals in Ireland or Scotland whose fluencies result from

14  Overview and Foundations growing up in a particular location, with those who, in Dublin, Glasgow, or Edinburgh, have more formally set themselves (or been set) to become bilingual. Lumping these two groups together, under a single ‘bilingual’ rubric might be quite inappropriate at personal levels and, as well, might give a rather inaccurate picture of the state of health of Irish and Scots Gaelic. Some have attempted to capture the last distinction by referring to ‘élite’ and ‘folk’ bilingualism. The former has typically involved two (or more) prestigious languages, and often had as much to do with social-status marking as it did with a thirst for knowledge and cultural boundary crossing. In earlier times, not to have known Latin or Greek or French in addition to one’s vernacular would have been unthinkable for educated people – but often unthinkable, perhaps, in the same way that it would have been unthinkable not to have had servants. ‘Folk’ bilingualism, on the other hand, is generally suggestive of a more informal and necessity-driven expansion. But a moment’s reflection will reveal that both varieties may be driven by necessity – different forms and levels of necessity, to be sure. Similarly, neither need be a capability learned at the maternal knee, nor is formal instructional always an ‘élite’ marker. The language-learning activities of immigrants might well be seen to fall under the ‘folk’ heading – after all, they are motivated by mundane but usually pressing considerations – but they (or their children) typically learn the new language at school. Formal education per se, then, seems not enough to elicit the élite label. Indeed, there are many real-life mixtures that should caution us against inaccurate and simplistic categorization. For example, before the passage of legislation strengthening the position of French at school, it was very common for Italians in Montreal to maintain Italian at home, to learn French dans les rues, and to put their children into English-medium schools. Trilingualism was the result of this interesting mixture of approaches; see Edwards (2010b).

The Bilingual or Multilingual Individual The fact that a majority of the world’s population, a great many of whom are poorly educated, have at least some level of multilingual competence surely indicates that language repertoire expansion is not a particularly rare feat. And yet, especially within dominant linguistic groups, it is common to find references to the difficulties involved or to the peculiar lack of language talents supposedly possessed. Today, for example, anglophones often complain that they have no aptitude for foreign-language learning. This may be accompanied by expressions of envy for those multilingual Africans, Asians, and Europeans, and sometimes (more subtly) by a linguistic smugness reflecting a deeply held conviction that, after all, it is those clever ‘others’ who don’t know English who will have to accommodate. All such attitudes, of course, reveal more about social dominance and convention than they do about aptitude. The literature on the specifics of second-language acquisition, both ‘natural’ and school-based, reveals that – with sufficient motivation and opportunity – all

Bilingualism and Multilingualism: Some Central Concepts  15 normally intelligent people can at least become functional in another variety. This is not to deny that there may exist individuals who have a greater innate or acquired aptitude: a good ‘ear’ may be helpful, as well as a good memory and a capacity for self-initiated application. Beyond these, adaptability and genuine interest in other cultures are no doubt important; an important difference between purely ‘instrumental’ and more thoroughgoing ‘integrative’ motivations has been frequently discussed in the literature (Edwards 2010c). All of these qualities (with, in some cases, the exception of the last) are of general value, however, and do not form a package uniquely implicated in language learning (see Butler, chapter 5, this volume, for review). Almost everyone, then, has the capacity to expand their linguistic repertoire, and evidence from both the laboratory and the street tells us that doing so exacts no cognitive price. But what of the notion that bilingualism can increase intellectual scope? While some have demurred – in the seventeenth century, for example, Samuel Butler suggested that ‘the more languages a man can speak, his talent has but sprung the greater leak’ (Hazlitt 1901: 92) – the historically common view was that personalities somehow expand with extra language competence. In the modern era, however, we find scholars beginning to associate bilingualism with lowered intelligence. Much of this early social-scientific work was conducted in America, at a time of great concern with the flood of immigrants from Europe (roughly, 1900–1920). The intelligence tests of the time were very culture-specific, and non-white, non-English-speaking, non-northern-European, non-educated individuals fared poorly. In this unenlightened climate, measured intelligence was seen to correlate strongly with competence in English, and Florence Goodenough –an important educational psychologist who worked with Lewis Terman, the developer of the Stanford–Binet intelligence test – actually wrote that ‘the use of a foreign language in the home is one of the chief factors in producing mental retardation’ (1926: 393). Later research tended to show essentially no relationship between intelligence and bilingualism, and this work was more carefully done than the earlier studies. Controlling for age, social class, gender, and so on became standard procedure, and the lack of such control was increasingly seen to have produced the negative associations found in previous work. A further turning-point came in the early 1960s, when findings showing a positive relationship between intelligence and bilingualism began to appear. In Montreal, Peal and Lambert (1962) examined ten-year-old middle-class bilingual and monolingual children; all of the former were assessed as having equal proficiency in French and English. The bilin­ guals were found to outperform their monolingual counterparts on both verbal and non-verbal intelligence tests. The authors were obliged to note, however, that their study could not determine directionality: do more intelligent children become bilingual, or does bilingualism enhance intellectual development? There are some important difficulties involved in attempting to show a relationship – positive or negative – between bilingualism and cognitive development, mental flexibility, and intelligence. In Peal and Lambert’s study, for instance, the unique reliance upon ‘balanced’ bilinguals may be thought to limit generalizability.

16  Overview and Foundations There are broader problems, too. The adequate representativeness of samples is a general difficulty, a specific variant of which is the problem of equating home backgrounds simply by controlling for gross measures of socioeconomic status. We have already noted the vagaries of bilingual definition, and these also apply to the accurate measurement of intelligence. And, as Peal and Lambert’s own caution reminds us, most investigations produce correlations – and correlation need not imply causation. There is very recent evidence suggesting that particular strengths are associated with bilingual competence, and the work of Bialystok and her colleagues is important here. Bialystok (2009) suggests, for example, that bilingualism is associated with improved cognitive functioning; a later paper (2010) finds that bilingual children solve certain tasks more quickly than do monolinguals, particularly those involving the processing of complex stimuli. Moreno et al. (2010) have extended such findings from children to adults and, indeed, Bialystok’s most recent work is concerned to investigate bilingualism across the lifespan. The idea that expanded competence may ameliorate the symptoms of Alzheimer’s disease has, unsurprisingly, attracted considerable media attention (see Bates 2010): here, Bialystok argues that bilinguals’ larger ‘cognitive reserves’ provide an advantage. Despite these sorts of results – and noting that Bialystok herself points to limitations and uncertainties, as well as to areas in which the performance of bilinguals is not better than that of their monolingual counterparts – it remains fair to say that strong conclusions about bilingualism and cognition are not warranted. It is clear that repertoire expansion does not lead to decreased or weakened capacities and, more positively, that it represents a useful addition to overall personal capacity. The most obvious bilingual benefit is of course language choice, but it is also common to find linguistic alteration occurring within one segment of speech. Weinreich (1953) considered such ‘code-switching’ as cross-language interference, but an abundance of studies has shown that switching may occur for emphasis, because the mot juste is found more readily in one language than in another, or because of a complicated network of perceptions of the speech situation, topic and content, the linguistic skills of interlocutors, degrees of intimacy and formality, and so on. The more neutral transference seems an more apt designation of behavior here; sometimes, as Poplack’s (1980) title has it, a speaker may ‘start a sentence in English y terminó en español’. Beyond this rather straightforward language switch, there are several types of language transfer typically found in bilingual and multilingual speech. (For a somewhat different use of the terms code-switching and transference or transfer, see Ritchie and Bhatia, chapter 15, this volume.) If a Brussels French speaker uses the Dutch vogelpik for a game of darts, for example, rather than the standard French fléchettes, this is an example of lexical transfer; vogelpik here constitutes a ‘loanword’ since it is a borrowing used in unchanged form. (If it were to be given a French pronunciation, this might suggest an attempt to bring it more closely into the maternal fold.) Another type of lexical transfer involves translation, as when the English skyscraper becomes the French gratte-ciel; these words are called ‘calques’ (i.e. copies). Morphological transfers occur when loanwords are rather more fully embraced by the receiving language:

Bilingualism and Multilingualism: Some Central Concepts  17 the Dutch heilbot (halibut) becomes un elbot in French. Syntactic transfer is revealed in such phrases as ‘She’s a nice girl, isn’t it,’ where the tag-words are slightly inappropriate in English but clearly reflect a more all-purpose phrase in another variety. Phonological transfer and ‘foreign’ accents are both common and obvious, as are the subtle variations in stress and intonation that constitute prosodic transfer. While some of these features are more clearly matters of choice than are others, they all represent some sort of borrowing. The degree to which the borrowed element is integrated (or can be integrated) into the other code may be of considerable interest for studies of group contact, of relative linguistic prestige, of the perceived or actual ease with which different languages deal with given topics. Borrowings may be on a ‘nonce’ basis or may represent more established practice, but the latter grows from the former and presumably reflects stronger and more widespread need. However, a further subdivision has been suggested for these established borrowings; some are indeed necessary – words filling lexical gaps in the other language, for example – but some seem gratuitous, since an equivalent item already exists. The motivation here is most often perceived status and prestige, and common examples include the use of foreign words or phrases. It is one thing, then, to refer to das Web-Design or der Cursor, and perhaps another to employ der Trend, or der Team or der Cash-Flow. More pointedly, consider packets of Japanese sweets labeled ‘Meltykiss Chocolate,’ or ‘Men’s Pocky Chocolate.’ To focus on the oddity of usage here is to miss the point – which is that any such oddity is meaningless. Very few of the purchasers of these sweets will decipher these English words, and they will neither know nor care that ‘meltykiss’ or ‘pocky’ chocolate would probably not be a popular treat in London or New York.

Perspectives on Theory and Practice There are some classic accounts in the literature of children being brought up bilingually (e.g. Ronjat 1913; Leopold 1939–49; see Wang, chapter 22, this volume, for review), and it seems clear that the practice need involve few risks. Where negative consequences have been observed, these are almost always due to social, personal, cultural, or other factors – and not to the bilingualism process itself. Indeed, most observers point to the advantages of an early-acquired bilingual competence; these tend to reflect, above all, the relative ease of early learning and the higher levels of fluency and vocabulary that often result. There is something to this, but arguments that the young brain is more ‘plastic’ or receptive than the older one have sometimes led to an overemphasis upon early acquisition. Older learners have cognitive experience lacking in small children and, providing the motivation is sufficient, can be better learners. If we could combine the maturity and articulated necessity of the older with the impressionability, imitativeness, spontaneity, and unselfconsciousness of the younger, we would surely have a recipe for rapid and proficient bilingual acquisition.

18  Overview and Foundations There are many formal methods for teaching languages; very generally, older ones tended to stress the memorization of grammatical rules and vocabulary, in the service of literary study; little attention was given to spoken language. In more contemporary school settings this has changed, although even high-tech language laboratories sometimes merely individualize older approaches, rather than signaling a change of course towards more conversational competence. Still, while it remains difficult for the classroom to become a representation of the street, the tendency has steadily been towards more and more conversation. Students are encouraged to speak before learning formal grammar, and the use of the maternal variety is often kept to a minimum; the idea is to have second-language acquisition resemble as far as possible first-language learning. Immersion classrooms provide the most recent and most important embodiment of this principle. It is in this contemporary context that the most useful theoretical perspectives have emerged; see Edwards (2010c). Most approaches to second-language acquisition reject a simplistic behaviorist model and endorse a cognitive conception in which (as in first-language learning) rules are formulated and tested. Learning occurs in a series of non-random stages, each of which is characterized by a sort of interlanguage. Theories within social psychology have paid particular attention to motivational features, and this makes a good deal of sense. If we agree that language is a social activity, and if we accept that almost everyone is cognitively capable of learning second (and subsequent) varieties, then it follows that the force of the situation, and the attitudes it provokes in potential learners, are central. A well-known and evolving framework for second-language learning is that of Gardner, whose work here extends from the late 1960s (see Gardner and Lambert 1972) to the present (Gardner 2010). He has consistently attempted to link the social context, and the cultural beliefs within it, to individual learner capacities – including, of course, motivational levels – and the formal/informal settings in which the language is to be learned. Throughout, he stresses the influence of integrative motivation upon positive outcomes. Clément’s model (see Noels and Clément 1998) embeds individual motivations still more deeply in the social setting and has particular relevance for those language learners who are also minority-group members, and whose first language is threatened by the forces of those speaking the second. In the formulations of Giles and his colleagues (see Giles and Coupland 1991), language learning is considered as an intergroup process, with more attention given to assimilative tendencies and apprehensions, to the preservation of ethnicgroup boundaries and identities. Spolsky’s ‘general theory’ (1989) attempts to bring together all aspects of language learning, and assumes learning to be an interactive and socially contextualized process. Dörnyei and Ushioda (2009) provide an excellent treatment, in eighteen chapters, by all the important current researchers in the field. Beyond the introductory and concluding sections, the two editors also contribute substantive chapters themselves; additionally, there are noteworthy pieces by MacIntyre, Clément, Kormos, Segalowitz, Noels, and many others. An important work, one that embeds discussions of attitude and motivation in the wider language literature is the 1,000-page encyclopedic

Bilingualism and Multilingualism: Some Central Concepts  19 book by Ellis (2008). Roughly speaking, then, we can see the steadily more nuanced emphasis upon social and motivational features. (See Butler, chapter 5, this volume, for further discussion.) Application and prediction are the acid test in all such theoretical models, and some have suggested that they have done little more than codify and formalize what has been known for a long time. Nonetheless, they all scotch the myth that some people, or some groups, have no ‘head’ for languages and that secondlanguage aptitude is a rare commodity usually best seen in non-anglophones. Instead, they stress the power of the setting and, within it, the desires, needs, attitudes, and motivations of ordinary people: the social factors impinging upon language learning are, quite simply, the most important ones. We should also bear in mind that, for those many millions who develop bilingual or multilingual fluencies in the informal realm of daily life, simple necessity is the great motivator and the great determiner of how far this competence develops. It can dwarf all other features and, in particular, can ride roughshod over personal attitudes and motivation. Most historical changes in language use have a bilingual element, and most owe much more to socioeconomic and political exigencies than they do to attitude. The adoption of English by the Irish population, for example, was not generally accompanied by favorable or ‘integrative’ attitudes. There may have been a grudging instrumentality at work, but it certainly was not of the type which pushes students to study French or German in the hopes of joining the diplomatic service.

Language and Identity Language clearly intertwines powerfully with conceptions and definitions of allegiance and ‘belonging.’ It possesses more than instrumental value; it is the vehicle of tradition and culture, and the medium of group narrative. Issues of psychology and sociology, of symbol and subjectivity, become important whenever we observe language in society. When more than one language is involved, then, we should expect ramifications in terms of identity and ‘groupness.’ Much of interest here rests upon the degree to which bilinguals possess either two (theoretically) separately identifiable systems of language – from each of which they can draw, as circumstances warrant – or some more intertwined linguistic and, perhaps, cognitive duality. As Hamers and Blanc (2000) point out, we are far from having compelling empirical data here. There is a difficult circularity at work, one that confounds all scientific attempts to link the observable to the intangible: the ambiguous or unclear results of the relatively few studies of the non-verbal repertoires (for example) of bilinguals do not provide clear indications of likely underlying mechanisms; on the other hand, plausible variations in rational accounts of these mechanisms make the interpretation of subtle behavioral differences hard to assess. Whether we are interested in verbal communication, its paralinguistic accompaniments, or the broader reaches of personality traits generally, we find very little experimental evidence. It is interesting that, in their

20  Overview and Foundations massive study of bilingualism, Baker and Jones (1998) give only six pages (out of more than 750) to a section on personality. Consider, for instance, the ‘popular’ (and, sometimes, academic) view that bilinguals must have some sort of split mentality – two individuals in one, as it were. Grosjean (1982) and others have reported that bilinguals sometimes feel, themselves, that language choice draws out, and draws upon, different person­ alities. But, as Baker and Jones (1998) and Hamers and Blanc (2000) note, the evidence here is anecdotal at best. Indeed, we could go a bit further, and point to the large logical and rational difficulties which some two-in-one arrangement would create. There is certainly, however, evidence that language choice may implicate different aspects of the personality: bilinguals responding to interviews and questionnaires are liable to give slightly different pictures of themselves, depending upon the language used. They may make different responses to objective or projective probes, responses may be more emotional through one variety (typically, but not inevitably, their maternal language – see Martinovich and Altarriba, chapter 12, this volume), they may more strongly affirm their sense of ethnic identity in one language than in another, and so on (see, for example, studies by Ervin, Guttfreund, Bond, and others, usefully summarized in Hamers and Blanc 2000). The fact that different social settings and variations in language–affect linkages lead to different patterns of self-presentation clearly does not imply separate personalities, although it does suggest an enhanced repertoire of possibility. People belong to many groups, and all of them – all, at least, that have boundaries possessing some degree of permanence – have characteristics which mark their identity. This marking is, of course, more or less visible at the level of the individual member. The implication is that each of us may carry the tribal markings of many groups, that our ‘group identity’ is itself a mosaic rather than a monolith. Still, it is clear that, where language issues are central, the pivotal group is the ethnocultural community: overlaps of importance may occur because of simultaneous membership in gender, socioeconomic, educational, occupational, and many other categories, but the base here is an ethnic one. The point at issue, then, is the significance of a bilingualism which links an individual to more than one ethnocultural community. How does it feel, we might ask, to have a foot in more than one camp? Is it this that could lead to that dubious conception of psychological duality? Or is such duality the origin of the expanded acuity and awareness that some have claimed for bilinguals? The short answers to these sorts of questions are all positive, or potentially positive, in a world where complicated patterns of social relations are made more intricate still by a very wide (theoretically infinite, in fact) range of linguistic capabilities. Of course, a great deal of bilingualism has very little emotional significance: the purely instrumental fluencies needed to conduct simple business transactions do not, after all, represent much of an excursion from one’s ethnic base camp. This is probably a rather larger category than is often thought. For example, breadth of multiple fluencies does not, per se, imply emotional or psychological depth; it may, more simply, reflect the exigencies of a complicated public life. On the other hand, it is certainly possible to hold dual (or multiple) allegiances, involving different-

Bilingualism and Multilingualism: Some Central Concepts  21 language groups, in the absence of personal bilingualism. The attachment felt by the English-speaking Irish or Welsh to a culture and an ancestry whose language they no longer possess is a psychologically real one, and demonstrates the continuing power of what is intangible and symbolic. It is those many bilinguals and multilinguals whose competence is more deepseated and whose abilities go beyond a prosaic instrumentality that comprise the category of interest when considering the relationship between bilingualism and identity. It is surely the case that the deeper the linguistic and cultural burrowing into another community the greater the impact upon identity. This in turn suggests that those whose bilingual competence is nurtured early will, other things being equal, have a firmer foot in the two (or more) camps. One camp will usually have psychological and emotional primacy. But there are some cases in which bilingual or multilingual capacities, linked to their several cultural bases, develop so early and/or so deeply that a primary allegiance is hard to discover. Steiner (1992), mentioned in the opening paragraph, is by his own account maternally and perfectly trilingual. Further, he has suggested that such ‘primary’ multilingualism is an integral state of affairs in itself. There has been virtually no research on the consequences for identity of multilingual tapestries so closely woven, but one imagines that there are subtleties here that go far beyond simple additive relationships. Since, as we have seen, it is difficult to define and assess perfectly and fully balanced bilingualism, even polyglots like Steiner might fall short under the most rigorous examination; nonetheless, more attention to deep-seated multiple fluencies is indicated. As we move towards the bilingualism of more ordinary individuals, we move more obviously towards the idea of a unitary identity – woven from several strands, to be sure, but inevitably influenced by one language and culture more than by others. But, if we move from the Steiners (and Conrads, Nabokovs, Kunderas, Stoppards, and all the rest), whose literary power, and the ability to reflect in meaningful ways upon its multifaceted origin, are simply unavailable to most people, we must not imagine that we have moved away from enlarged identities per se. It is both the obligation and the fulfillment of intellectual life, after all, to express what less articulate souls may somehow feel or possess. When we consider that the language competences of most bilinguals are shallower than those of the Steiners of the world – broader, sometimes, but rarely as deep – and that neither the capacity nor the inclination to think much about identity is a widely distributed quantity, we realize again what important questions remain to be asked, what research, more psychological than linguistic, still needs to be undertaken. The intellectuals can look after themselves here: Steiner (1975) has written famously about the ‘extraterritoriality’ of multilingual writers; Ilan Stavans argues that monolingualism is a form of oppression (see Kellman 2000); others, from Goethe to Eliot, have argued over the ability – particularly the poet’s ability – to be fully expressive beyond the muttersprache. We need more and fuller reports from less elevated quarters, too. As it is, we rely largely upon inference to support the contention that it is the identity components, the symbols of the tribe, that energize languages beyond

22  Overview and Foundations their instrumental existences. One large and obvious example here is the powerful association between language and nationalism. Since the latter is, among other things, a pronounced and often mobilizing sense of groupness, it follows that any language component will be carefully delineated. And so, historically, it is. The language in which you do your shopping, and which (if you thought much about it) is also the variety in which your group’s tradition is inscribed, can become a symbol of your oppressed state, a rallying-point, a banner under which to assemble the troops. Would people be so ready to sacrifice for something that was of purely mundane importance? We might regret that circumstances encourage us to put aside a familiar tool, and learn to use another – but we go to war over histories, not hammers. Beyond the bridging (discussed above) that multilingualism necessitates, the contact of different languages and groups can also generate moat-building in the service of identity protection. An interesting form of this defensive strategy is linguistic prescriptivism or purism which, given free rein, would often lead to proscription. Concern about the ‘contamination’ of one language by another, about infiltration and borrowing and about the bullying of small languages by larger ones is an historically longstanding worry; the desire to keep one’s language ‘pure’ has been strong, at least since the time of the decline of Latin in Europe, the rise of standardized vernaculars, the development of printing, and the growth of literacy. Linguistic standardization does, of course, require some prescriptivism, and the classic instruments here have been academies and dictionaries. The first institution devoted specifically to language was the Accademia della crusca of Florence, given royal blessing in 1572. The Académie française was established in 1634, and the Real Academia Española followed in 1713. Many more such bodies followed: in fact, language academies (or similar entities) are very much the rule.3 Beyond the needs of standardization, however, we often find the initial prescriptivist impulses becoming more blatantly protective ones. At a purely linguistic level, this seems both unrealistic and undesirable. After all, languages have always borrowed from one another, and history reveals that attempts to erect walls between them generally turn out to be misguided and fruitless. It is the powerful link between language and identity that is crucial here, however – prescriptivism writ large is more a psychosocial matter than a linguistic one. Returning to multilingual rubrics, the important associations of a particular language with a particular base camp are made clearer when we think about translation. This is an exercise driven by obvious necessity and, if language were not invested with emotion and association, its operation would be unremarkable. While employing them, we might applaud those whose expertise allows them the access denied the rest of us, but we would rarely be suspicious. And yet the old proverb says traduttori, traditori. We would hardly equate translation with treason unless we feared that ‘hoarded dreams, patents of life are being taken across the frontier’ (Steiner 1992: 244). And what are these ‘patents of life,’ if not the psychological collections of past and present that are felt to belong to ourselves alone? An informal Whorfianism tells us that every language interprets and presents the world in a somewhat different way, that the unique wellsprings

Bilingualism and Multilingualism: Some Central Concepts  23 of group consciousness, traditions, beliefs, and values are intimately related to a given variety. Translation may thus mean the revealing of deep matters to others, and cannot be taken lightly. The translator, the one whose multilingual facility permits the straddling of boundaries, is a necessary quisling. But necessity is not invariably associated with comfort, and not even their employers care very much for traitors. For monolingual majority-group members in their own ‘mainstream’ settings, the instrumentality and the symbolism of language are not split and, for most such individuals, the language–identity linkage is not problematic: indeed, it is seldom considered. Minority-group members, however, rarely have this luxury; for them, matters of language and culture are often more immediate. Now, while it is true that no simple equation exists between bilingualism and minority-group membership, it is also true that many bilinguals are found in the ranks of ‘smaller’ or threatened societies. The implication is that a link will often exist between bilingualism and a heightened awareness of, and concern for, identity. Specific linguistic manifestations include attempts at language maintenance or revival, the use of language in ethnic or nationalist struggles, the efforts to sustain at least some domains in the face of external influence, and so on. A more general consequence is that the position and the responses of minority groups focus attention on the possibility – and, in many instances, the inevitability – of a split between the communicative and the symbolic functions of language: you may have to live and work in a new language, a medium that is not the carrier of your culture or the vehicle of your literature. In these sorts of settings we see, in fact, an extended value to the study of bilingualism and identity. First, the attitudes and actions of bilinguals in situations of risk and transition have a special poignancy and visibility: identities, like everything else, are thrown into sharper relief when threats are perceived. Second, these same attitudes and actions can galvanize others, and can remind a larger and often unreflective society that matters of language and identity are not relevant for those multilingual ‘ethnics’ and ‘minorities’ alone. The study of bilingualism and multilingualism, with all its many ramifications and technicalities, is clearly an intrinsically important sociolinguistic area, but it can also illuminate much larger patches of social life. The importance of being multilingual is, above all, social and psychological rather than linguistic. Beyond types, categories, methods, and processes is the essential animating tension of identity.

NOTES 1  In an earlier version of this chapter (Edwards 2004), I gave more attention to facets of individual bilingualism and to the technicalities of definition and measurement than I do in this contribution, in which more room must be made for broader societal issues. Readers should also note that, throughout this piece, I often refer simply to ‘bilingualism’ for ease and expediency; in most cases ‘multilingualism’ may also be understood, mutatis mutandis. I acknowledge with thanks the support of the St. Francis Xavier University Council for Research in the preparation of this chapter.

24  Overview and Foundations 2  Given present space limitations, I do not reproduce the tabular sketches that accompany the much fuller description of this typological exercise; these may be found in Edwards (2010a). 3  Conspicuous by its absence is any comparable English-language institution. There was certainly interest in one, dating at least from the time of the establishment of the continental bodies. The English requirements for language standardization were met by lexicographers – essentially oneman academies: Samuel Johnson in England and Noah Webster in America.

REFERENCES Baker, C. (1988). Key Issues in Bilingualism and Bilingual Education. Clevedon: Multilingual Matters. Baker, C. and Jones, S. (1998). Encyclopedia of Bilingualism and Bilingual Education. Clevedon: Multilingual Matters. Bates, D. (2010). Learn another language to help protect yourself from Alzheimer’s disease. Daily Mail, October 13. Bialystok, E. (2009). Bilingualism: The good, the bad and the indifferent. Bilingualism: Language and Cognition 12: 3–11. Bialystok, E. (2010). Global-local and trailmaking tasks by monolingual and bilingual children. Developmental Psychology 46: 93–105. Bloomfield, L. (1933). Language. New York: Holt. Clyne, M. (2003). Dynamics of Language Contact. Cambridge: Cambridge University Press. Conan Doyle, A. (1966). A study in scarlet. In The Complete Sherlock Holmes Long Stories. London: John Murray. Dörnyei, Z. and Ushioda, E. (eds.) (2009). Motivation, Language Identity and the L2 Self. Bristol: Multilingual Matters. Edwards, J. (1995). Multilingualism. London: Penguin. Edwards, J. (2004). Foundations of bilingualism. In T. Bhatia and W. Ritchie (eds.), The Handbook of Bilingualism. 7–31. Oxford: Blackwell. Edwards, J. (2009). Language and Identity. Cambridge: Cambridge University Press. Edwards, J. (2010a). Minority Languages and Group Identity: Cases and Categories. Amsterdam: John Benjamins. Edwards, J. (2010b). The treason of translation? Bilingualism, linguistic borders and identity. In H. Tonkin and M. Frank (eds.), The Translator as Mediator of Cultures. 89–105. Amsterdam: John Benjamins. Edwards, J. (2010c). Language Diversity in the Classroom. Bristol: Multilingual Matters.

Ellis, Rod (2008). The Study of Second Language Acquisition, 2nd edn. Oxford: Oxford University Press. Extra, G. and Gorter, D. (2008). The constellation of languages in Europe: An inclusive approach. In G. Extra and D. Gorter (eds.), Multilingual Europe: Facts and Policies. 3–60. Berlin: Mouton de Gruyter. Ferguson, C. (1962). The language factor in national development. In F. Rice (ed.), Study of the Role of Second Languages in Asia, Africa, and Latin America. 8–14. Washington: Center for Applied Linguistics. Ferguson, C. (1966). National sociolinguistic profile formulas. In W. Bright (ed.), Sociolinguistics. 309–24. The Hague: Mouton. Ferguson, C. (1991). Diglossia revisited. Southwest Journal of Linguistics 10: 214–34. Gardner, R. (2010). Motivation and Second Language Acquisition. Bern: Peter Lang. Gardner, R. and Lambert, W. (1972). Attitudes and Motivation in Second-Language Learning. Rowley, MA: Newbury House. Giles, H., Bourhis, R., and Taylor, D. (1977). Towards a theory of language in ethnic group relations. In H. Giles (ed.), Language, Ethnicity and Intergroup Relations. 307–48. London: Academic Press. Giles, H. and Coupland, N. (1991). Language: Contexts and Consequences. Milton Keynes: Open University Press. Goodenough, F. (1926). Racial differences in the intelligence of school childen. Journal of Experimental Psychology 9: 388–97. Grenoble, L. and Whaley, L. (1998). Toward a typology of language endangerment. In L. Grenoble and L. Whaley (eds), Endangered Languages: Current Issues and Future Prospects. 22–54. Cambridge: Cambridge University Press. Grosjean, F. (1982). Life with Two Languages. Cambridge, MA: Harvard University Press.

Bilingualism and Multilingualism: Some Central Concepts  25

Haarmann, H. (1986). Language in Ethnicity: A Spolsky, B. (1989). Conditions for Second-Language View of Basic Ecological Relations. Berlin: Learning. Oxford: Oxford University Press. Mouton de Gruyter. Steiner, G. (1975). Extraterritorial: Papers on Hamers, J. and Blanc, M. (2000). Bilinguality and Literature and the Language Revolution. Bilingualism. Cambridge: Cambridge Harmondsworth: Penguin. University Press. Steiner, G. (1992). After Babel. Oxford: Oxford Haugen, E. (1953). The Norwegian Language in University Press. America: A Study in Bilingual Behavior. Stewart, W. (1962). An outline of linguistic Philadelphia: University of Pennsylvania typology for describing multilingualism. Press. In F. Rice (ed.), Study of the Role of Second Haugen, E. (1972). The ecology of language. In Languages in Asia, Africa, and Latin America. A. Dil (ed.), The Ecology of Language: Essays by 15–25. Washington: Center for Applied Einar Haugen, 325–39. Stanford: Stanford Linguistics. University Press. Stewart, W. (1968). A sociolinguistic typology Hazlitt, W. (1901). Table-Talk: Essays on Men and for describing national multilingualism. In J. Manners. London: Oxford University Press. Fishman (ed.), Readings in the Sociology of Kellman, S. (2000). The Translingual Imagination. Language. 531–45. The Hague: Mouton. Lincoln: University of Nebraska Press. Tancock, L. (1954). Introduction to Germinal Ladegaard, H. (2006). ‘I like the (Émile Zola). Harmondsworth: Penguin. Americans . . . but I certainly don’t aim for an Tsunoda, T. (2006). Language Endangerment and American accent’: Language attitudes, vitality Language Revitalization. Berlin: Mouton de and foreign language learning in Denmark. Gruyter. Journal of Multilingual and Multicultural Vail, P. (2006). Can a language of a million Development 27: 91–108. speakers be endangered? Language shift and Leopold, W. (1939–49). Speech Development of a apathy among Northern Khmer speakers in Bilingual Child (4 volumes). Evanston, IL: Thailand. International Journal of the Sociology Northwestern University Press. of Language 178: 135–47. Moreno, S., Bialystok, E., Wodniecka, Z., and Weinreich, U. (1953). Languages in Contact. The Alain, C. (2010). Conflict resolution in Hague: Mouton. sentence processing by bilinguals. Journal of White, P. (1991). Geographical aspects of Neurolinguistics 23: 564–79. minority language situations in Italy. In C. Noels, K. and Clément, R. (1998). Language in Williams (ed.), Linguistic Minorities, Society and education. In J. Edwards (ed.), Language in Territory. 44–65. Clevedon: Multilingual Canada. 102–24. Cambridge: Cambridge Matters. University Press. Yag˘ mur, K. and Kroon, S. (2003). Ethnolinguistic Paulston, C., Haragos, S., Lifrieri, V., and vitality perceptions and language Martelle, W. (2007). Some thoughts on revitalisation in Bashkortostan. Journal of extrinsic linguistic minorities. Journal of Multilingual and Multicultural Development 24: Multilingual and Multicultural Development 28: 319–36. 385–99. Yag˘ mur, K. and Kroon, S. (2006). Objective and Peal, E. and Lambert, W. (1962). The relation of subjective data on Altai and Kazakh bilingualism to intelligence. Psychological ethnolinguistic vitality in the Russian Monographs 76: 1–23. Federation Republic of Altai. Journal of Poplack, S. (1980). Sometimes I’ll start a Multilingual and Multicultural Development 27: sentence in English y terminó en español. 241–58. Linguistics 18: 581–616. Zamenhof, L. (1887). Meждунapoдньιŭ язьικь пpeдucлoвue u noлньιŭ yч язьικь пpeдucлoвue u noлньιŭ yчeбнuκь (por Rusoj). Ronjat, J. (1913). Le développementMeждунapoдньιŭ du langage observé chez un enfant bilingue. Paris: [International Language: Preface and Complete Champion. Textbook (for Russians).] Warsaw: Kelter.

2  Conceptual and Methodological Issues in Bilingualism and Multilingualism Research LI WEI

Bilingualism and multilingualism refer to the coexistence, contact, and interaction of different languages. The coexistence may take place at the societal level or the individual level. A society can have many different languages spoken and used by different groups of individuals. The individuals themselves may not necessarily be bilingual or multilingual. Likewise, individuals in a community may be bilingual or multilingual, while the community as a whole recognizes only one language for public use through legislation and other forms of regulatory control. For a society or an individual to become bilingual or multilingual, contacts and interactions between languages are essential. Over the years, bilingualism and multilingualism have become a major topic of multidisciplinary research, attracting the attention of scholars in disciplines ranging from linguistics, psychology, neurology, and computer science, to sociology, education, public policy, and management. Scholars with such different disciplinary backgrounds and research interests often approach issues of bilingualism and multilingualism with very different, sometimes contradictory, views of what bilingualism and multilingualism consist in, which, in turn, affect their choice of research methods and design. In this chapter, I will outline the various conceptualizations of bilingualism and multilingualism, the key research themes and methodological perspectives, as well as the connections between the different perspectives that have emerged from the existing literature.

The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

Conceptual and Methodological Issues  27

The Fundamentals for Bilingualism and Multilingualism The coexistence of different languages in society or in an individual presupposes three things: 1 there are different languages; 2 different languages have the opportunity to get into contact with each other; and 3 human beings are capable of learning and using multiple languages and human societies are capable of accommodating and managing multiple languages.

Linguistic diversity: What does it mean? The claim that there are different languages in the world is in fact more complex than it seems. On the one hand, we are all very used to terms such as Arabic, English, Japanese, Spanish, or Swahili, and consider these as different languages. On the other hand, all languages share the same basic design features. Indeed, the most significant scientific discovery of modern linguistics is that there are common principles amongst all human languages, and in the grand scheme of things, the above named languages are nothing more than minor variations on a theme. Mark Baker in his popular book The Atoms of Language (2001) compares English with Mohawk, a polysynthetic language spoken by a few thousand Native Americans in Quebec, Ontario, and New York. Whereas English speakers communicate complex ideas by putting words into sentences, Mohawk speakers just construct bigger words. ‘Katerihwaiénstha’ in Mohawk translates, roughly, as ‘I am a student’ in English. Yet, as Baker argues, the two languages are really very similar. He lists eight basic grammatical properties that Mohawk and English share, such as ‘Agents are subjects, and undergoers are objects.’ The contrast between Mohawk and English is due to just a few subtle differences, most of them can be captured in terms of a single ‘polysynthesis parameter.’ So, are languages different from each other? There are, broadly speaking, three ways to answer this question, one essentially negative and two affirmative from historical and current perspectives, respectively. The essentially negative answer is often known as the language universal approach. It does not deny the fact that different groups of speakers use different codes to express their meanings and intentions in everyday life, and that these different codes have come to adopt different names and labels over time and are therefore known as different languages. But, as has been said earlier, all languages share the same basic design features. Scholars who adopt the language universal perspective argue that language arises from an innate ‘language organ’ that all neurologically normal humans possess (Chomsky 1986). The different languages, as we know them, are therefore different representational systems of the universal and biologically rooted design features.

28  Overview and Foundations From this perspective, the focus of linguistic science should be on delineating these universal features or principles for all human languages. The two affirmative answers do not deny the existence of the shared features of human languages. However, in contrast to the language universal perspective, attention is given to the differences and diversity amongst human languages, how they came into being, and how they manifest themselves in everyday human interaction. The first, historical approach considers linguistic diversity as an outcome of the evolution of languages, which is assumed to be akin to the evolution of species, complete with competition, adaptation, and extinction. A key process in the evolution of languages is contact. In another popular book, The Power of Babel, John McWhorter (2001) describes how contact, or lack of it, between speaker groups can lead to interesting and complex patterns of linguistic diversity. For instance, creoles – languages that are built up as an outcome of contacts between people – often show fairly simple morpho-syntactic structures, while isolated languages spoken by small hunter-gatherer groups often have extremely complicated morphologies. From this perspective, languages that have developed a writing system, such as English and Chinese, have a further complication in that the writing system can instill stilted and unnatural patterns into people’s speech and help freeze languages in time, creating grammatical and lexical fossils. The historical, evolutionary approach to linguistic diversity has led to a number of classification systems for human languages: (i) genetic classification, which is based on historical relatedness of languages and groups different languages into families that are believed to have a common ancestor language, such as the IndoEuropean, the Afro-Asiatic, the Austronesian, and the Sino-Tibetan languages; (ii) typological classification, based on similarity of structural features, especially word order of the verb, the subject and the object in a sentence, resulting in SVO, SOV, VSO languages and so on. This classificatory approach shares with the universal perspective on linguistic diversity an interest in common features, or language universals, but differs in the sense that it does not argue for a biological base in the human brain; and (iii) areal classification, based on geographical closeness and contacts between language-speaking communities, such as the Balkan languages, the Caucasian languages, the East Asian languages. Although the members of each of these areal language groups are not closely related genetically, they have converged in the course of history due to prolonged contacts with each other, resulting in shared ‘areal features.’ A rather different approach to linguistic diversity that pays more attention to the differences between human languages than their commonalities is a current, practice-based approach that focuses on how linguistic diversity manifests itself in everyday, here-and-now interaction and how human beings manage the differences between the languages they use. Scholars adopting this particular perspective are usually highly skeptical about the commonsense notion of language. Indeed, there is a temptation to treat categories such as English and Chinese as bona fide entities, distinguishable from so-called ‘dialects’ and from mixedlanguage varieties. But in fact, what gets counted as a language is often a matter of social and political considerations and has little to do with linguistic facts. The

Conceptual and Methodological Issues  29 so-called Chinese dialects such as Mandarin, Cantonese, or Hokkien are mutually unintelligible and are far more different from each other than the Scandinavian languages, Danish, Norwegian, and Swedish, which are counted as separate languages even though they share a considerable number of common features. Differences between languages are therefore socio-historically constructed. In this sense, the approach shares with the universal perspective the belief that language acquisition and use are common human capacities, and with the historical perspective that linguistic diversity is an outcome of language evolution. But the crucial difference is that the current, practice-based approach emphasizes the emerging nature of linguistic diversity and the role of everyday social interaction in creating, maintaining, and resolving differences between different groups of language users. A language, from this particular perspective, is a property of the group, and is constructed through individuals’ practices in everyday social interaction, and is a culturally and historically accumulated value system. So, we have three rather different conceptualizations of a language: as a particular representational system based on the biologically rooted language faculty; as complex and historically evolved patterns of structures, as revealed in genetic, typological, and areal classifications of languages; or as a social practice and a culturally loaded value system. These different conceptualizations of language have a fundamental impact on the way scholars of different disciplinary backgrounds and interests view bilingualism and multilingualism, which are essentially about the contacts and interactions between languages and language users.

Language contact There is a popular metaphor in linguistics that a language is a living organism, which is born, grows, and dies. However, a language is a human faculty: it coevolves with us, homo sapiens; and it is we who give a language its life, change it, and, if so desired, abandon it. So, when we speak of ‘language contact,’ we are really talking about people using different languages coming into contact with one another. There are many reasons for users of different languages to come into contact. Some do so out of their own choosing, while others are forced by circumstances. For instance, political or military acts such as colonization, annexation, resettlement, and federation can have an immediate impact on language users. People may become refugees, either in a new place or in their homeland, and have to learn the language of their new environment. After a successful military invasion, the indigenous population may have to learn the invader’s language in order to prosper. Famine, floods, volcanic eruptions, and other natural disasters can also be the cause of major movements of population. New language contact situations then emerge as people are resettled. Some of the Irish and Chinese resettlements in North America were the result of natural disasters. People may wish to live in a country because of its religious significance, or to leave a country because of its religious oppression. In either case, a new language may have to be learned. Russian speakers in Israel are a case in point. Similarly, a desire to identify with

30  Overview and Foundations a particular ethnic, cultural, or social group usually means learning the language of that group. Minority ethnic and cultural groups may wish to maintain their own languages, which are different from the languages promoted by the governing state or institution. Nationalistic factors are particularly important. Very large numbers of people across the world have migrated to find work and to improve their standard of living. Economic factors account for most of the linguistic diversity of the United States and an increasing proportion of the bilingualism and multilingualism in present-day Europe. Of course, one does not have to move to a different place to come into contact with people using a different language. There are plenty of opportunities for language contact in the same country, the same community, the same neighborhood, or even the same family. Indeed, globalization has made it easier for people from different parts of the world to experience different cultures, lifestyles, and languages. The desire for education, learning, and access to knowledge also plays an important role in language contact. It led to the universal use of Latin in the Middle Ages, and today motivates the international use of English. The availability of information and communication technologies (ICT), such as the internet, has led to a further expansion of the use of English across the world so that the vast majority of ICT users are nonnative speakers of English. Many other factors can be added to this list. For further discussions of the various causes of language contact, see Li Wei, 2007, 2010a.

