How Languages are Learned PDF

Fourth Edition How Languages are Learned  OXFORD ACKNOWLEDGEMENT S U N I V E R S I T Y PRESS The authors and publisher are grate.ful to those who have given...

Fourth Edition How Languages are Learned  OXFORD ACKNOWLEDGEMENT S U N I V E R S I T Y PRESS The authors and publisher are grate.ful to those who have given permission to reproduce thefo!lowing extracts and Great Clarendon Street, Oxford, ox2 6DP, adaptations of copyright material: p.17 Extract from United Kingdom Language Development and Language Disorders by lois Oxford University Press is a department of the Bloom and Margaret lahey (1978). Macmillan University of üxford. It furthers the University's Publishers; p.47 Figure from 'Sorne issues relating objective of excellence inresearch, scholarship, to the Monitor Model' by Stephen Krashen, On and education by publishing worldwide. Oxford TESOL(1977). Reprinted by permission of TESOl is a registered trade mark of Oxford University International Association; p.49 Extract from Press in the UKand in certain other countries 'Constructing an acquisition-based procedure © Oxford University Press 2013 for second language assessment' by Manfred Pienemann, MalcolmJohnston, and Geoff Brindley The moral rights of the author have been asserted inStudies in SecondLanguage Acquisttion, Volume 10/2, First published in 2013 pp.217-43 (1988). Reproduced by permission of 2017 2016 2015 2014 2013 Cambridge University Press; p.53 Extract from 10 9 8 7 6 5 4 3 2 1 'Speeding up acquisition of hisfher: Explicit L1/L2 Al! rights reserved. No part of this publication contracts help' in Second Language Acquisttion and may be reproduced, stored in a retrieval system, the Younger Leamer: Child's Play? by Joanna White or transmitted, in any form or by any means, (2008) pp.193-228. With kind permission ofJohn without the prior permission in writing of Oxford Benjamins Publishing Company, Amsterdam/ University Press, or as expressly permitted by law, Philadelphia; p.54 Extract from 'Second language by licence or under terms agreed with the instruction does make a difference' by Catherine appropriate reprographics rights organization. Doughty in Studies in Second Language Acquisition, Enquiries concerning reproduction outside the Volume 13/4, pp.431-69 (1991). Reproduced by scope of the above should be sent to the ELT permission of Cambridge University Press; p.136 Rights Department, Oxford University Press, Reprinted from Intemationa!Jouma! of Educational at the address above Research, Volume 37 by Merrill Swain and Sharon Lapkin 'Talking it through: two French immersion You must not circulate this work in any other learners' response to reformulations' pp.285-304 form and you must impose this samecondition (2002) with permission from Elsevier; p.139 Extract on any acquirer from 'Corrective feedback and learner uptake' Links to third party websites are provided by by Roy lyster and Leila Rauta in Studies in Second Oxford in good faith and for information only. Language Acquisition, Volume 19/1 pp.37-66 (1997). Oxford disclaims any responsibility for the Reproduced by permission oíCambridge materials contained in any third partywebsite University Press. referenced in this work Cartoons by: Sophie Grillet © Oxford University Photocopying Press 1993, 2005, and 2012. The Publisher grants permission for the photocopying of those pages marked 'photocopiable' according to the following conditions. Individual purchasers may make copies for their own use or for use by classes that they teach. School purchasers may make copies for use by staff and students, but this permission does not extend to additional schools or branches Under no circumstances may any part of this book be photocopied for resale I S BN : 978 O 19 454126 8 Printed in China This book is printed on paper from certified and well-managed sources. L  To the teachers and students from whom we have learned so much 1 t CONTENTS Acknowledgements x1 Preface to the fourth edition xiii lntroduction 1 Before we begin... 2 1 Language learning in early childhood 5 Preview 5 First language acquisition 5 The first three years: Milestones and developmental sequences 6 The pre-school years 12 The school years 13 Explaining first language acquisition 14 The behaviourist perspective 15 The innatist perspective 20 lnteractionist/developmental perspectives 24 Language disorders and delays 29 Childhood bilingualism 30 Summary 33 Suggestions for further reading 34 2 Second language learning 35 Preview 35 Learner characteristics 36 Learning conditions 38 Studying the language of second language learners 40 Contrastive analysis, error analysis, and interlanguage 41 Developmental sequences 45 More about first language influence 57 Vocabulary 60 Pragmatics 65 Phonology 68 Sampling learners' language 72 Summary 72 Suggestions for further reading 73 viii Contents 3 Individual differences in second language learning 75 Preview 75 Research on learner characteristics 77 Intelligence 79 Language learning aptitude 80 Learning styles 83 Personality 84 Attitudes and motivation 87 Motivation in the classroom 88 Identity and ethnic group affiliation 89 Learner beliefs 90 Individual differences and classroom instruction 92 Age and second language learning 92 The critica! period: More than just pronunciation? 94 Intuitions of grammaticality 95 Rate oflearning 96 Age and second language instruction 96 Summary 99 Suggestions for further reading 100 4 Explaining second language learning 103 Preview 103 The behaviourist perspective 103 Second language applications: Mimicry and memorization 103 The innatist perspective 104 Second language applications: Krashen's 'Monitor Model' 106 The cognitive perspective 108 Information processing 108 Usage-based learning 110 The competition model 111 Language and the brain 113 Second language applications: Interacting, noticing, processing, and practising 113 The sociocultural perspective 118 Second language applications: Learning by talking 119 Summary 120 Suggestions for further reading 121  Contents lX 5 Observing learning and teaching in the second language classroom 123 Preview 123 Natural and instructional settings 123 In natural acquisition settings 124 In structure-based instructional settings 126 In communicative instructional settings 127 Observation schemes 129 Classroom comparisons: Teacher-student interactions 129 Classroom comparisons: Student-student interactions 135 Corrective feedback in the dassroom 139 Q uestions in the dassroom 145 Ethnography 149 Summary 151 Suggestions far further reading 152 6 Second language learning in the classroom 153 Preview 153 Proposals far teaching 153 1 Get it right from the beginning 154 2 Just listen... and read 159 3 Let's talk 165 4 Get two far one 171 5 Teach what is teachable 177 6 Get it right in the end 182 Assessing the proposals 194 Summary 197 Suggestions far further reading 198 7 Popular ideas about language learning revisited 201 Preview 201 Reflecting on the popular ideas: Learning from research 201 Conclusion 212 Glossary 213 Bibliography 227 lndex 249 PREFACE TO THE FOURTH EDITION How Languages Are Learned (HLAL) started out as a series of professional development workshops for teachers in Quebec, Canada, where we both worked for many years. Three editions of the book have now travelled far from those origins. When we were working on the first edition in the 1980s and 1990s we were still in the early days of remarkable growth of research in second language acquisition. In updating the research for each new edition, the decisions about what to include have grown more diflicult. Keeping the book to a reasonable length has often meant choosing between classics in the field and