Three is a Crowd? Acquiring Portuguese in a Trilingual Environment

Three is a Crowd? Child Language and Child Development: Multilingual–Multicultural Perspectives Series Editor: Professor Li Wei, University of Newcastle-upon-Tyne, UK Editorial Advisors: Professor Gina Conti-Ramsden, University of Manchester, UK Professor Kevin Durkin, The University of Western Australia Professor Susan Ervin-Tripp, University of California, Berkeley, USA Professor Jean Berko Gleason, Boston University, USA Professor Brian MacWhinney, Carnegie Mellon University, USA Children are brought up in diverse yet specific cultural environments; they are engaged from birth in socially meaningful and appropriate activities; their development is affected by an array of social forces. This book series is a response to the need for a comprehensive and interdisciplinary documentation of up-to-date research on child language and child development from a multilingual and multicultural perspective. Publications from the series will cover language development of bilingual and multilingual children, acquisition of languages other than English, cultural variations in child rearing practices, cognitive development of children in multicultural environments, speech and language disorders in bilingual children and children speaking languages other than English, and education and healthcare for children speaking non-standard or non-native varieties of English. The series will be of particular interest to linguists, psychologists, speech and language therapists, and teachers, as well as to other practitioners and professionals working with children of multilingual and multicultural backgrounds. Recent Books in the Series Culture-Specific Language Styles: The Development of Oral Narrative and Literacy Masahiko Minami Language and Literacy in Bilingual Children D. Kimbrough Oller and Rebecca E. Eilers (eds) Phonological Development in Specific Contexts: Studies of Chinese-Speaking Children Zhu Hua Bilingual Children’s Language and Literacy Development Roger Barnard and Ted Glynn (eds) Developing in Two Languages: Korean Children in America Sarah J. Shin Other Books of Interest Foundations of Bilingual Education and Bilingualism Colin Baker Learning to Request in a Second Language: A Study of Child Interlanguage Pragmatics Machiko Achiba Third Language Learners: Pragmatic Production and Awareness Maria Pilar Safont Jordà Language Acquisition: The Age Factor (2nd edition) David Singleton and Lisa Ryan For more details of these or any other of our publications, please contact: Multilingual Matters, Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon, BS21 7HH, England http://www.multilingual-matters.com CHILD LANGUAGE AND CHILD DEVELOPMENT 6 Series Editor: Li Wei, University of Newcastle Three is a Crowd? Acquiring Portuguese in a Trilingual Environment Madalena Cruz-Ferreira “So again the seeds of things are of much latent virtue, and yet of no use except in their development.” Francis Bacon (1620), Novum Organon, 1.121 MULTILINGUAL MATTERS LTD Clevedon • Buffalo • Toronto Library of Congress Cataloging in Publication Data Cruz-Ferreira, Madalena Three is a Crowd? Acquiring Portuguese in a Trilingual Environment Madalena Cruz-Ferreira. Child Language and Child Development: 6 Includes bibliographical references and index. 1. Multilingualism in children. 2. Language acquisition. 3. Language and culture. 4. Portuguese language–Acquisition. I. Title. II. Series. P115.2.C78 2005 306.44' '6 09469–dc22 2005014803 British Library Cataloguing in Publication Data A catalogue entry for this book is available from the British Library. ISBN 1-85359-838-0 / EAN 978-1-85359-838-8 (hbk) Multilingual Matters Ltd UK: Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon BS21 7HH. USA: UTP, 2250 Military Road, Tonawanda, NY 14150, USA. Canada: UTP, 5201 Dufferin Street, North York, Ontario M3H 5T8, Canada. Copyright © 2006 Madalena Cruz-Ferreira. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. Typeset by Datapage Ltd. Printed and bound in Great Britain by the Cromwell Press Ltd. To my children, for showing me how three languages can be managed so smoothly within a family, and for putting up with a data-collecting mother throughout so many years. To Peter, with whom a 'different language' was never a problem. For Tio Zé, in memoriam. Contents Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x 1 Preview . . . . . . . . . . . . . . . . . . What : Layout and Contents . . How : Data and Data Analysis Why : Purposes of this Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Part 1: Becoming Multilingual 2 Issues in Bilingualism and Bilingual Acquisition . Bilingual and Monolingual Lingualism . . . . . . . Bilingualism versus dual monolingualism . . . Mixed speech and bilingual fluency . . . . . . . . The language(s) of bilinguals . . . . . . . . . . . . . The ‘Buffet Effect’ in Bilingual Acquisition. . . . . 3 The Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . Family Background . . . . . . . . . . . . . . . . . . . . . . A Few Issues in Development . . . . . . . . . . . . . . The sibling effect . . . . . . . . . . . . . . . . . . . . . . Developmental feature or idiosyncrasy? . . . . . 4 Data Collection and Analytical Choices . . . . . . . . The Database . . . . . . . . . . . . . . . . . . . . . . . . . . . Modes of data collection. . . . . . . . . . . . . . . . . Choices in data collection: limitations and purposes . . . . . . . . . . . . . . . . . . . . . . . . . Analytical Choices . . . . . . . . . . . . . . . . . . . . . . . Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The choice of target forms . . . . . . . . . . . . . . . Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . Format of examples . . . . . . . . . . . . . . . . . . . . Phonetic transcription . . . . . . . . . . . . . . . . . . . Intonational transcription . . . . . . . . . . . . . . . . 5 Speaking Languages, and Talking about Them . . Signalling Different Languages . . . . . . . . . . . . . Refuse to accept one particular language . . . . Use carriers of language-specific prosody . . . . Make each language maximally different . . . . Use any language-specific device . . . . . . . . . . vii . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 . .2 . .3 . 11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 . 13 . 14 . 19 . 21 . 27 . 30 . 30 . 34 . 35 . 38 . 40 . 40 . 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 . 49 . 49 . 50 . 52 . 55 . 57 . 57 . 60 . 61 . 62 . 63 . 64 . 66 Three is a Crowd? viii Referring to Language(s) . . . . . . . . . . . . . . . . . . Use the language to refer to it . . . . . . . . . . . . Use speakers of each language as symbols for each language . . . . . . . . . . . . . . . . . . . . . . Make use of translation . . . . . . . . . . . . . . . . . Use the name of the language, or the word ‘language’ itself . . . . . . . . . . . . . . . . . . . . . . . Discussing Language and Language Use . . . . . . Characterising language varieties . . . . . . . . . . Commenting on non-native uses of language . Part 2: Making Sense of Portuguese 6 Shaping Sound . . . . . . . . . . . . . . . . . . . . Laying the Foundation for Sounding . . . Practising the Components . . . . . . . . . Mastering intonational fluency . . . . . . The basics of intonational choices and beyond . . . . . . . . . . . . . . . . . . . . Speaking in Sentences . . . . . . . . . . . . . . Sentence fluency . . . . . . . . . . . . . . . . . Filling in the fluent whole . . . . . . . . . The prosodic role of fillers . . . . . . . . . 7 Probing for Constituency . . . . . . . . . . . . . Words in Inventories . . . . . . . . . . . . . . . The word ‘word’. . . . . . . . . . . . . . . . . Child inventories . . . . . . . . . . . . . . . . Child Vocabularies . . . . . . . . . . . . . . . . . Passive vocabulary . . . . . . . . . . . . . . . Active vocabulary. . . . . . . . . . . . . . . . Words in Context . . . . . . . . . . . . . . . . . . Multi-word meanings . . . . . . . . . . . . . Agreement . . . . . . . . . . . . . . . . . . . . . Compounding . . . . . . . . . . . . . . . . . . 8 Probing for Meaning . . . . . . . . . . . . . . . . Words and Concepts . . . . . . . . . . . . . . . Word referents . . . . . . . . . . . . . . . . . . Word meanings . . . . . . . . . . . . . . . . . Strategies to Approach Word Meanings . Semantic manipulation . . . . . . . . . . . . Querying around words . . . . . . . . . . . Replacive words . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 . . . . . . . . . . . 76 . . . . . . . . . . . 78 . . . . . . . . . . . 81 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 . 88 . 88 . 91 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 . . 99 . 100 . 105 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 . 119 . 119 . 122 . 131 . 139 . 139 . 139 . 141 . 145 . 145 . 149 . 154 . 154 . 157 . 174 . 181 . 181 . 182 . 184 . 186 . 186 . 191 . 197 Contents Part 3: Acquiring a Third Language 9 A New Language: Intruder or Guest? . . . . . . . . . . . . . Learning Languages for Communicative Purposes. . . Learning Languages for Curricular Purposes . . . . . . . Attitudes Towards the Children’s Multilingualism . . . From Adults. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . From Peers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Language Input and Language Management in a Multilingual Environment . . . . . . . . . . . . . . . . . . . . . Who Speaks What to Whom, When, Where and Why One person !/ which language? . . . . . . . . . . . . . . . The multilingual nature of language interaction in the family . . . . . . . . . . . . . . . . . . . . . The Children’s Apportioning of Linguistic Space among their Languages . . . . . . . . . . . . . . . . . . . . . . . The languages of the home. . . . . . . . . . . . . . . . . . . Making a home for a new language . . . . . . . . . . . . Defining language territories . . . . . . . . . . . . . . . . . Language dominance? . . . . . . . . . . . . . . . . . . . . . . 11 Balancing Culture and Identity . . . . . . . . . . . . . . . . . . . Languages and Cultural Mindsets . . . . . . . . . . . . . . . Precursors to Socialisation . . . . . . . . . . . . . . . . . . . . . Cultural Idiomacy . . . . . . . . . . . . . . . . . . . . . . . . . . . Unravelling cultural roles . . . . . . . . . . . . . . . . . . . . Biculturalism and cultural modes . . . . . . . . . . . . . . Identity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Identity and cultural camouflage . . . . . . . . . . . . . . Identity and cultural shelter . . . . . . . . . . . . . . . . . . 12 Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appropriating Language: Acquisitional Strategies . . . Make do with what is available . . . . . . . . . . . . . . . One thing at a time. . . . . . . . . . . . . . . . . . . . . . . . . To learn how to use something, use it . . . . . . . . . . The Role of Acquisitional Strategies in Child Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Three is a Crowd? . . . . . . . . . . . . . . . . . . . . . . . . . . . ix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 . 210 . 217 . 224 . 224 . 228 . . . . . . 233 . . . . . . 233 . . . . . . 233 . . . . . . 237 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 . 244 . 249 . 258 . 268 . 278 . 278 . 279 . 288 . 289 . 291 . 294 . 295 . 297 . 300 . 301 . 302 . 303 . 304 . . . . . . 306 . . . . . . 