Learning, using, and managing multiple languages The usual consequence of language contact is bilingualism, or even multilin­ gualism, which can be found either in an individual language user or in a society as a whole. The different socio-historical processes of language contact can lead to many different types of bilingual and multilingual language users and societies. The following three are typical cases of societal bilingualism and multilingualism: 1 Territorial bilingualism and multilingualism. Groups of speakers of specific languages find themselves mostly within their own geographically and politically defined territories. Each group’s language, or languages, are protected by legislation and institutional support, while other languages are also available though without official status. Territorial bilingualism and multilingualism can be found in countries such as Belgium, Canada, India, Spain, and Switzerland, all of which apply the principle of territorial bilingualism and multilingualism in different ways. 2 Diglossia. Two (or more) languages coexist at the community level but are used by sections of the society in a complementary way, i.e. one language having a higher status than the other and being used in certain functions and domains that are different from those in which the other language is used. It is important to remember that diglossia is a construct to describe the relationship between two or more languages; it is not a static state of language use

Conceptual and Methodological Issues  31 (Ferguson 1959; Hudson 1992). Over time, the relationships between the different languages may change, due to various social factors, and the diglossic configuration can become highly complex. Examples of simple binary diglossia include Guarani vs. Spanish in Paraguay; double-overlapping diglossia of African vernacular vs. Swahili and Swahili vs. English in Tanzania; doublenested diglossia of Sanskrit vs. Hindi vs. Khalapur in India; linear polyglossic diglossia of English, Mandarin Chinese, Malay, Tamil, other Chinese, other Malay, other Indian languages in Singapore; and complex triglossia of standard French and Dutch vs. Belgian French and Belgian Dutch vs. local dialects (Walloon and Flemish) in the Belgian capital of Brussels. 3 Widespread multilingualism. In many countries of Africa, Asia, and Latin America, many different kinds of native languages of indigenous ethnic groups or nations coexist with one or more languages of wider communication. The majority of the people of these communities are also highly multilingual. In all bilingual and multilingual societies, how to manage the relationships across the different languages is a critical question. This gives rise to issues such as language maintenance and language shift, language planning, and language policy. As has been mentioned earlier, it is an interesting fact that not all individuals in a bilingual or multilingual society are bilingual or multilingual themselves. Individuals become bilingual or multilingual in rather different ways. For instance, some people may become bilingual or multilingual because they are born into bilingual or multilingual families and are exposed to multiple languages simultaneously from birth, while others learn and acquire additional languages at a later age in life due to schooling or changes in life experience. In theory, these different routes to bilingualism should result in different bilingual behavior and cognitive organizations of the bilingual brain. But at the moment, there is very little systematic research linking the process of language contact with the outcome of bilingualism or multilingualism at the individual level. Instead, researchers have spent a great deal of time investigating the various factors that lead to the success or failure of becoming bilingual and multilingual, addressing the intriguing question of why every individual in a bilingual or multilingual society isn’t bilingual or multilingual. Some of the factors appear to be learner-internal, i.e. are conditioned by the learner’s biological and cognitive capacities, while others are more social. With regards to the learner-internal factors, research evidence suggests that human beings’ ability to learn languages is closely related to such biological factors as age and related memory capacity. Infants seem to be able to acquire new languages with great ease, whereas adults often have difficulty in achieving the same level of fluency or accuracy in the later-learned languages. Such facts gave rise to the ‘critical period hypothesis’ which claims that there is an ideal age to acquire language in a linguistically rich environment, after which the individual will never achieve a full command of language (Lenneberg 1967; Penfield and

32  Overview and Foundations Roberts 1959). However, there is little agreement as to when that critical period is. Empirical findings are varied: some demonstrate prepubescent children acquire language easily, and some that older learners have a certain advantage and can achieve native-level proficiency (see studies in Singleton and Lengyel 1995). Brain plasticity is often cited as the factor underlying the critical period. More recently, it has been suggested that the delayed development of the prefrontal cortex in children and a concomitant delay in the development of cognitive control may facilitate convention learning, allowing young children to learn language far more easily than cognitively mature adults and older children (Ramscar and Gitcho 2007; Thompson-Schill et al. 2009). Nevertheless, this is a highly controversial and live topic for research. Advances in brain imaging technologies are generating new findings all the time. The conceptualization of language as a mental faculty clearly underlies much of the research on the critical period hypothesis. Many researchers do not believe in the existence of the critical period for language learning at all. Instead, they point to social and environmental factors such as attitude, motivation, input, and interaction, as well as pedagogical effect and language policy. Individuals learn languages for very different reasons and with very different motivations: some learn new languages because they are married to speakers of other languages, while others seek employment that requires the use of certain languages. Community attitudes toward the language being learned can also have a profound impact on individual learners. Where the community has a broadly negative view of a particular language and its users, or a negative view of its relation to them, learning is typically much more difficult. In the meantime, the learner’s direct contact with the language users, i.e. input, has a crucial impact on the learning process and learning outcome. Other factors that researchers have examined in terms of success or failure of becoming bilingual and multilingual include motivation, personality, and gender. These factors are at the interface between learner-internal and -external dimensions. For instance, evidence from neuroscience suggests that boys and girls use different parts of their brains to process some basic aspects of grammar (Burman et al. 2008), while sociolinguists have shown that men and women use languages in different ways for communicative purposes (Holmes and Meyerhoff 2004). However, there is very little research into these gender differences in bilingual or second-language acquisition. Some researchers have suggested that there are gender differences in motivations for language learning (Kissau 2006; MacIntyre et al. 2002). It certainly seems to be the case that more girls than boys prefer to learn modern languages in schools and have somewhat better attainment in school-based language examinations. Yet none of the apparent difference can be solely attributed to biological differences between the two genders. It is more likely to be a sociocultural effect in that language learning and use is closely related to the different roles of girls and boys and women and men in society. Similarly, some personality traits have been linked to the relative success or failure in second or additional language learning. Extraverts or unreserved and outgoing people tend to acquire a second language better than introverts or shy people

Conceptual and Methodological Issues  33 (Kezwer 1987; MacIntyre and Charos 1996). But extraversion and introversion are sociocultural constructs, rather than biologically determined. One of the key questions that concern researchers of bilingualism and multilingualism is what it means to be bilingual or multilingual. For example, does a bilingual speaker’s brain function differently from that of a monolingual’s brain? In the majority of right-handed adults, the left hemisphere of the brain is dominant for language processing. There is some evidence to suggest that adult secondlanguage acquisition involves the right hemisphere more than first-language acquisition in language processing (Seliger 1982). As proficiency in a second language grows, right hemisphere involvement decreases and left hemisphere involvement increases. However, quantitative analyses of the existing data often show that the left hemisphere strongly dominates language processing for both monolinguals and bilinguals, and that differences between them are the exception rather than the rule (Galloway and Scarcella 1982). Bilinguals and multilinguals do not seem to vary from monolinguals in neurological processes; the lateralization of language in the brains of the two groups of speakers is similar (Fabbro 1999; Obler and Gjerlow 1998). A related issue concerns the mental representation of a bilingual’s two languages and the processing emanating from such representation. Evidence exists for both separate storage and shared storage of the two languages in the bilingual’s brain, resulting in the suggestion that bilinguals have a language store for each of their two languages and a more general conceptual store. There are strong, direct interconnecting channels between each of these three separate stores. The interconnections between the two languages comprise association and translation systems, and common images in the conceptual store act as mediators. Furthermore, speakers of different proficiency levels or at different acquisitional stages vary in the strength and directness of the interconnections between the separate stores in language processing; e.g., those highly proficient in two languages may go directly from a concept to the target language, while those whose second language is weaker than their first tend to use the first language to mediate (Heredia and Brown chapter 11, this volume; Kroll and Dussias chapter 9, this volume; Kroll and Tokowicz 2005; Runnqvist et al. chapter 10, this volume). Although the more general definitions of bilingualism and multilingualism would include people who understand multiple languages – in either spoken or written form or both – but do not necessarily speak or write all of them, a more common usage of the term refers to someone who can function in two or more languages in conversational interaction. Bilingual and multilingual speakers choose their languages according to a variety of factors, including the type of person addressed (e.g. members of the family, schoolmates, colleagues, superiors, friends, shopkeepers, officials, transport personnel, neighbors), the subject matter of the conversation, location or social setting, and relationship with the addressee. However, even more complex are the many cases when a bilingual talks to another bilingual with the same linguistic background and changes from one language to another in the course of conversation. This is what is known as code-switching (MacSwan chapter 13, this volume; Ritchie and Bhatia chapter 15, this volume).

34  Overview and Foundations There is a widespread impression that bilingual and multilingual speakers code-switch because they cannot express themselves adequately in one language. This may be true to some extent when a bilingual is momentarily lost for words in one of his or her languages. However, code-switching is an extremely common practice among bilinguals and takes many forms. A long narrative may be divided into different parts which are expressed in different languages; sentences may begin in one language and finish in another; words and phrases from different languages may succeed each other. Linguists have devoted much attention to the study of code-switching. It has been demonstrated that code-switching involves skilled manipulation of overlapping sections of two or more grammars, and that there is virtually no instance of ungrammatical combination of languages in codeswitching, regardless of the linguistic ability of the speaker. Some suggest that code-switching is itself a discrete mode of speaking (Gardner-Chloros 2009). There is another group of bilinguals who engage in cross-modality language production. This is the case with speech-sign bilinguals who, in addition to the oral modality, use the manual–visual modality in everyday communication. They are special in one aspect, i.e. the two different modalities allow for the simultaneous production of the two languages. In other words, one can speak and sign at the same time. Research has shown that such simultaneous bi-modal production is typically exemplified by the use of lexical items from both languages but only one set of grammatical rules, which is usually from the spoken language. Right now we know relatively little about how the two linguistic systems interact in the language production and processing of speech–sign bilinguals (for a review of existing studies of sign bilingualism, see Berent chapter 14, this volume and Dufour 1997). Indeed, much more work needs to be undertaken before we can fully appreciate the complexity of the language behavior of bilinguals generally. Bilingualism and multilingualism can change over the lifespan of an individual. Language attrition refers to the loss of a first- or later-acquired language, or a portion of that language, by individual speakers. It can happen at any point in time during an individual’s lifespan, because the necessary contact with speakers of that language is lacking over a lengthy period. It can also happen to elderly bilinguals who experience a cognitive decline of some kind. As in the study of bilingual and multilingual acquisition, what linguistic structures pose specific challenges and why they do so are central issues of investigation in language attrition research (Schmid et al. 2004).

Research Traditions and Methodological Perspectives Contrary to popular assumption, bilingualism and multilingualism research has a very long history and has always been highly multidisciplinary. Detailed documentation of language contacts in Europe, for example, dates back to the seventeenth century; Whitney’s analysis of the grammatical structure of codemixing was published in 1881; and Cattell’s experiments which compared word association and reaction times of bilingual and monolingual speakers were pub-

Conceptual and Methodological Issues  35 lished in 1887. Nevertheless, bilingualism and multilingualism have become a major focus of scientific research only in the last century, especially since the 1970s. Three disciplines that have influenced much of the research are psychology, linguistics, and sociology, which happen to coincide to a large extent with the three conceptualizations of language that have been discussed earlier in this chapter, as a mental faculty, as a system of structures, and as a social practice (see further discussions in Auer and Li Wei 2007 and Li Wei and Moyer 2008).

Psycholinguistic approaches In one of the early pioneering studies of bilingualism in the modern era, Weinreich (1953) proposed three types of bilinguals representing three types of relationship between the linguistic sign (or signifier) and its semantic content (its signified). In Type A (see below), the individual combines a signifier from each language with a separate unit of signified. Weinreich called them ‘coordinative’ (later often called ‘coordinate’) bilinguals. In Type B, the individual identifies two signifiers but regards them as a single compound, or composite, unit of signified; hence ‘compound’ bilinguals. Type C relates to people who learn a new language with the help of a previously acquired one. They are called ‘subordinative’ (or ‘subordinate’) bilinguals. His examples were from English and Russian (see also Heredia and Brown chapter 11, this volume): (A)

‘book’

‘kniga’

/buk/

/kn’iga/



Coordinative bilingual (B)

‘book’ = ‘kniga’

/buk/

/kn’iga/

Compound bilingual (C)

‘book’



/buk/ /kn’iga/

Subordinative bilingual

36  Overview and Foundations Weinreich’s typology is often interpreted in the literature as referring to differences in the degree of proficiency in the languages. But in fact the relationship between language proficiency and cognitive organization of the bilingual individual, as conceptualized in Weinreich’s model, is far from clear. Some ‘subordinate’ bilinguals demonstrate a very high level of proficiency in processing both languages, as evidenced in grammaticality and fluency of speech, while some ‘coordinative’ bilinguals show difficulties in processing two languages simultaneously (i.e. in code-switching or in ‘foreign’ word identification tasks) (Lambert 1969; MacNamara 1967; MacNamara and Kushnir 1971). It must also be stressed that using Weinreich’s distinctions, bilingual individuals are distributed along a continuum from a subordinate or compound end to a coordinate end, and can at the same time be more subordinate or compound for certain concepts and more coordinate for others, depending on, among other things, the age and context of acquisition (see Hamers and Blanc 2000: 7.1.1 for a discussion). Weinreich’s model, although not always explicitly acknowledged, has clearly influenced much of the thinking behind the later psycholinguistic modeling of the bilingual lexicon. Potter et al. (1984) proposed a theory of the manner in which bilingual lexical knowledge could be represented in the mind in terms of two competing models: the concept mediation model and the word association model. In the concept mediation model, words of both L1 and L2 are linked to an underlying, amodal conceptual system. In the word association model, on the other hand, words in a second language are understood through L1 lexical repre­ sentations. As can be seen in Figure 2.1(A) and (B), the models are structurally equivalent to Weinreich’s distinction between coordinative and subordinative bilingualism. At the same time, several researchers (e.g. Hummel 1986; Kolers and Gonzalez 1980; Paivio and Desrochers 1980) presented evidence for the so-called dual-store model, which poses separate but interconnected bilingual memory stores. This latter model also generated considerable research on the existence of the putative ‘bilingual language switch’ which has been postulated to account for the bilingual’s ability to switch between languages on the basis of environmental demands (e.g. MacNamara 1967; MacNamara and Kushnir 1971). Subsequent studies found conflicting evidence in favor of different models. Some of the conflicting evidence could be explained by the fact that different types

Conceptual store

L1 words

L2 words (a)

Conceptual store

L1 words

L2 words (b)

Figure 2.1  (A) Concept mediation model. (B) Word association model Source:  Adapted from Potter et al. (1984)

Conceptual and Methodological Issues  37

Conceptual store

L1 words

L2 words

Figure 2.2  The revised hierarchical model Source:  Adapted from Kroll and Stewart (1994)

of bilingual speakers were used in the experiments in terms of proficiency level, age, and context of acquisition. It is possible that lexical mediation is associated with low levels of proficiency and concept mediation with higher levels especially for those who have become bilingual in later childhood or adulthood. Some researchers called for a developmental dimension in the modeling of bilingual knowledge (Hamers 2004). Indeed, although the various psycholinguistic models were initially proposed without reference to bilingual acquisition, they clearly have important implications for acquisitional research and need to be validated with acquisition data. Kroll and Stewart (1994), for example, proposed the revised hierarchical model, which represents concept mediation and word association not as different models but as alternative routes within the same model (see Figure 2.2) (see also Heredia and Brown chapter 11, this volume). As has been mentioned earlier, an important distinctive feature of being multilingual is to be able to make appropriate language choices. Multilingual speakers choose to use their different languages depending on a variety of factors, including the type of person addressed, the subject matter of the conversation, location or social setting, and relationship with the addressee. However, even more complex are the many cases when a multilingual talks to another multilingual with the same linguistic background and changes from one language to another in the course of conversation. On the basis of such observations, Grosjean (2001) proposed a situational continuum that induces different language modes. At one end of the continuum, bilinguals are in a totally monolingual language mode, in that they are interacting with monolinguals of one – or the other – of the languages they know. At the other end of the continuum, bilinguals find themselves in a bilingual language mode, in that they are communicating with bilinguals who share their two (or more) languages and with whom they normally mix languages (i.e. code-switch and borrow). These are endpoints, but bilinguals also find themselves at intermediary points. Figure 2.3 is a visual representation of the continuum. The base languages (A or B) are located in the top and bottom parts of the figure, and the continuum is in the middle. Additional dimensions can be introduced when more than two languages are involved. At the monolingual end

38  Overview and Foundations LANGUAGE A Speaker X

MONOLINGUAL LANGUAGE MODE

BILINGUAL LANGUAGE MODE Speaker Y

LANGUAGE B

Figure 2.3  Language mode. The speaker’s position on the continuum is represented by the discontinuous vertical lines and the level of languages activation by the shade of the circles (black is active, gray is partly active, and white is inactive) Source:  Adapted from Grosjean (2001, Figure 1.1, p. 3)

of the continuum, bilinguals adopt the language of the monolingual interlocutor(s) and deactivate their other language(s) as much as possible. When a bilingual is in a bilingual mode, he or she will access or select words from two languages to produce sentences. How words are accessed or selected in speech production has been a central issue in psycholinguistics. In psycholinguistic research on bilingualism, the question becomes how different lexical items in different languages may be accessed or selected differently. Following earlier psycholinguistic models of speech production and more recent work by Clahsen (1999), Pinker (1999) and Jackendoff (2002). Myers-Scotton (2005) proposes a differential access hypothesis for bilingual production. The hypothesis assumes what is known as the 4-M model, which differentiates four types of morphemes – content morphemes and three types of system morphemes: early system morphemes, bridge late system morphemes, and outsider late system morphemes. It is suggested that the different types of morpheme under the 4-M model are differentially accessed in the abstract levels of the production process. Specifically, content morphemes and early system morphemes are accessed at the level of the mental lexicon, but late system morphemes do not become salient until the level of the formulator, as in Levelt’s speaking model (1989) and De Bot’s adaptation of that model (De Bot 1992). The hypothesis has received considerable attention in the literature and is being tested with a range of language contact phenomena. Other more general theoretical models of bilingual speech production (e.g. Green 1986, 1998) examine the way in which bilinguals control the use of their different languages. In view of the fact that bilingual speakers can choose which language they want to use in a specific context, it is suggested that the bilinguals’ languages are organized in separate subsystems which can be activated to different extents. A bilingual speaker who wishes to speak a particular language must ensure that activation exceeds that of competing languages. The regulation and control of the activation level requires specific cognitive resources.

Conceptual and Methodological Issues  39 As well as developing new models of bilingual behavior, psycholinguists have used the latest functional neuroimaging technologies to investigate the cognitive organization of languages in the bilingual brain (see Abutalebi et al. 2005 for a summary). The key research question here is the relationship between the pre-wired neurobiological substrate for multiple languages and environmental, time-locked influences such as age of acquisition, exposure, and proficiency. It has been found that whilst the patterns of brain activation associated with tasks that engage specific aspects of linguistic processing are remarkably consistent across different languages and different speakers, factors such as proficiency seem to have a major modulating effect on brain activity: more extensive cerebral activations associated with production in the less-proficient language and smaller activations with comprehending the less-proficient language. Another active area of psycholinguistic research on bilingualism and multilingualism has been bilingual and multilingual first-language acquisition (BAMFLA) (Li Wei 2010b). In an early study, Volterra and Taeschner (1978) suggested that bilingual acquisition went through three key stages: Stage I: the child has one lexical system composed of words from both languages. Stage II: the child distinguishes two different lexicons, but applies the same syntactic rules to both languages. Stage III: the child speaks two languages differentiated both in lexicon and syntax, but each language is associated with the person who uses that language. Although some studies both before and after Volterra and Taeschner’s had evidence supporting the model, there has been much criticism particularly of the claims made regarding the first two stages (e.g. De Houwer 1990; Genesee 1989; Meisel 1989). This is generally known as the ‘one-system-or-two’ debate, i.e. do bilingual children begin with a fused linguistic system and gradually differentiate the two languages, or do they start with a differentiated system? Part of that debate centers around the question: What counts as evidence for differentiation or fusion? Volterra and Taeschner (1978), for instance, based their decision on whether the child made appropriate sociolinguistic choices, i.e. whether the child spoke the ‘right’ language to the ‘right’ person. It was argued that awareness of the two languages as distinct plays a crucial role in deciding the issue of differentiation, and a child’s ability to make appropriate language choices reflects that awareness. However, as McLaughlin (1984) points out, the argument that bilingual children separate the languages when they are aware there are two systems is circular unless some criterion is provided for assessing what is meant by awareness other than that children separate the languages. In any case, we need to bear in mind that a child’s apparent (in)ability to choose the right language for the right addressee is a rather different issue from whether the child has one or two linguistic systems. Part of the problem is the familiar one of what we can infer about competence from performance. There now exists a large body of literature rebutting the ‘fused’ system hypothesis, arguing instead that bilinguals have two

40  Overview and Foundations distinct but interdependent systems from the very start (see De Houwer 2009; Genesee 2002; and Meisel 2004 and for summaries of recent studies). In terms of acquisition of linguistic knowledge, a more interesting question than the one-or-two-systems debate has emerged. Specifically, is bilingual acquisition the same as monolingual acquisition? Theoretically, separate development is possible without there being any similarity with monolingual acquisition. Most researchers argue that bilingual children’s language development is by and large the same as that of monolingual children. In very general terms, both bilingual and monolingual children go through an initial babbling stage, followed by the one-word stage, the two-word stage, the multi-word stage, and the multi-clause stage. At the morpho-syntactic level, a number of studies have reported simi­ larities rather than differences between bilingual and monolingual acquisition. Nevertheless, one needs to be careful in the kinds of conclusions one draws from such evidence. Similarities between bilingual and monolingual acquisition do not mean that (i) the two languages a bilingual child is acquiring develop in the same way or at the same speed; (ii) the two languages a bilingual child is acquiring do not influence and interact with each other. Paradis and Genesee (1996), for example, found that while the 2–3-year-old French–English bilingual children they studied displayed patterns that characterize the performance of monolingual children acquiring these languages separately and they acquired these patterns within the same age range as monolingual children, they (i) used finite verb forms earlier in French than in English; (ii) used subject pronouns in French exclusively with finite verbs but subject pronouns in English with both finite and nonfinite verbs, in accordance with the status of subject pronouns in French as clitics (or agreement markers) but full noun phrases in English; and (iii) placed verbal negatives after lexical verbs in French (e.g. ‘n’aime pas’) but before lexical verbs in English (‘do not like’). There is one area in which bilingual children clearly differ from monolingual children, namely, code-mixing. Studies show that bilingual children mix elements from both languages in the same utterance as soon as they can produce two-word utterances (e.g. De Houwer 1990; Deuchar and Quay 2000; Lanza 1997). As with adult code-switching, bilingual children’s language mixing is highly structured. The operation of constraints based on surface features of grammar, such as word order is evident from the two-word/-morpheme stage onward, and the operation of constraints based on abstract notions of grammatical knowledge is most evident in bilingual children once they demonstrate such knowledge overtly (e.g. verb tense and agreement markings), usually around 2;6 (that is, two years and six months) of age and older (see further Meisel 1994; Koppe and Meisel 1995). As Genesee (2002) points out, these findings suggest that in addition to the linguistic competence to formulate correct monolingual strings, bilingual children have the added capacity to coordinate their two languages online in accordance with the grammatical constraints of both languages during mixing. While these studies provide further evidence for the separate development, or two systems, argument, they also suggest that there are both quantitative and qualitative differences between bilingual acquisition and monolingual acquisition.

Conceptual and Methodological Issues  41 Psycholinguistic approaches to bilingualism have offered deeper insights into how multiple languages are simultaneously acquired and represented by the bilingual individual. The typical methods psycholinguists use tend to be laboratory based, using carefully designed experiments or standard assessments. The identifiability and separability of different languages is assumed rather than tested. The interface between linguistic and cognitive capacities of the human being is most often at the heart of the psycholinguistic inquiry.

Linguistic approaches If the psycholinguistic approaches to bilingualism and multilingual focus primarily on the consequence of language contact on the cognitive capacity of the individual, the linguistic approaches typically focus on the consequences of the contact for the structural patterns of languages. Winford (2003: 23–4) provides a taxonomy of contact situations, based partly on Thomason and Kaufman (1988), in which three key processes – language maintenance, shift, and creation – are identified. Each of these processes has distinctive linguistic outcomes. In language maintenance, one could expect, for example, lexical and structural borrowing (e.g. English borrowings from French; Sanskrit influence on Dravidan languages), or convergence in the form of structural diffusion (e.g. Tibetan influence on Wutun and languages of Arnhem Land, Australia). Linguists working on areal classification of languages have a particular interest in the situation. In language shift, little or no substratum interference in target language or moderate to heavy substratum interference may be expected, as in the case of urban immigrant groups shifting to English in the United States or shift to English by Irish speakers in Ireland. In language creation, the linguistic outcomes would be bilingual mixed languages, pidgins, or creoles. Code-switching, which is not dealt with in Winford’s taxonomy, may be an outcome in either language maintenance or language shift. In language maintenance, code-switching often involves inter-speaker and inter-turn alternation, which helps to keep the languages apart. In language shift, on the other hand, code-switching may involve more complex structural integration of the different languages, leading to an eventual base or matrix language turnover (Fuller 2000; Myers-Scotton 2002). Although the diversity of the phenomena seems to invite different approaches, most of the existing linguistic studies of bilingualism and multilingualism focus on the structural constraints of lexical borrowing, diffusion, and code-switching. Opinions differ amongst linguists as to the nature of the structural constraints on borrowing, diffusion, and switching in language contact. There are two contrasting positions. One assumes bilingual language users have two discrete grammatical systems. In a language contact situation, they bring their knowledge of the two different grammars into language production. Usually one of the grammars is more dominant and constrains the ways items from the weaker, lessdominant language incorporate into the strong, dominant language. Based on such assumptions, linguists look for aspects of the grammars of the languages in contact that are either compatible or incompatible. Structural constraints are

42  Overview and Foundations developed based on congruence, equivalence, or markedness (e.g. Muysken 2000; Myers-Scotton 1993; Poplack 1980). Borrowing, diffusion, and switching can all be facilitated by compatible structures, while incompatible structures constrain or block borrowing, diffusion, or switching. While accepting basic assumptions of interacting grammars, MacSwan (e.g. 1999) questions the need to posit specific structural constraints on switching, other than the ones already afforded by universal grammar. This does not mean, of course, that bilingual language users do not differentiate the languages they know and use. The central argument regards the roles of surface grammars of specific languages versus the principles and parameters of universal grammar. MacSwan (2004: 298) goes so far as to say that ‘Nothing constrains code switching apart from the requirements of the mixed grammars.’ Put differently, the generativeuniversalist position is that all of the facts of bilingual code-switching may be explained in terms of principles and requirements of the specific grammars used in each specific utterance (see also MacSwan chapter 13, this volume). MacSwan also questions the status and explanatory power of the notion of ‘matrix language,’ a concept that is widely believed by code-switching researchers to exist and is central to models such as the one proposed by Myers-Scotton (1993). While it is generally accepted that the two languages involved in code-switching tend to play different roles – one, the matrix language, providing the morpho-syntactic frame while the other, the embedded language, providing specific items, usually open-class, content morphemes, the concept of matrix language is not theoretically motivated and probably not needed for explaining the structural patterns or constraints, according to MacSwan. In contrast, some linguists reject the assumption that bilingual language users have two discrete grammars. They argue for a reconceptualization of the bilingual user’s grammatical knowledge, which should include not only the totality of all the surface grammars of specific languages they know and use but also their knowledge of how the different grammatical systems interact with each other. They believe that the attempt to identify structural constraints using abstract notions of grammar, either specific grammars or universal grammar, misses the point, as borrowing, diffusion, and switching are linguistic strategies of the bilingual language user who uses them in different ways in different situations for different purposes (e.g. Gardner-Chloros and Edwards 2003, 2007). A considerable amount of attention in linguistic studies of contact phenomena has been given to so-called contact languages, pidgins, creoles, and other bilingual mixed languages. In terms of methodology, studies of contact languages tend to adopt a much broader perspective than theoretical linguistics. They are typically, and necessarily, situated in historical and social contexts, and are based on detailed, community-based fieldwork. Research on contact languages has theoretical implications for understanding the genesis of human languages generally as well as other linguistic phenomena such as linguistic change, language acquisition, and language attrition. Indeed many of the structural processes of bilingual mixed languages, pidgins, and creoles, such as transfer, simplification, and fossilization,

Conceptual and Methodological Issues  43 find parallels in language change, acquisition, and attrition. (See also Siegel, chapter 21, this volume.)

Sociolinguistic approaches The sociolinguistic approaches to bilingualism and multilingualism differ from the psycholinguistic and linguistic perspectives outlined above both in terms of research methodologies and fundamental concerns. As has been discussed earlier, sociolinguists see bilingualism and multilingualism as a socially constructed phenomenon and the bilingual and multilingual person as a social actor. For the multilingual speaker, language choice is not only an effective means of communication but also an act of identity (Le Page and Tabouret-Keller 1985). Every time we say something in one language when we might just as easily have said it in another, we are reconnecting with people, situations, and power configurations from our history of past interactions and imprinting on that history our attitudes towards the people and languages concerned. Through language choice, we maintain and change ethnic-group boundaries and personal relationships, and construct and define ‘self’ and ‘other’ within a broader political economy and historical context. So, the issues of language use that linguists and psycholinguists are concerned with become issues of social identity and identification for the sociolinguist. The notion of identity has gone through considerable changes in sociolinguistics. In the earlier variationist sociolinguistic work, as exemplified by the work of Labov (1972), identity was taken to mean the speaker’s social economic class, gender, age, or place of origin. It was assumed that speakers express, rather than negotiate, identities through their language use. Several scholars, such as Cameron (1990) and Johnstone (1996), later criticized such assumptions and argued instead that identities are negotiated through social interaction. Linguistic forms and strategies have multiple functions and cannot be directly linked to particular identities outside of interactional contexts. Work by Rampton (1995, 1999) and Lo (1999), for example, demonstrated that identities are locally constructed. More recent work by Pavlenko and Blackledge (e.g. Blackledge and Pavlenko 2001; Pavlenko and Blackledge 2004) emphasizes the negotiation of identities. The idea that identity is negotiable can be traced back to the work of social psychologists who were interested in group processes and intergroup relations (e.g. Tajfel 1974, 1981). Identity, from this particular perspective, is reflective selfimage, constructed, experienced, and communicated by the individual within a group. Negotiation is seen as a transactional process, in which individuals attempt to evoke, assert, define, modify, challenge, and/or support their own and others’ desired self-images (Ting-Toomey 1999: 40). Identity domains such as ethnic, gender, relational, and facework, are seen as crucial for everyday interaction. Speakers feel a sense of identity security in a culturally familiar environment, but insecurity in a culturally unfamiliar environment. Satisfactory identity negotiation

44  Overview and Foundations outcomes would include the feelings of being understood, valued, supported, and respected. There are two major problems with the intergroup social-psychological approach to identity and identity negotiation. First, the categories used in the analysis are often rigid and ill-defined and have a monolingual and unicultural bias. The world is often seen as consisting of ‘them’ and ‘us,’ ‘in-group’ and ‘outgroup,’ or ‘we code’ and ‘they code.’ The so-called negotiation, in this particular perspective, is unidirectional – the native speaker abandoning (or at least modifying) his or her first language and culture in order to learn the language of the host culture. This process is often known as ‘convergence’ or ‘acculturation.’ The second major problem concerns the approach’s static and homogeneous view of culture and society. It does not take into account the historical, ideological, economic processes that led to the present social grouping or stratification (see Brown 2000 for a summary – see also Benwell and Stokoe 2006; De Fina et al. 2006; Gumperz 1982). Adopting a post-structuralist approach to the notion of identity, Pavlenko and Blackledge (e.g. Blackledge and Pavlenko 2001; Pavlenko and Blackledge 2004) argue that the relationship between language and identity is mutually constitutive and that identities are multiple, dynamic, and subject to change. For them, negotiation of identities is the interplay between reflective positioning (after Davies and Harré 1990), that is, self-representation, and interactive positioning, whereby others attempt to reposition particular individuals or groups. Their analyses of multilingualism and identities in a variety of social contexts demonstrate that languages are appropriated to legitimize, challenge, and negotiate particular identities, and to open new identity options. Identity options as constructed, validated, and performed through discourses available to individuals at particular times and places – that is, certain linguistic resources may be available to certain groups of speakers, while others may not be (Tabouret-Keller 1997). Parallel to the work on multilingualism and negotiation of identities, sociolinguists critically examine some of the concepts and notions commonly used by other researchers in the field of bilingualism and multilingualism. For example, the very idea of code-switching raises questions as to what a language is. Instead of thinking of languages as discrete systems, sociolinguists tend to see multilingual speakers as actors in social life who draw on complex sets of communicative resources which are unevenly distributed and unevenly valued. The linguistic systematicity therefore appears to be at least as much a function of historically rooted ideologies (of nation and ethnicity) and of the ordering practices of social life as of language per se (Gal and Irvine 1995). This perspective goes beyond a focus on mental representation of linguistic knowledge and opens up the possibility of looking at bilingualism and multilingualism as a matter of ideology, communicative practice, and social process. This particular sociolinguistic perspective has important implications for the way researchers collect, analyze, and interpret data. Informed by developments in anthropology, sociology, and cultural studies, sociolinguists have examined communicative practices within and across sites that can be ethnographically

Conceptual and Methodological Issues  45 demonstrated to be linked. Working with the ideas of trajectories (of speakers, linguistic resources, discourses, institutions) across time and space and of discursive spaces which allow for, and also constrain, the production and circulation of discourses, Heller (e.g. 1995, 1999) has examined multilingual practices in a number of communities and argued that they contribute to the construction of social boundaries and of the resources those boundaries regulate. They therefore also raise the question of the social and historical conditions which allow for the development of particular regimes of language, for their reproduction, their contestation, and eventually, their modification or transformation (see also studies in Heller 2007). A further, closely related, area in which sociolinguists have extended the work by linguists and psycholinguists on bilingualism is that of the acquisition of linguistic knowledge. Building on earlier research on language socialization, which focused on young children acquiring their first language in culturally specific ways, scholars such as Crago (Crago et al. 1993), Kulick (1992), Zentella (1997), and Schecter and Bayley (2002) examine bilingual and multilingual children’s developing competence in various speech and literacy events. Particular attention is given to the range of linguistic resources available (or not) in bilingual and multilingual communities and the ways in which children, as well as adolescents and adults, learn to choose among these resources for their symbolic value. The researchers emphasized language socialization as an interactive process, in which those being socialized also act as agents rather than as mere passive initiates. This line of inquiry also demonstrates how domains of knowledge are constructed through language and cultural practices, and how the individual’s positioning affects the process of knowledge acquisition and construction (see further Bayley and Schecter 2003).

Towards a Transdisciplinarity: Challenges and Future Directions There is no doubt that a much more nuanced picture of the human language faculty, and indeed of the human mind, has emerged as a result of extensive research on bilingualism and multilingualism over many decades. We understand more about the human capacity for language through such research than the monolingual perspective can ever offer. Theories of human language and mind have become informed in new and essential ways by research on bilingualism and multilingualism. New research questions have been asked, hypotheses formulated, and paradigms constructed. The multidisciplinary nature of bilingualism and multilingualism research as evidenced in the above discussion of the various research perspectives has been a clear strength of the field. Nevertheless, the future of the field requires a more comprehensive framework that transcends the narrow scope of disciplinary research. So, what are the main challenges to bilingualism and multilingualism research as it moves forward to a transdisciplinary future?

46  Overview and Foundations First, there is the issue of terminology. Each discipline develops its own jargon. Communication across disciplines may prove to be difficult since it requires the use of technical terms that are not well understood by colleagues in the other relevant disciplines. Even when the same terms are used, the intended meanings and connotations may be misinterpreted due to lack of common background. For example, the term ‘resources’ has been used by psychologists meaning cognitive capabilities as well as by sociologists meaning sociocultural capital or material goods. So when different researchers talk about a resource as a factor contributing to the success or failure of becoming bilingual or multilingual, they could potentially mean very different things. Yet, there are potential connections between these different things too. An individual’s cognitive capabilities could be enhanced by environmental, affective, and even political factors, as research on learner attitude and motivation has shown. Similarly, how is a ‘bilingual’ person defined? Researchers from different disciplinary backgrounds may come up with very different answers. Some may insist on having no monolingual experience at all; others are happy to include adult second or foreign language learners as bilinguals. Still others may argue that language proficiency and dominance are determining factors. A common understanding across the disciplines requires, as a first step, explicit information about the language configuration and status (e.g. typologically different/similar languages, heritage language, minority language), learning history (first/second/foreign), usage (professional/habitual), modality (sign, scriptal/reading) of the language users and the community that they come from. Second, research methods. Disciplines are often devoted to their own methods of investigation. This may lead to misunderstandings and opposition to research findings. It is important to remember that research methods are chosen for a purpose and have to be appropriate for the research questions. Yet, the research questions are not at all value-free. They are often posed with particular disciplinary, even ideological, biases. Even apparently neutral, scientific terms, such as ‘accommodation,’ ‘variation,’ can be used to serve particular biases and require in-situ explanations. Certain research questions favor certain research methods. Consequently, evidence from studies that employ different methods may be brushed aside as irrelevant. Third, there is a confusion of ‘multi-/interdisciplinarity’ and ‘innovation.’ A comprehensive understanding of any complex psychological and social phenomenon such as bilingualism and multilingualism requires contributions from a variety of disciplines. The multidisciplinary and interdisciplinary approach has definitely generated research outcomes that challenge the received wisdom about the human mind and society. But being multidisciplinary or interdisciplinary does not in itself mean being innovative. Innovation requires thinking creatively, breaking new ground, adding value, and making a difference. Innovation often results in the constitution of a new approach or discipline. Fourth, the tension between ‘basic’ research and ‘applied’ research. At a time when research funding and resources are limited, applied research that has more direct and immediate impact on policy and practice receives more attention and

Conceptual and Methodological Issues  47 support than studies that address basic research questions. There is also a tendency to mistakenly regard applied research as naturally interdisciplinary and basic research as narrowly disciplinary. Yet, without advances in basic research, there would be no firm basis for knowledge transfer which is the key to applied research. Basic research can address social concerns. In fact, it can be argued that the majority of the research questions in the bilingualism and multilingualism field come from the concerns of individuals and their communities. They can be, and have been, turned into basic research questions. The increased amount of bilingualism and multilingualism at both the individual and societal levels offers the research community new opportunities to evaluate its knowledge base and develop its theories and models of language and communication. Society’s interest in bilingualism and multilingualism is also growing. A challenge to the research community is to make what may be viewed as scientific research socially relevant as well. Academic researchers working in the field of bilingualism and multilingualism feel rightly proud of the fact that they not only have a lot to say about the relevant linguistic and psychological theories and models, but also make significant contributions to the socio-political debates about the world we are living in today. Researchers should look forward to moving away from a narrow focus on one’s individual discipline, to learn from each other’s perspectives, and create new ideas. It is the responsibility of researchers to lay the irrational fears of bilingualism and multilingualism to rest through good science. It is equally important that bilingual and multilingual researchers address socio-political issues head-on.

REFERENCES Abutalebi, Jubin, Cappa, Stefano F., and Perani, Daniela (2005). What can functional neuroimaging tell us about the bilingual brain? In J. F. Kroll and A. M. B. de Groot (eds.), Handbook of Bilingualism: Psycholinguistic Approaches. 497–515. Oxford: Oxford University Press. Auer, Peter and Li Wei (eds.) (2007). Multilingualism and Multilingual Communication. Handbook of Applied Linguistics Vol. 5. Berlin: Mouton de Gruyter. Baker, Mark (2001). The Atoms of Language. New York: Basic Books. Bayley, Robert and Schechter, Sandra R. (2003). Language Socialization in Bilingual and Multilingual Societies. Clevedon: Multilingual Matters. Benwell, Bethan and Stokoe, Elizabeth (2006). Discourse and Identity. Edinburgh: Edinburgh University Press.

Blackledge, Adrian and Pavlenko, Aneta (2001). Negotiation of identities in multilingual contexts. Special Issue of International Journal of Bilingualism 5(3). Brown, Rupert (2000). Social identity theory: Past achievements, current problems and future challenges. European Journal of Social Psychology 30(6): 745–78. Burman, Douglas D., Bitan, Tali, and Booth, James R. (2008). Sex differences in neural processing of language among children. Neuropsychologia 46(5): 1349–62. Cameron, Deborah (1990). Demythologizing sociolinguistics: Why language does not reflect society. In J. Joseph and T. Taylor (eds.), Ideologies of Language. 79–93. London: Routledge. Cattell, James McKeen (1887). The time taken up by cerebral operations. Mind 11: 524–38.

48  Overview and Foundations Chomsky, Noam (1986). Knowledge of Language: Its Nature, Origin and Use. New York: Praeger. Clahsen, Harold (1999). Lexical entries and rules of language: A multidisciplinary study of German inflection. Behavioral and Brain Sciences 22: 991–1060. Crago, Martha B., Annahatak, Betsy, and Ningiuruvik, Lizzie (1993). Changing patterns of language socialisation in Inuit homes. Anthropology and Education Quarterly 24: 205–23. Davies, Bronwyn and Harré, Rom (1990). Positioning: The discursive production of selves. Journal for the Theory of Social Behaviour 20(1): 43–63. De Bot, Kees (1992). A bilingual production model: Levelt’s ‘speaking’ model adapted. Applied Linguistics 13: 1–24. De Fina, Anna, Schiffrin, Deborah, and Bamberg, Michael (eds.) (2006). Discourse and Identity. Cambridge: Cambridge University Press. De Houwer, Annick (1990). The Acquisition of Two Languages from Birth. Cambridge: Cambridge University Press. De Houwer, Annick (2009). Bilingual First Language Acquisition. Bristol: Multilingual Matters. Deuchar, Margaret and Quay, Suzanne (2000). Bilingual Acquisition. Oxford: Oxford University Press. Dufour, Robert (1997). Sign language and bilingualism: Modality implications for bilingual language representation. In A. M. B. de Groot and J. F. Kroll (eds.), Tutorials in Bilingualism, Psycholinguistic Perspectives. 301–30. Mahwah, NJ: Erlbaum. Fabbro, Franco (1999). Neurolinguistics of Bilingualism. Hove: Psychology Press. Ferguson, Charles, (1959). Diglossia. Word 15: 325–40. Fuller, Janet M. (2000). Morpheme types in a matrix language turnover: The introduction of system morphemes from English into Pennsylvania German. International Journal of Bilingualism 4(1): 45–58. Gal, Susan and Irvine, Judith (1995). The boundaries of language and disciplines: How ideologies construct difference. Social Research 62: 967–1001. Galloway, Linda M. and Scarcella, Robin (1982). Cerebral organization in adult second language acquisition: Is the right hemisphere more involved? Brain and Language 16: 56–60.

Gardner-Chloros, Penelope (2009). CodeSwitching. Cambridge: Cambridge University Press. Gardner-Chloros, Penelope and Edwards, Malcolm (2003). When the blueprint is a red herring: Assumptions behind grammatical approaches to code-switching. Transactions of the Philological Society 102: 103–29. Gardner-Chloros, Penelope and Edwards, Malcolm (2007). Compound verbs in code-switching: Bilinguals making do? International Journal of Bilingualism 11: 73–91. Genesee, Fred (1989). Early bilingual language development: One language or two? Journal of Child Language 16: 161–79. Genesee, Fred (2002). Rethinking bilingual acquisition. In J.-M. Dewaele, A. Housen, and Li Wei (eds.), Bilingualism: Beyond Basic Principles. 204–28. Clevedon: Multilingual Matters. Green, David (1986). Control, activation, and resource: A framework and a model for the control of speech in bilinguals. Brain and Language 27: 210–23. Green, David (1998). Mental control of the bilingual lexico-semantic system. Bilingualism: Language and Cognition 1: 67–81. Grosjean, François (2001). The bilingual’s language modes. In Janet Nicol (ed.), One Mind Two Languages. 1–22. Oxford: Blackwell. Gumperz, John J. (ed.) (1982). Language and Social Identity. Cambridge: Cambridge University Press. Hamers, Josiane F. (2004). A sociocognitive model of bilingual development. Journal of Language and Social Psychology 23: 70–98. Hamers, Josiane F. and Blanc, Michel (2000). Bilinguality and Bilingualism, 2nd edn. Cambridge: Cambridge University Press. Heller, Monica (1995). Language choice, social institutions, and symbolic domination. Language in Society 24: 373–405. Heller, Monica (1999). Linguistic Minorities and Modernity. A Sociolinguistic Ethnography. Harlow: Addison-Wesley Longman. Heller, Monica (ed.) (2007). Bilingualism: A Social Approach. Basingtoke: Palgrave. Holmes, Janet and Meyerhoff, Miriam (2004). The Handbook of Language and Gender. Oxford: Blackwell. Hudson, Alan (1992). Diglossia: A bibliographic review. Language in Society 21: 611–74.