important new studies, of which there are now so many. In this edition, we have annotated sorne 'Suggestions for further reading' at the end of each chapter. We encourage readers to follow these readings and the refer ence list to deepen their understanding of topics that we can only introduce here. In this fourth edition of HLAL, we have added 'Questions for reflection' at the end of each chapter, and we have included sorne new 'Activities' that give readers opportunities to explore sorne of the topics. Another new feature of this edition is a companion website which contains additional activi ties, readings, and other web-based material and resources to enhance your reading and understanding of the contents of the book. It will also provide opportunities for readers to interact with others and to share their ideas for teaching and learning languages. The website for How Languages are Learned can be accessed at www.oup.com/elt/teacher/hlal. We are currendy working on a new series of books for teachers, the Oxford Key Conceptsfar the Language Classroom. Each volume, written by a different author, will focus on a specific topic (such as assessment, content-based lan guage teaching, literacy, and oral interaction), reviewing the relevant research and linking the findings to classroom practice. We hope that the books in this series will encourage teachers to continue learning about sorne of the topics that are introduced in HLAL. We hope that both new readers and those who have read the previous edi tions of HLAL will find ideas and information that will challenge and inspire them to make their own contributions to second language learning, teach ing, and research. Patsy M. Lightbown, Harwich, MA, USA Nina Spada, Toronto, ON, Canada INTRODUCTI ON When new methods and textbooks for second and foreign language teach ing are introduced, they are often said to be based on the latest research in psychology, linguistics, or pedagogy. Teachers are told that they will be more effective than those that have gane befare. I n many cases, the new approaches are prescribed for immediate implementation in a school or region. Sometimes, the new materials come with opportunities for extensive training in their implementation. Sometimes, they are simply ordered and distributed to teachers who have to do their best to use them effectively. Many approaches to language teaching have been proposed and imple mented. One approach requires students to learn rules of grammar and lists of vocabulary to use in translating literary texts. Another emphasizes the value of having students imitate and practise a set of correct sentences and memorize entire dialogues. Yet another encourages 'natural' communication between students as they engage cooperatively in tasks or projects while using the new language. In sorne classrooms, the second language is used as the medium to teach subject matter, with the assumption that the language itself will be learned incidentally as students focus on the academic content. How are teachers to evaluate the potential effectiveness of different instruc tional practices? To be sure, the most important influence on teachers' decisions is their own experience with previous successes or disappointments, as well as their understanding of the needs and abilities of their students. We believe that ideas drawn from research and theory in second language acquisition are also valuable in helping teachers to evaluate claims made by proponents of various language teaching methods. The goal of this book is to introduce teachers-both novice and experienced-to sorne of the language acquisition research that may help them not only to evaluate existing text books and materials but also to adapt them in ways that are more consistent with our understanding of how languages are learned. The book begins with a chapter on language learning in early childhood. This background is important because both second language research and second language teaching have been influenced by our understanding of how children acquire their first language. Several theories about first language (LI) learning are presented in this chapter and they are revisited later in the book in relation to second language (L2) learning. 2 !ntroduction In Chapter 2 we look at second language learners' developing knowledge, their ability to use that knowledge, and how this compares with L1 learning. InChapter 3, we turn our attention to how individual learner characteristics may affect success. I n Chapter 4, several theories that have been advanced to explain second language learning are presented and discussed. Chapter 5 begins with a comparison of natural and instructional environments for second language learning. We then examine sorne different ways in which researchers have observed and described teaching and learning practices in second language classrooms. In Chapter 6, we examine six proposals that have been made for second language teaching. Examples of research related to each of the proposals are presented, leading to a discussion of the evidence available for assessing their effectiveness. The chapter ends with a discussion of what research findings suggest about the most effective ways to teach and learn a second language in the classroom. In Chapter 7, we will provide a general summary of the book by looking at how research can inform our response to sorne 'popular opinions' about lan guage learning and teaching that are introduced below. A Glossary provides a quick reference for a number of terms that may be new or have specific technical meanings in the context of language acquisition research. Glossary words are shown in bold letters where they first appear in the text. For readers who would like to find out more, an annotated list of suggestions for further reading is included at the end of each chapter. The Bibliography provides full reference information for the suggested readings and all the works that are referred to in the text. We have tried to present the information in a way that does not assume that readers are already familiar with research methods or theoretical issues in second language learning. Examples and case studies are included through out the book to illustrate the research ideas. Many of the examples are taken from second language classrooms. We have also included a number of activi ties for readers to practise sorne of the techniques of observation and analysis used in the research that we review in this book. At the end of each chapter are 'Questions for reflection' to help readers consolidare and expand their understanding of the material. Before we begin... lt is probably true, as sorne have claimed, that most of us teach as we were taught or in away that matches our ideas and preferences about howwe learn. Take a moment to reflect on your views about how languages are learned and what you think this means about how they should be taught. The statements in the activity below summarize sorne popular opinions about language  Introduction 3 learning and teaching. Think about whether you agree or disagree with each opinion. Keep these statements and your reactions to them in mind as you read about current research and theory in second language learning. AC T I V I T Y Give your opinion on these statements Indicare the extent to which you agree with each statement by marki ng an X in the box associated with your opinion: SA-strongly agree A-agree somewhat D-disagree somewhat SD-strongly disagree SA A D SD 1Languages are learned mainly through imitation. 2 Parents usually correct young child ren when they make grammatical errors. 3 Highly intelligent people are good language learners. 