309 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330 Acknowledgements My greatest debt goes to Prof. Li Wei, who first persuaded me to write this book. Along the many ups and downs in the writing itself, I always had the reassurance of his timely encouragement and his judicious advice, which were there for me, any time. Thank you for believing in this project, and for believing that I could bring it to port. I also owe my gratitude to the many extra informants who contributed to this study. Several children and adults outside of the core family took part in exchanges involving the three children in the study, or are named by them. Permission to use the data and names of all participants who, at times unwittingly, found themselves involved in recording sessions was duly requested and obtained, from children, their parents and other adults. Their role in the children’s life is introduced where relevant for the discussion of the data. Particularly in what concerns the children’s playmates, they did know that they might be recorded, but they did not know where or when. To all of them, my sincerest thanks, for their willingness to participate and for allowing me to quote them by their real names. MCF Singapore, December 2004 x Chapter 1 Preview This book is about three children and their breakthrough into language. The children are siblings, primary bilinguals in Portuguese and Swedish, and acquired English as the language of schooling. The book presents a preliminary description of the children’s mastery of one of their languages (European) Portuguese, against the background of their multilingual environment. The aim is to provide an account of the processes that enable the children’s competent allocation of linguistic, social and cultural space to one language alongside two others, not despite their multilingualism, but because of it. In its as yet unforeseeable complete form, the ongoing study on these children has the ambitious aim of providing as detailed a picture as possible of their linguistic and social development throughout the first ten years of their lives, as competent users of each language and as multilingual users of each. The book does not aim at giving a full account of the children’s linguistic development, not even in Portuguese. Rather, the material presented in the book is the result of a number of choices, detailed below, that aim instead at giving a first glimpse into a much larger developmental picture. The focus is on the investigation of how the children came to be able to talk about what interests them, i.e. how they made language work for them. In the process, extensive discussion is given of a central issue in child language studies, namely, what is it that children are saying and in what way are we able to interpret what they say. Language is new for children, though not for their linguistic role models. It is therefore interesting to investigate what strategies children devise, from the very beginning, in order to navigate their way along their languages. These strategies can be linguistic or, for example, gestural, used as discovery procedures for linguistic and cultural patterns. From within the perspective of the acquisition of Portuguese, this book gives an overview of how the children are learning to make sense of their languages, their surroundings and themselves. The study concerns the children’s spontaneous, everyday use of language, and what it reveals about their linguistic development. The role played by monolingual or multilingual adults and peers in the children’s development is analysed. The matter of the children’s choice of language deserves particular attention, both in the sense of language being chosen over other available means of communication to achieve a particular effect, and in the sense of a particular language being chosen over another in a particular situation. Not least, special attention is paid 1 2 Three is a Crowd? to the children’s attitudes towards each of their languages, as well as to their own insight into their multilingualism and multiculturalism, and how to express it in different languages. Other issues that are central to the understanding of language use are also surveyed, such as language input and language management within the family and among the children’s peers, as well as strategies used by monolingual speakers around the children to assist effective communication. The rationale behind the three major sets of choices in the present study is detailed below. What: Layout and Contents The book is divided into three main parts. The first one, Becoming Multilingual , provides background information along four chapters. Chapter 2 reviews a number of issues in studies on bilingualism, relating these to the children’s acquisitional strategies and to patterns of language mixing and of monolingual usage in the children’s speech. The chapter concludes with the presentation of the research questions guiding the present study. Chapter 3 focuses on the children, including personality traits, perceived and practised sibling hierarchy, parent!