Conceptual and Methodological Issues  49 Hummel, Kirsten M. (1986). Memory for bilingual prose. In Jyotsna Vaid (ed.), Language Processing in Bilinguals: Psycholinguistic and Neuropsychological Perspectives. 47–64. Hillsdale, NJ: Erlbaum. Jackendoff, Ray. (2002). Foundations of Language. Oxford: Oxford University Press. Johnstone, Barbara (1996). The Linguistic Individual: Self-expression in Language and Linguistics. New York: Oxford University Press. Kezwer, Paula (1987). The extroverted vs. the introverted personality and second language learning. TESL Canada Journal 5: 45–58. Kissau, Scott (2006). Gender differences in motivation to learn French. Canadian Modern Language Review 62: 401–22. Kolers, Paul and Gonzalez, Esther (1980). Memory for words, synonyms, and translation. Journal of Experimental Psychology, Human Learning and Memory 6: 53–65. Koppe, Regina and Meisel, Jurgen (1995). Code-switching in bilingual first language acquisition. In Lesley Milroy and Pieter Muysken (eds.), One Speaker Two Languages. 276–301. Cambridge: Cambridge University Press. Kroll, Judith F. and Tokowicz, N. (2005). Models of bilingual representation and processing. In J. F. Kroll and A. M. B. de Groot (eds.), Handbook of Bilingualism: Psycholinguistic Approaches. 531–53. New York: Oxford University Press. Kroll, Judith and Stewart, E. (1984). Category interference in translation and picture naming. Journal of Memory and Language 33: 149–74. Kulick, Don (1992). Language Shift and Cultural Reproduction: Socialization, Self, and Syncretism in a Papua New Guinean Village. Cambridge: Cambridge University Press. Labov, William (1972). Sociolinguistic Patterns. Philadelphia: Pennsylvania University Press. Lambert, Wallace E. (1969). Psychological studies of interdependencies of the bilingual’s two languages. In J. Puhvel (ed.), Substance and Structure of Language. 99–126. Berkeley, CA: University of California Press. Lanza, Elizabeth (1997). Language Mixing in Infant Bilingualism. Oxford: Oxford University Press. Le Page, Robert and Tabouret-Keller, Andree (1985). Acts of Identity: Creole-Based Approaches to Language and Ethnicity. Cambridge: Cambridge University Press.

Lenneberg, Eric H. (1967). Biological Foundations of Language. Chichester: Wiley. Levelt, Willem (1989). Speaking: From Intention to Articulation. Cambridge, MA: MIT Press. Li Wei (2007). Dimenions of bilingualism. In Li Wei (ed.), The Bilingualism Reader, 2nd edn. 3–22. London: Routledge. Li Wei (2010a). Contact. In Jürgen Jaspers, Jan-Ola Östman, and Jef Verschueren (eds.), Society and Language Use. 127–39. Amsterdam: Benjamins. Li Wei (2010b). BAMFLA: Issues, methods and directions. International Journal of Bilingualism 14(1): 3–9. Li Wei and Moyer, Melissa (eds.) (2008). The Blackwell Handbook of Research Methods on Bilingualism and Multilingualism. Oxford: Blackwell. Lo, Adrienne (1999). Codeswitching, speech community membership, and the construction of ethnic identity. Journal of Sociolinguistics 3: 461–79. MacIntyre, Peter, Baker, Susan, Clément, Richard, and Donovan, Leslie (2002). Sex and age effects on willingness to communicate, anxiety, perceived competence, and L2 motivation among junior high school French immersion students. Language Learning 53: 137–65. MacIntyre, Peter D. and Charos, Catherine (1996). Personality, attitudes, and affect as predictors of second language communication. Journal of Language and Social Psychology 15: 3–26. MacNamara, John (1967). The linguistic independence of bilinguals. Journal of Verbal Learning and Verbal Behavior 6: 729–36. MacNamara, John and Kushnir, Seymour (1971). Linguistic independence of bilinguals: The input switch. Journal of Verbal Learning and Verbal Behavior 10: 480–7. MacSwan, Jeff (1999). A Minimalist Approach to Intrasentential Code-Switching. New York: Garland. MacSwan, Jeff (2004). Code-switching and grammatical theory. In T.K. Bhatia and W.C. Ritchie (eds.), The Handbook of Bilingualism. 283–311. Oxford: Blackwell. McLaughlin, Barry (1984). Second Language Acquisition in Childhood. Hillsdale, NJ: Erlbaum. McWhorter, John (2001). The Power of Babel. New York: Times Books.

50  Overview and Foundations Meisel, Jürgen (1989). Early differentiation of languages in bilingual children. In Kenneth Hyltenstam and Loraine K. Obler (eds.), Bilingualism Across the Lifespan: Aspects of Acquisition, Maturity and Loss. 13–40. Cambridge: Cambridge University Press. Meisel, Jürgen (ed.) (1994). Bilingual First Language Acquisition. Amsterdam: Benjamins. Meisel, Jürgen (2004). The bilingual child. In Tej Bhatia and William Ritchie (eds.), The Handbook of Bilingualism. 91–113. Oxford: Blackwell. Muysken, Pieter (2000). Bilingual Speech. Cambridge: Cambridge University Press. Myers-Scotton, Carol (1993). Duelling Languages: Grammatical Structure in Codeswitching. Oxford: Oxford University Press. Myers-Scotton, Carol (2002). Contact linguistics: Bilingual Encounters and Grammatical Outcomes. Oxford: Oxford University Press. Myers-Scotton, Carol (2005). Supporting a differential access hypothesis: Code-switching and other contact data. In J. F. Kroll and A. M. B. de Groot (eds.), Handbook of Bilingualism: Psycholinguistic Approaches. 326–48. Oxford: Oxford University Press. Obler, Loraine K. and Gjerlow, Kris (1998). Language and the Brain. Cambridge: Cambridge University Press. Paradis, Johanne and Genesee, Fred (1996). Syntactic acquisition in bilingual children: Autonomous or interdependent? Studies in Second Language Acquisition 18: 1–25. Paivio, Allan and Desrochers, Alain (1980). A dual-coding approach to bilingual memory. Canadian Journal of Psychology 34: 388–99. Pavlenko, Aneta and Blackledge, Adrian (eds.) (2004), Negotiation of Identities in Multilingual Settings. Clevedon: Multilingual Matters, Penfield, Wilder and Roberts, Lamar (1959). Speech and Brain Mechanisms. Princeton: Princeton University Press. Pinker, S. (1999). Words and Rules: The Ingredients of Language. New York: Basic Books. Poplack, Shana (1980). Sometimes I’ll start a sentence in Spanish y termino en español: Toward a typology of code-switching. Linguistics 18: 581–618. Reprinted with postscript in Li Wei (ed.), The Bilingualism Reader, 2nd edn. 213–43. London: Routledge. Potter, Mary C., So, Kwok-Fai, Von Eckardt, Barbara, and Feldman, Laurie B. (1984).

Lexical and conceptual representation in beginning and proficient bilinguals. Journal of Verbal Learning and Verbal Behavior 23(1): 23–38. Rampton, Ben (1995). Crossing: Language and Ethnicity among Adolescents. London: Longman. Rampton, Ben (1999). Styling the other. Thematic issue of Journal of Sociolinguistics 3(4). Ramscar, Michael and Gitcho, Nicole (2007). Developmental change and the nature of learning in childhood. Trends in Cognitive Science 11(7): 274–9. Schecter, Sandra R. and Bayley, Robert (2002). Language as Cultural Practice. Mahwah, NJ: Erlbaum. Schmid, Monika S., Köpke, Barbara, Keijzer, Merel and Weilemar, Lina (2004). First Language Attrition: Interdisciplinary Perspectives on Methodological Issues. Amsterdam: John Benjamins. Seliger, Herbert W. (1982). On the possible role of the right hemisphere in second language acquisition. TESOL Quarterly 16: 307–14. Singleton, David and Lengyel, Zsolt (1995). The Age Factor in Second Language Acquisition. Clevedon: Multilingual Matters. Tabouret-Keller, Andree (1997). Language and identity. In F. Coulmas (ed.), The Handbook of Sociolinguistics. 315–26. Oxford: Blackwell. Tajfel, Henri (1974). Social identity and intergroup behaviour. Social Science Information 13: 65–93. Tajfel, Henri (1981). Human Groups and Social Categories. Cambridge: Cambridge University Press. Thomason, Sarah and Kaufman, Terrance (1988). Language Contact, Creolization and Genetic Linguistics. Berkeley, CA: University of California Press. Thompson-Schill, Sharon, Ramscar, Michael, and Chrysikou, Evangela (2009). Cognition without control: When a little frontal lobe goes a long way. Current Directions in Psychological Science 8(5): 259–63. Ting-Toomey, Stella (1999). Communicating Across Cultures. New York: Guilford Press. Volterra, Virgnia and Taeschner, Traute (1978). The acquisition and development of language by bilingual children. Journal of Child Language 5: 311–26.

Conceptual and Methodological Issues  51 Weinreich, Uriel (1953). Languages in Contact: Findings and Problems. New York: The Linguistic Circle of New York. Whitney, William Dwight (1881). On mixing in language. Transactions of the American Philological Association 12: 5–26.

Winford, Donald (2003). An Introduction to Contact Linguistics. Oxford: Blackwell. Zentella, Ana Celia (1997). Growing up Bilingual: Puerto Rican Children in New York. Oxford: Blackwell.

Part II  Neurological and Psychological Aspects of Bilingualism and Multilingualism

Introduction WILLIAM C. RITCHIE

The chapters in Part II are dedicated to research on the plurilingual individual – ranging from what the study of plurilingual aphasia can tell us about the representation of the plurilingual’s languages in the brain to the role of the plurilingual’s social identity in the choice of language in a given set of circumstances. Part II is divided into five sections; each devoted to an aspect of the plurilingual’s knowledge and use of his/her languages. The first section is concerned with the neurology of plurilingualism, the second with the process of second- or third-language acquisition, and the third with the interaction of the plurilingual’s languages. The fourth and fifth sections treat, respectively, memory and general cognition in relation to plurilingualism on the one hand and the plurilingual’s capacity and motivation for switching from one of his/her languages to another on the other. The single chapter in the first section (The Neurology of Bilingualism and Multilingualism) is Chapter 3, ‘Bilingual Aphasia: Theoretical and Clinical Considerations,’ by Elizabeth Ijalba, Loraine Obler, and Shyamala Chengappa. After a historical overview and a discussion of methodological issues, the authors evaluate three models (or kinds of models) of bilingual brain functions – the declarative/procedural-memory model, the inhibitory control model, and hierarchical (translation) models – with respect to the available evidence from the research literature on aphasia. The first model addresses the issue of which of the bilingual’s two languages is more impaired and in what ways depending on the location of the lesion. The second model concerns the bilingual aphasic’s ability (or inability) to switch languages under appropriate circumstances. The significance of the third kind of model is that it allows the testing of the patient’s proficiency in each language, thereby helping to determine what strategies to adopt in therapy with bilingual and polyglot patients. The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

56  Neurological and Psychological Aspects of Bilingualism and Multilingualism The second section in Part II, Approaches to Bilingualism, Multilingualism, and Second-Language Acquisition, consists of four chapters. In ‘The Bilingual Child’ (Chapter 4), Ludovica Serratrice reviews the literature on the development of the bilingual child with respect to speech perception and production, word learning, morphology, syntax, language differentiation, and cross-linguistic influence. Yuko Butler (Chapter 5, ‘Bilingualism/Multilingualism and Second-Language Acquisition’) begins by tackling the difficult problem of characterizing bilinguals and multilinguals. She then explores in some detail the complexities involved in coming to an understanding of bilingualism and multilingualism and, in particular, in arriving at a satisfactory conceptualization and assessment of a notion that is central to the study of second-language acquisition – language proficiency. The chapter then addresses the issue of age and second- (or third-) language acquisition and the interaction between the learner’s first language on the one hand and subsequently learned languages on the other. Chapter 6, ‘Multilingualism: New Perspectives on Syntactic Development’ by Éva Berkes and Suzanne Flynn, reviews the growing body of research on the acquisition of a third language. The authors differentiate among three positions on the nature of third-language acquisition – the L2 status factor model, the typological primacy model, and the cumulative enhancement model (CEM) – and argue for the CEM on the basis of existing published work and a new study of theirs, which they report in the chapter. In Chapter 7, ‘Bilingualism and the Heritage Language Learner,’ Silvina Montrul reviews the research literature on the highly complex instances of second-language acquisition and bilingualism that result from cases in which the learner, usually a bilingual member of a minority culture, has been raised in a home where the language of that culture (the heritage language) is spoken and who has learned the language of the majority culture outside the home. Understanding of these cases requires the making of a number of crucial distinctions: first, between majority and minority language (in the surrounding society); second, between first and second language; and, third, between primary and secondary language (these latter distinctions in the individual). Having drawn these distinctions, Montrul describes the systems of linguistic competence that result from these learning conditions and traces the change of the heritage language from the learner/user’s first and primary language in the early stages of development to his/her secondary language in later stages. She then addresses the issue of whether a heritage language is like a second language or not. The third section of Part II (Bilingual and Multilingual Language Use: Knowledge, Comprehension, and Production) treats the interaction between the grammars of the two languages as well as the deployment of the grammar in speech comprehension and production in bilinguals. Pieter Muysken (Chapter 8, ‘Two Linguistic Systems in Contact: Grammar, Phonology, and Lexicon’) proposes that the structural influence between the two languages on the one hand and the form of language alternation referred to as code-switching or code-mixing on the other differ only in that the latter involves the lexicon of the two languages and the former does not. He provides a thorough description of the conditions under which code-mixing occurs in terms of a system of hierarchies of such conditions.

Introduction  57 Judith Kroll and Paola Dussias (in Chapter 9, ‘The Comprehension of Words and Sentences in Two Languages’) review the research on bilingual comprehension and argue on the basis of their review that, in the process of comprehension, the bilingual’s two languages are remarkably open to each other – that the bilingual’s language user’s second language influences performance in the first language and conversely. In Chapter 10 (‘An Appraisal of the Bilingual Language Production System: Quantitatively or Qualitatively Different from Monolinguals’), Elin Runnqvist, Ian Fitzpatrick, Kristof Strijkers, and Albert Costa examine the research on the interaction of the bilingual’s two languages in the production of speech. They conclude that bilingual production differs quantitatively but that the question of qualitative differences is not clear from the research. Bilingualism and Multilingualism: Memory, Cognition, and Emotion, the fourth section of Part II, consists of two chapters. Chapter 11, ‘Bilingual Memory’ by Roberto Heredia and Jeffrey Brown, reviews past attempts to model the memory of bilinguals including the distinction between compound and coordinate bilinguals, which predicts shared and separate memory systems, respectively, as well as more recent models such as the revised hierarchy model and the distributed model. Heredia and Brown also touch on the character of autobiographical memory in bilinguals, a topic taken up in more detail in Chapter 12 (‘Bilingualism and Emotion: Implications for Mental Health’) by Ines Martinovic and Jeanette Altarriba. Martinovic and Altarriba examine issues of bilingualism and biculturalism in psychotherapy – how the bilingual’s memories of past experiences may be more accessible in one language than the other and the effect this may have on the success or failure of psychotherapy for bilinguals. The final section in Part II (The Bilingual’s and Multilingual’s Repertoire: CodeMixing, Code-Switching, and Speech Accommodation) is concerned with one of the more extensively researched areas in the study of bilinguals. It includes four chapters, the first of which (Jeff MacSwan’s Chapter 13, ‘Code-Switching and Grammatical Theory’) reviews the previous attempts to characterize restrictions on intrasentential language alternation (code-switching) in terms of independently motivated theories of grammatical structure and sentence production. He then provides explanations within the minimalist program for several constraints on switching that have been proposed in the descriptive research literature. Gerald Berent (Chapter 14, ‘Sign-Language/Spoken-Language Bilingualism and the Derivation of Bimodally Mixed Sentences’) examines the phenomenon of code-mixing in the special case of cross-modal bilingualism and argues for the existence of a condition on such mixing that has been proposed independently for code-mixing involving spoken languages. Three chapters round out this section on code-mixing and -switching by looking at the social and psychological circumstances under which switching/mixing occurs. Chapter 15 by William Ritchie and Tej Bhatia (‘Social and Psychological Factors in Language-Mixing’) addresses the issues of whether or not languagemixing is random, what motivates bilinguals to mix (given that mixing is not random), what the social evaluation of mixing is, and, finally, how bilinguals themselves view mixing. Itesh Sachdev, Howard Giles, and Anne Pauwels (Chapter

58  Neurological and Psychological Aspects of Bilingualism and Multilingualism 16, ‘Accommodating Multilinguality’) offer an account of language-mixing and -switching within the communication accommodation theory, an account of language choice that has been developed over a number of years within social identity theory. Finally, Chapter 17, ‘Bilingualism and Gesture,’ Marianne Gullberg reviews the growing body of research on the relationship between linguistic performance and manual gesture – itself a dynamic area of research – specifically in the case of bilinguals.

The Neurology of Bilingualism and Multilingualism

3  Bilingual Aphasia: Theoretical and Clinical Considerations ELIZABETH IJALBA, LORAINE K. OBLER, AND SHYAMALA CHENGAPPA

Introduction The questions surrounding bilingual aphasia are numerous – from theoretical frameworks of language organization and cognitive processing to clinical applications and even educational planning and interventions. A number of bilingual or polyglot speakers will experience aphasia at some point in their lives – yet much confusion exists regarding their language treatment. Likewise, many bilingual children with language impairment are caught in the uncertainty of what language(s) should be the focus of interventions. The study of aphasia – the set of language disorders that arise after a brain lesion – provides us with a window to understand brain organization and cognitive demands for language functions. The study of aphasia in bilinguals or polyglots seeks to answer questions regarding recovery patterns, i.e., which language (or languages) is recovered first following a lesion. Most crucially, the study of aphasia seeks to answer questions about bilingualism per se such as: Does our brain provide separate stores for each language we know? To what extent do languages share conceptual areas? How do linguistic constraints shape language organization and language processing? How are we able to switch between languages and use them appropriately and without interference? Are aphasic deficits due to a disruption in language representation or to a disruption in control mechanisms? In this chapter we consider the discussion of language breakdown in polyglots, which has its roots in nineteenth-century Europe and spans across the globe in the twenty-first century. We discuss the questions that have been addressed and

The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

62  Neurological and Psychological Aspects of Bilingualism and Multilingualism explore two models that focus on different but complementary aspects: one model emphasizes different memory systems that influence language processing in bilinguals; the other model focuses on executive control of language functions. We briefly review hierarchical models and also consider recent investigations focused on cross-linguistic clinical findings, such as transfer and interference among languages. This structure permits us to discuss the literature on clinical cases of bilinguals with aphasia focusing on the phenomenology of disorders seen in them, including language-switching, translation disorders, and order of language return during recovery from aphasia.

Issues in the Case Literature on Bilingual or Polyglot Aphasias The late nineteenth century marked a period of intense debate focused on answering questions of language representation in the brain. Initial findings were the result of clinical observation and pathological correlates discovered in the brain during autopsy. Paul Broca (1861/1977) is credited with labeling the inferior part of the third frontal convolution in the left hemisphere as the ‘articulation center’ for spoken words (see Figure 3.1). He based his conclusions on the fact that one patient, Leborgne, was unable to speak, except to say the single syllable ‘tan,’ and that an autopsy of Leborgne and then of six additional similar patients later revealed this part of the brain to be damaged. A few years later, Carl Wernicke (1874) complemented Broca’s findings by revealing the posterior part of the first temporal gyrus as the area responsible for the ‘sensory image of words’ or comprehension for language (see Figure 3.1). Wernicke also created the first model drawn right onto an outline of the brain, delineating a ‘motor speech area’ and a ‘sensory speech area,’ both connected by a bundle of fibers transmitting information between the two centers. This simple model explained many of the known

Arcuate Fasciculus

Broca’s area

Wernicke’s area

Figure 3.1  Broca’s area involved in speech production and Wernicke’s area involved in the comprehension of speech. The two areas are connected by the arcuate fasciculus

Bilingual Aphasia: Theoretical and Clinical Considerations  63 aphasic symptoms of the time, including speaking difficulties concomitant with preserved ability to understand spoken language, the preserved ability to speak concomitant with difficulties in understanding language, and the preserved ability to both speak and understand, yet failure to repeat back a word or phrase. Because these initial observations on brain function took place in Europe, where most people spoke more than one language or dialect, the debates on brain organization of language were soon enriched with the question of how multiple languages are represented in the brain. This question arose naturally as a result of counter-intuitive symptoms demonstrated during recovery from aphasia by some bilingual and polyglot speakers. Neurologists began to report unusual recovery patterns in patients who had spoken two or more languages before their aphasia but showed differential recovery patterns for each language after an aphasiaproducing incident. The earliest known theory from this period was that each language has separate brain representation or ‘language centers’ and that recovery is determined in accordance with which language centers were disrupted, explaining why some languages of the polyglot were regained and others were not, due to damage of their specific representation in the brain (Scoresby-Jackson 1867). In contrast, Albert Pitres, a neurologist who was Dean of the Faculté de Medecine of Bordeaux, was the first to propose (1895, translated to English in 1953) a view that persists to this time. He reviewed the clinical studies of bilinguals with aphasia published by his contemporaries and rejected the notion of separate language representation in the brain for each language by pointing out that each language would need at least four separate centers in the brain: two for reception of auditory and visual images; and two motor ones for producing phonetic and written images, making it very unlikely that lesions would be distributed in such coordinated fashion. Pitres’ strategy, like that of other aphasiologists at the end of the nineteenth century, was to replace the neuroanatomical perspective with an analysis of the functional quality of the disorder. He pointed out that during recovery from aphasia, patients initially go through a stage of ‘inertia’ during which they fail to understand or use any known languages and that this is due to disruption but not total destruction of their language centers. Gradual return of understanding the most familiar language follows, continued in turn by regaining the ability to speak the most familiar language. This same pattern is repeated for any or all of the other languages that the patient spoke prior to onset of the lesion; for each, language comprehension returns first followed by production. Pitres reasoned that the more familiar language would have more extensive associations (neural connections), thus ‘functional inertia’ would most likely inhibit languages that were less well established through these neural associations. In other words, Pitres conceptualized that a polyglot’s languages were not organized in distinct brain areas, but rather were part of a neural network built upon the speaker’s experiences with the language(s). The extent of the associations would explain the order of return of the known languages. Mieczyslaw Minkowski (1927/1983), a Swiss neurologist, supported Pitres’ rejection of separate neuroanatomical centers for each language in the bilingual’s brain. He proposed that within a common area, active elements from known

64  Neurological and Psychological Aspects of Bilingualism and Multilingualism languages combine and interact at a linguistic level. Minkowski also pointed out, systems should not be viewed as destroyed in aphasia but as ‘weakened’; therefore, aphasic symptoms are a reflection of the language system functioning at a reduced level of activation. This position had been strongly advocated by pioneers such as John Hughlings-Jackson (1879) and Sigmund Freud (1891, translated to English in 1953) in the study of monolingual aphasia. Indeed, the idea that laterlearned languages are superimposed on the first-learned language and that destruction through lesion results in greater impairment of the later-learned language was a position asserted by Freud (1891: 60–1). The degree to which a later-learned language is used and the age of its acquisition were the two factors considered by Freud as primary in determining which language is preserved in aphasia. If all the languages of a multilingual speaker have a common or shared neural representation, aphasic symptoms should be evident across languages and language recovery should also be similar among languages. However, if neural representation for each language is separate, then each language known by the speaker may show different levels of impairment and the order of return may also vary among languages. We can see that in its inception, the discussion regarding language breakdown in polyglots was demarcated into two camps: a highly localizationist perspective that argued for specific language areas to represent each of a polyglot’s language; and a more unified perspective that argued for shared neural networks among languages. Thanks to modern imaging techniques, we can now better understand that bilingual representation for speech/language production is not only determined by neural networks, but is also influenced by multiple factors, including proficiency, age of acquisition, education, and shared characteristics among the languages (Abutalebi et al. 2009; Chee et al. 2001; Illes et al. 1999; Kim et al. 1997; Klein et al. 2001; Tan et al. 2001; Wang et al. 2009). Pitres’ report opened the gates to a century of research on bilingual aphasia aimed at addressing which language is first to return subsequent to a cerebral lesion. Albert and Obler (1978) pointed out that in most cases of bilingual aphasia, the two (or more) languages recover proportionally to their pre-morbid proficiency. In those instances where differential recovery was evidenced, however, the language most frequently used at the time of the patient’s lesion was the most likely to be recovered first (Obler and Albert 1977). This pattern of recovery follows the rule of Pitres rather than that of Ribot (e.g., Ribot 1881), whose general theory of memory predicted that the first-learned language should return first. Of course, Paradis (1977) pointed out, predicting recovery in the individual polyglot speaker is still not easy. He created a taxonomy of patterns of recovery influencing degree of impairment and language return in polyglots who show differential recovery (each language may show different impairment and recovery at the same or different rate); parallel recovery (languages may show similar impairment and rate of recovery); antagonistic recovery (there is regression in one language but progress in another); successive recovery (one language is first restored followed only then by progress in the other language); selective recovery (one or more of the languages previously known remains selectively impaired);

Bilingual Aphasia: Theoretical and Clinical Considerations  65 and mixed recovery (the speaker combines all languages to communicate). Individual cases have been reported of each of these types of recovery, with the first two being by far the most frequent. In addition to the reports about recovery from aphasia, to which we return in the Models section of this chapter, two other issues have been discussed in a set of papers on bilingual aphasia over the past half-century. One is the question of whether crossed aphasia (a lesion to the right hemisphere causing language deficits) occurs with greater frequency among bilinguals and polyglots, and the second is what can we learn from patterns of generalization (transfer) across languages of aphasia therapy performed in one of the languages of the bilingual. Clinical cases of crossed aphasia raised the question of whether there is increased incidence of crossed aphasia in bilinguals or polyglots as compared to monolinguals. This is a relevant question because it would point to increased righthemisphere participation in language processing among healthy bilingual speakers when compared with monolinguals. Since the first report on four bilingual patients with crossed aphasia (Gloning and Gloning 1965), a small set of studies has argued both for and against disproportional right-hemisphere involvement in bilingualism. Khateb et al. (2004) present the case of a right-handed monolingual patient with a right frontal meningioma who presented with crossed aphasia. Neuropsychological testing and fMRI revealed bilateral representation for language. Kho et al. (2006) report two cases in which language-switching was induced in two bilingual patients, one undergoing a Wada test and the other through intra-operative electrocortical stimulation. In both cases, the researchers found bilateral representation of both the patients’ first languages (L1s) and their second languages (L2s), although the left hemisphere was described as dominant. In spite of reports of this nature, there is still no convincing methodologically sound study supporting increased incidence of crossed aphasia in bilinguals. There is a growing amount of literature focusing on the role of therapy in recovery from aphasia. A key question in rehabilitation is what language to use in speech-language therapy with a polyglot speaker. Practical considerations are the availability of bilingual clinicians, particularly when the L1 (often the home language) is different from the mainstream language (often the L2) and therapy is only available in the L2. The question of interest is whether therapy provided in one language can generalize (i.e., transfer) to another language. In a systematic review of the literature which included 14 studies and 45 participants with aphasia, Faroqi-Shah et al. (2010) reviewed three essential questions involved in clinical decision making: What are the outcomes when therapy is provided in a less proficient language; what is the extent of cross-language transfer; and what are the outcomes when therapy is provided through a language broker? As for receptive language outcomes, Faroqi-Shah et al. found that in five studies that examined cross-language transfer from L2 to L1, three showed significant improvements in the untrained L1. As for expressive language outcomes, of 11 studies that examined cross-language transfer from L2 to L1, five reported significant improvement in the untrained L1. Four of these five studies focused on word retrieval and one study focused on sentence production. These findings indicate that transfer effects

66  Neurological and Psychological Aspects of Bilingualism and Multilingualism do take place across languages, even in patients with chronic aphasia. However, it should be noted that transfer effects to the untrained language only occurred in approximately 50% of the studies reviewed. Even when transfer to the untrained language of a polyglot speaker is reported in the literature, an essential question is what aspects of language transfer. One factor to consider is similarity between two languages. Koda (2004) points out that distance between two languages (e.g., reflected in the orthography, phonology, syntax, and/or lexicon) can predict the degree of transfer between two languages. For example, there is more likelihood of transfer facilitating learning between English and Spanish than between English and Chinese. If that is the case, the use of cognate or borrowed words (words that share meaning and phonology, e.g., telephone and teléfono) would facilitate learning across languages. A cognate-based treatment is reported by Kohnert (2004) in a highly educated Spanish–English bilingual speaker, D.J., who was diagnosed with transcortical motor aphasia (deficits in naming and expressive language but preserved repetition). Treatment included the use of 20 Spanish–English translated word pairs that were high frequency nouns. The word pairs used were cognates and non-cognates (e.g., plate/plato and box/caja). The treatment in this study was remarkably short, only four one-hour intervention sessions over a two-week period. Therapy in the first week was conducted in Spanish and in the second week in English. D.J. showed greater gains in naming for cognate word pairs than for non-cognate word pairs. More importantly, during the second week of treatment, these gains were only maintained for the cognates and were transferred from Spanish to English. Clearly, the shared language features between Spanish–English must have influenced transfer between the languages. A relevant aspect that may influence overall language improvement and transfer among languages is the amount of therapy provided and the emotional attachment toward a specific language. Miertsch, Meisel, and Isel (2009) report on the improvement experienced by B.L., a highly educated multilingual speaker (L1 German, L2 English, L3 French, and highly proficient in Spanish and Portuguese) with Wernicke’s aphasia. B.L. was 56 years old and was treated at eight years postonset using intensive therapy (twice a day for 45 minutes for three and a half weeks) in the L3 (French). It should be noted that during the first two and a half years post-onset of aphasia, B.L. was treated in his L1 (German), which was his everyday language. Because B.L. had already experienced considerable progress and had established compensatory strategies in his L1, therapy was then attempted in a different language. B.L. reported a strong emotional attachment to the French language, having lived in Paris for three years and having professional relationships there. The French language was thus chosen to work on lexical-semantic aspects and became the focus of intensive therapy. Whereas in the prior therapy focused on L1 there was limited or no transfer to L2 and L3, when intensive therapy was conducted in L3, improvement was also noted in L2 but not in L1. These findings are consistent with those of Goral et al. (2006), who reported that more interference was seen between L2 and L3 in a multilingual recovering from agrammatic aphasia than between either of these languages and the patient’s L1.

Bilingual Aphasia: Theoretical and Clinical Considerations  67 In that study, Goral et al. (2006) addressed questions regarding patterns of language use, proficiency post-stroke, and similarity between the languages (lexical and orthographic) and their effect on interference in E.C., a trilingual speaker (L1 Hebrew, L2 English, L3 French) diagnosed with non-fluent aphasia, agrammatism, and good comprehension, seen at four years post-onset. The researchers were interested in measuring the amount of interference (inter-language activation or intrusion of one language upon another) among the patient’s known languages. In order to maintain all the languages active, their participant was engaged in naturalistic conversations for 12 hours (four hours in each of the three languages) and instructed to not suppress any inter-language activation he might experience, but to simply produce the words that came to mind, even if different from the target word. All instances of inter-language activation were tabulated, showing that E.C. displayed greatest interference during conversations in French (23 instances per hour). Analysis of these inter-language activations showed that 93% of the instances were English words and only 7% of inter-language activation came from Hebrew, his L1. Nearly half of the English interference instances were cognate words (e.g., hôpital–hospital). A decreased number of inter-language activation events were present during conversations in English (4.5 events per hour, mostly words in French) and even less interference was present in Hebrew (3.5 events per hour, mostly words in English). It is interesting to note that E.C. experienced interference among cognate words in his L2 and L3, since using cognate words has been described as an effective therapy in bilingual aphasia (Kohnert 2004). Goral et al. interpret the inter-language interference errors as evidence for the activation of the equivalent word in the non-target language. Moreover, the pattern of interference lends evidence for the strength of lexical connections among the three languages. That is, Hebrew and English were better recovered and French was the least recovered language. Whereas all three languages were active, the strength of the activation was stronger in Hebrew and English than in French. The fact that most of the inter-language interference errors in French arose from English and vice versa is easily explained by the large amount of shared vocabulary between the two languages. Additional literature has suggested that generalization is more likely to occur for comprehension than for production (e.g., Watamori and Sasanuma, 1978, for Japanese–English-speaking bilinguals). Faroqi and Chengappa (1997), for example, investigated the trace deletion hypothesis and its implications for intervention in a multilingual agrammatic aphasic patient with knowledge of Kannada, Hindi, English, and Telugu. They found that disruption of traces in D-structure representations was a reasonable explanation for the deficits in patients with agrammatism. Generalization of the trained concepts was facilitated only when the patient was a compound bilingual and the languages were historically related (Kannada and Telugu) and when both the languages were used to an equal extent in the patient’s pre-morbid environment (Hindi and English). Cross-language generalization was better for comprehension than for production. This suggests that the perceptual system of bilinguals is unified whereas production is dual. Parallel recovery was evidenced with respect to Kannada and Telugu and differential

68  Neurological and Psychological Aspects of Bilingualism and Multilingualism recovery for English and Hindi, which was explained based on differences in language structure and on pre-morbid language use and proficiency. Even more specific studies have been performed on groups of bilinguals with aphasia. Bose and Chengappa (2000), for example, investigated naming deficits in three Kannada–English bilinguals with aphasia for three different naming tasks and compared their performance with that of age- and gender-matched controls. Parallel deficits were observed in Kannada and English for aphasics except for results on confrontation naming where differential impairment was observed. As to comprehension disturbances, in a master’s dissertation with Chengappa, Sridevi (1999) studied eight Tamil–English bilinguals with aphasia and eight typical bilinguals where Tamil was their mother tongue and English was their second language. Their performance on the revised Token Test did not reveal any significant difference between languages in normal subjects and patients with aphasia performed poorly on all subtests compared to typical bilinguals. As expected, among different types of aphasics, those with anomia had better comprehension in both languages, followed by those with Broca’s, Wernicke’s, and global aphasia. In sum, the recovery pattern was differential, with no consistent pattern of L1 or L2 sparing for either naming or comprehension. Sunilkumar et al. (2010) studied the action-naming deficits in 10 Kannada– English bilinguals with aphasia and they found a significant difference in reaction times for L1 and L2 in this group of aphasics. Analysis of response-time results revealed no significant difference between L1 and L2, supporting the revised hierarchical model (Kroll and Stewart 1994; see also Heredia and Brown, chapter 11, this volume). By contrast, analysis of errors suggested that the lexical processing of L2 depends on the L1 lexical system, i.e., supporting a word association model like that of de Groot and Kroll (1997). The orthographic differences among languages become highly relevant both for assessment and for intervention in bilingual aphasia. Chengappa, Bhat, and Padakannaya (2004), discuss the differences between the syllabic scripts of the Indian languages and alphabetic ones like English. The manifestation of surface vs. deep dyslexia and characteristics of phonological dyslexia in Kannada are carefully reviewed by Karanth (2002). In the twentieth century, certain psychological notions contributed to our understanding of brain organization for bilingualism. Age of acquisition for each language, for example, has also been found to influence hemispheric representation, with earlier acquired languages being more left lateralized and later acquired languages being more bilaterally represented (Albert and Obler 1978; Isel et al. 2010; Neville et al. 1997; Vaid and Hall 1991; Vaid and Hull 2002). However, age of acquisition, to the extent the data were available, was not statistically linked to the patterns of recovery from aphasia (Albert and Obler 1978). Language models that differentiate how bilinguals acquire a first and a second language provided an early bridge between language function and neural correlates. Weinreich (1953) suggested that manner of acquisition of a second or later language would influence the functional organization of the two languages of the bilinguals. He posited that there are coordinate bilinguals, who learned different

Bilingual Aphasia: Theoretical and Clinical Considerations  69 languages at the same time but in separate environments, and compound bilinguals, who learned different languages within the same context and used them concurrently, as well as subordinate bilinguals who interpret words of the weaker language through words in the stronger language. Weinreich posited separate lexical stores for coordinate bilinguals, as the languages are learned in different contexts. Compound bilinguals were described as having a single lexical store common to both languages, reflecting that the languages are interdependent. Subordinate bilinguals have primary representation in their native language and an additional lexical store for the other language that is dependent on the primary language. Therefore, words in the second language have to be interpreted through words in the first language. Evidence supporting this model of functional representation of bilinguals’ language was found by Lambert and Fillenbaum (1959), who pointed out that coordinate bilingual aphasic patients tended to show damage that was more localized with respect to either language, while the disturbance in compound bilinguals was more generalized across both languages. See Appel and Muysken (1997) and Romaine (1995) for reviews of early literature in bilingualism. Cross-linguistic research has also shown that characteristics of each language may influence the way in which language is represented in the brain. Most of this work has been done on differing orthographies, rather than on structural differences in languages per se (see Obler 1984). Reading and writing of ideographic scripts, one can conclude, may include more occipital areas and right-hemisphere involvement than that of alphabetic scripts more associated with temporal areas in the left hemisphere (Tan et al. 2001; Vaid and Genesee 1980). A body of research on non-aphasic bilinguals, however, has demonstrated the influence of differing linguistic patterns across one’s two languages in influencing the performance in those languages (e.g., Bates et al. 1984; Koda 2004). Some of those findings have been extended to the bilingual aphasia population. Vaid and Chengappa, (1988), for example, demonstrated that bilingual aphasic speakers of Kannada and English showed the same patterns of preferring animacy cues over word-order cues in comprehending ambiguous phrases as did non-aphasic control bilinguals.

Methodological Concerns in the Bilingual Aphasia Case Literature We must note at this point that much of the bilingual aphasia case literature mentioned to this point has been flawed in that methodological differences arise at all levels to further complicate matters. Many studies report the site of lesion whereas other investigations have a different focus and primarily provide clinical profiles based on symptoms. Most reported cases of aphasia have a wide range of ages. For example in the Faroqi-Shah et al. (2010) review of 14 studies, the participants ranged in age from 42 to 80 years. Patterns of attrition in L1 and L2 particularly

70  Neurological and Psychological Aspects of Bilingualism and Multilingualism affect naming and word retrieval in healthy aging adults, deficits that are characteristic of aphasia and thus may be confounded in reports of anomia (see Goral, 2004, for a review of first-language decline in healthy aging bilinguals). Age of acquisition in each language plays an important role when analyzing language breakdown in patients, but it is not always mentioned, nor is it easy to determine (e.g., when one has studied a language ill-taught in school and then immigrates to a country in which it is spoken, which age is to be used? Some consider ‘age of arrival’ the crucial date, but clearly the question is vexed). It is similarly unclear which age or age interval is the crucial one distinguishing early from late acquisition. Lenneberg (1967) would have nominated puberty, Johnson and Newport (1989) age four, and Fabbro (2001) age seven. Moreover, earlier and later acquired languages may be subserved by different processing mechanisms with correspondingly different neural correlates (Neville et al. 1997). Paradis (1994), for example, argued that languages learned at different times result in different types of knowledge: Those learned earlier involve implicit knowledge whereas later acquired languages involve explicit types of knowledge. Additionally, attained language proficiency achieved and the linked factor, language exposure, as opposed to age of acquisition have been identified by many researchers as more determinant of cerebral representation in the languages of adult bilingual and polyglot speakers (e.g., Abutalebi, Cappa, and Perani, 2001; Illes et al. 1999; Perani et al. 1998). Increased language proficiency is associated with the activation of common neural networks within the language areas of bilingual and polyglot speakers. In addition to these variables, the contextual demands and actual use patterns for each language and the speaker’s attitudes toward each language tend to be omitted in studies of bilingual aphasia. Much of the early literature on aphasia relied on clinical case studies of one individual with aphasia or a small group of individuals with aphasia. Current research, by contrast, more frequently looks at group data. Differences in design and methodology distinguish these studies. We can find experimental vs. clinical methods, differences in the type of stimuli and variables measured, variety in diagnostic instrumentation and interpretation, and differences in the time at which patients are evaluated (e.g., at the acute vs. later stage post-lesion and onset of symptoms). Even the selection of participants in studies varies with factors such as age, gender, and handedness that are not necessarily reported. For a comprehensive review of differences in methodology and the problems that this lack of uniformity presents in interpreting and applying research findings, see Vaid and Hull, 2002, and concerning neurolinguistic studies of bilingualism more generally, see Hyltenstam and Obler, 1989. In summary, left-hemisphere cortical specialization for language and localization for language functions within specific regions in frontal and temporal areas were elucidated in the nineteenth century. These findings in turn generated questions regarding the organization of language in polyglot speakers. Questions regarding bilingual language representation have pivoted on the concept of separate brain centers for each language or shared neural networks that may be variably disrupted by lesion. Much of the current research is focused on translat-

Bilingual Aphasia: Theoretical and Clinical Considerations  71 ing theoretical findings into clinical applications that can better serve polyglots. An increase in bilingual aphasia might be expected with the rise of a worldwide multilingual population. Current research methodology carefully attempts to account for variables such as sequence of language acquisition, intrinsic and extrinsic characteristics influencing the speaker (e.g., individual differences and also environmental factors), and even cross-linguistic differences. Moreover, brain regions outside those mentioned during the nineteenth century are now being discussed, as their role in cognitive abilities underlying language behavior seems important for understanding the phenomena of bilingual aphasia. Neural support for the languages in polyglot speakers may be associated with different memory systems, a theory proposed by the declarative/procedural model. Effective use of the known languages by bilingual and polyglot speakers may in turn be determined by control over inhibition or activation of each language, a theory posited by the inhibition control model.