4 The most im portant predictor of success in second language acquisition is motivation. 5 The earlier a second language is introduced in school programmes, the greater the likelihood of success in learni ng. 6 Most of the mistakes that second language learners make are due to interference from their first language. 7 The best way to learn new vocabulary is th rough reading. 8 lt is essential for learners to be able to pronou nce all the individual sounds in the second language. 9 Once learners know 1,000 words and the basic structu re of a language, they can easily participate in conversations with native speakers. 10 Teachers should present grammatical rules one at a time, and learners should practise examples of each one befare going on to another. 11Teachers should teach simple language structu res befare complex ones. 4 Introduction 12 Learners' errors should be corrected as soon as they are made i n order to prevent the formation of bad habits. 13 Teachers should use materials that expose students only to language structures they have already been taught. 14 When learners are allowed to interact freely (for example, in group or pai r activities), they copy each other's mistakes. 15 Students learn what they are taught. 16 Teachers should respond to students' errors by correctly reph rasing what they have said rather than by explicitly pointi ng out the error. 17 Students can learn both language and academic content (for example, science and history) simultaneously in classes where the subject matter is taught in their second language. 18 Classrooms are good places to learn about language but not for learning how to use language. Photocopiable © Oxford University Press 1 LAN GUAGE LEARNING IN EARLY CHI L D H O O D Preview In this chapter, we will look briefly at the language development of young children. We will then consider several theories that have been offered as explanations for how language is learned. There is an immense amount of research on child language. Although much of this research has been done in middle-class North American and European families, there is a rich body of cross-linguistic and cross-cultural research as well. Our purpose in this chapter is to touch on a few main points in this research, primarily as a prepa ration for the discussion of second language acquisition (SLA), which is the focus of this book. First language acquisition Language acquisition is one of the most impressive and fascinating aspects of human development. We listen with pleasure to the sounds made by a three-month-old baby. We laugh and 'answer' the conversational 'ba-ba-ba' babbling of older babies, and we share in the pride and joy of parents whose one-year-old has uttered the first 'bye-bye'. Indeed, learning a language is an amazing feat -o n e that has attracted the attention oflinguists and psycholo gists for generations. How do children accomplish this? What enables a child not only to learn words, but to put them together in meaningful sentences? What pushes children to go on developing complex grammatical language even though their early simple communication is successful for most pur poses? Does child language develop similarly around the world? How do bilingual children acquire more than one language? 6 Langu,age learning in early childhood The.first threeyears: Milestones and developmentalsequences One remarkable thing about first language acquisition is the high degree of similarity in the early language of children all over the world. Researchers have described developmental sequences far many aspects of first language acquisition. The earliest vocalizations are simply the involuntary crying that babies do when they are hungry or uncomfartable. Soon, however, we hear the cooing and gurgling sounds of contented babies, lying in their beds looking at fascinating shapes and movement around them. Even though they have little control over the sounds they make in these early weeks of life, infants are able to hear subde differences between the sounds of human languages. Not only do they distinguish the voice of their mothers from those of other speakers, they also seem to recognize the language that was spoken around their mother befare they were born. Furthermore, in cleverly designed experiments, researchers have demonstrated that tiny babies are capable of very fine auditory discrimination. For example, they can hear the difference between sounds as similar as 'pa' and 'ha'. Janet Werker, Patricia Kuhl, and others have used new technologies that allow us to see how sensitive infants are to speech sounds. What may seem even more remarkable is that infants stop making distinctions between sounds that are not phonemic in the language that is spoken around them. Far example, by the time they are a year old, babies who will become speakers of Arabic stop reacting to the difference between 'pa' and 'ha' which is not pho nemic in Arabic. Babies who regularly hear more than one language in their environment continue to respond to these differences far a longer period (Werker, Weikum, and Yoshida 2006). One important finding is that it is not enough far babies to hear language sounds from electronic devices. In arder to le a rn -o r retain -the ability to distinguish between sounds, they need to interact with a human speaker (Conboy and Kuhl 2011). The Internet abounds with remarkable videos of infants reacting to language sounds. Whether they are becoming monolingual or bilingual children, however, it will be many months befare their own vocalizations begin to reflect the characteristics of the language or languages they hear and longer still befare they connect language sounds with specific meaning. However, by the end of - their first year, most babies understand quite a few frequently repeated words j in the language or languages spoken around them. They wave when someone F says 'bye-bye'; they clap when someone says 'pat-a-cake'; they eagerly hurry to the kitchen when 'juice and cookies' are mentioned. At 12 months, most babies will have begun to produce a word or two that everyone recognizes. By the age of two, most children reliably produce at least 50 different words and sorne produce many more. About this time, they begin to combine words into simple sentences such as 'Mommy juice' and  Language learning in earlychildhood 7 'baby fall down'. These sentences are sometimes called 'telegraphic' because they leave out such things as anides, prepositions, and auxiliary verbs. We recognize them as sentences because, even though function words and gram matical morphemes are missing, the word order reflects the word order of the language they are hearing and the combined words have a meaningful relationship that makes them more than just a list of words. Thus, for an English-speaking child, 'kiss baby' does not mean the same thing as 'baby kiss'. Remarkably, we also see evidence, even in these early sentences that children are doing more than imperfectly imitating what they have heard. Their two- and three-word sentences show signs that they can creatively combine words. For example, 'more outside' may mean 'I want to go outside again.' Depending on the situation, 'Daddy uh-oh' might mean 'Daddy fell down' or 'Daddy dropped something' or even 'Daddy, please do that funny thing where you pretend to drop me off your lap.' As children progress through the discovery of language in their first three years, there are predictable patterns in the emergence and development of many features of the language they are learning. For sorne language features, these patterns have been described in terms of developmental sequences or 'stages'. To sorne extent, these stages in language acquisition are related to children's cognitive