/child interaction, schooling and settling in among the different countries and cultures where the family has lived. Chapter 4 deals with the database used in this study, including choices in data collection and transcription. The final section in this chapter lists the conventions used in the presentation of results. Chapter 5 gives an overview of the children’s early and later differential uses of their languages, as well as of their ways of discussing language itself with the different means at their disposal along their linguistic development. Part 2, Making Sense of Portuguese , contains three chapters that detail issues in the children’s progressive mastery of Portuguese in a multilingual context. Chapter 6 discusses the children’s grappling with the phonology of the language, focusing on its prosody through the children’s use of sounds and tunes. Chapter 7 deals with matters pertaining to structure at word, phrase and utterance level, whereas Chapter 8 discusses linguistic meanings, whether carried by single-word or multiword utterances. Part 3, Acquiring a Third Language , has three chapters dealing with the role played by each of the children’s three language in their approach to communication and to life. Chapter 9 covers the emergence of English in the children’s linguistic lives, and the related turning point in their family environment from bilingual to trilingual. It also addresses the children’s first encounters with school-bound second-language learning, as well as attitudes towards the children’s multilingualism, from the children themselves and from others. Chapter 10 focuses on the Preview 3 apportioning of linguistic space and usage resulting from the appearance of English in the children’s linguistic repertoire, including the children’s perception and practice of each language’s individuality and scope. It discusses language choice among the children, according to factors such as interlocutor or topic of conversation, and addresses matters of language input and language management in a multilingual environment, reflected in modes of parent!/child, child !/parent and child!/child linguistic interaction. Chapter 11 situates the children’s uses of Portuguese within the broader issues of multilingual culture and identity, by addressing matters such as values and mindsets conveyed through language, as well as attitudes and interaction mediated by the use of different languages. The final chapter in the book articulates an overview of findings with issues in child bilingualism, cognitive and social development, and related language use. How: Data and Data Analysis Any presentation of results involves at least two kinds of choices, one concerning the type and amount of data, the other concerning the analytical framework from within which discussion of the data can make sense. Decisions about what to say and how to say it are personal, because they serve the purposes that the researcher set for the research. Choices in data are compounded in large databases of spontaneous speech, like the one on which this study draws. Raw data are besides necessarily untidy, which may further muddle up the analysis because there are so many things to say that can be said in so many different ways. Choices in analytical framework are often no choice, in that they often pursue a habit that was shaped by the preference of academic mentors, who therefore made the choices for the novices under their supervision. The rationale behind specific choices made in this study is as follows. Major choices concerned the language to be analysed and, within it, which aspects of it to focus on. A study on Portuguese can be interesting on several counts. Two of these are that Portuguese is widely spoken as a native and foreign language, including by large groups of immigrant communities, and that it is a language that has given rise to several pidgins and creoles around the world. A more specific reason behind the choice of this language is that Portuguese is the first peer language of the children in this study, i.e. the one that they started using among themselves. There is no lack of material on child Portuguese, particularly due to the surge of interest in this area in the last three decades (see CruzFerreira (2000b) for an edited and annotated bibliography on the acquisition of Portuguese, both European and Brazilian). The bulk of these studies concerns the monolingual acquisition of specific aspects of 4 Three is a Crowd? the language. On the other hand, most studies on bilingual children pay exclusive attention to the children’s bilingualism, often to ascertain matters related to language imbalance or features of mixed speech. This book purports to give a different story, by including discussion of the children’s monolingual production in Portuguese. That is, by reporting on how, faced with several languages, the children make sense of Portuguese by making Portuguese make sense. The descriptive chapters dealing with Portuguese in this book cover selected features of the children’s speech, concerning issues in their acquisition of phonology, grammar and vocabulary that were deemed novel in the literature on child language, child bilingualism and child Portuguese. Phonetic features of the children’s productions deserve extensive discussion throughout the book, because sound is the medium of speech, and because it is through clues present in the sound signal that evidence for the emergence of words, phrases and parsing in general can be sought. Particular emphasis is given to the acquisition of intonation, a linguistic system commonly neglected in acquisitional studies, whether monolingual or multilingual. The three children appear in fact to use features of intonation as the prime strategy guiding their search for both meaning and identity among the languages in their environment. The study also concerns the acquisition of discoursal and pragmatic features characterising the use of Portuguese, as well as the acquisition, through language, of the culture and modes of thinking associated with the language. One consequence of the choices of language and issues pertaining to it is that the book deals with what its title suggests, i.e. the acquisition of Portuguese in a multilingual environment, and not with the children’s multilingualism per se . However, what the children do with their Portuguese relates to what they do with their multilingualism and multiculturalism in ways that, to me, are striking for our