Models of Bilingual Processing that Address Bilingual Aphasia The declarative/procedural memory model The learning of language seems to be mediated by two different memory mechanisms as they interact with age of acquisition (Paradis 1994; Ullman 2001). The procedural memory system is responsible for motor and cognitive functions acquired early in life. This system is mediated by frontal and basal ganglia structures with contributions from inferior parietal regions in the left hemisphere (see Figures 3.2 and 3.3). The learning of a first language – or even of a second language that is learned early in life – relies primarily on this system for grammatical computation and lexical processing. The declarative memory system mediates semantic and episodic memory and is supported by medial and temporo-parietal neural structures in both left and right hemispheres. The learning of a second language later in life involves greater reliance on this system for the learning of grammar, if not lexicon. Therefore, later exposure to a second language implies a shift from procedural to declarative memory systems and increased righthemisphere representation. According to the exponents of this model (Fabbro 2001; Paradis 1994; Ullman 2001) the study of language recovery in aphasic polyglots is closely related to the order of acquisition of the languages (simultaneous vs. sequential) and the memory systems involved in language acquisition. A first language tends to be learned informally and in natural settings, relying more on implicit/procedural memory mechanisms and unconscious cognitive processes rooted in subcortical structures. Languages that are acquired later in life tend to rely more on declarative memory, are often learned in more structured and less natural settings, such as school, and are more exclusively represented in the cerebral cortex than the first lan­ guage. The role of what neuropsychologists call ‘executive processes’ influencing

Parietal Lobe Thalamus and Basal Ganglia

Corpus Callosum

Frontal Lobe

Occipital Lobe

Temporal Lobe

Pons

Figure 3.2  Medial view of the human brain showing basal ganglia in relation to corpus callosum and cortical lobes

Parietal Lobes

Right Pre-Frontal Lobe

Left Pre-Frontal Lobe

Thalamus Basal Ganglia Occipital Lobe

Cerebellum

Temporal Lobe

Figure 3.3  Approximate structural relationship of the thalamus and basal ganglia to cortical areas within the human brain

Bilingual Aphasia: Theoretical and Clinical Considerations  73 inhibition may also explain why during the recovery process of some aphasic polyglots, fluctuation in the reappearance of languages is seen. Many studies lend support to the declarative/procedural model and link the differential language recovery in aphasics to subcortical lesions involved in automatic language processes (Aglioti and Fabbro 1993; Aglioti et al. 1996; Fabbro 2001; Moretti et al. 2001). Moretti et al. (2001) point out the role of subcortical language structures and implicit memory in explaining the recovery of a patient who developed an impairment of the mother tongue following an infarct of the caudate. The authors further describe how during follow-up, improvement of the mother tongue was accompanied by worsening of the second language as the ischemic lesion extended to the cortex. The declarative/procedural model posits that brain damage in left neocortical temporal and temporo-parietal regions will be associated with increased difficulty in a later-learned and less-used L2 than in L1 or in an earlier-learned or wellpracticed L2. Alternatively, lesions to the left-hemisphere frontal lobe or basal ganglia would result in greater impairment to grammatical L1 structures or to an earlier-learned and well-practiced L2 than to a less-used or later-learned L2. Lexical performance should be affected for all cases. The case of a 16-year-old Chinese immigrant who learned English after the age of 10 upon his arrival in the United States presents with interesting evidence linking the left temporal lobe with L2 learning (Ku, Lachmann, and Nagler 1996). This patient developed herpes simplex encephalitis evident only in the left temporal lobe, and subsequently lost the ability to speak or comprehend English but retained these abilities in his native language, Chinese. This selective impairment of the second language with relative preservation of the mother tongue effectively links left temporal structures with a later-learned language (English) that was the most impaired subsequent to the lesions that this patient presented. Lesions that involve the left frontal lobe or basal ganglia structures are associated with increased difficulty in L1, since these structures mediate implicitmemory processes. Fabbro and Paradis (1995) present four cases of bilingual and polyglot patients with aphasia with lesions primarily circumscribed to the left basal ganglia. All four patients showed worse grammatical performance in their native or most proficient language than in their non-native or less proficient language. Cases such as these also lend support to the declarative/procedural model that associates native language performance to the left frontal lobe and basal ganglia. Lesions specific to these areas are more often associated with greater impairment in the first language than in later acquired languages. More difficulty in using the mother tongue than in the use of a second language subsequent to a subcortical lesion was reported by Agliotio et al. (1996). Their bilingual patient, E.M., whose mother tongue was Venetian and whose second language was Italian, suffered a subcortical lesion involving the left basal ganglia and subsequently presented increased difficulty spontaneously using her mother tongue, even though this was her most frequently used language. E.M. also had increased difficulty translating into her mother tongue as compared to translating into her second language. Her deficits were stable and sustained after five

74  Neurological and Psychological Aspects of Bilingualism and Multilingualism years from the time of stroke. E.M.’s symptomatology supports the declarative/ procedural model, in that her deficits and recovery patterns were related to lesions in the left basal ganglia with greater corresponding impairment in the first language with respect to the second language. The important role of the basal ganglia in automatized motor and cognitive performance affecting the first language was thus evident again in this case.

The inhibitory control model Polyglots are able to control and determine when to use each language, or when one language should be activated and the other suppressed or inhibited. The rules determining language choice and language-switching behaviors in normal speakers allow great flexibility, from intrasentential language switches to intersentential switches that may mark a shift in the language used. However, a general rule that is always observed by bilingual and polyglot speakers is that language-switching takes place only when the listener shares the language codes being used. This ability to switch language codes only when contextually appropriate is a skill that has been reported in bilingual speakers as young as two years old (Leopold 1949). There are three main components in bilingual and polyglot speakers’ language systems that must be accounted for in order to explain regulation in the use of two or more languages, according to Green (1986). Any language system must first address the issue of control. How are normal bilingual speakers with access to both language systems able to control which language they produce? If a bilingual speaker were not able to effectively separate language systems, code-switching (alternating use between one language and the other) would violate many of the rules observed by normal bilinguals in alternating between languages. Successful control of the language systems primarily implies the avoidance of error, such as blending two words in one language or across languages. In normal speakers, failure to exercise full control over an intact language system does occur, and may be due to a variety of reasons, such as temporary distraction, stress, fatigue, or the influence of toxic substances. In the aphasic speaker, however, brain damage itself provides the ‘stressful’ situation that may bring about failure to exercise full control over two or more language systems, even in otherwise non-stressful conditions. This lack of control by aphasics over their language systems is pervasive, often interfering with functional communication. The second component in the language system of bilingual and polyglot speakers involves the issue of activation determined by the internal representation of words from the known languages. Word frequency may be influential in determining activation or inhibition at both semantic and phonological levels of word production. More frequent words will generate larger activation patterns in lexical access than words that are less frequent or familiar. Finally, the third component described by Green involves the issue of resources linked to processes of control and regulation. Language representation systems must possess excitatory and inhibitory resources in the selection or de-selection of languages when polyglots are speaking.

Bilingual Aphasia: Theoretical and Clinical Considerations  75 Green (1998) explains that there are multiple levels of control in the language processing of bilingual speakers. In accordance, he elaborated his inhibitory control (IC) model by specifying one level of control involving language task schemas that compete to control output; the locus of word selection is the lemma level and involves the use of language tags; control at the lemma level is inhibitory and reactive. In his papers, Green delineates a functional-control circuit with three basic loci of control: an executive locus in charge of supervisory attention and critical in establishing and maintaining the speaker’s language goals; a locus at the level of language task schemas that the speaker selects; and a locus within the bilingual lexico-semantic system itself where lemmas are established. The bilingual speaker possesses a conceptualizer that builds conceptual representations from long-term memory and is driven by a goal to communicate through language. The speaker’s communicative and planning intention is mediated by a supervisory attentional system (SAS) and components of the language system itself, such as the lexico-semantic system and language task schemas. These language task schemas (e.g., translation or word production schemas) are in competition to control output (production) from the lexico-semantic system. The speaker’s selection of a word is determined through communication between the SAS and the task schemas and between the SAS and the conceptualizer. Another important aspect of this system is that once a speaker specifies a task schema to regulate behavior, the speaker must maintain the task as a goal in order to avoid competing task schemas becoming active and producing error behavior. The bilingual speaker must therefore first establish language task schemas to determine when to speak in one language or another, or when to translate between languages. These task schemas determine access into the lexico-semantic system within each language when polyglots select responses in one language or another. Therefore the lexico-semantic system and a set of language task schemas (such as translation or word production) both compete to control output from the lexicosemantic system. Language task schemas regulate responses from the lexico-semantic system by altering the activation levels of representation within the system and by inhibiting responses in the language that is not needed. In other words, when the speaker is responding in one of the languages, other languages are inhibited and cannot interfere by introducing responses into the output. Cases of paradoxical translation in bilinguals with aphasia are a good example of how language task schemas can be disrupted. In cases of paradoxical translation, the patient more efficiently translates from L1 to L2 whereas normal bilinguals typically translate best when going from the less proficient to the more proficient language, usually L2 to L1. Such a failure to inhibit one language over another can be explained as a breakdown in the supervisory attentional system and in the control of language task schemas. Code-switching between languages in bilingual and polyglot speakers becomes a pathological behavior when it is inappropriately used within a context where speakers do not share both language codes. The fact that this constitutes

76  Neurological and Psychological Aspects of Bilingualism and Multilingualism inappropriate ‘use’ of specific language(s), suggests the specificity of a pragmatic language disorder. Fabbro (2001) points out that pathological switching between languages is associated with lesions of the frontal lobe (left and right), whereas persistent mixing of elements from languages constitutes an aphasic disorder associated with fluent aphasias and tends to be correlated with left post-rolandic lesions (Fabbro 1999, 2001; Fabbro, Skrap, and Aglioti, 2000). The problem, of course, is that symptoms are not always clear-cut in patients and that pathological language-switching may coexist with language-mixing, thus reflecting impairment of executive control along with true language deficits affecting grammatical constructions or lexical search. Increased use of switching between languages by aphasic patients may also denote linguistic deficits in both languages and the need to rely on all available language codes to communicate more effectively. This kind of language-switching behavior should be distinguished from the subtle pragmatic deficits that can be present in bilinguals with aphasia. Muñoz, Marquardt, and Copeland (1999) compared discourse samples from four aphasic and four neurologically normal Hispanic bilinguals in monolingual-English, monolingual-Spanish, and in bilingual contexts, finding consistent matching of the language context by both the aphasic and normal subjects. However, the aphasic subjects demonstrated increased codeswitching not evident in the speech samples of the normal subjects, indicating increased dependence on both languages for communication subsequent to neurological impairment. A comparison of code-switching in Malayalam–English and Kannada–English bilinguals with aphasia with normals has also revealed that there is only increase in the quantity of code-switching and no difference in type of switches (Bhat and Chengappa, 2005; Chengappa, Daniel, and Bhat, 2004). It seems to be the case, then, that in bilingual speakers increased code-switching between languages will often take place as the result of compensatory strategies for deficits in lexical retrieval. These studies indicate the need to establish careful distinctions in discriminating between normal and pathological use of this communicative strategy. Support for this model can be found in cases of bilingual and polyglot aphasic patients where pathological switching and language-mixing are evidenced. Fabbro (1999) reports that some of the early cases describing pathological fixation on one language point to areas affecting the left supramarginal gyrus and the left temporal lobe. Cases involving pathological switching among languages involve the ventral anterior nucleus of the thalamus, the right hemisphere (reported in cases of right temporal lobe epilepsy), and the left prefrontal lobe. Paradis (1993) argues that language-switching mechanisms are part of a general system responsible for the selection of behaviors and that even though their localization is unclear, it should involve structures of the frontal system. Translation disorders also can be accounted for within Green’s model. According to Green (1986), aphasic individuals possess weakened language systems in which executive processes in charge of inhibiting or activating one language over the other fail to do so effectively, thus generating some of the symptoms often reported in the literature. Bilinguals with aphasia sometimes reveal disorders

Bilingual Aphasia: Theoretical and Clinical Considerations  77 associated with translation characterized by different features. One of these is when a bilingual or polyglot aphasic patient presents with an inability to translate. This kind of patient may thus be able to understand and use either language, yet be unable to translate from the first to the second language or from the second to the first language. This lack of flexibility in alternating between one language and the other in spite of comprehension and use of both languages may reflect the pervasiveness of the aphasic symptoms in the polyglot speaker. Patients presenting with paradoxical translation can translate only from the more proficient language into the language that is most difficult to speak spontaneously (however, normal bilinguals have an easier time translating from the less proficient language into their mastered language). Clinical cases exhibiting this pattern of language recovery were first described by Paradis, Goldblum, and Abidi in 1982. One of these patients, A.D., communicated only in Arabic some days and was subsequently able to express herself only in French for a number of days. On one day this patient was able to comprehend both languages but paradoxically translated from Arabic into French, while the next day she was only able to translate from French into Arabic. Similarly, De Vreese, Motta, and Toschi (1988) discuss paradoxical translation and spontaneous translation in a patient diagnosed with presenile dementia of the Alzheimer’s type who was able to translate correctly and without hesitation from his mother tongue into French, which he had difficulty speaking spontaneously. At the same time, he could not translate from French into his mother tongue. This patient also demonstrated involuntary translation of his own and other’s utterances. More curious yet is the symptomatology wherein patients are able to translate verbal information that they cannot understand. Cases involving translation without comprehension have been reported in patients who fail to understand commands that are given to them, but can nevertheless translate the sentences uttered by an interlocutor to express these commands (Fabbro 2001; Fabbro and Paradis 1995). The case of a 65-year-old woman who suffered a stroke with aphasia and right-sided hemiplegia was described by Veyrac in 1931. This patient’s native language was English; she acquired French after the age of 15. Subsequent to a stroke, she spoke a few utterances in French but did not understand even simple commands in French. When her physician gave her spoken instructions in English, the patient spontaneously translated these into French but was unable to carry them out, suggesting a dissociation between comprehension and translation. Language choice is a third area of phenomena that the IC model can account for. Cases of alternating antagonism refer to a patient’s alternating ability to control which language to use. Thus, the patient may be able to communicate in one language but not in another on some days and alternate to only using another language at a different time. Nilipour and Ashayeri (1989) reported such symptoms in a patient who developed a left fronto-temporal injury as a consequence of an explosion. This patient was reportedly equally proficient in Farsi and German and with good knowledge of English pre-morbidly. He was unable to speak during the first week post-trauma. He then only spoke limited Farsi for five days, but regained use of German for the following three weeks. The patient then

78  Neurological and Psychological Aspects of Bilingualism and Multilingualism switched to using only Farsi for four days until finally regaining control of all his languages, including English. Decreased use of switching among languages may also be seen in demented bilingual patients, who, unlike aphasic polyglots, will often address interlocutors in the patient’s native language despite the fact that the interlocutor speaks only the second language (De Santi et al. 1994; Hyltenstam and Stroud 1993). Dementias of the type such problems are reported in – namely, Alzheimer’s dementia – it should be noted, affect frontal lobe non-language structures primarily. In sum, several of the phenomena observed in bilingual aphasia (namely, pathological switching between languages and translation disorders) are consistent with the IC model that Green has developed.

Hierarchical (translation) models De Groot and Kroll (1997) propose a model of dual-language organization in bilinguals that focuses on sequential language acquisition. They argue that for those who have already learned an L1, an L2 is initially learned by translating newly acquired words through the L1. As the L2 learner acquires greater proficiency in the L2, direct lexical access takes place without having to mediate through the L1. This model predicts that translation from the L1 to the L2 will be more difficult than translating from the L2 to the L1. The significance of this model is that it allows testing how a bilingual speaker develops proficiency in the L2, simply by measuring changes in translation accuracy and processing time. The hierarchical model is consistent with the inhibition model proposed by Green and Abutalebi (2008) in that it points out that more cognitive resources are needed for a bilingual speaker to gain access and use their L2 if they have to process it through their L1. Inhibitory control may make it more difficult to access the L2 for an aphasic bilingual who, because of the brain lesion, may have difficulty in maintaining activation of the languages known before the aphasia, particularly for the language that was less proficient pre-morbidly or that is not as supported by the person’s everyday linguistic environment. Whereas the hierarchical model focuses only on bilingual speakers and not on multilingual speakers, the model nevertheless points to the need to check for proficiency levels in each of the languages of a polyglot speaker. When we look at the clinical research, we find that simply working on cognitive skills is not sufficient to improve language functions in an aphasic speaker. However, when therapy is conducted to improve inhibitory control and activation (as may be necessary to avoid patients employing translation equivalents between languages), remarkable gains can take place in the recovery of language functions. This was evident in D.J., the patient with whom Kohnert (2004) worked, who did not show language improvement when working on a cognitive task that required him to notice distractors on a computer screen. However, D.J. made rapid recovery (in two weeks) when working on translation equivalents using cognate words in English and Spanish. Findings along the same line were reported by Goral et al. (2006) when working with their trilingual patient E.C., who demonstrated differences in the level of

Bilingual Aphasia: Theoretical and Clinical Considerations  79 inter-language activation and interference in French, his less proficient language. This patient reported that he used translating among the languages he knew as a way to improve his language functions as he worked hard with dictionaries and grammars, along with standard speech therapy, to recover from his initially severe non-fluent aphasia in his three best languages. In summary, distinguishing between early bilinguals and sequential bilinguals is useful not only for theoretical purposes, but also to select strategies for clinical work with bilinguals or multilinguals who require language intervention. At this point, we can reasonably posit that working on inhibitory control and activation of the various languages known by a polyglot should have transfer effects and improve processing time.

Conclusions Over a century and a quarter, interest in bilingual and polyglot aphasia has progressively increased. Initially, questions concerning the order of recovery of the two or more languages predominated, along with questions of what the causes of observed recovery patterns were. Questions about the incidence of crossed aphasia, with its implications for lateralized participation in language production and processing, entered in the mid-twentieth century as language lateralization became a topic of great interest in neurolinguistics generally. In the past quartercentury linguists, speech language pathologists, and investigators within related disciplines have joined behavioral neurologists in the study of bilingual aphasia and its particular phenomenology (e.g., day-to-day fluctuations in language accessibility, disordered translation, language choice, code-switching and -mixing behaviors). In the past decades, psychologists have joined the discussion, contributing models of the ways in which frontal-system executive control and subcortical memory systems subserve healthy bilingual processes and can account for deficits documented in bilingual aphasia.

ACKNOWLEDGMENTS Thanks to Eve Higby for help with the references and proofing, and to Sara Weiss for help with proofing.

REFERENCES Abutalebi, Jubin, Rosa, Pascuale A., Tettamanti, Marco, Green, David W., and Cappa, Stefano F. (2009). Bilingual aphasia and language control: A follow-up fMRI and intrinsic connectivity study. Brain and Language 109: 141–56.

Abutalebi, Jubin, Cappa, Stefano F. and Perani, Daniela (2001). The bilingual brain as revealed by functional neuroimaging. Bilingualism: Language and Cognition 4(2): 179–90.

80  Neurological and Psychological Aspects of Bilingualism and Multilingualism Aglioti, Salvatore and Fabbro, Franco (1993). Paradoxical selective recovery in a bilingual aphasic following subcortical lesions. NeuroReport 4: 1359–62. Aglioti, Salvatore, Beltramello, Alberto, Girardi, Flavia, and Fabbro, Franco (1996). Neurolinguistic and follow-up study of an unusual pattern of recovery from bilingual subcortical aphasia, Brain 119: 1551–64. Albert, Martin L. and Obler, Loraine K. (1978). The Bilingual Brain: Neuropsychological and Neurolinguistic Aspects of Bilingualism. New York: Academic Press. Appel, René and Muysken, Pieter (1997). Language Contact and Bilingualism. New York: Arnold. Bates, Elizabeth, MacWhinney, Brian, Caselli, Cristina, Devescovi, Antonella, Natale, Francesco, and Venza, Valeria (1984). A cross-linguistic study of the development of sentence interpretation strategies. Child Development 55: 341–54. Bhat, Sapna and Chengappa, Shyamala K. (2005). Code switching in normal and aphasic Kannada–English bilinguals. In James Cohen, Kara T. McAlister, Kellie Rolstad, and Jeff MacSwan (eds.), ISB4: Proceedings of the 4th International Symposium on Bilingualism. 306–16. Somerville, MA: Cascadilla Press. Broca, Paul (1861/1977). Remarks on the seat of the faculty of articulate speech, followed by the report of a case of aphemia. Trans. Claude Wasterlain and David A, Rottenberg, In David A. Rottenberg and Fred H. Hochberg (eds.), Neurological Classics in Modern Translation. 136–49. New York: Hafner. Bose, Arpita and Chengappa, Shyamala (2000). Naming deficits in bilingual aphasic individuals. Paper presented at Canadian Association of Speech-language Pathologists and Audiologists Conference (CASLPA), Toronto, Canada. Chee, Michael W. L., Hon, Nicholas, Lee, Hwee Ling, and Soon, Chun Siong (2001). Relative language proficiency modulates BOLD signal change when bilinguals perform semantic judgments. Neuroimage 13: 1155–63. Chengappa, Shyamala, Bhat, Sapna, and Padakannaya, Prakash (2004). Reading and writing skills in multilingual/multiliterate aphasics: Two case studies. Reading and Writing: An Interdisciplinary Journal 17: 21–135. Chengappa, Shyamala, Daniel, Krupa Elizabeth, and Bhat, Sapna (2004). Language mixing and

switching in Malayalam–English bilingual aphasics. Asia Pacific Disability Rehabilitation Journal 15(2): 68–76. De Groot, Annette and Kroll, Judith (1997). Tutorials in Bilingualism. New Jersey: Erlbaum. De Santi, Susan, Koenig, Laura, Obler, Loraine K., and Goldberger, Joan (1994). Conversational discourse in demented bilinguals. In Ronald Bloom, Loraine K. Obler, Susan De Santi, and Jonathan Ehrlich (eds.), Discourse Analysis and Applications: Studies in Adult Clinical Populations. Hillsdale, NJ: Erlbaum. De Vreese, Luc Pieter, Motta, Massimo, and Toschi Andrea (1988). Compulsive and paradoxical translation behaviour in a case of presenile dementia of the Alzheimer type. Journal of Neurolinguistics 3: 233–41. Fabbro, Franco (1999). The Neurolinguistics of Bilingualism: An Introduction. Hove: Psychology Press. Fabbro, Franco (2001). The bilingual brain: Cerebral representation of languages. Brain and Language 79: 211–22. Fabbro, Franco and Paradis, Michel (1995). Differential impairments in four multilingual patients with subcortical lesions. In Michel Paradis (ed.), Aspects of Bilingual Aphasia. 139–76. Oxford: Pergamon Press. Fabbro, Franco, Skrap, Miran, and Aglioti, Salvatore (2000). Pathological switching between languages after frontal lesions in a bilingual patient. Journal of Neurology, Neurosurgery and Psychiatry 68: 650–2. Faroqi, Yasmeen and Chengappa, Shyamala (1997). Trace deletion hypothesis and its implications for intervention with a multilingual agrammatic aphasic patient. Osmania Papers in Linguistics 22–3: 79–106. Faroqi-Shah, Yasmeen, Frymark, Tobi, Mullen, Robert, and Wang, Beverly (2010). Effect of treatment for bilingual individuals with aphasia: A systematic review of the evidence. Journal of Neurolinguistics 23(4): 319–41. Freud, Sigmund (1891). Zur Auffassung der Aphasien. Eine kritische Studie. Leipzig: Franz, Deutsche. Trans. E. Stengel (1953). New York: International University Press. Gloning, I. and Gloning, K. (1965). Aphasien bei polyglotten. Wiener Zeitschrift für Nervenheilkunde 22: 362–97. Green, David W. (1986). Control, activation, and resource: A framework and a model for the

Bilingual Aphasia: Theoretical and Clinical Considerations  81 control of speech in bilinguals. Brain and Language 27: 210–23. Green, David W. (1998). Mental control of the bilingual lexico-semantic system. Bilingualism: Language and Cognition 1: 67–81. Green, David W. and Abutalebi, Jubin (2008). Understanding the link between bilingual aphasia and language control. Journal of Neurolinguistics 21: 558–76. Goral, Mira, Levy, Erika, Obler, Loraine K., and Cohen, Eyal (2006). Cross-language lexical connections in the mental lexicon: Evidence from a case of trilingual aphasia. Brain and Language 98(2): 235–47. Hughlings Jackson, John (1879). On affections of speech from disease of the brain. Brain I: 304–30. Hyltenstam, Kenneth and Obler, Loraine K. (1989). Bilingualism across the Lifespan: Aspects of Acquisition, Maturity and Loss. New York: Cambridge University Press. Hyltenstam, Kenneth, and Stroud, Christopher (1993). Second language regression in Alzheimer’s dementia. In Kenneth Hyltenstam and Ake Viberg (eds.), Progression and Regression in Language: Sociocultural, Neuropsychological and Linguistic Perspectives. Cambridge: Cambridge University Press. Illes, Judy, Francis, Wendy S., Desmond, John E., Gabrieli, John D. E., Glover, Gary H., Poldrack, Russell A., Lee, Christine J., and Wagner, Anthony D. (1999). Convergent cortical representation of semantic processing in bilinguals. Brain and Language 70: 347–63. Isel, Frederic, Baumgaertner, Annette, Thran, Johannes, Meisel, Jürgen M., and Buchel, Christian (2010). Neural circuitry of the bilingual mental lexicon: Effect of age of second language acquisition. Brain and Cognition 72(2): 169–80. Johnson, Jacqueline S. and Newport, Elissa L. (1989). Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language. Cognitive Psychology 21: 60–99. Karanth, Prathibha (2002). The search for deep dyslexia in a syllabic script. Journal of Neurolinguistics, 15(2): 143–55. Khateb, Said, Martory, Marie Dominique, Annoni, Jean Marie, Lazeyras, François, de Tribolet, Nicolas, Pegna, Alan J., Mayer, Eugene et al. (2004). Transient crossed aphasia

evidenced by functional brain imagery. NeuroReport 15(5): 785–90. Kho, Kuan H., Duffau, Hugues, Gatignol, Peggy, Leijten, Frans S. S., Ramsey, Nick F., van Rijen, Peter C., and Rutten, Geert-Jan M. (2007). Involuntary language switching in two bilingual patients during the Wada test and intraoperative electrocortical stimulation. Brain and Language 101(1): 31–7. Kim, Karl H. S., Relkin, Norman R., Lee, Kyoung-Min, and Hirsch, Joy (1997). Distinct cortical areas associated with native and second languages. Nature 388(10): 171–4. Klein, Denise, Zatorre, Robert J., Milner, Brenda, and Zhao, Viviane (2001). A cross-linguistic PET study of tone perception in Mandarin Chinese and English speakers. Neuroimage 13: 646–53. Koda, Keiko (2004). Insights into Second Language Reading: A Cross-Linguistic Approach. Cambridge: Cambridge University Press. Kohnert, Kathryn (2004). Cognitive and cognatebased treatments for bilingual aphasia: A case study. Brain and Language 91(3): 294–302. Kroll, Judith F. and Stewart, Erika (1994). Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language 33: 149–74. Ku, Aubrey, Lachmann, Elizabeth A. and Nagler, Willibald (1996). Selective language aphasia from herpes simplex encephalitis. Pediatric Neurology 15(2): 169–71. Lambert, Wallace E. and Fillenbaum, Samuel (1959). A pilot study of aphasia among bilinguals. Canadian Journal of Psychology 13: 28–34. Lenneberg, Eric H. (1967). Biological Foundations of Language. New York: Wiley. Leopold, Werner (1949). Speech Development of a Bilingual Child (4 vols.). Evanston, IL: Northwestern University Press. Miertsch, Barbara, Meisel, Jürgen M., and Isel, Frederic (2009). Non-treated languages in aphasia therapy of polyglots benefit from improvement in the treated language. Journal of Neurolinguistics 22(2): 135–50. Minkowski, Mieczyslaw (1927/1983). A clinical contribution to the study of polyglot aphasia especially with respect to Swiss-German. In Michel Paradis (ed.), Readings on Aphasia in Bilinguals and Polyglots. 205–32. Montreal: Didier.

82  Neurological and Psychological Aspects of Bilingualism and Multilingualism Moretti, Rita, Bava, Antonio, Torre, Paola, Antonello, Rodolfo M., Zorzon, Marino, Zivadinov, Robert, and Cazzato, Giuseppe (2001). Bilingual aphasia and subcorticalcortical lesions. Perceptual Motor Skills 92(3): 803–14. Muñoz, Maria L., Marquardt, Thomas P. and Copeland, Gary (1999). A comparison of the code switching patterns of aphasic and neurologically normal bilingual speakers of English and Spanish. Brain and Language 66: 249–74. Neville, Helen J., Coffey, Sharon A., Lawson, Donald S., Fischer, Andrew, Emmorey, Karen, and Bellugi, Ursula (1997). Neural systems mediating American Sign Language: Effects of sensory experience and age of acquisition. Brain and Language 57(3): 285–308. Nilipour, Reza, and Ashayeri, Hassan (1989). Alternating antagonism between two languages with successive recovery of a third in a trilingual aphasic patient. Brain and Language 36(1): 23–48. Obler, Loraine K. (1984). Knowledge in neurolinguistics: The case of bilingualism. In Alexander Z. Guiora (ed.), An Epistemology of Language Sciences. 159–91. Wayne, IN: Wayne State University Press. Obler, Loraine K. and Albert, Martin (1977). Influence of aging on recovery from aphasia in polyglots. Brain and Language 4: 460–3. Paradis, Michel (1977). Bilingualism and aphasia. In Harry Whitaker and Haiganoosh Avakian Whitaker (eds.), Studies in Neurolinguistics Vol. 3. 65–121. New York: Academic Press. Paradis, Michel (1993). Linguistic, psycholinguistic, and neurolinguistic aspects of ‘interference’ in bilingual speakers: The activation threshhold hypothesis. International Journal of Psycholinguistics 9: 133–45. Paradis, Michel (1994). Neurolinguistic aspects of implicit and explicit memory: Implications for bilingualism and second language acquisition. In Nick C. Ellis (ed.), Implicit and Explicit Language Learning. London: Academic Press. Paradis, Michel, Goldblum, Marie-Claire, and Abidi, Raouf (1982). Alternate antagonism with paradoxical translation behavior in two bilingual aphasic patients. Brain and Language 15(1): 55–69. Perani, Daniela, Paulesu, Eraldo, SebastianGalles, Nuria, Dupoux, Emmanuel, Dehaene,

Stanislas, Bettinardi, Valentino, Cappa, Stefano F. et al. (1998). The bilingual brain: Proficiency and age of acquisition of the second language. Brain 121: 1841–52. Pitres, Albert (1885/1983). Aphasia in polyglots. In Michel Paradis (ed.), Readings on Aphasia in Bilinguals and Polyglots. 26–49. Montreal: Didier. Ribot, Théophile. (1881). Les maladies de la mémoire. Paris: Baillère. Romaine, Suzanne (1995). Bilingualism. Oxford: Blackwell. Scoresby-Jackson, Robert E. (1867). Case of aphasia with right hemiplegia. Edinburgh Medical Journal 12: 696–706. Sridevi, Kasirajan (1999). Comprehension deficits in bilingual aphasics. Unpublished Master’s dissertation submitted to University of Mysore, Mysore. Sunilkumar, Ravi, Gnanavel, Kuppusamy, Vishnu, Kaleeckal, Krishnankutty, Nair, and Chengappa, Shyamala K. (2010). Action naming in bilingual (Kannada–English) aphasics: Implications to models of lexical organization. In Shyamala Chengappa (ed.), Proceedings of the International Symposium on Bilingual Aphasia. 134–42. Mysore: All India Institute of Speech and Hearing Publications. Tan, Li Hai, Liu, Ho Ling, Perfetti, Charles A., Spinks, John A., Fox, Peter T., and Gao, Jia Hong (2001). The neural system underlying Chinese logograph reading. Neuroimage 13: 836–46. Ullman, Michael T. (2001). The neural basis of lexicon and grammar in first and second language: The declarative/procedural model. Bilingualism: Language and Cognition 4(1): 105–22. Vaid, Jyotsna and Chengappa, Shyamala (1988). Assigning linguistic roles: Sentence interpretation in normals and aphasic Kannada – English bilinguals. Journal of Neurolinguistics 3: 61–83. Vaid, Jyotsna and Genesee, Fred (1980). Neuropsychological approaches to bilingualism: A critical review. Canadian Journal of Psychology 34: 417–45. Vaid, Jyotsna and Hall, D. Geoffrey (1991). Neuropsychological perspectives on bilingualism: Right, left, and center. In Allan G. Reynolds (ed.). Bilingualism, Multiculturalism, and Second Language Learning: The McGill Conference in Honor of

Bilingual Aphasia: Theoretical and Clinical Considerations  83 Wallace E. Lambert. 81–112. Hillsdale, NJ: Erlbaum. Vaid, Jyotsna and Hull, Rachel (2002). Reenvisioning the bilingual brain using functional neuroimaging: Methodological and interpretive issues. In Franco Fabbro (ed.), Advances in the Neurolinguistics of Bilingualism: Essays in Honor of Michel Paradis. 315–55. Udine Forum: Udine University Press. Veyrac, G. J. (1931). A study of aphasia in polyglot subjects. In Michel Paradis (ed.), Readings on Aphasia in Bilinguals and Polyglots. 320–38. Montreal: Didier.

Wang, Yapeng, Kuhl, Patricia K., Chen, Chunhui, and Dong, Qi (2009). Sustained and transient language control in the bilingual brain. Neuroimage 47(1): 414–22. Watamori, Toshiko S. and Sasanuma, Sumiko (1978). The recovery process of two English– Japanese bilingual aphasics. Brain and Language 6: 127–40. Weinreich, Uriel (1953). Languages in Contact. The Hague: Mouton. Wernicke, Carl (1874). The symptom complex of aphasia. In Boston Studies in the Philosophy of Science Vol. IV, 1966/1968, 34–97.

Approaches to Bilingualism, Multilingualism, and Second-Language Acquisition

4  The Bilingual Child LUDOVICA SERRATRICE

Introduction The study of bilingualism in general, and of language development in bilingual children in particular, has gained momentum over the last 20 years with a number of new international journals, a plethora of publications, new corpora, and conferences attracting increasing numbers of scholars worldwide. Important theoretical and methodological changes have taken place in the field from the early days of Ronjat (1913) and Leopold (1939), and, more recently, dating back to the seminal paper by Volterra and Taeschner (1978), which started a thriving line of research in the investigation of bilingual children’s lexical and morpho-syntactic development, and the relationship between their two languages. From a theoretical point of view the focus has now shifted from an emphasis on language differentiation to the issue of cross-linguistic influence. Currently the majority consensus is that simultaneous bilingual children who are regularly exposed to two languages can learn them as independent systems; independence, however, does not necessarily mean complete autonomy and researchers have been concentrating on the determinants of interaction between the two linguistic systems in bilingual infants and children. Much recent work has also been devoted to the relationship between language and cognition in bilinguals across the lifespan. In the following I will review recent advances in the field in terms of phonological, lexical, and morpho-syntactic development; the last section covers some of the more recent work on language and cognition, in particular executive function skills, in young bilinguals.

The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

88  Neurological and Psychological Aspects of Bilingualism and Multilingualism

Speech Perception The initial challenge facing a child exposed to two languages in her environment is the discrimination of the two systems. In order to acquire two languages the child must first of all discover that two such systems exist in her input, i.e. a child growing up with Norwegian and Russian must categorize input as either Norwegian or Russian. At a stage when infants cannot yet exploit lexical information for language discrimination they will have to rely on pre-lexical cues: prosody; phonetic repertoire; and phonotactic constraints (Ramus and Mehler 1999). At birth, monolingual infants can rely on rhythmic properties to discriminate between languages belonging to different classes (e.g., between stress-timed English, syllable-timed Italian, and mora-timed Japanese) but not between languages belonging to the same rhythmic class (e.g. stress-timed English and Dutch) (Nazzi, Bertoncini, and Mehler 1998). By the age of 5 months monolinguals can discriminate between their native dialect (American English) and another nonnative dialect (British English), between their nonnative dialect (British English) and another stress-timed language (Dutch), but not between two unfamiliar stresstimed languages (Dutch and German) (Nazzi, Jusczyk, and Johnson 2000). Nazzi et al. (2000) interpreted these findings in the light of the native-language hypothesis whereby children use their language-specific experience with their native language (e.g. American English) to classify speech input. Nazzi et al. (2000) proposed that infants relied on prosodic information (the longer duration of stressed syllables and shorter duration of unstressed syllables in British English) to discriminate between American and British varieties. A nonnative dialect like British English is both similar to the native language but different enough to be classified as other; two nonnative languages like Dutch and German are dissimilar enough from the native language to be classified as other, but too far removed from the nonnative dialect for the child to be able to differentiate them further. Research comparing bilinguals and monolinguals in their ability to discriminate rhythmically similar languages on the basis of prosodic information has shown that bilinguals are not delayed compared to monolinguals. Bosch and Sebastián-Gallés (1997, 2001) showed that by the age of 4 months both mono­ lingual Catalan and Spanish infants and Catalan–Spanish bilinguals rely on fine-grained prosodic cues (probably differential stress assignment patterns at word level) to discriminate between two syllable-timed languages. Even when bilingual children have accumulated experience with two native languages so that they are both familiar, they can still distinguish between them on the basis of prosody alone. In line with this argument, Byers-Heinlein, Burns, and Werker (2010) demonstrated that newborns whose mothers had spoken Tagalog and English during pregnancy showed equal preference for both languages while babies born to English-speaking mothers had a strong preference for English. Importantly, despite this familiarity effect for two languages in the case of the bilinguals, Byers-Heinlein et al. also showed that the bilingual newborns could reliably discriminate between Tagalog and English, two languages from different rhythmic classes, at birth.

The Bilingual Child  89 In addition to the well-established importance of prosodic cues, there is some emerging evidence for the role played by visual cues in language differentiation. Weikum et al. (2007) used silent clips of three bilingual English–French adults producing sentences in each language in a habituation phase and measured looking times on switch trials. Monolingual English infants were tested at 4, 6, and 8 months of age and compared to bilingual English–French infants at 6 and 8 months of age. The monolinguals looked significantly longer on switch trials at 4 and 6 months but not at 8 months, while the bilinguals did so at both 6 and 8 months of age. While this study shows that attention to redundant visual cues plays a part in early language identification in both monolingual and bilingual children, it also demonstrates that in the absence of sustained language exposure, these abilities decline over time for the monolingual infants. In addition to reliance on both innate and developing sensitivity to rhythmic properties of language there is evidence that, similarly to monolingual infants, bilinguals concurrently pay attention to phonetic cues. The monolingual literature shows a decline in infants’ ability to discriminate nonnative phonetic contrasts from the second half of the first year of life with a concomitant improvement in the discrimination of native contrasts between 6 and 12 months of age (SebastiánGallés 2006). When compared to monolinguals the findings from bilinguals have tended to show different developmental patterns depending on the contrasts under investigation. The frequency of the phonetic categories, their distribution, and their consonantal vs. vocalic status have all been shown to play a role in the ease with which bilingual infants can discriminate a contrast at the age of 8 months. In the case of the /e–ε/ pair, a contrast found in Catalan but not in Spanish where only /e/ is part of the vowel system, Bosch and Sebastián-Gallés (2003) found a U-shaped learning curve for Spanish–Catalan bilinguals. They could detect the contrast at the age of 4 months and 12 months, but not at 8 months; in contrast, the monolingual infants were consistently successful. The same pattern was replicated by Sebastián-Gallés and Bosch (2009) for the /o–u/ contrast in Spanish–Catalan bilinguals, but not for the more phonetically distant /e–u/ contrast, in this case the bilinguals, like the monolinguals, were successful in the discrimination task at both 4 and 8 months. Other studies have failed to show a U-shaped developmental pattern and have reported similar discriminatory abilities in monolinguals and bilinguals. Burns et al. (2007) compared monolingual English infants and French–English bilinguals at 6–8 months, 10–12 months, and at 14–20 months on the discrimination of the three-way distinction between /ba/ /pa/ and /pha/ testing the /p–b/ continuum. In French there is a shorter VOT boundary between /b/ and /p/ while in English the boundary separates /p/ and /ph/. In the younger age group both monolingual and bilingual infants successfully detected the /pa–ba/ and the /pa–pha/ contrasts, in the two older groups the monolingual children could only detect the native /pa–pha/ contrast while the bilinguals retained discrimination of the French contrast as well. Sundara, Polka, and Molnar (2008) showed similar results with English–French bilinguals and French and English monolinguals on the dental (French) and alveolar (English) contrast for the phoneme /d/. At 6–8 months all groups of children

90  Neurological and Psychological Aspects of Bilingualism and Multilingualism showed language-general abilities to discriminate between the two phones, by the age of 10–12 months the French children could no longer distinguish this contrast that adults also fail to perceive; only the bilinguals and the English monolinguals succeeded in this task. More recently the same /e–ε/ contrast tested by Bosch and SebastiánGallés (2003) with Spanish–Catalan bilinguals was tested with Spanish–English bilinguals by Sundara and Scutellaro (2011). Unlike in the study by Bosch and Sebastián-Gallés (2003), the infants in the Sundara and Scutellaro experiments successfully discriminated between the phones both at 4 and at 8 months of age. The bilinguals’ success in this pair of rhythmically different languages is likely due to what the authors call a strategy of language tagging on the basis of prosodic characteristics. Languages that belong to different rhythmic classes (e.g., English and Spanish) are discriminated earlier than languages from the same rhythmic class (e.g., Spanish and Catalan); if children tag the input on the basis of prosodic cues they are in a position to compute statistical regularities for each language independently and thus overcome the issue of overlapping distributions. Using the anticipatory eye movement paradigm (Aslin and McMurray 2004) AlbaredaCastellot, Pons, and Sebastián-Gallés (2011) went back to the anomalous /e–ε/ contrast and reported no differences between Spanish–Catalan bilinguals and Catalan monolinguals at the critical age of 8 months in this less demanding testing set-up not involving recovery of attention. The current evidence is increasingly pointing towards a model of speech perception development where simultaneous bilinguals do not lag behind monolinguals and perform just as accurately in their first year of life. Studies testing pairs of languages that are more or less typologically distant have also revealed interesting language-specific issues to bear in mind when comparing the per­ formance of bilingual children learning different language combinations (e.g., frequency of cognates, frequency of phones in each language, phonetic distance between phones, consonantal vs. vocalic status of phones). While building language-specific phonetic representations infants also learn about the statistical regularities and about the transitional probabilities of different sounds in their ambient language. For monolingual children there is evidence of emerging sensitivity to phonotactic constraints, i.e., to the legality of given strings of consonants and vowels, around 9 months of age. Jusczyk et al. (1993) found that American and Dutch 9 month olds had a preference for listening to lists of words in their own native language; Friederici and Wessels (1993) reported that 9-month-old Dutch infants could detect legal and illegal clusters at word boundaries. Paying attention to word boundaries is obviously extremely advantageous to infants’ developing word-learning abilities and it is significant that these skills seem to emerge in the second half of the first year of life when children start to produce their first words. In the case of bilingual infants Sebástian-Gallés and Bosch (2002) showed that language dominance, as measured by amount of exposure, interacted with Catalan–Spanish 10 month olds’ abilities to detect the legality of coda clusters in Catalan. Like the Catalan monolinguals, the Catalan-dominant bilinguals

The Bilingual Child  91 showed a significant preference for legal CVCC strings, i.e. codas that comply with the phonotactics of Catalan. As expected, the Spanish monolinguals had no preference as coda clusters are virtually nonexistent in Spanish; the infants were required to make a judgment between strings that were equally unfamiliar to them and their latencies reflected this. Like the Catalan-dominant bilinguals, the Spanishdominant bilinguals also preferred legal over illegal codas, but the difference in their case was nonsignificant. These findings were interpreted as being consistent with the hypothesis of early language dominance although the Spanish-dominant infants were not tested in Spanish and therefore no data was available to confirm that they would perform like Spanish monolinguals in their preferred language. Interestingly, there were nontrivial differences in the amount of exposure between the monolingual Catalan infants (95% Catalan) and the Catalan-dominant bilinguals (60% Catalan) and less of a difference between the Catalan-dominant infants and the Spanish-dominant infants (40% Catalan) suggesting that the magnitude of the difference is less important than the side of the dominance. In the same study, Sebástian-Gallés and Bosch (2002) also conducted two experiments testing adult Spanish–Catalan bilinguals who had been exposed to similar amounts of input in the two languages after the ages of 3 or 4. The pattern found with the infants was replicated to some extent in the adult experiments; the Catalan-dominant participants were better at processing legal codas than their Spanish-dominant counterparts, the differences between the two groups were very small and smaller than the ones observed in previous studies testing phonemic rather than phonotactic awareness where up to 50% of Spanish-dominant bilinguals failed to perceive Catalan phonemic contrasts (Bosch, Costa, and Sebástian-Gallés 2000; Sebástian-Gallés and Soto-Faraco 1999).