development. For example, children do not use temporal adverbs such as 'tomorrow' or 'last week' until they develop sorne under standing of time. In other cases, the developmental sequences seem to reflect the gradual acquisition of the linguistic elements for expressing ideas that have been present in children's cognitive understanding for a long time. For example, children can distinguish between singular and plural long before they reliably add plural endings to nouns. Correct use of irregular plurals (such as 'feet') takes even more time and may not be completely under control until the school years. Grammatical morphemes In the 1960s, severa! researchers focused on how children acquire grammati cal morphemes in English. One of the best-known studies was carried out by Roger Brown and his colleagues and students. In a longitudinal study of the language development of three children (called Adam, Eve, and Sarah) they found that 14 grammatical morphemes were acquired in a similar sequence. The list below (adapted from Brown's 1973 book) shows sorne of the mor phemes they studied. present progressive -ing (Mommy running) plural -s (two books) irregular past forms (Baby went) possessive -s (Daddy's hat) copula (Mommy is happy) anides the and a 8 Language learning in earlychildhood regular past -ed (she walked) third person singular simple present -s (she runs) auxiliary be (he is coming) Brown and his colleagues faund that a child who had mastered the gram matical morphemes at the bottom of the list had also mastered those at the top, but the reverse was not true. Thus, there was evidence far a 'developmen tal sequence' or arder of acquisition. However, the children did not acquire the morphemes at the same age or rate. Eve had mastered nearly all the mor phemes befare she was two-and-a-half years old, while Sarah and Adam were still working on them when they were three-and-a-half or faur. Brown's longitudinal work was confirmed in a cross-sectional study of 21 children. Jill and Peter de Villiers (1973) faund that children who correctly used the morphemes that Adam, Eve, and Sarah had acquired late were also able to use the ones that Adam, Eve, and Sarah had acquired earlier. The chil dren mastered the morphemes at different ages, just as Adam, Eve, and Sarah had done, but the arder of their acquisition was very similar. Many hypotheses have been advanced to explain why these grammatical morphemes are acquired in the observed arder. Researchers have studied the frequency with which the morphemes occur in parents' speech, the cognitive complexity of the meanings represented by each morpheme, and the difficulty of perceiving or pronouncing them. In the end, there has been no simple satis factory explanation far the sequence, and most researchers agree that the arder is determined by an interaction among a number of different factors. To supplement the evidence we have from simply observing children, sorne carefully designed procedures have been developed to further explore chil dren'sknowledge ofgrammatical morphemes. Oneo f the first and best known is the so-called 'wug test' developed by Jean Berko Gleason (1958). In this 'test', children are shown drawings of imaginary creatures with novel names or people perfarming mysterious actions. Far example, they are told, 'Here is a wug. Now there are two of them. There are two ' or 'Here is a man who knows how to bod. Yesterday he did the same thing. Yesterday, he '. By completing these sentences with 'wugs' and 'bodded', children demonstrate that they know the patterns far plural and simple past in English. By gener alizing these patterns to words they have never heard befare, they show that their language is more than just a list of memorized word pairs such as 'book/ books' and 'nod/nodded'. AC T I V I T Y Try out the 'wug' test A web search for 'wug test' will turn up many examples of the pictu res and the text created for this land mark research. lf you know sorne English-speaki ng child ren u nder the age of five years, try usi ng the test with them. '  Language learning in early childhood 9 What similarities and differences do you notice among the child ren at different ages? 2 Which grammatical morphemes do they find easy and which enes are more difficult? The acquisition of other language features also shows how children's language develops systematically, and how they go beyond what they have heard to create new forms and structures. Negation Children learn the functions of negation very early. That is, they learn to comment on the disappearance of objects, to refuse a suggestion, or to reject an assertion, even at the single word stage. However, as Lois Bloom's (1991) longitudinal studies show, even though children understand these func tions and express them with single words and gestures, it takes sorne time before they can express them in sentences, using the appropriate words and word order. The following stages in the development of negation have been observed in the acquisition of English. Similar stages have been observed in other languages as well (Wode 1981). Stage 1 Negation is usually expressed by the word 'no', either all alone or as the first r word in the utterance. No. No cookie. No comb hair. Stage2 1 Utterances grow longer and the sentence subject may be included. The neg s ative word appears just before the verb. Sentences expressing rejection or prohibition often use 'don't'. Daddy no comb hair. Don't touch that! Stage3 The negative element is inserted into a more complex sentence. Children - may add forms of the negative other than 'no', including words like 'can't' t and 'don't'. These sentences appear to follow the correct English pattern of attaching the negative to the auxiliary or modal verb. However, children do not yet vary these forms for different persons or tenses. I can't do it. He don't want it. Stage4 Children begin to attach the negative element to the correct form of auxiliary verbs such as 'do' and 'be'. 1O Language learning in earlychildhood You didn't have supper. She doesn't want it. Even though their language system is by now quite complex, they may still have difficulty with sorne other features related to negatives. I don't have no more candies. Questions The challenge oflearning complex language systems is also illustrated in the developmental stages through which children learn to ask questions. There is a remarkable consistency in the way children learn to form ques tions in English. For one thing, there is a predictable order in which the 'wh- words' emerge (Bloom 1991). 'What' is generally the first wh- question word to be used. It is often learned as part of a chunk. ('Whassat?') and it is sorne time before the child learns that there are variations of the form, such as 'What is that?' and 'What are these?'. 'Where' and 'who' emerge very soon. Identifying and locating people and objects are within the child's understanding of the world. Furthermore, adults tend to ask children just these types of questions in the early days of language learning, for example, 'Where's Mommy?' or 'Who's that?' 