understanding of an overall picture of language in the making. The greater attention given to features of Portuguese in the book does not, and in fact cannot, detract from the fact that the children are multilinguals, and are learning each of their languages as such. The book contains recurrent discussion of matters pertaining to bilingualism and bilingual acquisition, including multilingual acquisitional strategies, mixes and data from the children’s other two languages that may provide support for their command of features of Portuguese, or failure to do so. Because the children are multilinguals, the simultaneous acquisition of Swedish and the later acquisition of English must necessarily form part of an account of their language abilities. Because Portuguese is not the only linguistic tool at the disposal of these children, some insight may be gained into how multitooled children like these acquire language. The latter point is also one of the reasons why this study deliberately avoids comparison of these children’s data with data Preview 5 from monolingual peers: the children are not monolingual. The other reason is that comparative studies of this kind usually have a judgmental purpose, often that of showing that bilingual children are in some way lacking in linguistic competence, or are at least different. They are different, for the tautological reason that monolingual and bilingual experimental populations are different. Whether they are found lacking is a matter of analysis, and of what to count as the norm. Equally different results, and equally uninteresting, would be obtained if studies focused on asking questions the other way around, taking bilingual populations as the implicit norm and monolingualism as the variable. Given that the majority of the world’s population is multilingual, questions phrased in this way could at least claim some statistical legitimacy. Some of the data in this study raise questions about theoretical justification for normative stances such as the above in linguistic analysis. One, more general, concerns precisely the view that bilingualism is a kind of monolingualism that can only begin to be understood once monolingualism is accounted for. This is discussed in Chapter 2. Another, more particular, concerns for example, which of the grammatical genders in a gendered language like Portuguese should be taken as unmarked, or even whether a stance about marked versus unmarked linguistic features is useful at all. This is discussed in Chapter 7. Other issues in child language and child bilingualism, pertaining to particular topics presented in each chapter, will be discussed along the book. Much has been said about child data being taken as evidence for features ascribable to universal grammars and universal uses of language, and much has been said about features of particular languages being taken as features of language. These are assumptions, not necessities. Finding results that match an assumption, or hypotheses derived from it, is not difficult, and arguably not interesting either, because the answer is not in the data, but in the analysis. Because analyses depend on assumptions, the issue of empirical verification risks becoming fully circular. While the present study does not question these assumptions per se , it does keep in mind that they are assumptions and, as such, cannot be proven right or wrong. By presenting data from three children, this study hopefully avoids too broad generalisations sometimes found in studies of one single child or of one single feature of the speech of one child. Conversely, since the children are obviously individuals, taking some of their individual data as representative also shows that quite different conclusions may be drawn about the acquisition of Portuguese. It is in this sense that this study is exploratory. Discussion of theoretical matters is kept to a minimum for two additional reasons, related to whether child data and data from single languages mirror or suggest so-called universals of language, respectively. 6 Three is a Crowd? First, researchers are, naturally, conditioned by the research paradigm, in Kuhn’s (1970) sense of this term, in which they have been brought up, whose purpose, once established as a paradigm, is that of reproducing thinking. Currently popular, theory-driven analyses of child speech appear ensnared in ritual discussion of labels and thinking where ‘much of the work is argumentation that carries you further and further from the data’ (Napoli, 1996: 396!/397). Child data are probed for evidence of whatever constructs correspond to whatever labels in successive levels of abstraction, that end up having little to do with what the child actually says and all to do with what the analytical framework expects the child to be saying. In other words, what the child has acquired will depend on what the theory defines as acquirable and assumes is there to be acquired, a view that Donaldson (1978: 33) guards against: ‘To Western adults, and especially to Western adult linguists, languages are formal systems. A formal system can be manipulated in a formal way. It is an easy but dangerous move from this to the conclusion that it is also learned in a formal way.’ In my view, the formal elegance of a set of constructs said to have been acquired (or not), and said to reflect formal universal features of language matters far less than whether the child is successfully (or not) communicating by means of speech with other human beings. On the other hand, as Lass (1998: 6) rightly points out, ‘in a way, there are no facts without theories. One might even define a theory as !/ in part !/ a framework that tells you what a fact is.’ The framework that guides the analyses presented in this study assumes that language is a tool, whose successful use, like that of any tool, needs exploring. Exploration proceeds by means of strategies, devised according to the means, linguistic or otherwise, at the disposal of the child at different developmental stages, and probed for effectiveness. The study targets a description of these strategies, in what they reveal about the children’s progressive discovery of how language may serve their communicative needs. Second, it is often claimed that language, in collocations like ‘the language faculty’ or ‘the nature of language’, can only be accessed through its concrete manifestations, i.e. the different languages spoken by human beings. This is a commonsense claim, in that the term ‘language’ in these collocations refers to an abstract entity. Unfortunately, the word language is quite ambiguous in English, in that it also refers to each of the different tongues spoken by human beings. It is often not clear from the literature which of the two meanings of this word is being discussed, nor whether the ambiguity is being systematically explored or simply overlooked. By way of a lexical idiosyncrasy, one particular tongue, Saussure’s (1915/1969: 25ff.) ‘langue’, can easily be (mis)taken for the ability to use tongues !/ or to speak in them (Saussure’s, 1915/1969, ‘langage’). The two meanings of the English Preview 7 word are unambiguously differentiated in Portuguese too, through the words lı´ngua and linguagem , respectively. Researchers are, naturally, conditioned by the language that they use. The discourse of science varies not only according to the lexical and grammatical resources that are available in the language used by the writer (Fahnestock, 2003), but also according to the cultural traits associated with that language (Dahl, 2004). Given that more linguistic descriptions, as well as more detailed ones, are available for English than for any other language, and given that the bulk of linguistic thinking is published (and often worked out too) in English, regardless of which language constitutes the object of study, there is a clear risk of extrapolating features that are found in English to characterise universal uses of language (Sampson, 2001), and of mistaking English terminology for a neutral, default metalanguage in which to encapsulate them. In other words, English provides both the database and the tools for its analysis. Lucy and Gaskins (2001: 258) echo both these points, in the same paragraph: ‘In place of empirical research, the literature has been filled with a wide variety of speculative answers that inevitably confirm the initial theoretical predilections of the analyst’ and, they go on, ‘People have freely extrapolated from research data on our own language or some telling personal experience with another.’ Another important choice concerns the modelling of child data. Most studies on language acquisition assume the adult language not only as the target of child forms but as the model itself upon which to map and explain child utterances. Most studies do this implicitly, by interpreting the child’s utterances according to what the researcher (or the caregiver, or both) assume is the child’s target, taking into consideration context, mood of the child at the time of utterance and similarity of the child’s utterance to an acceptable adult form. Children are therefore assumed to be using the adult system, while acquiring it. This makes good sense, in that learning in general does not seem to depend on instruction as much as on use. It is not useful to wonder how children can use something that they have not acquired yet. The pertinent question is rather, how can children acquire something without using it. That is, the learning is in the use. We learn to walk, or to ride a bicycle by doing it, and the same is reasonably true of language. In addition, interpretation of utterances is largely instinctive, whether the speaker is a child, an adult, a native or non-native user of the language in question, for the reason that language users expect language uses to make sense. If adult listeners hear a child utterance that sounds like ‘shoe’ in a context where an association with footwear is likely, they will naturally infer that the child is attempting to replicate an adult form of a word. The child utterance will accordingly be taken as one child-way of saying the adult word, and described as exhibiting particular phonological, cognitive and other features that may be found to associate the child form to the adult form. Two issues arise 8 Three is a Crowd? here, one concerning what can we legitimately claim that the child is likely to be targeting, the other summarised in the more general question, ‘How do we know?’ The issue of deciding what a child means, particularly a very young child, is not easy, as has been noted even in early research on this subject (Stevenson, 1893). In particular, the identification of child meanings and associated forms will depend on what child productions are taken to target. Findings will of course vary according to assumptions, and so will, in turn, theorisation based on different findings. If the child who presumably said ‘shoe’ is assumed to have meant ‘shoe’ in the adult meaning of this word, then the child will be said to have acquired the adult word, complete with word class and associated meaning. But if the child’s use of the word is assumed to express pleasure at a new pair of shoes, or as a description of someone putting on their own or the child’s footwear, then the child word will be assigned a different meaning and a different word class. That is, the same child word will count as a token of a different vocabulary or grammatical acquisition