Speech Production Issues of language differentiation that have preoccupied researchers studying speech perception also loom large in the investigation of the acquisition of segmental and supra-segmental phonology in production. Cross-linguistic research on monolingual children has shown that maturational constraints operate on sound production and that phonetic accuracy is determined both by the children’s developing articulatory abilities and by their growing knowledge of the phonological system. For example, simple CV syllables emerge earlier than clusters (Lleó and Prinz 1996), and coronal and labial stops appear before fricatives (Macken and Ferguson 1981). Aside from the well-attested universal developmental schedule in the acquisition of sounds, the obvious issue to consider in childhood bilingualism is whether exposure to two languages has an effect over and beyond what would be expected in the course of typical monolingual development. Findings in the literature are somewhat mixed, partly because of some methodological limitations including the use of single case studies and the lack of monolingual controls (see Paradis 2001; Lleó and Kehoe 2002 for a critique). Individual differences are widespread,

92  Neurological and Psychological Aspects of Bilingualism and Multilingualism especially in the early years, and failure to observe cross-language interaction in a small number of bilingual children does not imply that all children, or even the majority, do not go through a stage where the two languages do interact in important ways. The absence of a monolingual group also prevents the researcher from drawing firm conclusions on whether the pattern observed can definitely be ascribed to the child’s bilingualism or whether it may be attested in the course of monolingual development as well. Over the last 10 years, however, a number of group studies including monolingual controls have investigated the issue of the independent and autonomous development of a phonological system in production. Paradis (2001) made the case for independence but not autonomy on the basis of a study of French–English 2 year olds’ knowledge of syllable patterns in a nonword repetition task. The findings showed that the bilinguals, similarly to the mono­ linguals, observed the language-specific trochaic production template in syllable deletion in English and the iambic template in French. Nevertheless, in terms of quantity there was an asymmetrical result; no differences in heavy and light syllable preservation in French between the two groups, but in English the bilinguals, unlike the monolinguals, failed to preserve heavy initial syllables. In a further analysis Paradis also showed a role for language dominance as a potential explanatory variable for the asymmetry of the finding. French-dominant children omitted more weak syllables in third position than strong syllables in second position in WS’WS nonwords, while the reverse pattern was observed for English-dominant bilinguals. In a study of prosodic word structures in Spanish monolinguals and Spanish–German bilinguals between the ages of 1 and 2, Lleó (2002) reported that the bilinguals produced monosyllables earlier than the monolinguals and that they observed the maximality constraint for a longer period of time. Input from German was held responsible for such a pattern of behavior. The emerging consensus in the field views supra-segmental and segmental acquisition as independent but not entirely autonomous. Although bilingual children can develop phonological systems and detailed phonetic representations that reflect the input of the ambient languages, nevertheless a degree of interaction has consistently been reported in a number of studies examining phonetic repertoires and accuracy of production. Fabiano-Smith and Barlow (2009), for example, found that by the age of 3 Spanish–English bilingual children had developed phonetic repertoires that were as complex as those of their monolingual peers in both English and Spanish; however, the bilingual children did not score as well as the Spanish monolinguals on accuracy measures. In another study with Spanish– English bilingual 3-year-olds, Fabiano-Smith and Goldstein (2010a) showed that the bilinguals were significantly less accurate in the percentage of correct consonants compared to the Spanish monolinguals but no significant differences were observed in English as a function of language background. In a more fine-grained analysis where they looked at accuracy of production of early-, middle- and late-acquired sounds in the two languages Fabiano-Smith and Goldstein (2010b) reported that in English accuracy in both groups of children correlated with age of acquisition whereby early-acquired sounds were more

The Bilingual Child  93 accurate than middle- and late-acquired sounds. In Spanish the monolinguals were more accurate on early- than middle-acquired sounds but no differences existed between middle- and late-acquired sounds. In contrast, for the bilinguals no significant differences in accuracy were observed depending on the age of acquisition showing an asymmetrical pattern that was tentatively ascribed to children’s use and proficiency of the languages according to parental assessment.

Word Learning The literature on bilingual infants’ receptive phonetic skills has shown that they can exploit subtle sub-phonemic cues in the input to discriminate sounds belonging to different languages. On the assumption that there is continuity between pre- and post-lexical representations in development (Werker and Curtin 2005), bilinguals’ ability to tune into language-specific phonetic differences should pave the way to building receptive vocabulary. Fine-grained phonetic differences may signal language-specific phonemic differences in one, but not in the other, of the bilingual’s two languages and children must pay attention to them. In Tamil, for example, there is a phonemic distinction between dental and alveolar nasals that is nonexistent in English. A child that is acquiring both English and Tamil can disregard this subtle difference in place of articulation in English, but must pay attention to it in Tamil as it signals a contrast in meaning. Moreover, while a child growing up in a monolingual English environment is unlikely to hear an alveolar /n/, a child growing up surrounded by bilingual Tamil–English speakers is likely to hear tokens of alveolar /n/ in English. Bilingual children’s emerging system must therefore be robust and flexible enough to deal with the inherent variability in the bilingual input that they receive. Evidence for such adaptive flexibility linking phonetic skills and word learning has been provided in a study by Mattock et al. (2010). The researchers tested French and English monolingual 17-month-olds and French–English bilingual 17month-olds in Canada on a word-learning task using the switch paradigm (Werker et al. 1998). In the first experiment both sets of monolinguals and the bilinguals were habituated to two nonce words (/bos/ and /gos/) paired with two novel objects. The stimuli were produced by an adult French–English bilingual using both the French and the English pronunciation so that half the tokens were more French sounding while the remaining tokens had an English pronunciation. The results of three experiments showed that only the bilingual children detected a change in the pairing between nonce word and novel object in the switch phase after the habituation phase, the monolingual children successfully did so only when they were exclusively exposed to stimuli that matched the pronunciation of their native language. Using the same switch paradigm Fennell, Byers-Heinlein, and Werker (2007) tested monolingual and bilingual infants at 14, 17, and 20 months of age on similar sounding nonwords differing only in place of articulation on the initial consonant

94  Neurological and Psychological Aspects of Bilingualism and Multilingualism (/b–d/ contrast). While the monolinguals succeeded at 17 months of age the bilinguals did not until 20 months. These results cannot easily be reconciled with those of Mattock et al., as Fennell et al. did not find any correlation between amount of exposure to English and success on the task. Werker, Byers-Heinlein, and Fennell (2009) proposed that although bilingual infants, and even monolingual infants, may have abstract representations of phonemes by the age of 17 months, nevertheless they may not yet be so well entrenched that children can easily ignore nonphonemic phonetic variation in the context of an experimental task. Bilingual toddlers’ efficiency with the processing of the speech signal has been linked to the size of their vocabulary. In a study examining Spanish–English 2year-olds’ lexical knowledge in a preferential looking task Marchman, Fernald, and Hurtado (2010) showed positive within-language correlations between children’s reaction times in the recognition of familiar words and vocabulary size but no significant across-language correlations. Speed of online lexical processing in Spanish was not predictive of vocabulary size in English and processing speed in English was not correlated with vocabulary size in Spanish. At the same time, processing speed in each language was moderately correlated with a composite vocabulary measure including both English and Spanish words, suggesting that general processing skills play a role in word recognition that are not purely language specific. Bilinguals’ flexibility in the linguistic categorization of the speech signal, and in the mapping between phonological form and referent, is shaped by their early experience with language; they are exposed to larger phonetic variability compared to monolingual children and their developing system must find a way to deal with it. Another area of word learning in which early experience with two languages has been shown to contribute to bilingual infants’ flexible learning strategies is in the application of the mutual exclusivity constraint. According to the mutual exclusivity constraint hypothesis (Markman 1989; Markman and Wachtel 1988) children have a bias to assume that a new label refers to an unknown object. So upon hearing ‘Give me the dax!’ in the presence of a familiar object (e.g., a cup) and of an unfamiliar object (e.g., a phototube) they are significantly more likely to choose the new object for which they have no name yet (Merriman and Bowman 1989). Recent work with monolingual children has shown that there is a maturational component to this bias; this tendency is not innate but develops over time in the second year of life (Halberda 2003; Mather and Plunkett 2009). The investigation of the application of the mutual exclusivity constraint in bilingual children is particularly revealing of the degree to which this bias is experience driven. Bilingual children will hear the same object being referred to by at least two different labels in each of their two languages (e.g., ‘ball’ in English and ‘palla’ in Italian) and will therefore need to overcome this bias and accept that the same referent can be identified by more than one label. Research on the composition of bilingual children’s early expressive vocabulary has provided evidence for the existence of translation equivalents, i.e., words in different languages that identify the same referent (e.g. ‘ball’/‘palla’) (Holowka, Brosseau-Lapré, and

The Bilingual Child  95 Petitto 2002; Junker and Stockman 2002; Pearson, Fernández, and Oller 1995; Quay 1995; Yip and Matthews 2007). Similar findings have also emerged for the presence of translation equivalents in bilingual toddlers’ receptive vocabulary (De Houwer, Bornstein, and de Coster 2006). More interestingly, Houston-Price, Caloghiris, and Raviglione (2010) showed that bilinguals between the ages of 17 and 22 months overcome the mutual exclusivity bias not only across languages but within the same language. In a preferential looking experiment Houston-Price et al. (2010) demonstrated that monolingual infants clearly relied on mutual exclusivity by preferentially fixating on a novel object, as opposed to a known object, when they heard ‘Look at the dax.’ In contrast, bilinguals matched in age and vocabulary size in English did not show this preferential looking pattern. As a result of their experience of hearing multiple labels for the same referent in their input, bilinguals fail to follow the mutual exclusivity strategy; they do not make the assumption that a new label will necessarily refer to a new object. Noncompliance with the mutual exclusivity constraints has an obvious adaptive advantage in the case of bilingual children as it allows them to build up a parallel lexicon in their two languages. In addition to the composition of early bilingual vocabulary, another issue that has preoccupied researchers in the field for a long time is the size of the lexicon, both in terms of how it compares across languages and to monolingual norms. Correlations between the size of bilingual children’s early vocabulary in each language is usually only moderate as most children tend to have more words in one language than in the other (Pearson, Fernández, and Oller 1993). The quantity (Marchman, Fernald, and Hurtado 2010; Patterson 2002; Pearson et al. 1997) and the quality of the input (Quiroz, Snow, and Zhao 2010) that bilingual children receive in their two languages have been shown to be predictive of the size and of the composition of their vocabularies. A typical finding in the literature is that bilingual children tend to have smaller receptive vocabularies than age-matched monolingual children in each of their languages (Nicoladis 2003a, 2006; Nicoladis and Genesee 1996; Pearson 1998; Pearson, Fernández, and Oller 1993). Most of this evidence is based on standardized measures, usually collected in only one of the languages (typically the Peabody Picture Vocabulary Test, PPVT; for English). In an aggregate analysis of 772 English-speaking monolinguals and 966 bilinguals between the ages of 3 and 10 Bialystok et al. (2010) reported that bilingual children’s scores on the PPVT were indeed significantly lower than monolinguals’ at all ages, although both groups were within the normal range. A further analysis in which items on the PPVT were divided into ‘home’ and ‘school’ vocabulary revealed that the school-age monolinguals obtained significantly higher scores only on ‘home’ items and that no significant differences existed in the school-related vocabulary where the bilingual children were receiving the same quantity and quality of input as the monolingual children. Because Bialystok et al. (2010) only analyzed data from English, the children’s language of instruction, we have no way of knowing whether children knew the ‘home’ items in their home language too; data from other studies suggest that this is indeed likely (Oller and Pearson 2002).

96  Neurological and Psychological Aspects of Bilingualism and Multilingualism In terms of the composition of bilinguals’ vocabularies there are words that children know in both languages (translation equivalents) and words that they only know in one of the two languages, most likely because they only have the opportunity and motivation to learn them in one of their languages. Over time the proportion of singlets (i.e., words that are known only in one language) is bound to decrease as bilinguals increasingly find themselves in situations in which they can and need to learn translation equivalents. Pearson et al. (1999) report that the proportion of singlets in the Miami Project (Oller and Eilers 2002) was around 70% at the age of 2, decreased to 40% by primary school age, and went further down to below 20% by college age. A 100% overlap between a bilingual’s vocabularies is almost unthinkable and the partial overlap that is typically found is a natural consequence of the distributed characteristic of the bilingual experience (Oller and Pearson 2002; Oller, Pearson, and Cobo-Lewis 2007). A more useful way of considering lexical knowledge in bilinguals is to think in terms of total conceptual vocabulary. Pearson and colleagues (Pearson 1998; Pearson et al. 1997; Pearson, Fernández, and Oller 1993, 1995) have proposed that, although bilingual children may have depressed vocabularies compared to agematched monolinguals, when one pools words from both languages the coverage of the semantic space (total conceptual vocabulary) is similar regardless of language background, and possibly larger for bilinguals. Although information is typically collected on children’s receptive vocabulary and in one language only (most often English), those studies that have assessed both receptive and expressive lexical skills (Oller et al. 2007; Yan and Nicoladis 2009) have reported a moderate disadvantage for expressive abilities, with bilinguals lagging further behind monolinguals in expressive vocabulary tests and showing more difficulties in lexical access than monolinguals in the context of naming tasks (Yan and Nicoladis 2009).

Morphology, Syntax, Language Differentiation, and Cross-Linguistic Influence Research on monolinguals has shown a tight relationship between the development of early lexical and grammatical skills over and beyond amount of language exposure and general cognitive abilities (Dale, Dionne, Eley, and Plomin 2000; Dionne, Dale, Boivin, and Plomin 2003). A significant positive correlation between grammar and the lexicon in the early stages of development (18–30 months) has now been demonstrated across a variety of languages for monolingual children. A few studies have started investigating the issue in bilingual populations showing that indeed the link holds, and that children’s early developing grammatical abilities in one language are not predicted by general cognitive abilities, maturational state or even amount of input, but by the size of the lexicon in that language. In a study of 113 simultaneous English–Spanish bilinguals between the ages of 17 and 30 months Marchman, Martínez-Sussmann, and Dale (2004) found

The Bilingual Child  97 that within-language vocabulary–grammar associations were stronger than those between languages even after partialing out the effects of age, amount of input, and general language abilities. Grammatical skills in Spanish were predicted by Spanish vocabulary and grammatical skills in English were predicted by English vocabulary. Similar results were obtained in a longitudinal study of English– Spanish bilinguals between 19 and 30 months by Conboy and Thal (2006). In their sample the proportion of different word types (nouns, predicates, closed class words, social words) in each language was correlated with the size of the lexicon in that language. Changes over time in the proportion of predicates and closed class word types (i.e., of those word types that correlate with the emergence of grammar in multiword utterances) were also predicted exclusively by samelanguage vocabulary and not by total conceptual vocabulary pooling words from both languages. Gains in the length and in the complexity of children’s utterances in English were significantly correlated with increases in English vocabulary size and not with total conceptual vocabulary. However, for the Spanish data there was a significant correlation between utterance length and Spanish vocabulary but also total conceptual vocabulary. Simon-Cereijido and Gutiérrez-Clellen (2009) similarly reported positive within-language correlations between vocabulary and grammar results for a group of 196 Latino 5-year-olds, but failed to find any between-language correlations. The findings from Marchman et al. (2004), Conboy and Thal (2006), and SimonCereijido and Gutiérrez-Clellen (2009) are in line with the hypothesis that simultaneous bilingual children treat their two languages as independent and largely autonomous systems, a view that has now been the majority consensus in the field for the last 20 years (De Houwer 1990; Genesee 1989; Meisel 1989). At the same time the Spanish utterance length data in Conboy and Thal (2006) suggest a degree of interaction across the two languages whereby the words that children learn in English have an effect on the length of their utterances in Spanish. Although the sample size was small and the significance of this result must be treated with some caution, Conboy and Thal speculate that this might constitute evidence for some form of cross-linguistic bootstrapping of the kind that has been observed elsewhere in the literature, a pooling of resources across languages to fill the gap in one of the languages (Gawlitzek-Maiwald and Tracy 1996). The authors did not conduct further qualitatively analyses but they concede that the English words that were responsible for this effect might be what Ninio (1999) called path-breaking verbs, i.e., semantically light verbs that children rely on to break into argument structure. Once children have some schematic representation of a construction in English they might apply this knowledge to Spanish thus increasing complexity and utterance length. A potential corollary of this hypothesis is that, although utterances in Spanish were longer than would have been predicted by the size of their Spanish vocabulary alone, they may also be modeled a little too closely on English and result in nontarget grammatical constructions. For example, as Conboy and Thal suggest, the higher cue validity of the VO ordering in English where the complement always follows the verb, regardless of its nominal or pronominal status (e.g., I want that toy/I want it) might lead children

98  Neurological and Psychological Aspects of Bilingualism and Multilingualism to incorrectly apply the same strategy in Spanish where word order depends on the status of the referential expression (e.g. Quiero ese jugete/Lo quiero, ‘I want this toy/I want it’). The pervasiveness of overt subjects in English may also have increased the frequency with which these bilingual children used an overt subject in Spanish, a null-subject language where the majority of sentential subjects are phonologically null. The status of weak or nonexistent between-language ties that have been reported between vocabulary and grammar in the early stages of acquisition may well change over the course of development and more longitudinal studies are needed to confirm whether this pattern changes over time. Moreover the pattern may be different in the case of sequential bilinguals where L1 semantic and grammatical skills have been shown to be predictive of the same set of skills in the L2 (Castilla, Restrepo, and Pérez-Leroux 2009). The extent to which the two independent grammatical systems of a bilingual child may actually interact has been the focus of much investigation since the late 1990s after more and more evidence had been gathered in support of language differentiation in simultaneous bilingualism. Hulk and Müller (2000) and Müller and Hulk (2001) originally formulated the cross-linguistic influence hypothesis whereby children may transfer the use of a grammatical construction from language A to language B if the construction in question is at the interface between two modules of grammar (syntax and pragmatics), and if the two languages overlap at the surface level. Since Hulk and Müller’s seminal work, the nature of the cross-linguistic influence has been extensively investigated in a variety of language combinations and in a range of constructions and further refined in terms of its domain of applicability and directionality. The attested instances of cross-linguistic influence fall into three broad categories: (a) higher omission rates than those observed in monolingual acquisition (e.g., higher rate of omissions in the Italian of a German–Italian bilingual child due to the pervasiveness of null topicalized objects in German); (b) use of a construction that is available in language A in contexts in which it is not semantically or pragmatically appropriate in language B (e.g., overproduction of overt pronominal subjects in the Spanish of an English–Spanish child because of the obligatoriness of overt subjects in English); (c) use of a construction in language A that is unattested in monolingual acquisition (e.g., the use of pre-nominal relative clauses in English, a language with post-nominal relative clauses, as the effect of transfer from Chinese). Evidence for some form of cross-linguistic influence has been reported in the domain of pronominal subjects (Argyri and Sorace 2007; Hacohen and Schaeffer, 2007; Paradis and Navarro 2003; Serratrice 2007; Sorace et al. 2009; Serratrice, Sorace, and Paoli 2004), pronominal objects (Paradis, Crago, and Genesee 2005/2006; Pérez-Leroux, Pirvulescu, and Roberge 2009; Serratrice et al. 2012), determiners (Bernardini 2003; Kupisch 2007; Serratrice et al. 2009), adjective placement (Nicoladis 2006), nominal compounds (Nicoladis 2003b), finiteness (Blom 2010), copula (Fernández-Fuertes and Liceras 2010; SilvaCorvalán and Montanari 2008), argument structure (Chan 2010; Chenjie Gu 2010), wh-questions (Yip and Matthews 2000) among others.

The Bilingual Child  99 Despite the evidence showing that a degree of interaction is indeed attested in the morpho-syntactic development of bilingual children, relatively little is yet known about the determinants of cross-linguistic influence in terms of language dominance, typological differences between the languages, development over time, and whether it is indeed only constructions at the interfaces that are vulnerable. Future research will need to address these issues to refine our understanding of cross-linguistic influence and on how this notion may be similar and/or different to the notion of transfer in older bilingual speakers where an L1 is already established.

Cognitive Correlates of Bilingualism The daily experience of dealing with two languages from infancy requires children to pay constant selective attention to the input as any incoming sound and word needs to be tagged according to the language to which it belongs. In production a similar mechanism of selection of the target language and suppression of the nontarget language must operate. The available evidence shows that from the second half of the first year of life bilingual children are starting to tune in to the language-specific properties of their ambient languages and that, similarly to monolinguals, they rely on prosodic, phonetic, and phonotactic information for successful language differentiation. Although the bilinguals’ path to word learning may be somewhat different from monolinguals’ (see for example the nonapplication of the mutual exclusivity constraint), bilingual children are nevertheless successful word learners. Their first words appear around the same time as for monolinguals (Genesee 2003), although their receptive and expressive vocabularies in each language are somewhat smaller compared to monolingual peers, they tend to be in the normal range (Bialystok et al. 2010), and their total conceptual vocabulary may be greater than monolinguals’. From the point of view of morphological and syntactic development bilinguals have been shown to be remarkably similar to monolingual children but they also exhibit characteristics that are peculiar to their experience of dealing with two languages. Regular exposure and use of two languages has obvious implications for mental representations and processing that have been of interest to developmental linguists and psycholinguists for some time. More recently, psychologists have started to investigate the relationship between language and cognition, in particular executive function skills, in bilinguals from infancy into adulthood and old age (Bialystok and Craik 2010). The executive function network comprises processes dedicated to planning and adjusting goals on the basis of the context requirements through managing conflicting demands and stimuli, ignoring irrelevant or distracting information, and inhibiting a dominant response. The findings that are emerging from this line of research indicate a significant role of early bilingual experience on the performance of some, but not all, executive function tasks. In a meta-analysis of 63 studies

100  Neurological and Psychological Aspects of Bilingualism and Multilingualism investigating the relationship between bilingualism and cognition, Adesope et al. (2010) reported a moderate positive overall effect of bilingualism on a number of cognitive measures such as abstract and symbolic representation, attentional control, and problem solving, with the greatest advantages for early bilinguals. In a study of young adult bilinguals, Hernández et al. (2010) refined our current understanding of the bilingual advantage in terms of attentional control. Their two experiments confirmed the results of previous research showing an advantage for bilinguals with respect to inhibitory control; at the same time they did not find significant differences in terms of orienting processes as a function of language background, thus showing that attentional control systems are selectively affected by the bilingual experience. With specific reference to the developmental literature the relationship between language and executive functions skills in preschoolers and school-age children has now been mapped out in some detail. The more consistent finding is that the benefits of bilingualism are rather task-dependent; young bilinguals tend to outperform their monolingual peers in tasks that tap into the management of conflicting attention demands, but not in tasks that require impulse control (Carlson and Meltzoff 2008). In essence, bilingual children are better than monolinguals at resolving perceptual conflict and at using nonsalient cues to attend to the task at hand. They can switch more easily than monolingual children between reliance on one perceptual cue (e.g., color) to another competing perceptual cue (e.g., shape) in sorting stimuli that vary along these perceptual dimensions. For instance, Chinese–English bilinguals between the ages of 3 and 6 matched to monolingual English-speaking peers on measures of receptive vocabulary and working memory performed significantly better in two tasks (the moving word task and the dimensional change card sort [DCCS] task, Zelazo, Reznick, and Pinon 1995) that require the resolution of a conflict (Bialystok 1999). In the moving word task children are presented with two pictures of two familiar referents (e.g., a king and a tree), after naming the pictures the experimenter produces a printed card with the name of one of the pictures (e.g. king), tells the child what the card says and places it under the named referent, the child is then asked to repeat what the card says. In the course of the game the card is misplaced and ends up below the other picture. The child is asked again what the card says and the same procedure is repeated a third time after the card appears for the last time in the original position below the first-named referent. If children recognize that the printed word has a symbolic meaning (Apperly, Williams, and Williams 2004) they should give the same answer on all three occasions (e.g., king), failure to do so signals that they have not yet developed this kind of understanding. Success in this task depends on children’s ability to inhibit a response whereby the printed card names whatever it is placed next to, children must persevere with the original response. In the DCCS task children are given a deck of cards and are asked to start sorting them into two piles according to the shape of the picture printed on the card. After they have sorted the cards using this rule the cards are shuffled and children are asked to sort the cards again, but this time not according to shape

The Bilingual Child  101 but to color. Unlike in the moving word task, where persevering with the original response was the right thing to do, to succeed in the DCCS task children have to inhibit their original response and switch to a new one. The rules of the games have changed and the children have to adapt to this change by ignoring the first rule and by following the new set of instructions. Importantly, the bilingual children’s performance in these two tasks shows that their superior cognitive flexibility relies both on perseverance in the use of an association – when it is symbolic and therefore stable (moving word task) – and in the ability to switch when the association is contingent on the current rule of the game (DCCS task). The DCCS consists of a bivalent display where two features (color and shape) coexist; success on the second sorting task requires participants to pay attention to a new feature (e.g. shape) and suppress interference from the previous feature (e.g. color) that is still present in the stimulus. Response inhibition, as opposed to interference suppression, is required in the contexts of univalent displays where an arbitrary response must override a familiar, entrenched response. In a series of experiments using adaptations of the Simon task (Simon and Craft 1970) and of the Stroop task (Stroop 1935) with bilingual 5- to 8-year-olds, MartinRhee and Bialystok (2008) reported an advantage for bilingual children in the ambivalent display of the Simon task where children were asked to press the red key on the computer keyboard when a red dot appeared on the screen, and the blue key when a blue dot was displayed. The red and blue keys were on the far left and on the far right of the keyboard, respectively. On congruent trials the position of the key and the position of the dot on the screen overlapped (e.g., both on the left), on incongruent trials the red dot appeared on the right side of the screen but the child had to press the key on the far left of the keyboard. The spatial cue (position of the dot and position of the key) conflicted on incongruent trials and the child had to pay attention to the color feature to succeed on the task. The bilingual advantage was, however, significant only in a condition in which children had to respond immediately to the stimulus appearing on the screen; in two conditions in which an interval of 500 ms or 1000 ms elapsed between stimulus presentation and response the monolinguals performed just as accurately as the bilinguals. In the same study Martin-Rhee and Bialystok also tested bilingual and monolingual 8-year-olds on a verbal and a nonverbal adaptation of the day– night Stroop task. In the verbal adaptation they were presented with two pictures of familiar entities (e.g., cats and dogs) and they simply had to label the picture using the other name (i.e. say dog when presented with a picture of a cat). In a second study the Simon task was adapted to include both a typical bivalent display using arrows appearing on the left or the right of the screen instead of dots, and a univalent condition in which an arrow appeared in the center of the screen and the child had to press the key on the opposite direction of the arrow (i.e. if the arrow was pointing to the left the child had to press the right key). Neither in the verbal, nor in the nonverbal univalent display tasks did the bilinguals show an advantage over the monolinguals. These results support the hypothesis that different aspects of inhibitory control are selectively affected by the experience of using two languages on a regular basis. Another notable

102  Neurological and Psychological Aspects of Bilingualism and Multilingualism finding of this study with child bilinguals was that these children were faster than the monolinguals in the Simon task experiments in both congruent and incongruent trials. This result suggests that there may be additional advantages of childhood bilingualism in addition to superior inhibitory control abilities in interference suppression. In a study with 6-year-old bilinguals and monolinguals Bialystok (2010) also reported a consistent bilingual advantage in three experiments including two tasks requiring interference suppression: a global–local task and the trail-making task (TMT). In the global–local task the stimuli are letters made up of smaller letters, in congruent stimuli the global level and the local level overlap (i.e., a big S letter is made up of smaller S letters), in incongruent stimuli the global and local level conflict (e.g., a big S letter made up of smaller H letters). In the trail-making task there are two conditions, in Trail A children have to join up numbers scattered on a page in an orderly ascending sequence (i.e., 1, 2, 3, etc.), in Trail B numbers and letters have to be ordered into an ascending cardinal and alphabetic sequence (i.e., 1A2B3C4D, etc.). Once again the bilinguals performed better than the monolinguals not only on the incongruent trials, where an inhibitory response was required, but also on the congruent trials. Speed of processing alone was ruled out as a potential explanatory variable as in two control experiments where the task demands were decreased there were no significant differences in reaction times as a function of language background. Bialystok’s argument is that even in the congruent trials, where no inhibition was required, children were faced with demands on the executive control aspects of monitoring and shifting. More specifically, the bilingual advantage on congruent trials was only found in mixed blocks in the global–local task where the responses alternated between global and local. Monolinguals experienced larger mixing costs than bilinguals showing that they found it more difficult to switch between the global and the local perceptual level. In the numbers-only Trial A condition in the TMT task children are not required to shift between numbers and letters, but in order to complete the sequence they have to monitor the stage at which they are at in the ordered sequence and find the next number while ignoring other distracting elements. These results open up the possibility that the effects of bilingualism are not confined to an advantage in inhibitory control due to early and regular practice with the inhibition of one language. The precise nature and the selectivity of the bilingual advantage are still in need of more targeted work.

More to Bilingual Acquisition: Similarities, Differences, Challenges, and Opportunities Learning two languages in childhood presents both similarities and differences to growing up with only one language. The literature on infant speech perception has shown that bilingual infants are capable of subtle distinctions both within and

The Bilingual Child  103 across languages, and that sustained exposure to more than one language protects them from losing those innate perceptual abilities that are relevant for speech discrimination in their two languages. At the same time speech production may be significantly different in young bilinguals both in terms of the size and accuracy of their phonetic repertoires and their phonological representations. Capitalizing on their receptive skills in two languages bilingual children build parallel vocabularies where translation equivalents are present from early on showing a degree of flexibility that may be somewhat superior to monolingual children’s. Over time their lexicon expands to include an increasingly larger proportion of translation equivalents although the distributed nature of the bilingual experience will inevitably lead to the existence of lexical pockets of language specificity where children have neither the opportunity nor the need to acquire the word in the other language. Ties between lexical and morpho-syntactic development are relatively strong in bilingual acquisition as they are in monolingual acquisition, but they are weak or nonexistent across languages inasmuch as vocabulary size in one language is not a strong predictor of grammatical skills in the other language. Correlations between the lexicon and grammar seem to be limited to the within-language context. In terms of milestones of morpho-syntactic development once again bilingual children follow the same developmental path of monolingual peers although use and knowledge of two languages may lead to systematic, but circumscribed, instances of cross-linguistic influence, particularly in phenomena that are at the interface between syntax and discourse pragmatics and/or semantics. The extent to which cross-linguistic influence extends to non-interface syntactic structures, and the precise determinants of the likelihood and of the magnitude of the phenomenon are yet poorly understood; more research assessing the role of typological distance, input, and individual differences will be necessary to shed light on this issue. Another aspect of the bilingual experience that sets young learners apart from their monolingual peers is the constant switching – in comprehension and in production – between two languages; bilinguals who regularly use two languages must routinely inhibit one when using the other. This inhibition process has repercussions both in terms of processing, and more in general in terms of inhibitory control. On the one hand we know from the adult psycholinguistic literature that there are processing costs associated with bilingual language use that affect both accuracy and speed. On the other hand there is also evidence that this processing, although costly, does confer young bilinguals a certain advantage in some tasks tapping into inhibitory control; once again being bilingual presents challenges but also affords opportunities that are not available to children who grow up with only one language. This selective overview of bilingual acquisition in early childhood has addressed some of the key topics of current research. Inevitably many other areas of interest in the field could not be covered. I direct the reader to research covering the linguistic and pragmatic skills involved in code-switching (Cantone 2007; Lanza 1997); the relationship between bilingualism and language impairment (Paradis

104  Neurological and Psychological Aspects of Bilingualism and Multilingualism 2010 and commentaries); similarities between early second-language acquisition and simultaneous bilingualism (Meisel 2008, 2010) to name but a few.

REFERENCES Adesope, Olusola, Lavin, Tracy, Thompson, Terri, and Ungerleider, Charles (2010). A systematic review and meta-analysis of the cognitive correlates of bilingualism. Review of Educational Research 80: 207–45. Albareda-Castellot, Bàrbara, Pons, Ferran, and Sebastián-Gallés, Núria (2011). The acquisition of phonetic categories in bilingual infants: New data from an anticipatory eye movement paradigm. Developmental Science 14: 395–401. Apperly, Ian, Williams, Emily, and Williams, Joelle (2004). Three-to four-year-olds’ recognition that symbols have a stable meaning: Pictures are understood before written words. Child Development 75: 1510–22. Argyri, Efrosyni and Sorace, Antonella (2007). Crosslinguistic influence and language dominance in older bilingual children. Bilingualism: Language and Cognition 10: 79–99. Aslin, Richard and McMurray, Bob (2004). Automated corneal-reflection eye tracking in infancy: Methodological developments and applications to cognition. Infancy 6: 155–63. Bernardini, Petra (2003). Child and adult acquisition of word order in the Italian DP. In Natascha Müller (ed.), (In)vulnerable Domains in Multilingualism. 41–81. Amsterdam: John Benjamins. Bialystok, Ellen (1999). Cognitive complexity and attentional control in the bilingual mind. Child Development 70: 636–44. Bialystok, Ellen (2010). Global-local and trail-making tasks by monolingual and bilingual children: Beyond inhibition. Developmental Psychology 46: 93–105. Bialystok, Ellen and Craik, Fergus (2010). Cognitive and linguistic processing in the bilingual mind. Current Directions in Psychological Science 19: 19–23. Bialystok, Ellen, Luk, Gigi, Peets, Kathleen, and Yang, Sujin (2010). Receptive vocabulary differences in monolingual and bilingual children. Bilingualism: Language and Cognition 13: 525–31.

Blom, Elma (2010). Effects of input on the early grammatical development of bilingual children. International Journal of Bilingualism 14: 422–46. Bosch, Laura and Sebastián-Gallés, Núria (1997). Native-language recognition abilities in 4-month-old infants from monolingual and bilingual environments. Cognition 65: 33–69. Bosch, Laura and Sebastián-Gallés, Núria (2001). Evidence of early language discrimination abilities in infants from bilingual environments. Infancy 2: 29–49. Bosch, Laura and Sebastián-Gallés, Núria (2003). Simultaneous bilingualism and the perception of a language-specific vowel contrast in the first year of life. Language and Speech 46: 217–43. Bosch, Laura, Costa, Albert, and SebastiánGallés, Núria (2000). First and second language vowel perception in early bilinguals. European Journal of Cognitive Psychology 12: 189–221. Burns, Tracey, Yoshida, Katherine, Hill, Karen, and Werker, Janet (2007). The development of phonetic representation in bilingual and monolingual infants. Applied Psycholinguistics 28: 455–74. Byers-Heinlein, Krista, Burns, Tracey, and Werker, Janet (2010). The roots of bilingualism in newborns. Psychological Science 21: 343–8. Cantone, Katja (2007). Code-Switching in Bilingual Children. Dordrecht: Springer. Carlson, Stephanie and Meltzoff, Andrew (2008). Bilingual experience and executive functioning in young children. Developmental Science 11: 282–98. Castilla, Anny Patricia, Restrepo, Maria Adelaida, and Pérez-Leroux, Ana Teresa (2009). Individual differences and language interdependence: A study of sequential bilingual development in Spanish–English preschool children. International Journal of Bilingual Education and Bilingualism 12: 565–80. Chan, Angel (2010). The Cantonese double object construction with Bei2 ‘give’ in

The Bilingual Child  105 bilingual children: The role of input. International Journal of Bilingualism 14: 65–85. Chenjie Gu, Chloe (2010). Crosslinguistic influence in two directions: The acquisition of dative constructions in Cantonese–English bilingual children. International Journal of Bilingualism 14: 87–103. Conboy, Barbara and Thal, Donna (2006). Ties between the lexicon and grammar: Crosssectional and longitudinal studies of bilingual toddlers. Child Development 77: 712–35. Dale, Philip, Dionne, Ginette, Eley, Thalia, and Plomin, Robert (2000). Lexical and grammatical development: A behavioural genetic perspective. Journal of Child Language 27: 619–42. De Houwer, Annick (1990). The Acquisition of Two Languages from Birth: A Case Study. Cambridge: Cambridge University Press. De Houwer, Annick, Bornstein, Marc, and de Coster, Sandrine (2006). Early understanding of two words for the same thing: A CDI study of lexical comprehension in infant bilinguals. International Journal of Bilingualism 10: 331–48. Dionne, Ginette, Dale, Philip S., Boivin, Michel, and Plomin, Robert (2003). Genetic evidence for bidirectional effects of early lexical and grammatical development. Child Development 74: 394–412. Fabiano-Smith, Leah and Barlow, Jessica (2010). Interaction in bilingual phonological acquisition: Evidence from phonetic inventories. International Journal of Bilingual Education and Bilingualism 13: 81–97. Fabiano-Smith, Leah and Goldstein, Brian (2010a.) Phonological acquisition in bilingual Spanish–English speaking children. Journal of Speech, Language, and Hearing Research 53: 160–78. Fabiano-Smith, Leah and Goldstein, Brian (2010b). Early-, middle and, late-developing sounds in monolingual and bilingual children: An exploratory investigation. American Journal of Speech and Language Pathology 19: 66–77. Fennell, Christopher, Byers-Heinlein, Krista, and Werker, Janet (2007). Using speech sounds to guide word learning: The case of bilingual infants. Child Development 78: 1510–25. Fernández-Fuertes, Raquel and Liceras, Juana (2010). Copula omission in the English developing grammar of English/Spanish bilingual children. International Journal of Bilingual Education and Bilingualism 13: 525–51.

Friederici, Angela and Wessels, Janine (1993). Phonotactic knowledge of word boundaries and its use in infant speech perception. Perception and Psychophysics 54: 287–95. Gawlitzek-Maiwald, Ira and Tracy, Rosemarie (1996). Bilingual bootstrapping. Linguistics 34: 901–26. Genesee, Fred (1989). Early bilingual development: One language or two? Journal of Child Language 16: 161–79. Genesee, Fred (2003). Educating Second Language Children: The Whole Child, the Whole Curriculum, the Whole Community. Cambridge: Cambridge University Press. Hacohen, Aviya and Schaeffer, Jeannette (2007). Subject realization in early Hebrew/English bilingual acquisition: The role of crosslinguistic influence. Bilingualism: Language and Cognition 10: 333–44. Halberda, Justin (2003). The development of a word-learning strategy. Cognition 87: 23–34. Hernández, Mireia, Costa, Albert, SebastiánGallés, Núria, Fuentes, Luis, and Vivas, Ana (2010). The impact of bilingualism on the executive control and orienting networks of attention. Bilingualism: Language and Cognition 13: 315–25. Holowka, Siobhan, Brosseau-Lapré, Françoise, and Petitto, Laura (2002). Semantic and conceptual knowledge underlying bilingual babies’ first signs and words. Language Learning 52: 205–62. Houston-Price, Carmel, Caloghiris, Zoe, and Raviglione, Eleonora (2010). Language experience shapes the development of the mutual exclusivity bias. Infancy 15: 125–50. Hulk, Aafke and Müller, Natascha (2000). Bilingual first language acquisition at the interface between syntax and pragmatics. Bilingualism: Language and Cognition 3: 227–44. Junker, Dörte and Stockman, Ida (2002). Expressive vocabulary of German–English bilingual toddlers. American Journal of Speech Language Pathology 11: 381–94. Jusczyk, Peter, Friederici, Angela, Wessels, Janine, Svenkerud, Vigdis, and Jusczyk, Anne Marie (1993). Infants’ sensitivity to the sound patterns of native language words. Journal of Memory and Language 32: 402–20. Kupisch, Tanja (2007). Determiners in bilingual German–Italian children: What they tell us about the relation between language influence and language dominance. Bilingualism: Language and Cognition 10: 57–78.

106  Neurological and Psychological Aspects of Bilingualism and Multilingualism Lanza, Elisabeth (1997). Language Mixing in Infant Bilingualism: A Sociolinguistic Perspective. Oxford: Clarendon Press. Leopold, Werner F. (1939). Speech Development of a Bilingual Child: A Linguist’s Record. Evanston, IL: Northwestern University. Lleó, Conxita (2002). The role of markedness in the acquisition of complex prosodic structures by German-Spanish bilinguals. International Journal of Bilingualism 6: 291–313. Lleó, Conxita and Kehoe, Margaret (2002). On the interaction of phonological systems in child bilingual acquisition. International Journal of Bilingualism 6: 233–7. Lleó, Conxita and Prinz, Michael (1996). Consonant clusters in child phonology and the directionality of syllable structure assignment. Journal of Child Language 23: 31–56. Macken, Marlys and Ferguson, Charles (1981). Phonological universals in language acquisition. In Harry Winitz (ed.), Native Language and Foreign Language Acquisition, Vol. 379. 110–29. New York: The New York Academy of Sciences. Marchman, Virginia, Fernald, Anne, and Hurtado, Nereyda (2010). How vocabulary size in two languages relates to efficiency in spoken word recognition by young Spanish– English bilinguals. Journal of Child Language 37: 817–40. Marchman, Virginia, Martínez-Sussmann, Carmen, and Dale, Philip (2004). The language-specific nature of grammatical development: Evidence from bilingual language learners. Developmental Science 7: 212–24. Markman, Ellen (1989). Categorization and Naming in Children: Problems of Induction. Cambridge, MA: MIT Press. Markman, Ellen and Wachtel, Gwyn (1988). Children’s use of mutual exclusivity to constrain the meaning of words. Cognitive Psychology 20: 121–57. Martin-Rhee, Michelle and Bialystok, Ellen (2008). The development of two types of inhibitory control in monolingual and bilingual children. Bilingualism: Language and Cognition 11: 81–93. Mather, Emily and Plunkett, Kim (2009). Learning words over time: The role of stimulus repetition in mutual exclusivity. Infancy 14: 60–76. Mattock, Karen, Polka, Linda, Rvachew, Susan, and Krehm, Madelaine (2010). The first steps

in word learning are easier when the shoes fit: Comparing monolingual and bilingual infants. Developmental Science 13: 229–43. Meisel, Jürgen (1989). Early differentiation of language in bilingual children. In Kenneth Hyltenstam and Loraine Obler (eds.), Bilingualism across a Lifespan: Aspects of Acquisition, Maturity and Loss. 13–40. Cambridge: Cambridge University Press. Meisel, Jürgen (2008). Child second language acquisition or successive first language acquisition? In Belma Haznedar and Elena Gavruseva (eds.), Current Trends in Child Second Language Acquisition. 55–80. Amsterdam: John Benjamins. Meisel, Jürgen (2010). Age of onset in successive acquisition of bilingualism: Effects on grammatical development. In Michèle Kail and Maya Hickmann (eds.), Language Acquisition across Linguistic and Cognitive Systems. 225–47. Amsterdam: John Benjamins. Merriman, William and Bowman, Laura (1989). Mutual Exclusivity Bias in Children’s Word Learning. Chicago: University of Chicago Press. Müller, Natascha and Hulk, Aafke (2001). Crosslinguistic influence in bilingual language acquisition: Italian and French as recipient languages. Bilingualism: Language and Cognition 4: 1–53. Nazzi, Thierry, Bertoncini, Josiane, and Mehler, Jacques (1998). Language discrimination by newborns: Toward an understanding of the role of rhythm. Journal of Experimental Psychology Human Perception and Performance 24: 756–66. Nazzi, Thierry, Jusczyk, Peter, and Johnson, Elizabeth (2000). Language discrimination by English-learning 5-month-olds: Effects of rhythm and familiarity. Journal of Memory and Language 43: 1–19. Nicoladis, Elena (2003a). Cross-linguistic transfer in deverbal compounds of preschool bilingual children. Bilingualism: Language and Cognition 6: 17–31. Nicoladis, Elena (2003b). What compound nouns mean to preschool children. Brain and Language 84: 38–49. Nicoladis, Elena (2006). Cross-linguistic transfer in adjective-noun strings by preschool bilingual children. Bilingualism: Language and Cognition 9: 15–32. Nicoladis, Elena and Genesee, Fred (1996). A longitudinal study of pragmatic

The Bilingual Child  107 differentiation in young bilingual children. Language Learning 46: 439–64. Ninio, Anat (1999). Model learning in syntactic development: Intransitive verbs. International Journal of Bilingualism 3: 111–31. Oller, D. Kimborough and Eilers, Rebecca (2002). Language and Literacy in Bilingual Children. Clevedon: Multilingual Matters. Oller, D. Kimborough and Pearson, Barbara (2002). Assessing the effects of bilingualism: A background. In D. Kimborough Oller and Rebecca Eilers (eds.), Language and Literacy Development in Bilingual Children. 3–21. Clevedon: Multilingual Matters. Oller, D. Kimborough, Pearson, Barbara, and Cobo-Lewis, Alan (2007). Profile effects in early bilingual language and literacy. Applied Psycholinguistics 28: 191–230. Paradis, Johanne (2001). Do bilingual two-yearolds have separate phonological systems? International Journal of Bilingualism 5: 19–38. Paradis, Johanne (2010). The interface between bilingual development and specific language impairment. Applied Psycholinguistics 31: 227–52. Paradis, Johanne, Crago, Martha, and Genesee, Fred (2005/2006). Domain-general versus domain-specific accounts of Specific Language Impairment: Evidence from bilingual children’s acquisition of object pronouns. Language Acquisition 13: 33–62. Paradis, Johanne and Navarro, Samuel (2003). Subject realization and crosslinguistic interference in the bilingual acquisition of Spanish and English: What is the role of the input? Journal of Child Language 30: 371–93. Patterson, Janet (2002). Relationships of expressive vocabulary to frequency of reading and television experience among bilingual toddlers. Applied Psycholinguistics 23: 493–508. Pearson, Barbara (1998). Assessing lexical development in bilingual babies and toddlers. International Journal of Bilingualism 2: 347–72. Pearson, Barbara, Fernández, Sylvia, Lewedeg, Vanessa, and Oller, D. Kimbrough (1997). The relation of input factors to lexical learning by bilingual infants. Applied Psycholinguistics 18: 41–58. Pearson, Barbara, Fernández, Sylvia, and Oller, D. Kimbrough (1993). Lexical development in bilingual infants and toddlers: Comparison to monolingual norms. Language Learning 43: 93–120.