'Why' emerges around the end of the second year and becomes a favourite for the next year or two. Children seem to ask an endless number of questions beginning with 'why', having discovered how effectively this little word gets adults to engage in conversation, for example, 'Why that lady has blue hair?' Finally, when the child has a better understanding of manner and time, 'how' and 'when' emerge. In contrast to 'what', 'where', and 'who' questions, chil dren sometimes ask the more cognitively difficult 'why', 'when', and 'how' questions without understanding the answers they get, as the following con versation with a four-year-old clearly shows. CHILD When can we go outside? PARE N T In about five minutes. CHILD 1-2-3-4-5! Can we go now? The ability to use these question words is at least partly tied to children's cog nitive development. It is also predicted in part by the questions children are asked and the linguistic complexity of questions with different wh- words. Thus it does not seem surprising that there is consistency in the sequence of their acquisition. Perhaps more surprising is the consistency in the acquisi tion of word order in questions. This development is not based on learning n new meanings, but rather on learning different linguistic patterns to express meanings that are already understood. E w  Langu,age learning in early childhood 11 Stage 1 Children's earliest questions are single words or simple two- or three-word sentences with rising intonation: Cookie? Mommy book? At the same time, they may produce sorne correct questions-correct because they have been learned as chunks: Where's Daddy? What's that? Stage2 As they begin to ask more new questions, children use the word arder of the declarative sentence, with rising intonation. You like this? I have sorne? They continue to produce the correct chunk-learned forms such as 'What's that?' alongside their own created questions. Stage3 Gradually, children notice that the structure of questions is different and begin to produce questions such as: Can I go? Are you happy? Although sorne questions at this stage match the adult pattern, they may be right for the wrong reason. To describe this, we need to see the pattern from the child's perspective rather than from the perspective of the adult grammar. We call this stage 'fronting' because the child's rule seems to be that questions are formed by putting something (a verb or question word) at the 'front' of a sentence, leaving the rest of the sentence in its statement form. Is the teddy is tired? Do I can have a cookie? Why you don't have one? Why you catched it? Stage 4 At Stage 4, sorne questions are formed by subject-auxiliary inversion. The questions resemble those of Stage 3, but there is more variety in the auxilia ries that appear befare the subject. f Are you going to play with me? At this stage, children can even add 'do' in questions in which there would be , no auxiliary in the declarative version of the sentence. Do dogs like icecream? Even at this stage, however, children seem able to use either inversion or a wh word, but not both (for example, 'Is he crying?' but not 'Why is he crying?' 12 Language learning in early childhood Therefore, we may find inversion in yes/no questions but not in wh- ques tions, unless they are formulaic units such as 'What's that?' Stage5 At Stage 5, both wh- and yes/no questions are formed correctly. Are these your boots? Why did you do that? Does Daddy have a box? Negative questions may still be a bit too difficult. Why the teddy bear can't go outside? And even though performance on most questions is correct, there is still one more hurdle. When wh- words appear in subordinate clauses or embedded questions, children overgeneralize the inverted form that would be correct for simple questions and produce sentences such as: Ask him why can't he go out. Stage 6 At this stage, children are able to correctly form all question types, including negative and complex embedded questions. Passage through developmental sequences does not always follow a steady uninterrupted path. Children appear to learn new things and then fall back on old patterns when there is added stress in a new situation or when they are using other new elements in their language. But the overall path takes them toward a closer and closer approximation of the language that is spoken around them. Thepre-schoolyears By the age of four, most children can ask questions, give commands, repon real events, and create stories about imaginary ones, using correct word order and grammatical markers most of the time. In fact, it is generally accepted that by age four, children have acquired the basic structures of the language or languages spoken to them in these early years. Three- and four-year-olds continue to learn vocabulary at the rate of several words a day. They begin to acquire less frequent and more complex linguistic structures such as passives re. and relative clauses. ·ca IS Much of children's language acquisition effort in the late pre-school years is spent in developing their ability to use language in a widening social environ of ment. They use language in a greater variety of situations. They interact more mt rid, often with unfamiliar adults. They begin to talk sensibly on the telephone to invisible grandparents (younger children do not understand that their telephone partner cannot see what they see). They acquire the aggressive or  Language learning in earlychildhood 13 ;- cajoling language that is needed to defend their toys in the playground. They show that they have learned the difference between how adults talk to babies and how they talk to each other, and they use this knowledge in elaborare pretend play in which they practise using these different 'voices'. In this way, they explore and begin to understand how and why language varies. In the pre-school years, children also begin to develop metalinguistic aware ness, the ability to treat language as an object separare from the meaning it conveys. Three-year-old children can tell you that it's 'silly' to say 'drink the chair', because it doesn't make sense. However, although they would never say 'cake the eat', they are less sure that there's anything wrong with it. They may show that they know it's a bit odd, but they will focus mainly on the fact 1e that they can understand what it means. Five-year-olds, on the other hand, d know that 'drink the chair' is wrong in a different way from 'cake the eat'. :t They can tell you that one is 'silly' but the other is 'the wrong way around'. Language acquisition in the pre-school years is impressive. lt is also note worthy that children have spent thousands of hours interacting with language-participating in conversations, eavesdropping on others' con versations, being read to, watching television, etc. A quick mathematical g exercise will show you just how many hours children spend in language-rich environments. If children are awake for ten or twelve hours a day, we may estimate that they are in contact with the language of their environment for y k 20,000 hours or more by the time they go to school. :y Although pre-school children acquire complex knowledge and skills for s language and language use, the school setting requires new ways of using n language and brings new opportunities for language development. Theschoolyears Children develop the ability to use language to understand others and to rt express their own meanings in the pre-school years, and in the school years, r this ability expands and grows. Learning to read gives a major boost to meta d linguistic awareness. Seeing words represented by letters and other symbols ;e on a page leads children to a new understanding that language has form as is well as meaning. Reading reinforces the understanding that a 'word' is sepa :o rare from the thing it represents. Unlike three-year-olds, children who can s read understand that 'the' is a word, just as 'house' is. They understand that 'caterpillar' is a longer word than 'train', even though the object it represents is substantially shorter! Metalinguistic awareness also includes the discovery is 1- of such things as ambiguity. Knowing that words and sentences can have re multiple meaning gives children access to word jokes, trick questions, and te riddles, which they love to share