depending on what is taken as the model target. Target forms need therefore to be incorporated in the description, not as abstract models taken from available accounts of the languages in question, but through surveying of which forms the child is in reality exposed to. Obviously, children do eventually acquire adult uses of language (they eventually become adults too), because if they did not, they would not be able to function properly in the adult-geared communities that characterise human social organisation. There is progress in acquiring language, in the sense that child utterances become more and more like adults’. This being so, one reasonable way of describing the progressive changes in child uses along time is to assume that children are attempting to follow a model, the one provided by pre-existing language uses around them. But within any developmental ‘stages’ that one might want to postulate, progress is not necessarily forwards in the sense of ‘steadily more like the adult target’. Progress is rather a process of exploration, of back-and-forth probing, whose visible results may at first sight suggest un-learning of forms that earlier appeared safely in place, or confusion between them. The human strategy of overgeneralising from previous experience in order to accommodate new information is naturally used by children approaching their language(s) too. Strategies may include devices that do not play a linguistic role in the particular language under observation but that may make sense crosslinguistically. The children’s own strategies cannot therefore always be said to follow an adult target. In fact, it is when children are judged to have stopped using their own exploratory strategies that language acquisition is deemed complete. By then, their strategies to approach a language are common to adult users of that language. Language Preview 9 acquisition is then a process of conforming to a pre-existing norm, both in acquiring target forms and in the ways of getting at them, as is discussed in different chapters along this book. Adult forms, however, are not unmoveable and they do not constitute a closed repertoire on which both children and adults can draw for their different communicative needs. Language is open-ended, because so are the purposes that it serves. It is in fact not clear at all from the literature whether the label ‘acquisition’, applied to language or languages and to children’s linguistic development, is a misnomer. Language learning, in all of its aspects, is of course a lifelong endeavour (Foley & Thompson, 2003). If the adults who serve as models for the child are also acquiring uses of language, children must be targeting forms whose use continually shifts. In addition, if language mastery is a lifelong process, either there is no cut-off age for its acquisition, because there is no point in time where one can say that acquisition is complete, or there are as many different cut-off ages as researchers choose to establish, depending on the different aspects of language mastery that interest them. Methodological issues of this kind are discussed in Chapter 4. A clear statement of assumptions in studies of child language is therefore crucial. It is the only legitimate way in which researchers can come clean about the unavoidable truth that we simply do not know what children are doing in order to get at language. The crucial questions are not what children have acquired, or not, that reflects the model that we assume. It is clear that the data will bear interpretation according to the model, because the model was designed to query the data according to it. The crucial question seems to me to be ‘How do we know?’, because this is the question on which all others depend. Child data are not free from analytical manipulation in order to corroborate the latest version of the theory of the day. For example, many studies on the acquisition of Portuguese are geared towards satisfying current generative-spawned theorisation about language acquisition. However, by their own nature, child data do escape the introspective activities said to be the crucial mode of enquiry in some research on language (see Sampson (2001) for discussion of this issue), despite the often unstated assumption of taking the results of adult introspection as the model of child forms. This being so, there are no certainties in child language; there is only a stronger or weaker likelihood that child data mean what we say they do. Bearing in mind these provisos, there are nevertheless well established ways of safeguarding interpretation of child data. As said above, I will adopt the common assumptions that child forms target adult uses and can be interpreted according to these. I will also give all the information that is available to me, with as much detail as possible on context and situation of child utterances, as well as on what prompted me to provide the interpretations that I offer. Phonetic transcription of child utterances 10 Three is a Crowd? is also given wherever relevant. By doing so, it is hoped that the likelihood of my interpretations can be assessed. It is also my hope that readers will want to come back to me with feedback on interpretations that I may have got wrong, or may have missed altogether. Besides assumptions, clarification is also needed in what concerns definitions of terms. Being the science of language, linguistics provides a way to talk about language. In this sense, linguistics can be said to be the language of language. Language is not only common to all human beings, it is also a topic that interests human beings, given its centrality in any human endeavour. Language is talked about by means of language, in everyday situations that