Pearson, Barbara, Fernández, Sylvia, and Oller, D. Kimbrough (1995). Cross-language synonyms in the lexicons of bilingual infants: One language or two? Journal of Child Language 22: 345–68. Pearson, Barbara, Umbel, Vivian, Andrews de Flores, Pam, and Cobo-Lewis, Alan (1999). Measuring Cross-Language Vocabulary in Childhood Bilinguals at Different Stages of Development. Austin, TX: Texas Research Symposium on Language Diversity. Pérez-Leroux, Ana Teresa, Pirvulescu, Manuela, and Roberge, Yves (2009). Bilingualism as a window into the language faculty: The acquisition of objects in French-speaking children in bilingual and monolingual contexts. Bilingualism: Language and Cognition 12: 97–112. Quay, Suzanne (1995). Bilingual lexicon: Implications for studies of language choice. Journal of Child Language 22: 369–87. Quiroz, Blanca, Snow, Catherine, and Jing, Zhao (2010). Vocabulary skills of Spanish–English bilinguals: Impact of mother–child language interactions and home language and literacy support. International Journal of Bilingualism 14: 379–99. Ramus, Franck and Mehler, Jacques (1999). Language identification with suprasegmental cues: A study based on speech resynthesis. Journal of the Acoustical Society of America 105: 512–21. Ronjat, Jules (1913). Le développement du langage observé chez un enfant bilingue. Paris: Librairie Ancienne H. Champion. Sebastián-Gallés, Núria (2006). Native-language sensitivities: Evolution in the first year of life. Trends in Cognitive Sciences 10: 239–41. Sebastián-Gallés, Núria and Soto-Faraco, Salvador (1999). Online processing of native and nonnative phonemic contrasts in early bilinguals. Cognition 72: 111–23. Sebastián-Gallés, Núria and Bosch, Laura (2002). Building phonotactic knowledge in bilinguals: Role of early exposure. Journal of Experimental Psychology Human Perception and Performance 28: 974–89. Sebastián-Gallés, Núria and Bosch, Laura (2009). Developmental shift in the discrimination of vowel contrasts in bilingual infants: Is the distributional account all there is to it? Developmental Science 12: 874–87. Serratrice, Ludovica (2007). Cross-linguistic influence in the interpretation of anaphoric

108  Neurological and Psychological Aspects of Bilingualism and Multilingualism and cataphoric pronouns in English–Italian bilingual children. Bilingualism: Language and Cognition 10: 225–38. Serratrice, Ludovica, Sorace, Antonella, Filiaci, Francesca, and Baldo, Michela (2009). Bilingual children’s sensitivity to specificity and genericity: Evidence from metalinguistic awareness. Bilingualism: Language and Cognition 12: 239–57. Serratrice, Ludovica, Sorace, Antonella, Filiaci, Francesca, and Baldo, Michela (2012). Pronominal objects in English-Italian and Spanish-Italian bilingual children. Applied Psycholinguistics. DOI: 10.1017/ S0142716411000543 Serratrice, Ludovica, Sorace, Antonella, and Paoli, Sandra (2004). Crosslinguistic influence at the syntax-pragmatics interface: Subjects and objects in English-Italian bilingual and monolingual acquisition. Bilingualism: Language and Cognition 7: 183–205. Silva-Corvalán, Carmen and Montanari, Simona (2008). The acquisition of ser and estar by a Spanish–English bilingual child: The early stages. Bilingualism: Language and Cognition 11: 341–60. Simon-Cereijido, Gabriela and Gutiérrez-Clellen, Vera (2009). A cross-linguistic and bilingual evaluation of the interdependence between lexical and grammatical domains. Applied Psycholinguistics 30: 315–37. Simon, Richard and Craft, John (1970). Effects of an irrelevant auditory stimulus on visual choice reaction time. Journal of Experimental Psychology 86: 272–4. Sorace, Antonella, Serratrice, Ludovica, Filiaci, Francesca, and Baldo, Michela (2009). Discourse conditions on subject pronoun realization: Testing the linguistic intuitions of older bilingual children. Lingua 119: 460–77. Stroop, J. Ridley (1935). Studies of Interference in Serial Verbal Reactions. Nashville, TN: George Peabody College for Teachers. Sundara, Megha, Polka, Linda, and Molnar, Monika (2008). Development of coronal stop perception: Bilingual infants keep pace with

their monolingual peers. Cognition 108: 232–42. Sundara, Megha and Scutellaro, Adrienne (2011). Rhythmic distance between languages affects the development of speech perception in bilingual infants. Journal of Phonetics 39: 505–13. Volterra, Virginia and Taeschner, Traute (1978). The acquisition and development of language by bilingual children. Journal of Child Language 5: 311–26. Weikum, Whitney, Vouloumanos, Athena, Navarra, Jordi, Soto-Faraco, Salvador, Sebastián-Gallés, Núria, and Werker, Janet (2007). Visual language discrimination in infancy. Science 316: 1159. Werker, Janet, Byers-Heinlein, Krista, and Fennell, Christopher (2009). Bilingual beginnings to learning words. Philosophical Transactions of the Royal Society of London 364: 3649–63. Werker, Janet, Cohen, Leslie, Lloyd, Valerie, Casasola, Marianella, and Stager, Christine (1998). Acquisition of word-object associations by 14-month-old infants. Developmental Psychology 34: 1289–309. Werker, Janet and Curtin, Suzanne (2005). Primir: A developmental framework of infant speech processing. Language Learning and Development 1: 197–234. Yan, Stephanie and Nicoladis, Elena (2009). Finding le mot juste: Differences between bilingual and monolingual children’s lexical access in comprehension and production. Bilingualism: Language and Cognition 12: 323–35. Yip, Virginia and Matthews, Stephen (2000). Syntactic transfer in a Cantonese–English bilingual child. Bilingualism: Language and Cognition 3: 193–208. Yip, Virginia and Matthews, Stephen (2007). The Bilingual Child: Early Development and Language Contact. Cambridge: Cambridge University Press. Zelazo, Philip, Reznick, Steven, and Pinon, Denise (1995). Response control and the execution of verbal rules. Developmental Psychology 31: 508–17.

5  Bilingualism/ Multilingualism and Second-Language Acquisition YUKO GOTO BUTLER

Introduction Rome in the first century ad had a form of bilingual education and there was a heated debate regarding which language, Greek or Latin, should be introduced first and how they should be introduced (Harris and Taylor 1997). A distinguished teacher of rhetoric at that time, Marcus Fabius Quintilianus, discussed how children should be educated bilingually as follows: I prefer that a boy should begin with Greek, because Latin being in general use, will be picked up by him (perbibet) whether we will or not; while the fact that Latin learning is derived from Greek is a further reason for his being first instructed in the latter. I do not however desire that this principle should be so superstitiously observed that he should for long speak and learn only Greek, as is done in the majority of cases. Such a course gives rise to many faults of language and accent; the latter tends to acquire a foreign intonation, while the former through force of habit becomes impregnated with Greek idioms, which persist with extreme obstinacy even when we are speaking another tongue. The study of Latin ought therefore to follow at no great distance and in a short time proceed side by side with Greek. The result will be that, as soon as we begin to give equal attention to both languages, neither will prove a hindrance to the other (Institutio Oratoria I, i, 12–14, cited in Harris and Taylor 1997: 63).

It is interesting to see that some of the same issues regarding bilingualism that were debated in ancient Rome are still a concern among researchers and educators

The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

110  Neurological and Psychological Aspects of Bilingualism and Multilingualism 2000 years later. We continue to encounter a number of ‘folk beliefs’ about acquiring or learning two or more languages. Two thousand years after Quintilianus, how much do we know about bilingualism and multilingualism? What does it mean to acquire proficiency in more than two languages? As has often been believed to be the case, is it true that ‘native-like’ proficiency cannot be acquired if one reaches a certain age? How does acquiring a second language (L2) at an early age influence one’s development of both languages and cognition? How does knowledge of a first language (L1) influence the development of L2 knowledge? What if a third or fourth language is acquired? The present chapter discusses the key issues at the intersection of bilingualism, multilingualism, and L2 acquisition. These issues have continued to be debated and have become some of the core issues in the field. As we shall see below, growing interest in multilingual acquisition in recent years has challenged some of the existing approaches and premises that heretofore were widely accepted in bilingualism and second-language acquisition (SLA) research. The field of SLA itself has undergone various changes; such changes include increased attention on the role of socio-contextual factors in L2 language acquisition, as well as growing interest in viewing language acquisition as a complex and dynamic system (Larsen-Freeman 2007; Larsen-Freeman and Cameron 2008). I begin by defining bilingualism/multilingualism and looking at the variability and complexity of their characteristics as they are nested in wider social contexts. Next, I discuss the related theoretical, methodological, and applied issues that contribute to our understanding of bilingual/multilingual acquisition. Due to the limited space available, this chapter focuses on the following three major issues in SLA: the construct of language proficiency; the effect of age of exposure to L2 acquisition and additional language acquisition; and cross-linguistic influences among languages. Bilingualism and multilingualism are both individual and group processes. Hamers and Blanc (2000) called individual bilingualism bilinguality and distinguished it from societal bilingualism. While individual and societal bilingualism/ multilingualism are closely connected with each other (e.g., individuals’ attitudes towards a certain language may lead to language shifts within a given community) (Baker 2001), the present chapter limits its discussion to individual bilingualism (i.e., bilinguality) and individual multilingualism (i.e., multilinguality).1

Definitions of Bilinguals and Multilinguals The terms bilingual and multilingual are widely used to refer to individuals who have obtained the ability to use more than one language. However, bilingualism and multilingualism (however they are defined) are highly complex social, psychological, and linguistic phenomena and need to be understood from a multidimensional aspect. Moreover, in recent years, there has been a growing number of researchers who have argued for the importance of making a clear distinction between bilinguals and multilinguals, as opposed to using bilinguals as a blanket

Bilingualism/Multilingualism and Second-Language Acquisition  111 term entailing multilinguals as one variation (e.g., Aronin and Hufeisen 2009; Cenoz and Genesee 1998; De Angelis 2007). In the literature on bilingualism, one can find a wide range of definitions for bilinguals. As is often believed, bilinguals can be defined as individuals who have ‘native-like control of two languages’ (Bloomfield 1933: 56). However, such a narrow definition eliminates the vast majority of people from being categorized as bilinguals, not to mention the difficulty of operationalizing ‘native-like control.’ Haugen defined bilinguals as individuals who are fluent in one language but who ‘can produce complete meaningful utterances in the other language’ (1953: 7). While this definition can include people who have various degrees of language abilities in their L2, it neglects the fact that the ability and use of the dominant language (or L1) can also vary by context and can change over time. Currently, many researchers employ a broader definition of bilinguals that includes individuals who have various degrees of language abilities in different domains in both languages, such as ‘those people who need and use two or more languages (or dialects) in their everyday lives’ (Grosjean 2010: 4). Another example can be seen in the definition of a bilingual as ‘an individual [who] possesses more than one language competence’ (Valdés and Figueroa 1994: 8). These types of broader definitions of bilinguals have an advantage in that they can bring the developmental processes of SLA into the scope of studies of bilingualism (Hakuta 1986). They also correspond better to the recent shift in focus among SLA researchers to communicative skills rather than the acquisition of formal rules of language per se (Mohanty and Perregaux 1997; Purpura 2008). Moreover, a broader definition of bilinguals allows researchers to examine the dynamics of language abilities and language use across contexts and times, which is again better aligned with recent calls for more socially oriented approaches to SLA (Lasen-Freeman 2007). While these broader definitions of bilinguals have a number of advantages, they have also been criticized for implying that bilinguals include any individuals who are not monolinguals. In other words, these definitions ‘effectively equates bilingualism with multilingualism’ (De Angelis 2007: 8). In recent years, there has been a growing body of studies on multilingual language acquisition, and these studies have indicated that there are a number of differences between bilingual and multilingual acquisition as well as similarities. Such differences include the influence of the prior experience of knowing more than one language (as opposed to one) when learning additional language(s), greater cognitive flexibility, higher degrees of metalinguistic awareness, and various strategies used among multilinguals (Aronin and Hufeisen 2009; De Angelis 2007; Hoffmann 2001; Jessner 2008). Due to the limited number of studies on multilingual acquisition/learning to date, especially in the psycholinguistic domain, the magnitude or significance of such differences is still unclear. However, it would appear to be important that we do not blindly assume that multilinguals are the same as bilinguals. In the present chapter, multilanguage users refers to both bilinguals and multilinguals, and they are contrasted with non-multilanguage users (namely, monolinguals). This conceptualization reflects the fact that multilanguage users, not monolinguals, are the norm in the world. Multilanguage users are defined

112  Neurological and Psychological Aspects of Bilingualism and Multilingualism here as individuals or groups of people who obtain communicative competence in more than one language, with various degrees of proficiencies, in oral and/or written forms, in order to interact with speakers of one or more languages in a given society. The term bilingual is used to refer to one type of multilanguage user who uses two languages, whereas multilingual refers to users of more than two languages such as trilinguals, quadrilinguals, and so forth. Although it is important to make a distinction between bilinguals and multilinguals conceptually for the reasons mentioned above, one also has to keep in mind that the boundary between them is not always as clear in reality. To begin with, we don’t have a good consensus about what counts as ‘a language’ (e.g., languages versus dialects or varieties of a given language), and as a result it is frequently very complicated to count the number of languages that multilingual individuals speak (Kemp 2009). Moreover, there appears to be substantial variability in language use and acquisition among multilinguals; potential differences among them may be as big as the differences seen between bilinguals and multilinguals (Kemp 2009). Therefore, the boundary between bilinguals and multilinguals can be rather unclear when compared to the boundary between non-multilanguage users (monolinguals) and multilanguage users.

The Complexity of Understanding Bilingualism and Multilingualism Bilinguals and multilinguals thus entail quite heterogeneous groups of individuals. Their diverse individual and linguistic conditions are nested in larger societal contexts which cause varying degrees and types of language contact. In the literature on bilingualism, there are a number of terminologies that have been used to characterize different types of bilinguals. While these terms are often used among authors without agreed-upon definitions, it is useful to summarize some of the major ones in order to illustrate the critical dimensions that contribute to the diversity of the concept of bilingualism itself (see Table 5.1). Reflecting the multidimensionality of bilingualism, we can see that different classifications have been proposed focusing on different dimensions of bilingualism. Such dimensions include: the relationship between language proficiencies in two languages (as seen in balanced and dominant bilinguals); the functional ability (receptive and productive bilinguals); the age of acquisition (simultaneous, sequential, and late bilinguals); the organization of linguistic codes and meaning units (compound, coordinate, and subordinate bilinguals); language status and learning environments (elite/elective and folk/circumstantial bilinguals); the effect of L2 learning on the retention of L1 (additive and subtractive bilinguals); cultural identity (L1 monocultural, L2 accultural, and deculturated bilinguals), and so forth. While these terms often appear in bilingual literature (and multilingual literature as well), it is also important to note that such classifications fail to capture the complex and dynamic nature of bilingualism. First, many of these dimensions

Greater diversity in the acquisition order; can have multiple L1s and/or L2s

Greater complexity and diversity in multilingual memory organization according to typological differences among languages and proficiency levels

Neurolinguistic differences; critical period hypothesis

Difficulties with operationalizing distinctions and testing differences

Maturational differences; schooling differences

Functional differences; differences in form-meaning mapping

Early Simultaneous Sequential Late (Genesee et al. 1978)

Compound Coordinate Subordinate (Weinreich 1953)

Age of acquisition

Organization of linguistic codes and meaning unit(s)

(Continued)

Greater diversity in functional differences across domains and across languages

Receptive Productive

Functional ability

Functional and motivational differences

Greater complexity in conceptualizing and measuring multilingual competences

Conceptualizing and assessing language proficiency; Cummins’ threshold hypothesis and interdependent hypothesis; Semi-lingualism Language use irrespective of proficiency levels and identity

Functional differences; related to age factors (?)

Achieving equal level of proficiency in L2 with L1 (balanced); L2 proficiency varies but not the same as L1 (dominant)

Balanced Dominant (Peal and Lambert 1962)

Relationship between proficiencies in two languages

Understand but not produce L2 either in oral and/or written domains (receptive); understand and produce L2 (productive) Exposed to two languages from birth (simultaneous); Exposed to L2 after L1 has some foundation (sequential); became bilinguals during adulthood (late) Two sets of linguistic codes stored in one meaning unit (compound); stored separately (coordinate); L2 is accessed through L1 (subordinate)

Additional complications in multilingual contexts

Related issues and educational implications

Characteristics of SLA

Definitions

Typology

Point of focus (Dimension)

Table 5.1  Classical typology of bilingualism

Multiple cultural identities coexist irrespective of competences High bilingual competence does not necessary coincide with dual identity

Cultural identity

Cultural identity shaped by two cultures (bicultural); identity in one culture (L1 monocultural); loss of L1 culture (L2 accultural); identity in neither cultures (deculturated)

Greater complexity of learning an additional language from previously acquired languages; greater diversity of status across languages

Social value of L1 greatly influences the retention of L1; support for literacy in L1 and L2 literacy development

L2 as enrichment with or without loss of L1; status of a language in a given context Differences in acculturation process

L2 as enrichment without loss of L1 (additive); L1 is replaced by L2 (subtractive)

Bicultural L1 monocultural L2 accultural Deculturated (Hamers and Blanc 2000)

Greater diversity in social values attached to multiple languages

Support for literacy in L1 and L2 literacy development

Differences in language status and value of bilingualism

No or little additive value of L1 as a language minority status (folk/ circumstantial); additive value of L2 (elite/ elective)

Elite Folk (Fishman 1977); Circumstantial Elective (Valdés and Figueroa 1994) Additive Subtractive (Lambert 1974);

Language status and learning environment; literacy support of L1

Effect of L2 learning on the retention of L1

Additional complications in multilingual contexts

Related issues and educational implications

Characteristics of SLA

Definitions

Typology

Point of focus (Dimension)

Table 5.1  (Continued)

Bilingualism/Multilingualism and Second-Language Acquisition  115 are not categorical but continuous constructs. For example, balanced bilinguals refers to bilinguals who have an equal level of proficiency in both languages, but this term usually refers only to individuals who have equally ‘high’ proficiencies in both languages. Dominant bilinguals are individuals whose proficiency in one language (usually in their dominant language, L1) is higher than in their other language. However, these terminologies do not specify the degree of proficiency. How much ‘high’ proficiency does one need to obtain in order to be qualified as a balanced bilingual? As we shall see below, defining and measuring bilingual proficiency is a very complex business and ‘high proficiency’ itself is frequently determined arbitrarily. Similarly, there is no consensus as to what cut-off point should be used to distinguish early bilinguals (individuals who acquired two languages during their childhood, further divided into simultaneous and sequen­ tial bilinguals) and late bilinguals (those who became bilinguals during their adulthood). Second, the categorizations above fail to capture the role of the context in which language is used, which adds another layer of complexity to understanding bilingualism. Bilinguals’ use of each of their languages varies greatly depending on the content, purpose, interlocutors, psychological conditions (such as when they are under stress), formality/informality of the social setting, and so forth. For example, an individual may exclusively use the language of instruction in her science class because she does not have enough scientific vocabulary to convey meaning in another language. However, the same individual may alternate between two languages when she carries on a casual conversation with her sister who is also a bilingual. When the sisters’ mother, who is limited in the host country’s language, joins the conversation, the sisters may exclusively use the language which is understood by their mother as well. The same family language may be used at their local church, where she reads and talks about the Bible in the family language with her fellow church members. Context can shape an individual’s proficiency levels in the two languages used in this example in each of the skill domains. This is referred to as the complementary principle (Grosjean 1997). As a consequence, bilinguals are rarely proficient in all domains in both languages. Moreover, context also defines the extent to which bilinguals can alternate languages and to what extent they can manage to separate the two languages. Mackey (1967) referred to this as alternation and interference respectively, and suggested that both factors should be included in describing bilingualism. Grosjean introduced the notion of bilingual mode, which he defined as ‘a state of activation of the bilinguals’ language and language processing mechanism’ (1998: 136), to explain how the degree of activation of each of the two languages differs depending on the context. According to Grosjean, the bilingual mode is a continuum, ranging from a monolingual mode to a bilingual mode. During the bilingual mode, both languages are activated to varying degrees, and during the monolingual mode, one language is inhibited but it is still available in case it is needed.2 Grosjean proposed that since the bilingual mode influences bilinguals’ language perception (e.g., their speed of processing) and production (e.g., the amount of code-mixing), this factor needs to be accounted for when describing bilingualism.

116  Neurological and Psychological Aspects of Bilingualism and Multilingualism The third, but not least, complication stems from the fact that the bilingual profile is constantly changing. An individual may lose her dominant language proficiency as she engages in school or daily activities more heavily in her L2, and eventually her L2 may take on the dominant language status. Societal values placed on each language an individual speaks may be different from one place to another; for instance, an individual’s bilingual profile can drastically change as a result of immigration to a new country. So far, we have seen that bilinguals are a highly heterogeneous group of people and can be classified into different groups based on multiple dimensions. We have also talked about the limitations of applying such classifications to capture the context-dependent and dynamic nature of bilingualism. However, if we consider multilingualism, we realize that the picture shows even greater complexity and diversity. Apparently the categories are no longer mutually exclusive (see the last column in Table 5.1). Take the age of acquisition, for example. In the case of early bilinguals, as we have seen above, there are two acquisition patterns: simultaneous (having two L1s) and sequential (having an L1 and an L2). For early trilinguals, we can come up with four possible patterns. A trilingual individual may acquire three languages in order (i.e., L1, L2, and L3), or may acquire two languages simultaneously after having some foundation in his/her L1 (i.e., one L1 and two L2s). A trilingual individual may also start as a simultaneous bilingual and then add a third language later on (i.e., two L1s and an L3), or may have three languages simultaneously from birth (i.e., three L1s). The possible acquisition patterns become even more complicated in the case of individuals who have more than three languages (Cenoz 2000). Each of these patterns may exhibit different processes and outcomes as a result of multiple language acquisition (e.g., Paradis 2007). Typological differences among languages (i.e., distances between languages being spoken by a multilingual individual) and the individual’s proficiency level in each language also influence multilinguals’ lexical-semantic organizations, which are thought to be partially integrated and partially separated (De Angelis 2007). Greater complexities apply to other categorizations as indicated in Table 5.1. While the empirical evidence on multilingualism is still limited, the classical typology used in the bilingual literature no longer appears sufficient enough to capture the complexity of multilingualism.

Conceptualizing and Assessing Language Proficiency As we can clearly see above, bilingualism and multilingualism are highly complex and multidimensional linguistic, psychological, and social behaviors. Accordingly, there is substantial diversity among individuals with respect to their acquisition of more than one language. One of the central issues in understanding language acquisition among bilinguals/multilinguals is how to conceptualize and measure such individuals’ language proficiency. Language proficiency has been conceptualized and meas-

Bilingualism/Multilingualism and Second-Language Acquisition  117 ured in various ways, reflecting different theoretical views that exist with regards to what counts as ‘language.’ In SLA research, which has been strongly influenced by theories and methodologies in L1 acquisition research, one can identify three major views towards language acquisition: (i) the formal linguistic view; (ii) the individual-cognitive view (psycholinguistic view); and (iii) the socio-contextual view. Admittedly, such a classification oversimplifies reality; there are many researchers who integrate some parts of different views. However, it is also true that the different schools of thought have had a distinct influence over the history of SLA research since the middle of the twentieth century. The formal linguistic view, which is heavily influenced by Chomsky’s theoretical frameworks (Chomsky 1965, 1981, 1986, 1995), primarily focuses on the form of languages. According to this view, human beings are innately equipped with an unconscious knowledge of grammar (Universal Grammar, or UG). UG involves a set of principles, some of which are universal and some of which are parameterized (i.e. variable). Depending on the child’s exposure to a particular linguistic environment, different values are set for a given parameter. This system of principles and parameters is considered to be the basis for language acquisition, or in Jordan’s words, ‘the process of language acquisition is simply one of selecting, rather than learning’ (2004: 132). For the purposes of the present discussion, there are three points related to Chomsky’s frameworks that are important to mention. First, Chomsky (1965: 4) distinguished between competence (defined as ‘the speaker-hearer’s knowledge of language’) and performance (defined as ‘the actual use of language in concrete situations’). This was later developed into a distinction between I-language (competence) and E-language (language use) (Chomsky 1986). Importantly, some have argued that the focus of inquiry should be the speaker’s knowledge of abstract rules of grammar (i.e., grammatical competence or I-language). Second, what Chomsky is concerned with is the knowledge of grammatical form of ‘an ideal speakerlistener, in a completely homogeneous speech-community, who knows its language perfectly’ (1965: 3). That is to say, Chomsky is describing a normalized, monolingual, native speaker’s intuition regarding linguistic form. Variation of language use within and across individuals has not fallen within his scope of interest. Finally, UG was considered to be specific to language, or what Chomsky (1965) referred to as having task-specificity; this was not a general cognitive system. While Chomsky did not directly address the mechanism of L2 acquisition, one of the central issues in a UG-based SLA approach is whether or not adult L2 learners can access UG. Different proposals have been made, ranging from a full-access position (e.g., Flynn 1991, 1996) to a no-access position (e.g., Bley-Vroman 1990).3 In his fundamental difference hypothesis, Bley-Vroman (1990) argues that adult L2 learners can no longer access UG but can develop abstract rules of grammar in L2 through their general problem-solving skills. Other researchers take a partialaccess position (e.g., Clahsen and Muysken 1986; Schachter 1988) or a dual-access position (Felix 1985). Both positions claim that L2 learners can still access UG but with some limitations. The former position holds that L2 learners can access only

118  Neurological and Psychological Aspects of Bilingualism and Multilingualism those parameters that are operative in their L1, and the latter position argues that learners’ general problem-solving abilities may block their access to UG. One of the leading methods of inquiry in the UG-based SLA approach is to compare the UG properties of L2 learners’ interlanguage grammar with the UG properties of native speakers. It has become apparent that the formal linguistic view has inherited some inadequacies when it comes to understanding language acquisition among bilingual and multilingual individuals. After all, UG ‘is not, of itself, a theory of acquisition’ (White 2003: 36). As we have already seen, bilinguals and multilinguals exhibit a wide range of language abilities depending on the context, and these are not limited to just their core grammatical knowledge. One cannot assume to have a ‘normalized’ bilingual or a ‘normalized’ multilingual. Bilinguals and multilinguals may have multiple ‘native’ languages and/or multiple ‘nonnative’ languages, and the language input that they receive in each of their languages can be drastically different from that of monolingual speakers. Interestingly, in a recent review on UG-based SLA research in the formal linguistic tradition, Liceras calls for acknowledging the important role of E-language (language use) in language acquisition ‘since it accommodates language contact and input variability in a way that leads the child to interpret triggers differently from his or her ancestors so that a different I-language may be created’ (2010: 263). While the formal linguistic view has restricted its scope of inquiry to grammatical form, others have taken broader approaches to the study of language. Hymes’ (1972) notion of communicative competence, which was originally proposed in the context of L1, refers to both knowledge (including not only grammatical but also sociolinguistic knowledge) and ability for use (referring to one’s potential to use language that is socially acceptable in a given context). Importantly, ability for use, which is considered as part of performance in Chomsky’s model, is now included in one’s competence in Hymes’ model. We should also note that ability for use is not exclusive to language performance; it can include various cognitive and noncognitive elements such as motivation (McNamara 1996). Such a broad view towards language has been adapted by theories of communicative competence that are widely subscribed to in SLA research. Canale and Swain (1980) identify three components of communicative competence (grammatical competence, sociolinguistic competence, and strategic competence), and discourse competence was added as a fourth component in the model later proposed by Canale (1983). Bachman (1990) further elaborated on the previous models and proposed a model of communicative language ability which contains not only language competence but also strategic competence (covering some aspects of Hymes’ ability for use) and psychophysiological mechanisms (psychophysiological processing for executing language). Although these models are not free from criticism, they have served as a basis for both cognitive and socio-contextual approaches to SLA, as we shall cover in more detail below. As we have seen, whereas the formal linguistic view sees language acquisition as partially innate and as a specialized operation that is distinct from general cognition, the individual-cognitive view considers language acquisition to be

Bilingualism/Multilingualism and Second-Language Acquisition  119 data-driven and a part of general cognition. Drawing from cognitive and learning theories such as information-processing models (e.g., Anderson 1983; Gagné 1985, Shiffrin and Schneider 1977), schema theory (e.g., Gagné 1985; Rumelhart and Ortony 1977), research on experts versus novices (e.g., Chi, Glaser, and Farr 1988; Simon and Chase 1973), and connectionist modeling (Rumelhart and McClelland 1986), L2 learning in the individual-cognitive camp is thought to consist of changes in individuals’ mental states as a result of cognitive problem solving. The aspects of language investigated via this perspective are not limited to rule-governed structures of language (primarily syntax) but extend to such areas as vocabulary, the comprehension and production of oral and written languages, strategies, metalinguistic awareness, and translation. Inquiries in this view focus on cognitive factors underlying learners’ knowledge representation, processing, production, and control. Examples of such inquiries include: how learners’ lexicons are organized and stored in memory (e.g., De Groot 1992; Heredia and Brown 2004; Kroll and Stewart 1994; Kroll and Tokowicz 2005); how learners utilize knowledge-based and control-based skills, and the role of metacognition therein (Bialystok 1991, 2001; Bialystok and Ryan 1985); the mechanisms whereby cognitively demanding control processes eventually become automatic processes (e.g., McLaughlin, Rossman, and McLeod 1983); the restructuring of internal representations (e.g., McLaughlin 1990); the mechanisms of form-meaning mapping processes (Bates and MacWhinney 1982; MacWhinney 1997); how the focus of attention shifts (e.g., Segalowitz 2000); and so on. Interactionist theories (e.g., Long 1996) may be broadly classified in this group; while the interactionists view the input that learners receive through interaction as a critical component of L2 acquisition, they maintain the idea that L2 acquisition is fundamentally an internal mental process. L2 acquisition in the individual-cognitive view is considered as the greater complexity of learners’ L2 knowledge and their greater control over such knowledge. In order to do so, in most cases, learners’ proficiency is measured by locating their linguistic performance along the path towards their target linguistic performance (i.e., native speaker-level performance). This approach towards measuring L2 proficiency has raised serious concerns. One of the major concerns is the validity of locating monolingual native speakers and bilinguals on the same spectrum. In essence, some have questioned the concept of determining the proficiency of bilinguals by comparing them to the proficiency of monolinguals if the linguistic abilities between these two groups are qualitatively different. This prevailing practice in cognitive-based SLA research has been referred to as the monolingual bias (De Angelis 2007). Grosjean (1982, 1998, 2010) has criticized the monolingual view of bilinguals (i.e., the view that bilinguals are essentially two monolinguals coexisting in one individual), and argues that bilinguals are not simply a juxtaposition of two monolinguals but should be qualitatively distinguished from monolingual speakers. Under this ‘holistic’ view of bilingualism, it does not make sense to use the monolingual norm as the guideline for bilingual proficiency. Similar concerns have also been raised by other researchers (e.g., Cook 1992; Herdina and Jessner 2000).

120  Neurological and Psychological Aspects of Bilingualism and Multilingualism In addition to the monolingual bias, some researchers have pointed to a bilingual bias in cognitive-based SLA and multilingual research (e.g., De Angelis 2007). The bilingual bias refers to the assumption that the psycholinguistic mechanisms and processing between bilinguals and multilinguals are essentially the same, whereas the latter may only be more complicated and contain more noise than the former. Such a perception is often found in studies where researchers treat participants as ‘L2 learners’ without carefully controlling for the possibility that the participants might know additional language(s) and the possible effects of this on their psycholinguistic mechanisms and processing. The same point can be made with regards to the treatment of ‘native speakers.’ Native-speaking counterparts may in fact know more than one language, and yet they may be treated as if they were monolinguals. We should take this criticism more seriously, given the fact that there is some evidence indicating that a minimum exposure to additional language(s) can significantly influence one’s L2 (or other nonnative languages) as well as one’s L1 (De Angelis 2007). To determine the ‘native standard’ is another complicated issue. As acknowledged by Bialystok, a challenge is that ‘the prescriptive rules provide an incomplete and somewhat inaccurate account of what native speakers really do with language’ (2001: 13). Moreover, even if one can find a group of monolingual speakers in order to identify a target level of proficiency, variations across and among such ‘native speakers’ have rarely been considered in most cognitive-based SLA research. Some native speakers may know more vocabulary than others. One can observe great variability in domains such as acquisition of pronunciation and speech styles across different speech communities. Native speakers’ literacy skills may differ substantially among themselves. SLA research has focused on educated adult learners; illiterate L2 learners and young learners have been given scant attention in SLA research until recently (Nicholas and Lightbown 2008; Tarone, Bigelow, and Hansen 2009). Researchers who take a socio-contextual view of language acquisition have frequently criticized the cognitive view for its lack of sufficient consideration of individual variability across different contexts. This third view provides a broader definition of language than the other two views discussed above. The sociocontextual view stresses the social and interpersonal communicative aspects of language, and considers ‘language’ to be a social construction. The sociocontextual approaches to L2 acquisition are concerned with both the variation of language use and the social and interactional factors affecting language use (Tarone 2007). It has been claimed that socio-contextual variables constrain L2 acquisition in both systematic and nonsystematic ways (Ellis 2008). Inspired by William Labov’s series of pioneering works on speech variation (Labov 1972), researchers have demonstrated how L2 learners show contextual variability depending on variables such as the linguistic environment (Dickerson 1975), social norms of L1 variation (Schmidt 1977), interaction partners (Beebe and Zuengler 1983), the function of interaction (Tarone 1983, 1985), discourse topics (Eisenstein and Starbuck 1989; Zuengler 1989), and so forth.

Bilingualism/Multilingualism and Second-Language Acquisition  121 Socio-contextual approaches to L2 acquisition also cover a wide range of social and contextual aspects of language that are related to the learner’s acquisition of communicative competence. Since the 1990s, we can see increasing attention being paid to socially based theories and approaches in SLA including sociocultural theory (e.g., Lantolf 2000); language socialization (e.g., Duff 2007); and applications of conversational analysis in SLA (e.g., Gardner and Wagner 2004). As mentioned before, communicative competence is not limited to knowing linguistic codes or referential meanings, but also involves the ability to use language appropriate to a given socio-context, employing various interaction strategies, and understanding the various functions of language as used in speech acts defined by a given speech community. The assessments of language proficiency should include sociolinguistic and pragmatic knowledge as well as the ability to use such knowledge. Thus, the importance of employing communicative and pragmatic assessments using authentic material has been stressed. Although there is no dispute over the existence of variation, there is concern with regards to how such variation can be accounted for in theories of SLA (Gregg 1990; Schachter 1986). Some have argued against focusing on variations by claiming that SLA should be about how language is learned, not how language is used (Long 1997). This is rooted in some of the central disagreements in SLA: namely, how to conceptualize interlanguage (Selinker 1972) and the nature of the relationship among interlanguage, competence, and performance. Some sociolinguists have taken the position that all variation is systematic. To support this claim, they have attempted to develop models that incorporate probabilistic values for influences on various forms in the language produced by L2 learners (Preston 1993, 2002). Lafford (2006), based on her review of research on study-abroad programs and classroom learning of Spanish as a second language, has claimed that it is learners’ perceptions towards social variations, such as setting and interlocutors, rather than the actual variations themselves, that account for individual differences in L2 acquisition. As a way to ‘deconstruct’ the dichotomy between the individual-cognitive view and the socio-contextual view towards language, researchers such as LarsenFreeman (2007) proposed taking an ecological approach to language. That is, language is a complex and dynamic system which constantly adapts itself to the environment across time and space. According to Larsen-Freeman, ‘what endures is not a rule-based competence, but a structured network of dynamic languageusing patterns, stored in memory with specific information about instances of use retained in the representation’ (783–4). Such a view of language as a discursive and fluid construct presents SLA researchers with various conceptual and methodological challenges. At this point, it is perhaps fair to say that there are no concrete remedies for such challenges. In sum, reflected in these different views towards ‘language,’ there is no simple answer for how best to conceptualize language proficiency or how to measure it. A number of the leading conceptual and methodological foundations (or assumptions) in the formal linguistics and individual-cognitive approaches, such as ‘native speakers’ norms’ and ‘interlanguage as a step towards (or measuring the

122  Neurological and Psychological Aspects of Bilingualism and Multilingualism gap from) the target forms’ have come under scrutiny. In recent years, SLA research has paid greater attention to the roles of context in L2 acquisition, but how best to account for contextual factors in understanding and measuring bilinguals’/ multilinguals’ language acquisition is still an ongoing challenge.

Age and L2 Acquisition We now turn our attention to the age factor in L2 learning, which is considered to be one of the major factors in explaining variability of language acquisition among bilinguals and multilinguals. It is often believed among the general public that children are superior in learning an L2 when compared to adults. While there is a substantial amount of research on the effect of age over L2 language acquisition, many questions have yet to be answered. In examining the effects of age on L2 acquisition, we need to separately consider its effect on the route of acquisition, the rate of acquisition, and the ultimate attainment of L2 proficiency. Regarding the route of acquisition, it has long been believed that the age of acquisition does not seem to have much effect. Bailey, Madden, and Krashen (1974), for example, using the Bilingual Syntax Measure, showed that their adult L2 learners (regardless of their L1 backgrounds) acquired grammatical morphemes in an order similar to the L2 learning children that were observed in Dulay and Burt (1974). While this premise that morphological acquisition order is universal among adult L2 learners regardless of their L1 has been widely accepted (it is often referred to as the natural order, Krashen 1977), some researchers (e.g., Luk and Shirai 2009) have challenged this premise and have argued that learners’ L1s influence the order of their morphological acquisition. Within the formal linguistic tradition, Meisel (2008: 59) ‘tentatively’ classified bilinguals into three groups according to the age of onset of their L2 acquisition: he referred to individuals exposed to more than one language from birth or before 3 as simultaneous bilinguals, individuals whose age of onset was between 4 and 8 as child L2 learners, and to those whose age of onset is after 8 as adult L2 learners. Meisel argued that these three groups show qualitatively different routes of acquisition as well as ultimate attainment, as discussed further below. Such differences, if they exist, may be selective; one may find differences in certain domains but not in others. Schwartz (2003) proposed the domain-by-age model, which claims that child L2 acquisition shows a similar developmental pattern with child L1 acquisition with regards to inflectional morphology while child L2 acquisition is more like adult L2 acquisition in the domain of syntax. Empirical studies comparing developmental patterns of acquisition, however, have shown mixed results to date (Haznedar and Gavruseva 2008). Regarding the rate of acquisition, there is some evidence that adults outperform children when the amount of exposure to L2 is controlled (e.g., Lowenthal and Bull 1984; Olson and Samuels 1973). However, Snow and Hoefnagle-Höhle (1978) found that in a typical, unstructured exposure to L2 (Dutch) for 3 months, their teenage group (consisting of 12 to 15 year olds) showed the most rapid progress,

Bilingualism/Multilingualism and Second-Language Acquisition  123 followed by their adult group (15 years and older), and finally by their group of children (ages 3 to 10). Moreover, the advantage of older learners seemed to be only in the short term. After a 10-month exposure to L2, the children caught up with the adult group on most measures. In some specific domains such as phonology, advantages for child learners were reported (e.g., Cochrane 1980; Tahta, Wood, and Lowenthal 1981), while others have not found any differences in the rate of acquisition between children and adult L2 learners (e.g., Slavoff and Johnson 1995). A substantial amount of research has been conducted in order to investigate the effect of age on the ultimate attainment of an L2. It is generally agreed that there are some age-related effects in L2 acquisition. Researchers disagree, however, with respect to: (i) whether or not a critical period for language acquisition exists; (ii) if it does, the times of onset and offset of the critical period; (iii) the domains affected by age; and (iv) the causes underlying the differences in attainment by age (e.g., De Groot 2011; Singleton 2005). The disagreements in the first two issues largely depend on how one conceptualizes the critical period itself. The notion of a critical period for language acquisition was first proposed by Penfield and Roberts (1959) and later was popularized by Lenneberg (1967). The critical period is generally understood as being a limited developmental window during which native-like language attainment is possible. Based on clinical data on patients with brain injuries and children with Down syndrome, Lenneberg proposed that the critical period starts at around one year of age and ends at puberty, by which time the brain loses its plasticity. This proposal (i.e., the critical period hypothesis) has inspired substantial research. However, various conceptualizations and related terms, such as the sensitive period, maturational constraints, and the age factor, have been used rather confusingly in the literature. In order to substantiate the existence of a critical period, Hakuta claims that the following characteristics must be observed: ‘(1) there should be clearly specified beginning and end points for the period; (2) there should be a well-defined decline in L2 acquisition at the end of the period; (3) there should be evidence of qualitative differences in learning between acquisition within and outside the critical period; and (4) there should be a robustness to environmental variation inside the critical period’ (2001: 195–6). According to Hakuta and other supporters (e.g., Bialystok and Hakuta 1994; Birdsong 2005; Bongaerts 2005; Flege 1999), to meet the conditions for (2) above, a linear decline in performance is not sufficient to be considered evidence for the critical period. Others (e.g., DeKeyser and Larson-Hall 2005), however, take a linear decline as evidence for the existence of the critical period. The nature of constraints on language acquisition is also not agreed upon among researchers. Some researchers think that the probability of obtaining nativelike proficiency is constrained (e.g., Birdsong 2005), while others claim that the potentiality of obtaining native-like proficiency is what matters (e.g., Hyltenstam and Abrahamsson 2003; Long 2005). Such differences in view influence the methodologies that researchers can employ. For the former position, random sampling should be critical. For the latter position, screening of participants is acceptable

124  Neurological and Psychological Aspects of Bilingualism and Multilingualism and it may be sufficient to find only one exceptional individual who began to be exposed to his/her L2 after the critical period and yet obtained native-like proficiency, as counter-evidence for arguing against the existence of the critical period (Long 1990). To date, among those who argue for the existence of the critical period, there is no consensus as to when the onset and offset of the critical period take place. Researchers have suggested different times for closure, such as 5 years old (Krashen 1973), 6 years old (Pinker 1994), 12 years old or puberty (Lenneberg 1967). Scovel (1988) argues that the critical period exists only in the phonetics/ phonology domain (with the offset being at the age of 12), while others have proposed that there are multiple critical periods in different linguistic domains (e.g., multiple critical period hypothesis, Seliger 1978). Long (2005) extends the domain of inquiry to lexical-collocational abilities and suggests that offset times can range from some time around 6 years old to the mid-teens depending on the domain, whether that be phonetics/phonology (6 to12 years of age), morphosyntax (mid-teens), or lexical-collocational abilities (6 years old to mid-teens). Abrahamsson and Hyltenstam (2009) do not completely deny the existence of a critical period but have proposed that the offset can be very early in one’s life: namely, shortly after birth. Moreover, the literature has proposed a number of potential causes of a critical period (for a summary of such explanations see Singleton 2005). However, the literature, including neuropsychological research on the critical period hypothesis (e.g., De Groot 2011; Ellis 2008; Stowe and Sabourin 2005 for a review of such neuropsychological studies), has not yet offered a clear explanation as to the reasons for the specific offset timing for different domains. There are a number of methodological challenges in conducting research on the age effect on L2 acquisition. To begin with, there is a problem related to the linear dependency among three important variables: the age of onset; the length of exposure to the target language; and the chronological age of testing (Bialystok and Hakuta 1994; Stevens 2006). One of the popular designs employed among researchers is to see if there is significant negative correlation between the age of arrival (AOA) to the host country where the target language is spoken and the performance in targeted domains. In this design, if the researcher gives all the participants the same length of residence (LOR) in the host country, the age of testing (AOT) is inevitably varied. If the participants are examined at the same AOT, then their LOR is varied. The problem is that all three variables significantly influence learners’ L2 performance. Studies with this type of design often assume that learners can reach their ultimate attainment after 5 years (10 years in some studies) and that they can do so uniformly across different age cohorts. However, there is evidence showing a strong association between immigrants’ LOR and their L2 performance, and the relationship lasts well beyond 5 or 10 years’ stay in the host country (Stevens 2006). Moreover, one cannot ignore an age-related cohort effect; namely, since many of our life experiences such as schooling, marriage, and work are agerelated, such life experiences similarly influence many L2 users within the same

Bilingualism/Multilingualism and Second-Language Acquisition  125 AOA cohort with respect to their motivation to learn the L2, as well as the types and intensity of L2 input they receive during their duration of residence. While the age of testing (AOT) has been the least considered variable among the studies in the literature, the potential influence of AOT should not be underestimated. A substantial amount of research has indicated that people are affected by age-related changes in cognitive-processing speed and working memory, which in turn influence L2 users’ use of L1 and L2, L2 acquisition at different developmental stages, and L2 performance at the time of testing (Birdsong 2006). Such changes in cognitive abilities may influence learners differently depending on the linguistic domains being investigated. A second potential methodological challenge stems from the measurements employed. Various types of measurements have been used, but it is not clear what kind of measurements can sufficiently capture ‘native-like’ proficiency. If the measurement results in a ceiling effect, one may fail to detect differences in proficiency that may exist between learners and native speakers. Although native-speakers’ judgments are frequently employed as a method of analyzing language production data, possible biases in their judgments have been found as well (e.g., Long 2005). Some researchers have argued against the existence of the critical period in L2 acquisition by identifying cases of late adult learners and individuals who could successfully attain native-like proficiencies in a given L2 (e.g., Birdsong 1992, 2007; Bongaerts, Mennen, and van der Slik 2000). While these individuals’ performance may be indeed ‘native-like,’ some studies have reported that they still could locate subtle differences in performance in certain aspects between these successful L2 learners and ‘native speakers’ (Coppieters 1987; Ioup et al. 1994; Moyer 1999). Abrahamsson and Hyltenstam, when using ‘a battery of 10 highly complex, cognitively demanding tasks and detailed measurements of linguistic performance, representation, and processing’ (2009: 249), found that not only did none of their late learners (the age of onset was later than 13) show ‘native-range’ performance, but also that only a few early learners could fall into the ‘native-range.’ The authors suggested that there may be ‘much higher rates of nonnativelike early learners’ than have been previously assumed (290). One can argue that these results should not be too surprising because what was revealed here is exactly the nature of the differences between monolinguals and multilanguage users. As we have seen already, since bilinguals are not a juxtaposition of two monolinguals (e.g., Grosjean 1998), we can assume that their performance in L1 and L2 would not be identical to that of monolingual users of either languages. To the best of the present author’s knowledge, no studies concerning the critical period and/or age effects have examined both the learners’ L1 and L2 performance, or their performance in additional language(s) when the participants were multilinguals. The third, but not the least, potential methodological challenge is the selection of ‘L2 learners.’ As mentioned already, the age of arrival (AOA) has been frequently used in immigrant studies. However, AOA is not necessarily the same as the age of onset (AO). Similarly, the length of residence (LOR) does not necessarily

126  Neurological and Psychological Aspects of Bilingualism and Multilingualism equate with the length of exposure to the target language. Any possible exposure to the target language, including the formal instruction that participants may have received prior to their immigration, has also not been examined thoroughly. If we restrict ourselves to Lenneberg’s (1967) original claim, which was only concerned with the influence of the mere exposure to the target language on attainment, we could argue that the role of formal instruction can be excluded from the critical period discussion (Hyltenstam and Abrahamsson 2001). While it is true that Lenneberg did not deny the possibility of adults learning a foreign language, he did not make any specific claims as to whether or not ‘the matrix for language skill’ (1967, p. 176), which is supposedly also available for adults, allows them to attain native-like proficiency in their L2. It is up to SLA researchers to determine how best to apply Lenneberg’s claim about L1 acquisition to L2 acquisition. In reality, many participants in previous studies presumably received some form of formal instruction before as well as after their immigration to the host country. Given the fact that there is substantial research supporting the observation that formal instruction can result in an increase in accuracy among both young and adult learners, and among both beginners and advanced learners (Ellis 2003), the possible impact of formal instruction should not be underestimated. The bilingual bias addressed above is also evident among research on the ultimate attainment of L2; ‘L2 learners’ in previous studies may include multilinguals, but they are often treated as if they were bilinguals and the potential effect of having an additional language over the ultimate attainment of their L2 has gone unexamined in many cases. Similarly, ‘native speakers’ in some of the studies may contain bilinguals. By virtue of having another language, interaction between the phonetic elements in the L1 and L2 among bilingual individuals can influence their production of sounds in both languages from very early on, and can cause them to deviate from the sound production of monolingual native speakers of either language (Flege 1999). As we shall see in the next section, there is some evidence showing that language distances between L1s and L2s among bilinguals also influence their ultimate linguistic attainment as well as the rate of L2 acquisition in any given cohort with the same AOA (e.g., Birdsong and Molis 2001; Hakuta, Bialystok and Wiley 2003). Thus, it is important to place careful consideration on participants’ background factors in selecting study participants as well as interpreting data.