with their friends and family. ir >r 14 Language learning in earlychildhood One of the most impressive aspects of language development in the school years is the astonishing growth of vocabulary. Children enter school with the ability to understand and produce several thousand words, and thousands more will be learned at school. In both the spoken and written language at school, words such as 'homework' or 'ruler' appear frequently in situations where their meaning is either immediately or gradually revealed. Words like 'population' or 'latitude' occur less frequently, but they are made important by their significance in academic subject matter. Vocabulary grows at a rate of between several hundred and more than a thousand words a year, depending mainly on how much and how widely children read (Nagy, Herman, and Anderson 1985). The kind of vocabulary growth required for school success is likely to come from both reading for assignments and reading for pleasure, whether narrative or non-fiction. Dee Gardner (2004) suggests that reading a variety of text types is an essential part of vocabulary growth. His research has shown how the range of vocabulary in narrative texts is different from that in non-fiction. There are words in non-fiction texts that are unlikely to occur in stories or novels. In addition, non-fiction tends to include more opportunities to see a word in its different forms (for example, 'mummy', 'mummies', 'mummified'). The importance of reading for vocabulary growth is seen when observant parents report a child using a new word but mispronouncing it in a way that reveals it has been encountered only in written form. Another important development in the school years is the acquisition of dif ferent language registers. Children learn how written language differs from spoken language, how the language used to speak to the principal is different from the language of the playground, how the language of a science report is different from the language of a narrative. As Terry Piper (2006) and others have documented, sorne children will have even more to learn if they come to school speaking an ethnic or regional variety of the school language that is quite different from the one used by the teacher. They will have to learn that another variety, often referred to as the standard variety, is required for suc cessful academic work. Other children arrive at school speaking a different language altogether. For these children, the work oflanguage learning in the early school years presents additional opportunities and challenges. We will return to this topic when we discuss bilingualism in early childhood. Explaining first language acquisition These descriptions oflanguage development from infancy through the early school years show that we have considerable knowledge of what children learn in their early language development. More controversia!, however, are questions about how this development takes place. What abilities does the child bring to the task and what are the contributions of the environment?  Language learning in early childhood 15 Since the middle of the 20th century, three main theoretical positions have.e been advanced to explain language development: behaviourist, innatist, and ls interactional/ developmental perspectives. Lt LS 1he behaviouristperspective e Behaviourism is a theory of learning that was influential in the 1940s and Lt 1950s, especially in the United States. With regard to language learning, the best-known proponent of this psychological theory was B. F. Skinner a (1957). Traditional behaviourists hypothesized that when children imitated y the language produced by those around them, their attempts to reproduce y what they heard received 'positive reinforcement'. This could take the form r of praise or just successful communication. Thus encouraged by their envi e ronment, children would continue to imitate and practise these sounds and t patterns until they formed 'habits' of correct language use. According to this y view, the quality and quantity of the language the child hears, as well as the n consistency of the reinforcement offered by others in the environment, , would shape the child's language behaviour. This theory gives great impor t tance to the environment as the source of everything the child needs to learn. e a Analysing children'sspeech: Definitions and examples S The behaviourists viewed imitation and practice as the primary processes in language development. To clarify what is meant by these two terms, consider the following definitions and examples. - n Imitation: word-for-word repetition of all or part of someone else's utterance. t MOTHER Shall we play with the dolls? s LUCY Play with dolls s e Practice: repetitive manipulation of form. s C I N DY He eat carrots. The other one eat carrots. They both eat t carrots. t Now examine the transcripts from Peter, Cindy, and Kathryn. They were all e about 24 months old when they were recorded as they played with a visiting 1 adult. Using the definitions above, notice how Peter imitates the adult in the following dialogue. Peter (24 months) is playing with a dump truck while two adults, Patsy and Lois, look on. PETER Get more. LO IS You're gonna put more wheels in the dump truck? PETER Dump truck. Wheels. Dump truck. (later) PATSY What happened to it (the truck)? 16 Language learning in earlychildhood PETER (looking under chair for it) Lose it. Dump truck! Dump truck! Fall! Fall! LOIS Yes, the dump truck fell clown. PETER Dump truck fell clown. Dump truck. (Unpublished data from P. M. Lightbown) If we analysed a larger sample of Peter's speech, we would see that 30-40 per cent of his sentences were imitations of what someone else had just said. We would also see that his imitations were not random. That is, he did not simply imitare 30-40 per cent of everything he heard. Detailed analyses of large samples of Peter's speech over about a year showed that he imitated words and sentence structures that were just beginning to appear in his spontaneous speech. Once these new elements became solidly grounded in his language system, he stopped imitating them and went on to imitare others. Unlike a parrot who imitares the familiar and continues to repeat the same things again and again, children appear to imitare selectively. The choice of what to imitare seems to be based on something new that they have just begun to understand and use, not simply on what is available in the environ ment. For example, consider how Cindy imitares and practises language in the following conversations. Cindy (24 months, 16 days) is looking at a picture of a carrot in a book and trying to get Patsy's attention. CINDY Kawo? kawo? kawo? kawo? kawo? PATSY What are the rabbits eating? C I N DY They eating... kando? PATSY No, that's a carrot. CINDY Carrot. (pointing to each carrot on the page) The other... carrot. The other carrot. The other carrot. (A few minutes later, Cindy brings Patsy a stuffed toy rabbit.) PATSY What does this rabbit like to eat? CINDY (incomprehensible) eat the carrots. (Cindy gets another stuffed rabbit.) C I N DY He (incomprehensible) eat carrots. The other one eat carrots. They both eat carrots. (One week later, Cindy opens the book to the same page.) CINDY Here's the carrots. (pointing) Is that a carrot? PATSY Yes. (Unpublished data from P. M. Lightbown)  Language learning in earlychildhood 17 Cindy appears to be working hard on her language acquisition. She prac tises new words and structures in a way that sounds like a student in sorne foreign language classes! Perhaps most interesting is that she remembers the 'language lesson' a week later and turns straight to the page in the book she had not seen since