involve no necessary precision in the use of its terms. It is therefore not surprising that several technical terms of linguistics, particularly the more common ones that are also everyday words, are often quite nebulous in the literature, giving rise to ambiguous and confusing interpretations. One of these terms is the word language itself, discussed above. In this study, the non-count meaning of the word is intended, when used without preceding determiner and in collocations such as ‘ability to use language’ or ‘language acquisition’. Countable uses of the noun refer to particular tongues. The term word continues to defy definition, and I make no attempt to contribute to the controversy surrounding it, let alone to point at some solution. Discussion of the word word is given in Chapter 7. The term grammar applies in a broader sense than usual. In some literature, grammar is often used as an equivalent of the terms syntax and/or morphology, to mean the set of formal properties and relationships characterising the parsing patterns found for morphemes and phrases. Given that meaning-yielding units of languages are not exhausted in their morphemes and phrases, the term grammar is here taken to mean a formulation about the patterning of any linguistic items, from sounds through words to texts. The grammar of a language is a statement of the coding rules describing linguistic behaviour that enable communication in that language, including the rules governing particular exchanges among particular speakers, that is, particular ways of using a language. The terms bilingual and trilingual , and their variants, applied to the children in this study, are used in different ways. Bilingualism refers to their simultaneous acquisition of Portuguese and Swedish, whereas trilingualism refers to the fact that there are three languages in their childhood. The prefix bi- to characterise a type of ‘lingualism’ is sometimes taken in the literature as synonymous with the prefix multi-. These uses stem from the assumption that there may be a difference between the use of one single language and the use of more than one, but not between the use of two languages and the use of more than two, an assumption that I endorse. The various uses to which the label bilingual Preview 11 has been put in the literature are discussed in the next chapter. Other more specific terms will have their definitions discussed in the relevant sections of the book. Why: Purposes of this Study The purpose of this book is to give an overview of developing language ability, in a multilingual environment. No exhaustive analysis of any language, language level or developmental stage (or of any child) is attempted, pass the paradox in a formulation like ‘exhaustive analysis’. There is, however, detailed discussion of several issues pertaining to the children’s progressive mastery of Portuguese, as well as of what their trilingualism may reveal about their mastery of language. The book proposes a general frame that appears to allow the children’s progressive linguistic development, gleaned from a sample of their productions. Filling this frame with the detail of finer brushstrokes must come later, because detail can only be understood if the foundation makes good sense. The children’s productions discussed here concern a limited amount of their linguistic abilities, that is besides given from a restricted perspective that is self-imposed, due to space constraints. Insight into language is given as it emerges for the children, as far as we outsiders can tell from their productions and how we best see it fit to interpret these. At all times, we must bear in mind that descriptions of child uses of language concern a process of exploration, not a final product. Children are learning to appropriate a tool in a way that serves them and their needs as social beings, not providing researchers with ready-made uses of it. The everyday uses of language that this study draws on appear ideally suited to capturing child modes of linguistic appropriation, for at least two reasons. One is that these uses reveal spontaneous tackling of linguistic challenges and of metalinguistic abilities to deal with them, as and when they occur, thereby providing insight into how language is falling into place for the children. The other reason is that children’s everyday uses of language are those that parents and caregivers are exposed to, and react to as spontaneously, linguistically as otherwise. By capturing linguistic interaction and feedback of this kind, this study hopes to contribute to a broadening of the gauges that guide assessment and opinions about child linguistic development, among linguistically untrained parents and linguistics researchers alike. This book also attempts to contribute thoughts and data to several issues within child multilingualism, a research area that has gained momentum in recent years. In particular, by addressing the perceived strategies of children progressively working their way through several languages, it is hoped that a fresh insight into child language and into child multilingualism can be gained. Not least, this study will attempt to 12 Three is a Crowd? show that child multilingualism, with its necessary exploration of the accidental limits within which particular languages happen to vary, is the ideal ground in which to look for features that may define the nature of human language. One final note: the book will of course be fully enjoyable by readers with at least some knowledge of linguistics and of the three languages involved in this study. For the benefit of those who may not have such knowledge, descriptive technicalities will be kept to a minimum, and English glosses are given throughout.