The Interaction between L1 and L2 (or Additional Nonnative Languages) Another factor that has gained substantial attention in SLA is the influence of L1. In the last 50 years, researchers have taken different views towards the role of L1 in SLA. In the 1960s, employing contrastive analysis, it was believed that learners’ errors reflected the structure of their L1. In the 1970s, however, researchers became increasingly aware that many of the errors made by learners could not be simply explained by L1 transfer. The errors were not solely predictable based on

Bilingualism/Multilingualism and Second-Language Acquisition  127 contrastive analysis. Learners’ interlanguage (Selinker 1972) was considered to be a unique intermediate system that emerged during the process of L2 acquisition and that differed from their L1. The notion of transfer has been broadened in SLA since then, and it is no longer considered to be either uni-directional or uni-dimensional. As we can see in Odlin’s definition (namely, ‘transfer is the influence resulting from the similarities and differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired’ (1989: 27)), the notion of transfer can include not only the influence from L2 learners’ L1 but also from any other ‘imperfectly’ acquired additional language(s). Transfer not only manifests itself in interferences (negative transfer), but also in facilitation (positive transfer), avoidance, and overuse that result from intralinguistic influence. Learners’ L1 and L2 (and any other additional languages in the case of multilinguals) interact with each other. Thus, some authors prefer to describe such interaction as cross-linguistic influence rather than transfer (e.g., Kellerman and Sharwood-Smith 1986; Zobl 1984). Note that in the present chapter, transfer and cross-linguistic influence are used interchangeably. Researchers have taken different approaches towards transfer. Linguists who take the UG approach have been interested in investigating whether or not UG is accessible by L2 learners as well as what role L1 plays in interacting with input from L2. Others have been interested in the cognitive mechanisms under which transfer occurs. For example, in the competition model, researchers investigated how L2 learners’ information-processing mechanisms that function as part of their L1 can be ‘adjusted’ to L2 sentence interpretation (Bates and MacWhinney 1981; MacWhinney and Bates 1989). The domains of investigation have not been limited to the structure of languages (e.g., syntax, phonology, and lexicon), but have been expanded to other complex cognitive and psychological domains such as reading, writing, communicative and metacognitive strategies, and pragmatics. Inquiries on L1 transfer are not limited to examining its effects on L2 acquisition per se, but have also expanded to analyzing the influence of L1 transfer on learners’ cognitive processing in L2 (e.g., Koda 2008). Most recently, some researchers have further extended their inquiry of transfer to concept transfer, and have stimulated discussions on language and cognition (e.g., linguistic relativism) and cognitive linguistics (Jarvis 2011; Odlin 2005). A number of factors that promote or hinder transfer have been identified. One of the major factors is language distance, or the degree of linguistic differences and similarities between one L1 and an L2. We may assume that smaller linguistic differences lead to easier acquisition of the target features and processing of an L2. Indeed, one simulation study using a neural network model indicated that language distance affects the rate of learning, and that the smaller the distance between two languages, the shorter it takes to learn them. The model also showed an interaction between language distance and type of learning (i.e., simultaneous vs. sequential learning types) (Inoue 1996). A series of studies which compared Swedish-speaking and Finnish-speaking English learners confirmed that the smaller distance between Swedish and English did facilitate the rate of English

128  Neurological and Psychological Aspects of Bilingualism and Multilingualism acquisition among Swedish-speaking English learners (Ringbom 2007). The orthographic distance between an L1 and an L2 also affects the learner’s lexical processing in reading (Koda 2008). However, the processing efficiency facilitated by smaller language distances may be limited to features that require identical processing (Muljani, Koda, and Moates 1998). While there are some reports indicating that facilitation attributed to language distance appears to be most evident at an early stage of learning and that it may disappear as learning progresses (Ringbom 2007), it is not yet clear how and to what extent language distance influences learners’ types and processes of learning and their ultimate attainment. We also have to remember that it is not easy to quantify the degree to which a learner’s L1 contributes across different features; transfer appears to be more evident for features that occur more frequently than those that occur less frequently (Odlin 2003). Moreover, Kellerman (1977, 1983) argued that learners’ perceived distance (psychotypology), rather than the actual linguistic distance, is a more powerful trigger for transfer. To make the matter more complex, learners’ perceived distance changes over time. In addition to language distance, other major factors that promote or hinder transfer include the developmental stage, age, sociolinguistic factors, and prototypicality (i.e., the degree to which a form or meaning is perceived as ‘prototypical’). Again, however, the relationship between these factors and the exact nature of transfer is still largely unclear. Some transferred features from an L1 disappear as the learner becomes more proficient, while other features remain persistently. Yet other features only emerge at the later stages of learning. Identifying transfer among young learners’ interlanguage is often very challenging, in part due to the fact that many of the errors seem both interlingual and, at the same time, are similar to what can be observed in L1 acquisition (Ellis 2008). Various sociolinguistic factors such as formality of context, interlocutors, topics, and social status of the learners’ L1 as well as their L2 can all influence transfer in complicated and dynamic ways. In SLA, as Odlin rightly pointed out, while ‘there does not yet exist any comprehensive theory of transfer,’ we can see some advancements with respect to methodology (2003: 478). Jarvis (2000: 253) suggested that researchers have to show the following three types of evidence in order to claim the existence of L1 transfer: (i) there is a homogeneity in interlanguage performance among users who share the same L1; (ii) there is a heterogeneity in interlanguage performance among users who have different L1s; and (iii) there is congruity between L1 and interlanguage among users who share the same L1 (i.e., among L2 users with the same L1, the feature in question in their interlanguage can also be observed in their use of corresponding features in L1). In recent years, there have been an increasing number of studies on transfer among multilinguals. The studies on multilinguals tell us about the complexity of cross-linguistic influences as well as the methodological challenges that researchers face. In addition to the influence from L1 to L2, multilingual cases involve the cross-linguistic transfer from any other additional language(s) that the multilingual users know. The transfer may occur from multiple languages simultaneously

Bilingualism/Multilingualism and Second-Language Acquisition  129 in a complicated way (e.g., L1 and L2 jointly influencing L3 acquisition). Moreover, transfer among multilinguals can be multi-directional (e.g., L1 and/or L2 influencing L3 acquisition while L3 acquisition might also influence L1 and/or L2, and so on). We still have only a very limited understanding of such complicated mechanisms of multilingual transfer (De Angelis 2007). Language distance appears to be an influential factor in transfer in multilingual acquisition, as seen in the case of SLA mentioned above. It has generally been found that the target language appears to be most influenced by the closest language, irrespective of whether or not it is an L1 or nonnative language (De Angelis 2007; Jessner 2008; Ringbom 1987). However, as we can easily expect, it is much more complicated to detect the role of language distance among multiple languages due to interdependency across languages in multilinguals’ language use. In addition to language distance, other factors also have been found to influence the type of transfer as well as the frequencies of transfer among multilinguals. Such factors include proficiency levels of languages, recency of use, order of acquisition, the extent of use, age of acquisition, status of languages, and so forth (e.g., Cenoz, Hufeisen, and Jessner 2001; De Angelis 2007). The foreign language effect (Hammarberg 2001; Meisel 1983) is also frequently mentioned in the multilingual literature. This refers to the tendency whereby a learners’ L2 (or other nonnative language) becomes the major source of transfer to the target nonnative language. Under such complexity of transfer among multiple languages, methodological innovation is necessary. For example, we can no longer expect homogeneity in interlanguage among multilingual users with the same L1, as suggested in the case of SLA in Jarvis (2000) above.

Conclusion Bilingualism is a highly complex linguistic, psychological, and social phenomenon. In this chapter, we have sketched some flavor of its complexity. Bilingualism is multidimensional, context-dependent, and nonstatic. It is also very heterogeneous. The growing interest in multilingualism in recent years has revealed further complexities in language acquisition and processing among multilingual users. Classical typologies which have been used widely in the bilingual literature seem to have only limited application in capturing such complexity. The present chapter discussed three major issues at the intersection between SLA and bilingual/multilingual research: namely, issues related to conceptualizing and measuring one’s language proficiency; the role of age in acquiring nonnative language(s); and cross-linguistic influences on the acquisition of nonnative language(s). Until recently, researchers of SLA and bilingualism have conducted investigations of similar phenomena almost totally independent of each other (Jessner 2008). Researchers of bilingualism and multilingualism in particular have challenged some of the leading major premises and approaches that have been widely subscribed to in SLA research. One of the major repeated criticisms has been characterized as the monolingual bias: namely, the assumption that

130  Neurological and Psychological Aspects of Bilingualism and Multilingualism bilinguals’ language competencies are qualitatively the same as those of monolingual users of the target language. In fact, based on that assumption, many SLA studies have compared L2 learners’ performance with that of monolingual native speakers of the target language. Another major criticism of SLA research can be referred to as the bilingual bias: namely, the assumption that language acquisition and processing among multilinguals are fundamentally the same as with bilinguals. While we still have only limited research on multilinguals’ language acquisition, there is growing evidence that this assumption is not warranted. SLA research itself has undergone a lot of changes in recent years. Increasing attention has been paid to the role of context in L2 acquisition. While there are different conceptualizations of ‘language’ among researchers, more and more researchers are interested in a holistic view towards language. New conceptualizations of language, such as looking at language as a discursive construct, which Kramsch and Whiteside called the ‘poststructualist view’ (2007: 913), seem to call for innovative methodologies and approaches to language acquisition. Bilingual and multilingual research highlights the importance of the role of context and holistic approaches in understanding language acquisition. The more we understand bilingualism and multilingualism, the more complexity we realize and the more challenges we face. But this may be precisely the reason why we have long been attracted to the mysterious phenomena of language acquisition since ancient Roman times.

NOTES 1  Aronin and Ó Laoire (2004) had argued for a distinction between individual multilingualism and multilinguality. According to them, the former places greater emphasis on ‘the language system and language codes in use by an individual in a “language in contact” situation’ whereas the latter is more concerned with ‘personality and intrapersonal dynamics’ (18). 2  There are some who have challenged Grosjean’s bilingual mode hypothesis. Dijkstra and van Hell (2003), for example, conducted a series of lexical activation studies among multilinguals, and have argued, based on their findings, that nonlinguistic context has only a limited influence over multilinguals’ relative activation of lexical items in different languages. 3  It’s also important to note that terms such as ‘no-access’ and ‘full-access’ do not necessarily refer to the same thing among researchers (White 2003).

REFERENCES Abrahamsson, Niclas and Hyltenstam, Kenneth (2009). Age of onset and nativelikeness in a second language: Listener perception versus linguistic scrutiny. Language Learning 59(2): 249–306. Anderson, John R. (1983). The Architecture of Cognition. Cambridge, MA: Harvard University Press.

Aronin, Larissa and Hufeisen, Britta (2009). The Exploration of Multilingualism. Amsterdam: John Benjamins. Aronin, Larissa and Ó Laoire, Muiris (2004). Exploring multilingualism in cultural contexts: Towards a notion of multilinguality. In Charlotte Hoffmann and Jehannes Ytsma (eds.), Trilingualism in Family, School and

Bilingualism/Multilingualism and Second-Language Acquisition  131 Community. 11–29. Clevedon: Multilingual Matters. Bachman, Lyle F. (1990). Fundamental Considerations in Language Testing. Oxford: Oxford University Press. Bailey, Nathalie, Madden, Carolyn, and Krashen, Stephen D. (1974). Is there a ‘natural sequence’ in adult language learning? Language Learning 24: 235–43. Baker, Colin (2001). Foundations of Bilingual Education and Bilingualism. 3rd edn. Clevedon: Multilingual Matters. Bates, Elizabeth and MacWhinney, Brian (1981). Functionalism and the competition model. In Brian MacWhinney and Elizabeth Bates (eds.), The Crosslinguistic Study of Sentence Processing. 3–73. New York: Cambridge University Press. Bates, Elizabeth and MacWhinney, Brian (1982). Functionalist approaches to grammar. In Eric Wanner and Lila R. Gleitman (eds.), Language Acquisition: The State of the Art. 173–218. New York: Cambridge University Press. Beebe, Leslie and Zuengler, Jane (1983). Accommodation theory: An explanation for style shifting in second language dialects. In Nessa Wolfson and Elliot Judd (eds.), Sociolinguistics and Language Acquisition. 195–213. Rowley, MA: Newbury House. Bialystok, Ellen (1991). Metalinguistic dimensions of bilingual language proficiency. In Ellen Bialystok (ed.), Language Processing in Bilingual Children. 113–40. Cambridge: Cambridge University Press. Bialystok, Ellen (2001). Bilingualism in Development: Language, Literacy, and Cognition. Cambridge: Cambridge University Press. Bialystok, Ellen and Hakuta, Kenji (1994). In Other Words: The Science and Psychology of Second-Language Acquisition. New York: Basic Books. Bialystok, Ellen and Ryan, Ellen B. (1985). A metalinguistic framework for the development of first and second language skills. In Donna-Lynn Forrest-Pressley, G. E. Mackinnon, and T. Gary Waller (eds.), Metacognition, Cognition, and Human Performance: Vol. 1. Theoretical Perspectives. 207–52. Orlando: Academic Press. Birdsong, David (1992). Ultimate attainment in second language acquisition. Language 68: 706–55. Birdsong, David (2005). Nativelikeness and nonnativelikeness in L2A research. International Review of Applied Linguistics 43: 319–28.

Birdsong, David (2006). Age and second language acquisition and processing: A selective review. Language Learning 56: 9–49. Birdsong, David (2007). Nativelike pronunciation among late learners of French as a second language. In Ocke-Schwen Bohn and Murray J. Munro (eds.), Language Experience in Second Language Speech Learning. 99–116. Amsterdam: Benjamins. Birdsong, David and Molis, Michelle (2001). On the evidence for maturational constraints in second-language acquisition. Journal of Memory and Language 44: 235–49. Bley-Vroman, Robert (1990). The logical problem of foreign language learning. Linguistic Analysis 20: 3–49. Bloomfield, Leonard (1933). Language. London: Allen and Unwin. Bongaerts, Theo (2005). Introduction: Ultimate attainment and the critical period hypothesis for second language acquisition. International Review of Applied Linguistics 43: 259–67. Bongaerts, Theo, Mennen, Susan and von der Slik, Frans (2000). Authenticity of pronunciation in naturalistic second language acquisition. The case of very advanced late learners of Dutch as a second language acquisition. Studia Linguistica 54: 298–308. Canale, Michael (1983). From communicative competence to communicative language pedagogy. In Jack C. Richards and Richard W. Schmidt (eds.), Language and Communication. 2–27. London: Longman. Canale, Michael and Swain, Merrill (1980). Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics 1: 1–47. Cenoz, Jasone (2000). Research on multilingual acquisition. In Jasone Cenoz and Ulrike Jessner (eds.), English in Europe: The Acquisition of a Third Language. 39–53. Clevedon: Multilingual Matters. Cenoz, Jasone and Genesee, Fred (1998). Psycholinguistic perspectives on multilingualism and multilingual education. In Jasone Cenoz and Fred Genesee (eds.), Beyond Milingualism: Multilingualism and Multilingual Education. 16–32. Clevedon: Multilingual Matters. Cenoz, Jasone, Hufeisen, Britta, and Jessner, Ulrike (eds.) (2001). Cross-Linguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives. Clevedon: Multilingual Matters.

132  Neurological and Psychological Aspects of Bilingualism and Multilingualism Chi, Michelene T. H., Glaser, Robert, and Farr, Marshall J. (eds.) (1988). The Nature of Expertise. Hillsdale, NJ: Erlbaum. Chomsky, Noam (1965). Aspects of the Theory of Syntax. Cambridge, MA: MIT Press. Chomsky, Noam (1981). Lectures on Government and Binding. Dordrecht, Netherlands: Foris. Chomsky, Noam (1986). Knowledge of Language: Its Nature, Origin, and Use. New York: Praeger. Chomsky, Noam (1995). The Minimalist Program. Cambridge, MA: MIT Press. Clahsen, Harald and Muysken, Peter (1986). The availability of universal grammar to adult and child learners: A study of the acquisition of German word order. Second Language Research 2: 93–110. Cochrane, R. McCrae (1980). The acquisition of /r/ and /l/ by Japanese children and adults learning English as a second language. Journal of Multilingual and Multicultural Development 1: 331–60. Cook, Vivian (1992). Evidence for multicompetence. Language Learning 42(4): 557–91. Coppieters, René (1987). Competence differences between natives and near-native speakers. Language 63: 544–73. De Angelis, Gessica (2007). Third or Additional Language Acquisition. Clevedon: Multilingual Matters. De Groot, Annette M. B. (1992). Bilingual lexical representation: a closer look at conceptual representations. In Ram Frost and Leonard Katz (eds.), Orthography, Phonology, Morphology and Meaning. 389–412. Amsterdam: Elsevier. De Groot, Annette M. B. (2011). Language and Cognition in Bilinguals and Multilinguals. New York: Psychology Press. DeKeyser, Robert and Larson-Hall, Jennifer (2005). What does the critical period really mean? In Judith F. Kroll and Annette M. B. de Groot (eds.), Handbook of Bilingualism: Psycholinguistic Approaches. 88–108. Oxford: Oxford University Press. Dickerson, Lonna (1975). Interlanguage as a system of variable rules. TESOL Quarterly 9: 401–7. Dijkstra, Ton and van Hell, Janet G. (2003). Testing the language mode hypothesis using trilinguals. International Journal of Bilingual Education and Bilingualism 6(1): 2–16.

Duff, Patricia (2007). Second language socialization as sociocultural theory: insights and issues. Language Teaching 40: 309–19. Dulay, Heidi and Burt, Marina (1974). Natural sequences in child second language acquisition. Language Learning 24: 37–53. Eisenstein, Miriam and Starbuck, Robin (1989). The effect of emotional investment on L2 production. In Susan Gass, Carolyn Madden, Dennis Preston and Larry Selinker (eds.), Variation in Second Language Acquisition: Psycholinguistic Issues. 125–37. Clevedon: Multilingual Matters. Ellis, Rod (2003). The Study of Second Language Acquisition. Oxford: Oxford University Press. Ellis, Rod (2008). The Study of Second Language Acquisition, 2nd edn. Oxford: Oxford University Press. Felix, Sascha (1985). More evidence on computing cognitive systems. Second Language Research 1: 47–72. Fishman, Joshua. A. (1977). The social science perspective. In Bilingual Education: Current Perspectives. Social Science. 1–49. Arlington, VA: Center for Applied Linguistics. Flege, James E. (1999). Age of learning and second language speech. In David Birdsong (ed.), Second Language Acquisition and the Critical Period Hypothesis. 101–31. Mahwah, NJ: Erlbaum. Flynn, Suzanne (1991). Government-binding: Parameter setting in second language acquisition. In Charles Ferguson and Thom Huebner (eds.), Crosscurrents in Second Language Acquisition and Linguistic Theories. 143–67. Amsterdam: John Benjamins. Flynn, Suzanne (1996). A parameter-setting approach to second language acquisition. In William C. Ritchie and Tej K. Bhatia (eds.), Handbook of Second Language Acquisition. 121–58. San Diego, CA; Academic Press. Gagné, Ellen D. (1985). The Cognitive Psychology of School Learning. Boston, MA: Little, Brown. Gardner, Rod and Wagner, Johannes (eds.) (2004). Second Language Conversations. London: Continuum. Genesee, Fred, Hamers, Josiane, Lambert, Wallace E., Mononen, L., Seitz, M., and Starck, R. (1978). Language processing in bilinguals. Brain and Language 5: 1–12. Gregg, Kevin (1990). The variable competence model of second language acquisition and why it isn’t. Applied Linguistics 11: 364–83.

Bilingualism/Multilingualism and Second-Language Acquisition  133 Grosjean, François (1982). Life with Two Languages: An Introduction to Bilingualism. Cambridge, MA: Harvard University Press. Grosjean, François (1997). The bilingual individual. International Journal of Research and Practice in Interpreting 2(1/2): 163–87. Grosjean, François (1998). Studying bilinguals: methodological and conceptual issues. Bilingualism: Language and Cognition 1(2), 131–49. Grosjean, François (2010). Bilingual: Life and Reality. Cambridge, MA: Harvard University Press. Hakuta, Kenji (1986). Mirror of Language: The Debate on Bilingualism. New York: Basic Books. Hakuta, Kenji (2001). A critical period for second language acquisition? In Donald Bailey, John Bruer, Frank Symons, and Jeff Lichtman (eds.), Critical Thinking about Critical Periods. 193–205. Baltimore, MD: Paul H. Brookes. Hakuta, Kenji, Bialystok, Ellen, and Wiley, Edward (2003). Critical evidence: A test of the critical period hypothesis for second language acquisition. Psychological Science 14(1): 31–8. Hamers, Josiane F. and Blanc, Michel (2000). Bilinguality and Bilingualism. 2nd edn. Cambridge: Cambridge University Press. Hammarberg, Björn (2001). Roles of L1 and L2 in L3 production and acquisition. In Jasone Cenoz, Britta Hufeisen, and Ulrike Jessner (eds.), Cross-Linguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives. 21–41. Clevedon: Multilingual Matters. Harris, Roy and Taylor, Talbot (1997). Landmarks in Linguistic Thought I: The Western Tradition from Socrates to Saussure. 2nd edn. London: Routledge. Haugen, Einar (1953). The Norwegian Language in America. Philadelphia: University of Pennsylvania Press. Haznedar, Belma and Gavruseva, Elena (2008). Recent perspectives in child second language acquisition. In Belma Haznedar and Elena Gavruseva (eds.), Current Trends in Child Second Language Acquisition: A Generative Perspective. 3–15. Amsterdam: John Benjamins. Heredia, Roberto R. and Brown, Jeffrey M. (2004). Bilingual memory. In Tej K. Bhatia and William C. Ritchie (eds.), The Handbook of Bilingualism. 225–49. Malden, MA: Blackwell. Herdina, Philip and Jessner, Ulrike (2000). A Dynamic Model of Multilingualism: Changing the

Psycholinguistic Perspective. Clevedon: Multilingual Matters. Hoffmann, Charlotte (2001). Towards a description of trilingual competence. International Journal of Bilingualism 5(1): 1–17. Hyltenstam, Kenneth and Abrahamsson, Niclas (2001). Comments on Stefka H. MarinovaTodd, D. Bradford Marshall, and Catherine E. Snow’s ‘three misconceptions about age and L2 learning’. TESOL Quarterly 35(1): 151–76. Hyltenstam, Kenneth and Abrahamsson, Niclas (2003). Maturational constraints in SLA. In Catherine J. Doughty and Michael H. Long (eds.), The Handbook of Second Language Acquisition. 539–88. Oxford: Blackwell. Hymes, Dell H. (1972). On communicative competence. In John B. Pride and Janet Holmes (eds.), Sociolinguistics: Selected Readings. 269–93. Harmondsworth: Penguin. Inoue, Masayoshi (1996). Kaosu to fukuzastukei no kagaku [Science of chaos and complexity]. Tokyo: Nihon Jistugyo. Ioup, Georgette, Boustagui, Elizabeth, El Tigi, Manal, and Moselle, Martha (1994). Reexamining the critical period hypothesis: A case study in a naturalistic environment. Studies in Second Language Acquisition 16: 73–98. Jarvis, Scott (2000). Methodological rigor in the study of transfer: Identifying L1 influence in the interlanguage lexicon. Language Learning 50(2): 245–309. Jarvis, Scott (2011). Conceptual transfer: Crosslinguistic effects in categorization and construal. Bilingualism: Language and Cognition 14(1): 1–8. Jessner, Ulrike (2008). Teaching third languages: Findings, trends and challenges. Language Teaching 41(1): 15–56. Jordan, Geoff (2004). Theory Construction in Second Language Acquisition. Amsterdam: John Benjamins. Kellerman, Eric (1977). Towards a characterization of the strategy of transfer in second language learning. Interlanguage Studies Bulletin 2: 56–68. Kellerman, Eric (1983). Now you see it, now you don’t. In Susan Gass and Larry Selinker (eds.), Language Transfer in Language Learning. 112–34. Rowley, MA: Newbury House. Kellerman, Eric and Sharwood-Smith, Michael (eds.) (1986). Crosslinguistic Influence in Second Language Acquisition. New York: Pergamon Press.

134  Neurological and Psychological Aspects of Bilingualism and Multilingualism Kemp, Charlotte (2009). Defining multilingualism. In Larissa Aronin and Britta Hufeisen (eds.), The Exploration of Multilingualism. 11–26. Amsterdam: John Benjamins. Koda, Keiko (2008). Impacts of prior literacy experience on second-language learning to read. In Keiko Koda and Annette M. Zehler (eds.), Learning to Read across Languages: Cross-Linguistic Relationships in First- and Second-Language Literacy Development. 68–96. New York: Routledge. Kramsch, Claire and Whiteside, Anne (2007). Three fundamental concepts in second language acquisition and their relevance in multilingual contexts. The Modern Language Journal 91: 907–22. Krashen, Stephen (1973). Lateralization, language learning, and the critical period: Some new evidence. Language Learning 23: 63–74. Krashen, Stephen (1977). Some issues relating to the Monitor Model. In H. Douglas Brown, Carlos Yorio, and Ruth Crymes (eds.), On TESOL ’77. 144–58. Washington, DC: TESOL. Kroll, Judith F. and Stewart, Erika (1994). Category interference in translation in picture naming: Evidence for asymmetric connection between bilingual memory representations. Journal of Memory and Language 33: 149–74. Kroll, Judith F. and Tokowicz, Natasha (2005). Models of bilingual representation and processing. In Judith F. Kroll and Annette M. B. de Groot (eds.), Handbook of Bilingualism: Psycholinguistic Approaches. 531–53. New York: Oxford University Press. Labov, William (1972). Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Lafford, Barbara A. (2006). The effects of study abroad vs. classroom contexts on Spanish SLA: Old assumptions, new insights and future research directions. In Carol. A. Klee and Timothy. L. Face (eds.), Selected Proceedings of the 7th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages. 1–25. Somerville, MA: Cascadilla Proceedings Project. Lambert, Wallace E. (1974). Culture and language as factors in learning and education. In Frances E. Aboud and Robert D. Meade (eds.), Cultural Factors in Learning and Education. 91–122. Bellingham, WA: Western Washington State University.

Lantolf, James P. (2000). Sociocultural Theory and Second Language Learning. Oxford: Oxford University Press. Lasen-Freeman, Diane (2007). Reflecting on the cognitive-social debate in second language acquisition. The Modern Language Journal 91 (Focus Issue): 773–87. Larsen-Freeman, Diane and Cameron, Lynne (2008). Research methodology on language development from a complex systems perspective. The Modern Language Journal 92(2): 200–13. Lenneberg, Eric (1967). Biological Foundation of Language. New York: Wiley and Sons. Liceras, Juana Muñoz (2010). Second language acquisition and syntactic theory in the 21st century. Annual Review of Applied Linguistics 30: 248–69. Long, Michael H. (1990). Maturational constraints on language development. Studies in Second Language Acquisition 12: 251–85. Long, Michael H. (1996). The role of linguistic environment in second language acquisition. In William C. Ritchie and Tej K. Bhatia (eds.), Handbook of Second Language Acquisition. 413–68. San Diego: Academic Press. Long, Michael H. (1997). Construct validity in SLA research: A response to Firth and Wagner. The Modern Language Journal 81: 318–23. Long, Michael H. (2005). Problems with supposed counter-evidence to the Critical Period Hypothesis. International Review of Applied Linguistics 43: 287–317. Lowenthal, Kate and Bull, Deborah (1984). Imitation of foreign sounds: What is the effect of age? Language and Speech 27: 95–7. Luk, Zoe Pei-sui and Shirai, Yasuhiro (2009). Is the acquisition order of grammatical morphemes impervious to L1 knowledge? Evidence from the acquisition of plural –s, articles, and possessive ‘s. Language Learning 54(9): 721–54. Mackey, William F. (1967). Bilingualism as a World Problem. Montreal: Harvest House. MacWhinney, Brian (1997). Second language acquisition and the competition model. In Annette M. B. De Groot and Judith F. Kroll (eds.), Tutorials in Bilingualism: Psycholinguistic Perspectives. 113–42. Mahwah, NJ: Erlbaum. MacWhinney, Brian and Bates, Elizabeth (eds.) (1989). The Cross-Linguistic Study of Sentence Processing. Cambridge: Cambridge University Press.

Bilingualism/Multilingualism and Second-Language Acquisition  135 McLaughlin, Barry (1990). Restructuring. Applied Linguistics 11: 1–16. McLaughlin, Barry, Rossman, Tammi, and McLeod, Beverly (1983). Second-language learning: An information-processing perspective. Language Learning 33: 135–58. McNamara, Tim (1996). Measuring Second Language Performance. London: Longman. Meisel, Jürgen M. (1983). Transfer as a second language strategy. Language and Communication 3: 11–46. Meisel, Jürgen M. (2008). Child second language acquisition or successive first language acquisition? In Belma Haznedar and Elena Gavruseva (eds.), Current Trends in Child Second Language Acquisition: A Generative Perspective. 55–80. Amsterdam: John Benjamins. Mohanty, Ajit K. and Perregaux, Christian (1997). Language acquisition and bilingualism. In John W. Berry, Pierre R. Dasen, and T. S. Saraswathi (eds.), Handbook of Cross-Cultural Psychology: Basic Processes and Human Development, Vol. 2. 217–53. Moyer, Alene (1999). Ultimate attainment in L2 phonology: The critical factors of age, motivation, and instruction. Studies in Second Language Acquisition 21: 81–108. Muljani, D., Koda, Keiko, and Moates, Danny R. (1998). Development of L2 word recognition: A connectionist approach. Applied Psycholinguistics 19: 99–114. Nicholas, Howard and Lightbown, Patsy M. (2008). Defining child second language acquisition: Defining roles for L2 instruction. In Jenefer Philip, Rhonda Oliver, and Alison Mackey (eds.), Second Language Acquisition and the Younger Learner. 27–51. Amsterdam: John Benjamins. Odlin, Terence (1989). Language Transfer: Cross Linguistic Influence in Language Learning. Cambridge: Cambridge University Press. Odlin, Terence (2003). Cross-linguistic influence. In Catherine J. Doughty and Michael H. Long (eds.), The Handbook of Second Language Acquisition. 436–86. Malden, MA: Blackwell. Odlin, Terence (2005). Crosslinguistic influence and conceptual transfer: What are the concepts? Annual Review of Applied Linguistics 25: 3–25. Olson, Linda L. and Samuels, S. Jay (1973). The relationship between age and accuracy of foreign language pronunciation. Journal of Educational Research 66: 263–7.

Paradis, Johanne (2007). Early bilingual and multilingual acquisition. In Peter Auer and Li Wei (eds.), Handbook of Multilingualism and Multilingual Communication. 15–43. Berlin: Mouton de Gruyter. Peal, Elizabeth and Lambert, Wallace E. (1962). The relation of bilingualism to intelligence. Psychological Monographs 76: 1–23. Penfield, Wilder and Roberts, Lamar (1959). Speech and Brain Mechanisms. New York: Athenaeum. Pinker, Steven (1994). The Language Instinct. New York: Morrow. Preston, Dennis (1993). Variation linguistics and SLA. Second Language Research 9: 153–72. Preston, Dennis (2002). A variationist perspective on SLA: Psyholinguistic concerns. In Robert B. Kaplan (ed.), Oxford Handbook of Applied Linguistics. 141–59. Oxford: Oxford University Press. Purpura, James E. (2008). Assessing communicative language ability: Models and their components. In Elana Shohamy and Nancy H. Hornberger (eds.), Encyclopedia of Language and Education, Vol. 7. Language Testing and Assessment, 2nd edn. 53–68. New York: Springer. Ringbom, Håkan (1987). The Role of L1 in Foreign Language Learning. Clevedon: Multilingual Matters. Ringbom, Håkan (2007). Cross-Linguistic Similarity in Foreign Language Learning. Clevedon: Multilingual Matters. Rumelhart, David E. and McClelland, James L. (1986). On learning the past tense of English verbs. In James L. McCelland, David E. Rumelhart, and the PDP Research Group (eds.), Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Volume 2. Psychological and Biological Models, 216–71. Cambridge, MA: MIT Press. Rumelhart, David E. and Ortony, Andrew (1977). The representation of knowledge in memory. In Richard Anderson, Rand Spiro, and William Montague (eds.), Schooling and the Acquisition of Knowledge. 99–135. Hillsdale, NJ: Erlbaum. Schachter, Jacquelyn (1986). In search of systematicity in interlanguage production. Studies in Second Language Acquisition 8: 119–34. Schachter, Jacquelyn (1988). Second language acquisition and its relationship to Universal Grammar. Applied Linguistics 9: 219–35.

136  Neurological and Psychological Aspects of Bilingualism and Multilingualism Schmidt, Richard (1977). Sociolinguistic variation and language transfer in phonology. Working Papers on Bilingualism 12: 79–95. Schwartz, Bonnie D. (2003). Child L2 acquisition: Paving the way. In Barbara Beachley, Amanda Brown and Frances Conlin (eds.), Proceedings of the 27th Boston University Conference on Language Development. 26–50. Somerville, MA: Cascadilla. Scovel, Thomas (1988). A Time to Speak: A Psycholinguistic Inquiry into the Critical Period for Human Speech. New York: Newbury House. Segalowitz, Norman (2000). Automaticity and attentional skill in fluent performance. In Heidi Riggenbach (ed.), Perspectives on Fluency. 200–19. Ann Arbor: University of Michigan Press. Seliger, Herbert W. (1978). Implications for a multiple critical period hypothesis for second language learning. In William C. Ritchie (ed.), Second Language Acquisition Research: Issues and Implications. 11–19. New York: Academic Press. Selinker, Larry (1972). Interlanguage. International Review of Applied Linguistics 10: 209–30. Shiffrin, Richard M. and Schneider, Walter (1977). Controlled and automatic human information processing: II. Perceptual learning, automatic attending, and a general theory. Psychological Review 84: 127–90. Singleton, David (2005). The critical period hypothesis: A coat of many colors. International Review of Applied Linguistics in Language Teaching 43: 269–85. Simon, Herbert A. and Chase, William G. (1973). Skill in chess. American Scientist 61: 394–403. Slavoff, Georgina R. and Johnson, Jacqueline S. (1995). The effect of age on the rate of learning a second language. Studies in Second Language Acquisition 17: 1–16. Snow, Catherine E. and Hoefnagel-Höhle, Marian (1978). The critical period for language acquisition: Evidence from second language learning. Child Development 49: 1114–28. Stevens, Gillian (2006). The age-length-onset problem in research on second language

acquisition among immigrants. Language Learning 56: 671–92. Stowe, Laurie A. and Sabourin, Laura (2005). Imaging the processing of a second language: Effects of maturation and proficiency on the neural processes involved. International Review of Applied Linguistics 43: 329–53. Tahta, Sonia, Wood, Margaret, and Lowenthal, Kate (1981). Age changes in the ability to replicate foreign pronunciation and intonation. Language and Speech 24: 363–72. Tarone, Elaine (1983). On the variability of interlanguage systems. Applied Linguistics 4: 143–63. Tarone, Elaine (1985). Variability in interlanguage use: A study of style-shifting morphology and syntax. Language Learning 35: 373–95. Tarone, Elaine (2007). Sociolinguistic approaches to second language acquisition research – 1997–2007. The Modern Language Journal 91: 837–48. Tarone, Elsinr, Bigelow, Martha, and Hansen, Kit (2009). Literacy and Second Language Oracy. Oxford: Oxford University Press. Valdés, Guadalupe and Figueroa, Richard A. (1994). Bilingualism and Testing: A Special Case of Bias. Norwood, NJ: Ablex Publishing. Weinreich, Uriel (1953). Languages in Contact. The Hague: Mouton. White, Lydia (2003). On the nature of interlanguage representation: Universal grammar in second language. In Catherine J. Doughty and Michael H. Long (eds.). The Handbook of Second Language Acquisition. 19–42. Malden, MA: Blackwell. Zobl, Helmut (1984). Aspects of reference and the pronominal syntax preference in the speech of young child L2 learners. In Roger Andersen (ed.), Second Languages: A CrossLinguistic Perspective. 375–91. Rowley, MA: Newbury House. Zuengler, Jane (1989). Performance variation in NS-NNS interactions: Ethnolinguistic difference, or discourse domain? In Susan Gass, Carolyn Madden, Dennis Preston, and Larry Selinger (eds.), Variation in Second Language Acquisition: Discourse and Pragmatics. 228–44. Clevedon: Multilingual Matters.

6  Multilingualism: New Perspectives on Syntactic Development ÉVA BERKES AND SUZANNE FLYNN

Introduction Linguistic diversity is a topic of increasing interest due not only to its manifold practical effects on politics and even on the economy, but also to its theoretical importance for the human sciences. According to the Ethnologue (www. ethnologue.com) there are 6,909 known living languages in the world. Just five states are listed as having only one indigenous living language and merely six countries appear with the lowest possible value, 0, for the Greenberg’s diversity index (Greenberg 1956), indicating that any two persons in the country selected at random would have the same mother tongue (Lewis 2009, Table 7). ‘Bilingualism – more generally, multilingualism – is a commonplace fact of life in the world today . . . two thirds of the world’s children grow up in a bilingual environment (Bhatia and Ritchie 2004: 1).’ (See also Grosjean 2004; Chomsky as cited in Mukherji, Patnaik, and Agnihotri 2000 among others.) This state of affairs has far-reaching implications not only for society but for the individual as well. Therefore, it seems to be crucial to study multilingualism not only from the perspective of what its societal consequences are but research must also address the nature of the bi- or multilingual mind. The study of language as a unique human capacity helps to shed light on the way in which the human brain and cognition work; the study of the ability to use language deepens and enlightens our understanding of the nature of human cognitive processes. Hence understanding multilingualism provides an important window on one aspect of the human brain and cognition, viz. the mental processes involved in language learning and use.