Patsy's last visit. What is most striking is that, like Peter, her imitation and practice appear to be focused on what she is currendy Jer 'working on'. X'e The samples of speech from Peter and Cindy seem to lend sorne support to Jly the behaviourist explanation of language acquisition. Even so, as we saw, the ·ge choice of what to imitare and practise seemed determined by something ·ds us inside the child rather than by the environment. ge Not all children imitate and practise as much as Peter and Cindy did. The amount of imitation in the speech of other children, whose development ne proceeded at a rate comparable to that of Cindy and Peter, has been cal ce culated at less than 1O per cent. Consider the examples of imitation and practice in the following conversation between Kathryn and Lois. 1st n Kathryn (24 months) in LO IS Did you see the toys I brought? KATH RYN I bring toys? Choo choo? Lois brought the choo choo train? LOIS Yes, Lois brought the choo choo train. KATHRYN (reaching for bag) I want play with choo choo train. I want play with choo choo train. (taking out slide) Want play.What's this? LOIS Oh you know what that is. KATH RYN Put clown on floor. This. I do this. (Kathryn puts the slide on the floor.) KATH RYN (taking out two cars of train) Do this. I want do this. (trying to put train together) I do this. I do this. LOIS OK. You can do it. You can do it. Look I'll show you how. U.ois puts it together.) KATH RYN (searching in box) I get more. Get a more. No more choo choo train. Get truck. (taking out truck) Kathryn truck. Where? Where a more choo choo train? LOIS Inside. It's in the box. KATHRYN A choo choo? (taking out part of train) This is a choo choo train. tfrom Bloom and Lahey 1978: 135) like Cindy, Kathryn sometimes repeats herself or produces a series of related practice sentences, but she rarely imitares the other speaker. Instead, she asks md answers questions and elaborates on the other speaker's questions or statements. 18 Language learning in earlychildhood Thus, children vary in the amount of imitation they do. In addition, many of the things they say show that they are using language creatively, not just repeating what they have heard. This is evident in the following examples. Patterns inlanguage The first example shows a child in the process of learning patterns in lan guage, in this case the rules of word formation, and overgeneralizing them to new contexts. Randall (36 months) had a sore on his hand. MOTHER Maybe we need to take you to the doctor. RAN DALL Why? So he can doc my little bump? Randall forms the verb 'doc' from the noun 'doctor', by analogy with farmers who farm, swimmers who swim, and actors who act. Focus on meaning Even older children have to work out sorne puzzles, for example, when famil iar language is used in unfamiliar ways, as in the example below. When David (5 years, 1 month) was at his older sister's birthday party, toasts were pro posed with grape juice in stemmed glasses: FATHER I'd like to propose a toast. Severa! minutes later, David raised his glass: D A V ID I'd like to propase a piece of bread. Only when laughter sent David slinking from the table did the group realize that he wasn't intentionally making a play on words! He was concentrating  Language learning in earlychildhood 19 ny so hard on perfarming the fascinating new gesture and the farmulaic expres 1st sion Td like to propase...' that he failed to realize that the word he thought he knew-'toast'-was not the same toast and could not be replaced with its apparent near-synonym, 'a piece of bread'. n Questionformation :m Randall (2 years, 9 months) asked the fallowing questions in various situa tions over the course of a day. Are dogs can wiggle their tails? Are those are my boots? Are this is hot? Randall had concluded that the trick of asking questions was to put 'are' at rs the beginning of the sentence. His questions are good examples of Stage 3 in question development. Order of events Randall (3 years, 5 months) was looking far a towel. You took all the towels away because I can't dry my hands. He meant 'I can't dry my hands because you took all the towels away', but he made a mistake about which clause comes first. Children at this stage of language development tend to mention events in the order of their occur rence. In this case, the towels disappeared befare Randall attempted to dry his hands, so that's what he said first. He did not yet understand how a word like 'befare' or 'because' changes the order of cause and effect. These examples of children's speech provide us with a window on the process of language learning. Imitation and practice alone cannot explain sorne of \ the farms created by children. They are not merely repetitions of sentences that they have heard from adults. Rather, children appear to pick out patterns and generalize them to new contexts. They create new farms or new uses of words. Their new sentences are usually comprehensible and often correct. Behaviourism seems to offer a reasonable way of understanding how children learn sorne of the regular and routine aspects oflanguage, especially at the ear il liest stages. However, children who do little overt imitation acquire language id as fully and rapidly as those who imitate a lot. And although behaviourism 'O- goes sorne way to explaining the sorts of overgeneralization that children make, classical behaviourism is not a satisfactory explanation far the acquisi rion of the more complex grammar that children acquire. These limitations led researchers to look far different explanations far language acquisition. ize ng 20 Language learning in earlychildhood lhe innatistperspective Noam Chomsky is one of the most influential figures in linguistics, and his ideas about how language is acquired and how it is stored in the mind sparked a revolution in many aspects of linguistics and psychology, includ ing the study of language acquisition. The innatist perspective is related to Chomsky's hypothesis that all human languages are based on sorne innate universal principles. In his 1959 review of B. F. Skinner's book Verbal Behavior, Chomsky chal lenged the behaviourist explanation far language acquisition. He argued that children are biologically programmed far language and that language devel ops in the child in just the same way that other biological functions develop. Far example, every child will learn to walk as long as adequate nourishment and reasonable freedom of movement are provided. The child does not have to be taught. Most children learn to walk at about the same age, and walking is essentially the same in all normal human beings. Far Chomsky, language acquisition is very similar. The environment makes only a basic contribu t i o n - i n this case, the availability of people who speak to the child. The child, or rather, the child's biological endowment, will do the rest. Chomsky argued that the behaviourist theory failed to account far 'the logical problem oflanguage acquisition'-the fact that children come to know more about the structure of their language than they could reasonably be expected to learn on the basis of the samples of language they hear. The language children are exposed to includes false starts, incomplete sentences, and slips of the tangue, and yet they learn to distinguish between grammatical and ungrammatical sentences. He concluded that children's minds are not blank slates to be filled by imitating language they hear in the environment. Instead, he hypothesized, children are born with a specific innate ability to discover far themselves the underlying rules of a language system on the basis of the samples of a natural language they are exposed to. This innate endowment was seen as a sort of templare, containing the principles that are universal to all human languages. This universal grammar (UG) would prevent the child from pursuing all sorts of wrong hypotheses about how language systems might work. If children are pre-equipped with UG, then what they have to learn is the ways in which the language they are acquiring makes use of these principles. Consider the fallowing sentences, from a book by Lydia White ( 1989). These English sentences contain the reflexive pronoun 'himself'. Both the pronoun and the noun it refers to (the antecedent) are printed in italics. (An asterisk at the beginning of a sentence indicares that the sentence is ungrammatical.) a John saw himself. b *Himselfsawjohn.  