The Handbook of Bilingualism and Multilingualism, Second Edition. Edited by Tej K. Bhatia and William C. Ritchie. © 2013 Blackwell Publishing Ltd. Published 2013 by Blackwell Publishing Ltd.

138  Neurological and Psychological Aspects of Bilingualism and Multilingualism

Key Issues in First- and Second-Language Acquisition There have been outstanding attempts within the generative framework for linguistic theory to determine the character of the language faculty and its deployment in language learning. First-language (L1) acquisition studies have provided important insights into how the mind and language work. The pioneering work of Braine (1963), followed by that of Brown (1973), argued for the child developing its grammar using its own rules independent of those of an adult speaker, an assumption that paved the way to instituting L1 acquisition as a research field in its own right. L1 acquisition, however, encounters serious difficulties when attempts are made to explore the interaction of general cognitive and specifically linguistic processes in the learner’s mind, because limitations deriving from general developmental (that is, maturational) deficits bear upon the languagelearning process of children. In order to understand language and language learning as a crucial part of human cognitive processes, linguistic theory must accommodate the results provided by second-language (L2) acquisition studies, where L2 refers to a language acquired after infancy,1 for here maturational issues play practically no role in the process of language development.

The issue of access to Universal Grammar in L2 acquisition L2 acquisition research, then, provides an opportunity to examine language development independent of the influence of other developing cognitive processes, a necessary step in setting the basis for an adequate theory of language acquisition and competence (Epstein, Flynn, and Martohardjono 1996; Flynn, Martohardjono, and O’Neil 1998; Flynn and O’Neil 1988; Guasti 2002; Martohardjono 1993; Rizzi 2004). Accepting the Chomskyan theory of language acquisition, according to which it is the individual’s language faculty which generates knowledge of language by responding to language input, Universal Grammar (UG) is defined as the theory of this language faculty and of ‘the children’s pre-linguistic initial state’ (Chomsky 1981: 7). Consequently, L2 research had to meet the challenge of describing theoretically how L2 learners access UG in their language development of the L2. Some of the most representative theories are developed in Bley-Vroman (1989), Clahsen and Muysken (1986), Johnson (1988), Johnson and Newport (1989), Schwartz and Sprouse (1996), Hawkins and Chan (1997), Epstein et al. (1996), Flynn (1983 and subsequent work); see Eubank (1991) for arguments and counterarguments on these theories. For a detailed review on the different hypotheses about access to UG in L2 acquisition, see e.g., White (1989, 1998, 2003) and Cook and Newson (2007). For general treatments of the field of second language, see the chapters in Ritchie and Bhatia (1996, 2009). Two principal models were developed to capture possible ways in which UG may be represented in L2 language acquisition. The first, the maturation model,

Multilingualism: New Perspectives on Syntactic Development  139 which was based on the maturation hypothesis (Borer and Wexler 1987; Felix 1984), claims that for L2 acquisition, UG is available to the L2 learner only in the form of a language-specific grammar viz., the L1, which is supposed to be the beginning basis, the initial state S0, for the development of the L2 grammar. Thus, under this hypothesis, new language knowledge is acquired through L1 alone via some yet to be defined transfer mechanism. The strong continuity model (Flynn and Lust 2002; also in Flynn, 2009 as ‘constant model’) is based on the strong continuity hypothesis as articulated for the study of L1 acquisition (Lust, 1999, 2006; Boser et al. 1992; for examples see Boser et al. 1995; Whitman, Lee, and Lust 1991). This model claims that it is not UG which changes over time but the theory-building capacity of the L2 learner. According to this model, UG remains distinct from the developing language-specific grammar and is continuously available for the L2 learner, as implicit principles that guide and constrain language acquisition. Learners with the help of their UG, the actual initial state (S0), which triggers a language-specific grammar in view of new language input, map the language-specific grammar by dissociating and integrating grammatical components to construct their new language-specific grammars, a process referred to as ‘grammatical mapping’ (Flynn and Lust 2002; Flynn et al. 2005, also in Lust 2006). Accepting the strong continuity model of UG does not mean denying obvious differences between the processes of L1 and L2 acquisition, but, rather, claiming that these differences are not due to a change in UG. Furthermore, new languagespecific knowledge is not accessed only through L1, but benefits from the knowledge UG may provide. Given this model, the initial state is not absolute, rather ‘the state of the mind/brain prior to experience with particular data and a particular new acquisition task’ (Flynn and Lust 2002: 114) or in Chomsky’s words: the standard idealized model of language acquisition takes the initial state S0 to be a function mapping experience (primary linguistic data, PLD) to a language. UG is concerned with the invariant principles of S0 and the range of permissible variation (Chomsky 1995: 169).

Given this characterization of the S0, the testing of hypotheses about the initial state in L2 acquisition is crucial to the development of an understanding of the L2 learner’s representation of grammar at the point of encountering the L2 input for the first time (White 1989, 1998, 2003; see related discussions on the nature of the L1 in the L2 grammar in Gass and Selinker 1983, 1992).

The issue of ‘transfer’ from L1 to L2 Another issue that has given way to fervent debates is the relationship of the two language grammars (L1 vs. L2) in the learner’s mind, which raises the question of transfer between languages. The acquisition of Wh-structures may serve as illustration here. Consider the following English sentences (Martohardjono and Flynn 1995: 213):

140  Neurological and Psychological Aspects of Bilingualism and Multilingualism (1)  a  Whati did Tom fix ei [that the man had broken]? b  *Whoi did Tom fix the door [that ei had broken]? English manifests overt Wh-movements (1a), although constrained, as exem­ plified in (1b); whereas, in Indonesian and Chinese there is no such overt Wh-movement (see Martohardjono and Flynn 1995 and references given therein). Thus, knowledge of Indonesian and Chinese does not provide an L2 learner of English from these two languages any knowledge about movement constraints in English. According to the claim that L2 learners transfer syntactic knowledge from their L1 and thus construct their L2 language-specific grammar, L1 Indonesian or Chinese learners of L2 English are expected not to be able to derive Wh-movement constraints for English as they do not apply them in their L1s. The prediction would be that they would not differentiate (1a) and (1b) above. Before we evaluate how transfer is understood in the context of the aforementioned models, here we would like to give a short review of the evolution of this concept in second-language acquisition. ‘Language transfer’ is the term which has been most frequently used in L2 research, but there is an extensive literature using different terms to refer to some certain aspect of the same phenomenon. For illustration, Weinreich (1953) used the term ‘interference’ while Schachter and Rutherford (1979) and Ringbom (1987) employed ‘cross-linguistic influence’ to refer to the negative effects the native language has on L2 acquisition. Lado (1957), Selinker (1972), Kellerman (1983), and Odlin (1989), for example, use ‘language transfer’ in a wider, more neutral way to refer to both facilitative and negative influence of the L1, while Gass (1983, 1984) uses the same term to refer to the determinant role of L1 patterns in the course of L2 acquisition. Anderson’s (1983) research focused on transfer between languages exhibiting typological similarity with regard to some particular, mainly morpho-syntactic feature and which is perceived by the learner as similar to his L1. The following Transfer to somewhere principle, therefore, considers language transfer as a conscious process, assuming a certain linguistic awareness on the part of the learner. Kellerman (1983) suggested first the term ‘psychotypology’ to refer to the learner’s perception of the typological characterization of the new language. Kellerman’s (1995) Transfer to nowhere principle shifts the general focus on transfer of syntactic structures to a completely different perspective in L2 research. The principle indicates that it is rather the conceptual organization of the L1, which is likely to be superposed onto L2 patterns (see also Jarvis 1997, 1998; Jarvis and Pavlenko 2008; Pavlenko 2000; Pavlenko and Jarvis 2002; and Jarvis 2007 and Odlin 2005 on detailed reviews of conceptual transfer). Jarvis (1997) offers an example of conceptual transfer. In his study he observed that Finnish-speaking learners of English tend to use the same verb to refer to a ‘collision’ regardless of whether it applies to humans or objects. Finnish does not differentiate between humans and objects with respect to the verb ‘collision.’ In contrast, Swedish speakers use two different verbs to distinguish these two types of ‘collision.’ Jarvis argues that when Swedish speakers learn English as an L2, they have difficulty when attempting to apply the singular concept of ‘collision’ to both humans and objects. He

Multilingualism: New Perspectives on Syntactic Development  141 concluded that the Swedish L2 learners conceptually categorize these two types as typologically different and attempt to transfer these concepts into their English interlanguage. This later move to conceptual transfer shifts the research focus from syntactic development (see detailed reviews on variables that affect crosslinguistic influence, e.g., Murphy 2003; Jarvis and Pavlenko 2008; Cook 2010; Rothman and Cabrelli Amaro 2010). Returning to the two models of UG access in L2 acquisition, we would like to point out a fundamental difference in the predictions made by each with respect to syntactic transfer. Given the maturation model, it follows that adult L2 learners are guided by transfer from their L1, as UG has developed into the grammar of the L1, the initial state for L2 acquisition. In contrast, the strong continuity model hypothesizes that it is UG that guides the acquisition of an L2, as it does in L1 acquisition. Thus, at some level, both models predict some form of constraint by UG. In the maturation model it is only via the L1. It is not clear how this model predicts the development of the new target language (see extended discussion in Epstein et al. 1996). In the strong continuity model, UG in its entirety is available to the learner. Consequently, this model predicts that language learners do not ‘transfer’; that is, they do not build the new grammar using the language-specific grammatical features of L1, but rather they construct the new grammar with the help of ‘grammatical mapping,’ i.e., they ‘map from one primary structure to a more developed structure by dissociating modular grammatical components and integrating them in the “assembly” of new language-specific grammars’ (Flynn et al. 2005: 2). All in all, the strong continuity model rejects transfer on the level of fundamental computational mechanisms which characterize the language faculty, as a result of understanding language learning not as a maturational, but rather a computational process. However, the strong continuity model does recognize the role of other languages known in terms of subsequent language development as will be discussed in more detail below.

Multilingual Acquisition Yet, there is more to it. The manner in which a multilingual learner approaches subsequent language learning is widely accepted to be different from that of an L2 learner in certain respects,2 for example, the L3 learner may rely on certain language-learning strategies picked up in the course of acquiring an L2. It seems empirically that the acquired knowledge of at least one L2 may well play a role in subsequent acquisition (see e.g., Cenoz, Hufeisen, and Jessner 2001; Hammarberg 2009; Leung 2009). Cook (1992) refers to the putatively qualitatively different linguistic competence of a multilingual learner as ‘multi-competence.’ There are ongoing fervent debates on multilingual acquisition and its relation to cross-linguistic influence, for the knowledge of an L2 (or various L2s) and with it the experience of having acquired a nonnative language as well as having developed language-learning strategies add to the complexity of issues such as access to UG and transfer from prior languages in the initial state of L3 acquisition.

142  Neurological and Psychological Aspects of Bilingualism and Multilingualism Numerous studies and volumes in recent years are dedicated to the exploration of how L1 and/or L2 bear upon the acquisition of subsequent language(s), i.e., multilingual acquisition (see e.g., Cenoz et al. 2001; De Angelis 2007; Gabrys´Barker 2006; Gibson, Hufeisen, and Libben 2001; Hammarberg 2009; Odlin, Alonso, and Alonso-Vázquez 2006; Odlin and Jarvis, 2004; Leung 2005, 2009; Ringbom 1987). For simplicity’s sake we will use the term ‘L3 acquisition’ to refer to the acquisition process of any multilingual learner whether of a third language, a fourth language, etc. Generative multilingual research may render answers that the study of L2 acquisition alone cannot. For example, L3 provides insights about the role of the L1 in subsequent language acquisition. Is there a privileged role for the L1? An answer to this question would inform debates concerning the availability of UG for next-language acquisition in general. More specifically, the study of L3 acquisition raises issues concerning the maturation model itself. What is available to the L3 learner in terms of UG – an L1 that has partially morphed into an L2? What would this mean? Clearly, the study of L3 acquisition has the potential to inform current debates concerning the role of UG in subsequent language acquisition as well as confronting claims made about subsequent language learning (see the discussion of recent models for multilingual acquisition below). Development of explanatorily adequate models of language need to account for L3 acquisition and in doing so need to delineate the nature of the initial state for all subsequent language learning. Much research has been done to explore the role of previous lexical knowledge at the L3 initial state and has suggested that there might be some form of lexical influence from the L2 to the L3 (see e.g., Cenoz et al. 2001; Boot 2004; Dewaele 1998; Gibson et al. 2001; Hammarberg 2001; Williams and Hammarberg 1998), whereas there have been few such attempts in exploring the nature of prior syntactic knowledge. Nevertheless, the emerging research within the framework of generative morpho-syntax has inspired numerous empirical studies necessary for further exploration of how language learning occurs. In the following, we will introduce three of the most influential current models of multilingual acquisition: the L2 status factor; the typological primacy model; and the cumulative enhancement model for language acquisition. All of these models coincide in assuming that multilingual learners a priori have access to UG and that they process their L3 under altered conditions from that of their L2; that is, previously attained linguistic knowledge may act as some type of source for cross-linguistic transfer. The L2 status factor and typological primacy model aim to describe the initial state of L3 acquisition; they attempt to tease apart the intervening variables in order to determine the most likely source of syntactic transfer for L3 acquisition, whereas the cumulative enhancement model for language acquisition attempts to create the representation of language development in the mind of a multilingual learner and at the same time to inform us about the initial state for subsequent language learning. In the following, we will briefly introduce the mentioned models in light of the research focus reviewed here. As its name suggests, the L2 status factor (L2SF)

Multilingualism: New Perspectives on Syntactic Development  143 (Bardel and Falk 2007, in press; Falk and Bardel 2010; Williams and Hammarberg 1998) describes a general tendency on the part of L3 language learners to activate their last-learned language when acquiring an L3; that is, it proposes that, when it comes to recounting the variables that influence syntactic transfer, the lastlearned language plays a primary role in determining the L3 initial state due to the higher degree of sociolinguistic and cognitive congruence surrounding the acquisition of L2, L3, . . . , Ln. These factors include age of onset, proficiency, exposure and use of the new target language as well as learning strategies, metalinguistic knowledge, and so on. The model focuses on the learners’ tendency to activate the last-learned language rather than their L1, especially in the initial state of L3 acquisition (but see Falk and Bardel 2010, making the same claim for more advanced learners of L3), and predicts a general tendency to activate and transfer linguistic knowledge and/or elements of the last-learned language into subsequent L3 language acquisition. Bardel and Falk (2007) propose, furthermore, that the L2SF guides syntactic transfer as well; that is, they posit that syntactic development in subsequent acquisition is also affected by the specific syntactic features of the last-learned language (for a different view, see Håkansson, Pienemann and Sayehli 2002). This claim is based on the experimental study of the placement of negation at the initial stage of nonnative acquisition of L3 Swedish and Dutch by two groups of learners with different linguistic backgrounds. Bardel and Falk (2007) start from the theoretical assumption argued for within the generative framework that the verb-second (V2) phenomenon in the Germanic languages (except for English) is due to the raising of the verb past negation (see Vikner and Schwartz 1996 for a review). Bardel and Falk assume then that if learners produce a ‘verb + negator’ string in their L3 Swedish or Dutch, it clearly indicates that verb raising is part of the grammar of the interlanguage of these learners. Subjects were placed into two groups according to the V2 property of the languages they knew prior to experiment, having either an L1+V2/L2non-V2 or the opposite, an L1non-V2/L2+V2. Results showed that learners with L2+V2 outperformed learners with L1+V2 in producing post-verbal negation at their initial stage of L3+V2, concluding that learners activated their L2, rather than relying on their L1, at the moment of facing the linguistic input from the new language. Furthermore, none of the L1+V2 learners seemed to transfer systematically from their L1, although L1 matched L3 in this syntactic aspect, which led the authors to claim that L1 for these learners was, in fact, inaccessible. The predictive power of the L2SF was further tested in a subsequent study (Falk and Bardel 2010), which looked at the placement of object pronouns by intermediate L3 German learners collocated into two groups, a FrenchL1/EnglishL2 and an EnglishL1/FrenchL2 group. Results were similar to the previous study; significant differences were attested between the performances of the two groups in a way that the overall outcome confirmed the strong role of the L2 status factor. According to this, learners are driven by their cognition-based learning strategies acquired in the course of L2 acquisition, independent of any typological similarities among the involved languages.

144  Neurological and Psychological Aspects of Bilingualism and Multilingualism Consequently, the L2SF contradicts the typological primacy model (TPM) (Rothman 2010; Rothman and Cabrelli Amaro 2010), which does not reject L1 transfer. The model seeks to predict which set of language properties a multilingual learner is likely to transfer when learning a subsequent language Ln. Based on the theory of the (psycho)typologically motivated transfer (Kellerman 1983), the TPM suggests that the learner may choose, with the help of an internal parser, the typologically more similar system for multilingual transfer to facilitate the acquisition of Ln. The real typological proximity or the typological proximity as perceived by the learner is to guide the learner in selecting the system to be transferred to facilitate the acquisition of Ln. Two subject groups were tested on their interpretation of noun–adjective order in romance determiner phrases (DPs) at an intermediate level of L3: ItalianL1/EnglishL2 learners of L3 Spanish and EnglishL1/SpanishL2 learners of L3 Brazilian Portuguese. Romance adjectival constructions involve the raising of the head noun of a noun phrase (NP) to a position before an adjective, unlike English where the noun remains in its underlying position after the adjective; therefore, the interpretation of adjectival constructions can be regarded as an indicator of the underlying mental representations of the syntactic features involved in raising (or not raising) the head noun to the position before the adjective. Results indicated that all learners had an almost target-like knowledge of these constructions, both syntactic and semantic; none of the groups, nor any of the individual subjects, differed from the others significantly in their tacit knowledge of noun raising in romance. The experiment in its attempt to tease apart the variables that play the most crucial roles in syntactic transfer disconfirms the claim that L2 has a leading role as proposed by the L2SF and concludes by claiming that the most likely source of selective transfer in multilingual acquisition is (psycho)typological proximity. In a subsequent experiment (see Rothman and Cabrelli Amaro 2010), which tested the TPM against the cumulative enhancement model, this conclusion was reiterated. To sum up, the L2SF and the TPM models agree that it is precisely the already existing grammars in the mind of the learner at the beginning of the acquisition process that make L3 language learning essentially and qualitatively different from L1 or L2 acquisition. Hence, the fundamental question these two models seek to investigate is the role of previously attained language-specific morphosyntactic knowledge at the initial state of L33 in attempts to tease apart the variables that guide perceived syntactic transfer. Without doubt the variables of the L2 status factor or typological affinity may prove to be crucial for learning, as they provide a legitimate basis for strategies the multilingual learner has access to and may apply in order to proceed. Nevertheless, the assumption that reduces the S0 of the multilingual learner when facing new input of a specific language to the result of one or more intervening variables does not provide an explanatorily adequate language acquisition theory as it does not attempt to explain the manner in which UG maps this new input into linguistic knowledge of L3. Furthermore, still left unexplained is the degree to which such variables can be part of or relate to the language faculty. In short, in order to provide an explanatorily adequate

Multilingualism: New Perspectives on Syntactic Development  145 theory of language acquisition – that is, the theory of how multilingual language development is constrained by UG – we need to explore the way multilingual learners build the grammar of a specific target language, and filter out universal tendencies that may shed light on how UG constrains development of a languagespecific grammar. A recently emerging effort to explain multiple-language acquisition in an attempt to answer the requirement of explanatory adequacy is the cumulative enhancement model (CEM) (see Flynn, Foley, and Vinnitskaya 2004). Its primary focus is to investigate the structural development of a specific Ln language. The fundamental prediction the model makes is that language learning is cumulative, and it excludes redundancy in linguistic representation, a claim that contradicts Hufeisen (1998) (see also Gibson et al. 2001), according to whom, despite the richer basis of experience, there is no trace of novel conditions for S0 beyond trilingualism. In Hammarberg’s words ‘Hufeisen’s claim still relies on reasoning and the absence of direct counter-evidence’ (Hammarberg 2010: 96); hence, we believe that the results reported in this chapter may provide direct empirical counter-evidence against such a claim. Here we attempt to show that the accumulated linguistic knowledge an L3 learner has enhances subsequent language learning; therefore, we undertake to explore what the linguistic nature of this enhancement might be. As a corollary we hope to provide insights into the S0 with which the multilingual learner starts off to construct the L3 grammar.

Theoretical Background According to the theoretical claim made by Chomsky (2000), there is linguistic development in the process of constructing language-specific grammars by learners. Consistent with the current view of language proposed within the generative framework of linguistics, language acquisition necessarily involves the correct selection and development of formal features on functional heads, which then guide the mapping of language-specific constituents and/or clausal architecture. This section builds upon a series of language development studies focusing on the acquisition of relative clauses in English. These studies seek to elucidate language development by contemplating the development of language-specific properties of the complementizer phrase (CP). It is assumed that the interpretable vs. non-interpretable features of the CP are responsible for the head–complement relationship among syntactic elements on the sentence level – specifically, the order of head and complement within CP in a given language. According to this position, specific features of the CP, generated by UG, determine the directionality of embedding (head directionality or branching) and the constituent word order within the subordinate clause (see Flynn and Foley 2004; Lust 2006). Therefore, we may talk about right- or left-branching languages with either subject– verb–object (SVO) or subject–object–verb (SOV) word order in the embedded clause.

146  Neurological and Psychological Aspects of Bilingualism and Multilingualism

The complementizer phrase (CP) setup Experience shows that head directionality may overlap in certain languages, like English, Spanish or Russian, which are strictly head-initial languages in the sense that they not only embed the relative clause after the relativized NP (right-branching) but also exhibit an SVO constituent word order within the relative clause. The CP feature setup is independent of whether the gap within the relative clause is in subject or in object position. The examples given below in (2) show the full range of combinations of the function of the head NP in the matrix clause and the function of the gap in relative clauses in English, in (2a) the head NP is the subject in the matrix clause, in (2b) the head NP is the object in the matrix clause; in the (i) examples the gap in the relative clause is in the subject position, and in the (ii) examples it is in the object position. Examples (3) show the schematic representation of subject-gap and object-gap relative clauses in English. (2) Strictly head-initial languages (right-branching and SVO – e.g., English) – examples: a Subject-head (the boy) in matrix clause i. Subject-gap in relative clause (referred to as an ‘SS’ or ‘subject– subject relative clause’): The boyi [CP whoi [ei/subject found the man]] saw the girl ii. Object-gap in relative clause (an SO relative): The boyi [CP whoi [the man found ei/object]] saw the girl b Object-head (the boy) in matrix clause i. Subject-gap in relative clause (OS relative): The girl saw the boyi [CP whoi [ei/subject found the man]] iii. Object-gap in relative clause (OO relative): The girl saw the boyi [CP whoi [the man found ei/object]] (3) Strictly head-initial languages – general schemas a Subject-gap in relative clause (SS or OS) Head NPi/subject/object [CP wh-phrasei [ei/subject V NPobject]] b Object-gap in relative clause (SO or OO) Head NPi/subject/object [CP wh-phrasei [NPsubject V ei/object]] By contrast, Japanese and Kazakh are left-branching (head-final) languages with an SOV constituent word order in the relative clause. The example in (4a) shows an object-gap relative clause (Saito 1985), the schematic representation of such relative clauses in Japanese is given in (4b). (4) Strictly head-final languages (left-branching and SOV – e.g., Japanese) – Object-gap relative clauses. a Example (an OO relative) John-wa [[CP Mary-gasubject ei/object kaita] [Head NP honi-o]] yonda John-topic Mary-nom wrote book-acc read ‘John read the book that Mary wrote’

Multilingualism: New Perspectives on Syntactic Development  147 b Schema (SO or OO) [CP [NPsubject ei/object V]] Head NPi/subject/object

Background Studies Design Of particular importance to this section are Flynn’s (1983, 1987) studies of adult L2 and Flynn et al.’s (2004) study of L3 acquisition of English. The L2 studies investigated adult L1 Japanese and L1 Spanish speakers learning L2 English and the L3 study investigated KazakhL1/RussianL2 speakers learning L3 English. All of these studies used the same test sentences – an adaptation of the original study of Flynn and Lust (1980) of monolingual children acquiring English – and employed the same procedure, an elicited imitation task. Results of these studies revealed that in order to examine how the CP develops in L2 acquisition, it is necessary to test learners on their handling of free relatives, because free relatives appeared to be developmental precursors to headed relatives in the process of building a full-fledged, language-specific CP architecture (Flynn et al. 2004; Flynn, Vinnitskaya, and Foley 2008). Consequently, the homogeneous groups were tested on lexically headed and headless relatives, and the lexically headed relatives were further divided into relative clauses with specified vs. unspecified heads (i.e., a lexical NP head vs. person).4 Learners were given an elicited imitation task, the design varied along three factors. The stimulus sentences involved the relativization of a noun phrase object or subject; within the subordinate clause, the gap was either in subject or in object position. These variations were extended to three types of relative clause structures that were varied in terms of the semantic and syntactic status of the relativized head NP – lexical head NP, person as head NP, and no head (free relative) (see Table 6.1). Furthermore, a proficiency test was administered to the participants prior to the experiment in order to allocate them into three proficiency levels in English (low, mid, and high), as established by the Michigan test.

Results Table 6.2 provides a simplified summary of the results of the above-mentioned studies. Results of these studies indicated that the L1 Japanese learners of L2 English (study #2 in Table 6.2), like the English monolingual children (study #1 in Table 6.2), scored significantly higher on free relatives than on any of the lexically headed relative types, whereas the Spanish L1 group (study #3 in Table 6.2) did not do significantly better on any of the three types of relatives in their L2 acquisition of English, despite having been equated with the Japanese L1 speakers at all levels of English competence. It seems that the Spanish L1 learners could somehow draw upon their knowledge of CP structure, for Spanish and English match in terms of both CP properties (contrary to Japanese vs. English), and use it in

The lawyer who criticized the worker called the policeman. The person who criticized the engineer greeted the man. Whoever entered the office introduced the professor. Whoever the policeman greeted questioned the gentleman.

The student who the professor introduced answered the man. The person who the engineer answered criticized the man.

The professor introduced whoever greeted the lawyer.

The boss introduced the person who instructed the lawyer.

The boss introduced the gentleman who questioned the lawyer.

Gap position: Subject (OS)

Gap position: Subject (SS)

Gap position: Object (SO)

Head position: Object

Head position: Subject

Source:  Flynn, Foley, and Vinnitskaya (2004).

Free

Lexically headed, unspecified

Lexically headed, specified

Relative type

Table 6.1  Stimulus sentences used in study

The doctor answered whoever the policeman criticized.

The janitor questioned the person who the student greeted.

The woman instructed the lawyer who the policeman called.

Gap position: Object (OO)

Multilingualism: New Perspectives on Syntactic Development  149 Table 6.2  General summary of results of relative clause studies for L1, L2, and L3 English Target language

Group

Pattern

Abbr. name of study

1.  English as L1

Children

EnL1

2.  English as L2

Adults, L1 Japanese (head-final)

3.  English as L2

Adults, L1 Spanish (head-initial)

4.  English as L3

Adults, L1 Kazakh (head-final) and L2 Russian (head-initial)

Free relative precedes lexically headed relative clause Free relative precedes lexically headed relative clause Free relative does NOT precede lexically headed relative clause Free relative does NOT precede lexically headed relative clause

JaL1/EnL2 SpaL1/EnL2 KazL1/RuL2/ EnL3

Sources:  Flynn and Lust (1980); Flynn (1983, 1987, 1989); Flynn (2009).

subsequent learning. This observation led the authors to conclude that the free relative clause structure appears to be a developmental precursor to the lexically headed form. The primacy of free relatives in CP development of the target language was further tested by the mentioned L3 study, which investigated an adult group of KazL1/RuL2 learners of L3 English (study #4 in Table 6.2). Kazakh is a head-final language with an SOV constituent word order within the relative clause, like Japanese (see (4)). The difference between study #2 and this one was that these learners learned Russian – which is a head-initial language with primary SVO order in relative clauses so that it matches the English configuration represented in ((2)–(3)) – as L2 first and subsequently English as L3. Results of this study indicated that the learners’ development of CP structures in the target language patterned with that of the SpaL1/EnL2 rather than the JaL1/EnL2 group; in other words, the Kazakh L1 adult learners, with the help of their experience in acquiring L2 Russian, had developed a target-like CP structure upon which they could draw. Flynn et al. (2004) concluded, on the one hand, that the universal knowledge underlying the free relative seems to be fully available for the learner and, thus, acts as a developmental precursor in the acquisition of English, as was to be observed in the case of the JaL1/EnL2 group. On the other hand, prior CP development appears to influence the development of target-specific CP structure, as in the case of the KazL1/RuL2/EnL3 group or the Spanish L1 learners of L2

150  Neurological and Psychological Aspects of Bilingualism and Multilingualism English, who had already instantiated the CP setup for English through the acquisition of L2 Russian and their own L1, respectively. Summarizing the main findings so far, data seem to support the theoretical claim, according to which in L1, L2, L3 acquisition of relative clauses, there appears to be linguistic development, a process by which the learner constructs a specific language grammar, i.e., a theory of the specific language. Accepting the assumption that language learning is computational and that it is constrained by UG in the course of acquisition, the learner’s linguistic development consists of the learner trying to map from one primary structure to another more developed structure by dissociating modular grammatical components and integrating them in the ‘assembly’ of new language-specific grammars (Flynn et al. 2005; see also Foley 1996). Results seem to indicate that in L1, L2, L3 acquisition of English restrictive relative clauses development of the ‘headless’ relative is developmentally primary to that of the lexically headed relative clause, i.e., free relatives seem to precede the development of full-fledged CP structure. In contrast to the language-specific knowledge, the universal underlying the free relative appears to be fully available at all points in development and it seems that this knowledge is precisely what is needed to enable the acquisition process for lexically headed relatives. Furthermore, results from the L2 studies suggest that at least the directionality of the CP (branching) is critical for this development. The L3 study of Flynn et al. (2004) concluded with the postulation of the CEM for language acquisition. This model hypothesizes that language learning is cumulative, i.e., all previously known languages are available to the learner to constructively enhance subsequent language learning.5 According to the CEM any prior language can be strategically drawn upon in subsequent acquisition and may enhance the acquisition of further languages. Results of the L3 study investigating the development of language-specific CP features support the validity of the claims of the CEM in the following manner (5): (5)  a Development of the CP structures in a prior language or languages determines the course of future language-specific development. b Having integrated language-specific CP features with universal knowledge of CP in earlier language acquisition, the learner can draw upon that developmental process or template created by this earlier developmental experience in later acquisition. More informally stated, the model predicts that the architecture of previously developed CP structures may be available to the learners in terms of enhancing subsequent language development. Consequently, it appears that with respect to the CP, the mind does not redundantly represent CP clausal structure.

Berkes and Flynn (in press) In this section we continue with describing the results of an additional study using the same test design with the objective of testing the predictive strength of the

Multilingualism: New Perspectives on Syntactic Development  151 CEM and to refine further the claims supporting this model. What we were principally interested in was the linguistic nature of the proposed CP development, thus we had selected German L1 learners of L2 English (GeL1/EnL2) and compared their production with that of a group of Hungarian L1 with L2 German learners of L3 English (HuL1/GeL2/EnL3). Matrix clauses in German are head-initial (right-branching), like in English (and in Hungarian; see (2)–(3)), but with regard to CP-relevant word order (that is, in embedded clauses), it exhibits a general SOV word order, as do most of the headfinal (left-branching) languages, like Japanese; therefore, in this aspect, it matches neither English nor Hungarian. A relative sentence contains a relative pronoun, which introduces the relative clause and which is marked for case, gender, and person, as we can observe in the example in (6), where the relative pronoun der refers to a singular masculine subject: (6) Der Anwalt, der den Arbeiter kritisierte, rief den Polizisten The lawyer rel.pr. the-Ac worker criticized called the-Ac policeman ‘The lawyer who criticized the worker called the policeman’ Hungarian is a Uralic language belonging to the Ugric group; it is agglutinative with a complex conjugation system. As mentioned before, in Hungarian, relative clauses appear to the right of the relativized head; hence, it is a head-initial (rightbranching) language with a primary SVO word order. The stimulus sentences we are concerned about in our experiment include objects modified by an article, in which case the SVO word order is regarded to be the default (MacWhinney and Pléh 1988; see also É. Kiss, 1981), like English and in contrast to German. The Hungarian relative pronoun is marked for case and person, so the sentence in (6) translates as (7) into this language. (7) Az ügyvéd, aki kritizálta a munkást, hívta a rendőrt The lawyer rel.pr. criticized the worker-Ac called the policeman-Ac We must add here, that although Hungarian manifests an English-like CP structure in restrictive relative clauses, this language is considered to be organized around the concept of topic and focus rather than the concept of subject (Li 1976). There is an ongoing debate whether the underlying structure of the Hungarian verb phrase (VP) is hierarchical (e.g. Bródy 1995; Marácz 1989) or flat (e.g. É. Kiss 2002).6 Facts suggest that there is relative freedom in terms of constituent word order in the postverbal domain of the Hungarian sentence which seems to be related to the extractability of the verb into a functional head (É. Kiss 2008). Given that German manifests the same head directionality as English (headinitial) in matrix clauses but differs with regard to constituent word order within subordinate clauses by exhibiting a standard SOV order, the comparison between the two studies provided the unique possibility of teasing apart the influence of the two constitutional features of the CP, head directionality and constituent word order relevant to the CP, on the development of target language syntax. A

152  Neurological and Psychological Aspects of Bilingualism and Multilingualism careful analysis of the produced developmental patterns by the two groups learning English, where the L3 group had previously acquired German (the L1 of the L2 study group), promised to shed light on the nature of the acquisition process involving complex structures. The comparison between the L2 and L3 studies offered an additional chance to investigate the effect of enhancement on development as a result of multiple language acquisition experience. Furthermore, the design of this study allowed us to investigate the role of the last-learned language on the acquisition of language-specific syntactic knowledge. Here we refer back to the Flynn et al. (2004) study, which had strongly suggested that the L1 does not have a privileged role in L3 development on the level of complex sentence construction. KazL1/RuL2 learners of English L3 manifested a clearly distinct developmental pattern in their acquisition of English from that of the Japanese L1 speakers, even though the CP properties of Kazakh and Japanese match, i.e., they are both head-final languages with SOV constituent word order within the relative clause. Therefore, we expected that the language combination chosen for this study, where L1 and L3 match completely but not L2, could shed light on what, if any, effect the L2 has on subsequent language development. Additionally, we carried out error analyses as part of this study with the objective to reveal whether the errors made by the two groups at the matched proficiency levels exhibit traces of syntactic transfer from L2 to L3. In order to facilitate further reference to the feature arrangement relevant to CP in the languages involved in this series of experiments (see Table 6.2), Table 6.3 offers a simplified overview including German and Hungarian as well. The statistical analysis of the results of the L2 study group revealed significant differences between correct production of lexically headed and free relatives at the low and mid levels of their acquisition of L2 English. The developmental patterns at the early and intermediate stages of acquisition seemed to resemble those

Table 6.3  CP directionality and word order in subordinate clauses in tested languages Relevant study

Language involved

#1  English L1 #2  JaL1/EnL2 #3  SpaL1/EnL2 #4  KazL1/RuL2/ EnL3 #5  GeL1/EnL2 #6  HuL1/GeL2/EnL3

English Japanese Spanish Kazakh Russian German Hungarian

Source:  Berkes and Flynn (in press).

Headinitial (RB)     

Headfinal (LB)  

SVO    

SOV

  

Multilingualism: New Perspectives on Syntactic Development  153 isolated for the L2 acquisition of English by Japanese speakers and for the L1 acquisition of English by children, rather than the one reported for the SpaL1/ EnL2 and the KazL1/RuL2/EnL3 studies. This result suggested that free relatives act as developmental precursors to the lexically headed forms for the L1 German learners of L2 English as well, for the data revealed a free relative primacy at low and intermediate levels of English proficiency. We concluded, then, that we had found evidence to support the claim that the feature responsible for the constituent word order within subordinate clauses seems to play a role in the process by which the learner constructs subsequent language-specific grammars, i.e., it may suggest that due to CP-related word order differences they have to figure out the new values for the CP, which does not seem to be immediately evident, just like in the case of the JaL1/EnL2 learners. L1 speakers of Hungarian acquiring English as an L3, on the contrary, encountered a language that involves a similar type of integration of universal and language-specific properties concerning CP head directionality to their L1. Our prediction in support of the CEM were thoroughly fulfilled as the statistical analysis of the results indicated that HuL1/GeL2 learners of L3 English did not need to fall back on the primary use of free relatives in order to build up a target-like CP, since this group showed no evidence that they found the imitation of free relatives easier than that of the headed ones. These results matched those produced by the KazL1/RuL2/EnL3 group (Flynn et al. 2004). These findings prove to be relevant to our research on the role of the lastlearned language in subsequent language acquisition as well. The different developmental patterns exhibited by the L2 and the L3 studies provide one of the strongest pieces of evidence for rejecting an astructural transfer of the last-learned language into subsequent acquisition. If the syntactic setup of the last learned L2 German was what L1 Hungarian learners transfer into their L3 English, we would have expected to find almost identical developmental patterns. The lack of negative transfer from L2 to L3 was also corroborated by the error analyses. Concerning enhancement, the L3 group’s exceptionally good performance was also attested by the fact that their production at the high level was significantly better than the comparable results of the GeL1/EnL2 group.

Research Focus Along these lines we have come to show that the CEM may serve as a valid model for language acquisition, as it genuinely reflects the way that language-specific CP develops within the constraints of UG. Learners at comparable proficiency levels (L2 study vs. L3 study) seem to manifest distinct levels of CP knowledge. So, the question that is still to be explored is whether there is any connection between enriched CP knowledge and ease of subsequent language learning. Multilingual individuals can be frequently heard saying that language learning ‘gets easier’ the more languages one knows.7 But what is the linguistic explanation for this, i.e., what is the linguistic nature of this enhancement? Can it be that the

154  Neurological and Psychological Aspects of Bilingualism and Multilingualism representation of the various CP feature-setups in the mind triggers some ‘multilingual function’ that enhances language-learning capacity, i.e. they show how to draw upon this knowledge in the course of subsequent language acquisition? Or is it rather, that in the course of multiple language-learning experience the learner unconsciously acquires a boosted skill to figure out CP options, which has an impact on learning strategies, and it ultimately produces the effect of enhancement? These are the questions we turn to explore in the following section of this chapter.

Enhancement Study To explore this intriguing question was what motivated our study to compare the production data of two groups of learners of English with Hungarian L1, but one of the groups had learned German as L2 before acquiring L3 English. We have chosen these language combinations to investigate the linguistic nature of the ‘enhancement effect’ in the course of multiple language acquisition predicted by the CEM. Recall that the CEM was based on studies that investigated the production of groups with language combinations which provided them with the possibility to draw upon their knowledge acquired in the course of earlier CP development, hence the claim that CP does not seem to be represented redundantly in the mind of a speaker. KazL1/RuL2 and HuL1/GeL2 learners of L3 English did not need to fall back onto starting out to construct the languagespecific CP architecture with the help of the developmentally primary free relatives, since they had already developed an English-like CP architecture through their previous experience in language acquisition (Russian and Hungarian, respectively). This new study, however, compares production data of groups where the only difference is an additional L2 German in case of the L3 study group, but this L2 since it does not match the L3 with respect to CP architecture – does not contribute to subsequent language acquisition as a direct source of knowledge with respect to CP architecture. This means that, in this regard, the L3 study group has no advantage over the L2 group. Furthermore, we made sure to statistically equate the two groups on proficiency.

Predictions This design allows us to focus on our research question, that is, what is the nature of the enhancement the CEM predicts provided language proficiency is controlled? Taking the previously enumerated findings into account, we have generated the following predictions: (8) If there is no apparent difference between the production of the L2 and the L3 groups, we would find support for a straightforward application of the CEM. According to this model, Hungarian L1 learners of L2 and L3 English (the L2 and the L3 study, respectively) may draw upon their knowl-

Multilingualism: New Perspectives on Syntactic Development  155 edge of full-fledged CP, because Hungarian seems to manifest the same CP properties in restrictive relative constructions as English. (9) If the L2 study group clearly performs better than the L3 study group, we would find indices of negative transfer in syntactical development from L2 to L3, since in the L3 study L1 matches L3 but differs from L2 with regard to one of the CP-relevant features (word order within the subordinate clause is SOV in German, as opposed to the Hungarian/English SVO order). (10) If the performance of the L3 study group on relative clauses is undeniably better than that of the L2 group, we would find support for the CEM in a radically new way. Results would indicate that enhancement takes place in the learners’ syntactical knowledge due to multilingual experience.

Design and subjects’ data The design of our L2 and our L3 studies matched the ones reported above (see also Berkes and Flynn in press; Flynn 1983, 1987, 1989; Flynn et al. 2004; and Flynn and Lust 1980). Using an elicited imitation task, we tested two groups of young adults on their production of three types of relatives. Thirty-six Hungarian learners of L2 English and 36 Hungarian learners (with L2 German) of L3 English took part in the experiment. Members of both groups were mainly high school students preparing to enter university or attended university at the time of testing. They had also received 2 to 10 years of formal instruction in English. The L3 group had been exposed to German for at least 10 years. All of the subjects declared themselves nonnative speakers of German, although their exposure to German in school was very intensive. These students exhibited an advanced competency in this language, which was also tested independently with the help of an online test (www.testpodium.com). Those who did not score high enough to be on a C2 level (the ‘Mastery’ level of the European Language Portfolio, the highest level of language competence in a certain language, according to common criteria accepted throughout Europe) were excluded from the study prior to the experiment, because our aim was to see to what extent fully constructed, languagespecific CP structures influence the development of subsequent learning. A proficiency test was administered to the participants prior to the experiment in order to allocate them into three proficiency levels in English (low, mid, and high), as established by the standardized Michigan test. Subject information is summarized in Table 6.4. We carried out an ANCOVA statistical analysis on the data, a three-way mixed design 2*3*3, with study (L2 and L3 study) and level (low, mid, high) as betweensubjects factors and sentence type (specified headed, unspecified headed, free) as a within-subjects factor, using proficiency as a covariate to statistically equate the two groups on proficiency. The nonsignificant value for the two-way study*level interaction (p = .555) suggests that the proficiency effect was statistically equivalent for the two language groups. Moreover, the three-way study*level*type of relative interaction (p = .537) gave a nonsignificant value, which suggests that the

156  Neurological and Psychological Aspects of Bilingualism and Multilingualism Table 6.4  Subject information Level

L2 Study: HuL1/EnL2

L3 Study: HuL1/GeL2/EnL3

Number of learners

Mean ESL score

Number of learners

Mean ESL score

12 15 9

18.8 27.8 40.8

9 13 14

19.8 32.0 42.4

Low Mid High

Table 6.5  HuL1/EnL2 – Mean number of correct responses on three types of relatives by level (max. value = 8) Level

Specified

Unspecified

Free

Low Mid High Total

.3333 1.7333 3.6667 1.7500

.3333 1.6000 3.3333 1.6111

.8333 3.1333 5.6667 3.0000

effect of proficiency is not significantly different across levels of the two studies and sentence type.

Results L2 Study: Hungarian L1/English L2 group  Participants were tested on three types of relative clauses (lexically headed and specified, lexically-headed and unspecified, and free relatives); each type varied according to the grammatical function of the relativized head and its gap in the relative clause (SS, SO, OS, OO – see Table 6.1). The mean number of correct responses for the three types of relatives is presented in Table 6.5 (two measures were taken from all 3*4 sentences; max. value is 8). A two-way analysis of variance was carried out on the data with level as a between-subjects factor (low, mid, and high), and within-subjects repeated measures on type of relative clause (specified headed, unspecified headed, and free relatives). Type of relative proved to be a significant main effect (F2.66 = 17.646, p