Language learning in earlychildhood 21 In (a) and (b), it looks as if the reflexive pronoun must follow the noun it refers to. But (e) disproves this: md lnd e Looking after himself boresjohn. 1d If we consider sentences such as: l to ate d John said that Fred liked himse/f. e *John said that Fred liked himse/f. f John told Bill to wash himse/f. 1al g *John told Bill to wash himse/f. hat rel we might conclude that the noun closest to the reflexive pronoun is the ante op. cedent. However, (h) shows that this rule won't work either: ent h john promised Bill to wash himself. ave ing And it's even more complicated than that. Usually the reflexive must be in age the same clause as the antecedent as in (a) and (d), but not always, as in (h). bu Furthermore, the reflexive can be in the subject position in (i) but not in (j). Ihe i. ]ohn believes himself to be intelligent (non-finite clause). j. *John believes that himself is intelligent (finite clause). ical In sorne cases, more than one antecedent is possible, as in (k) where the ore reflexive could refer to either John or Bill: :ted age k ]ohn? showed BilP. a picture of himself. lips When we look at this kind of complexity, it seems it would be very hard t o md learn, and children do make errors along the way. Yet, most school-age mk children would be able to correctly interpret the grammatical sentences and :ad, recognize the ungrammaticality of the others. Researchers who study ,ver language acquisition from the innatist perspective argue that such complex the grammar could never be learned purely on the basis of imitating and practis 1ent ing sentences available in the input. They hypothesize that since all children.1to acquire the language of their environment, they must have sorne innate 1ild mechanism or knowledge that allows them to discover such complex syntax ms in spite of limitations of the input. They hypothesize furthermore that the e to innate mechanism is used exclusively for language acquisition. 1ese The innatist perspective emphasizes the fact that almost all children success fully acquire their native language-or more than one language if they live 1ese in a multilingual community. Children who are profoundly deaf will learn Un sign language if they are exposed to it in infancy, and their progress in the risk acquisition of that language system is similar to hearing children's acquisition :al.) of spoken language. Even children with very limited cognitive ability develop quite complex language systems if they are brought up in environments in which people interact with them. 22 Language learning in earlychildhood Children acquire the basic syntax and morphology of the language spoken to them in a variety of conditions, sorne of which would be expected to enhance language development (for example, caring, attentive parents who focus on the child's language), and sorne which might be expected to inhibir it (for example, abusive or rejecting parents). Children achieve different levels of vocabulary, creativity, social grace, and so on, but virtually all achieve the ability to use the patterns of the language or languages spoken to them. This is seen as support for the hypothesis that language is somehow separare from other aspects of cognitive development and may depend on a specific module of the brain. The Critica!Period Hypothesis The innatist perspective is often linked to the Critica! Period Hypothesis ( C P H ) -t h e hypothesis that animals, including humans, are genetically programmed to acquire certain kinds of knowledge and skill at specific times in life. Beyond those 'critica! periods', it is either difficult or impossible to acquire those abilities. With regard to language, the CPH suggests that chil dren who are not given access to language in infancy and early childhood (because of deafness or extreme isolation) will never acquire language if these deprivations go on for too long. lt is difficult to find evidence for or against the CPH, since nearly all children are exposed to language at an early age. However, history has documented a few 'natural experiments' where children have been deprived of contact with language. Two of the most famous cases are those of 'Victor' and 'Genie'. In 1799, a hoy who became known as Victor was found wandering naked in the woods in France. His story was dramatized in a 1970 film by Frarn;:ois Truffaut called L'enfantsauvage ( The Wild Child). When Victor was captured, he was about 12 years old and completely wild, apparently having had no contact with humans. Jean-Marc-Gaspard ltard, a young doctor accustomed to working with deaf children, devoted five years to socializing Victor and trying to teach him language. Although he succeeded to sorne extent in devel oping Victor's sociability, memory, and judgement, there was little progress in his language ability. Neirly 200 years later, Genie, a 13-year-old girl who had been isolated, neglected, and abused, was discovered in California. Because of the irrational demands of a disturbed father and the submission and fear of an abused mother, Genie had spent more than 11 years tied to a chair or a crib in a small, darkened room. Her father had forbidden his wife and son to speak to Genie and had himself only growled and barked at her. She was beaten when she made any kind of noise, and she had long since resorted to complete silence. Genie was undeveloped physically, emotionally, and intellectually. She had no language.  Language learning in earlychildhood 23 After she was discovered, Genie was cared for and educated with the par ticipation of many teachers and therapists, including Susan Curtiss ( 1977). After a brief period in a rehabilitation centre, she lived in a foster home and attended special schools. Genie made remarkable progress in becoming socialized and cognitively aware. She developed deep personal relationships and strong individual tastes and traits. Nevertheless, after five years of expo sure to language, Genie's language was not like that of a typical five-year old. There was a larger than normal gap between comprehension and produc tion. She used grammatical forms inconsistently and overused formulaic and routine speech. Although Victor and Genie appear to provide evidence in support of the CPH, it is difficult to argue that the hypothesis is confirmed on the basis of evidence from such unusual cases. We cannot know what other factors besides biological maturity might have contributed to their inability to learn language. It is not possible to determine whether either of them suffered from brain damage, developmental delays, or a specific language impair ment, even befare they were separated from normal human interaction. A more appropriate test of the CPH is the case of children who come from homes where they receive love and care from their parents, yet do not have access to language at the usual time. This is the case for sorne profoundly deaf children who have hearing parents. Only 5-1O per cent of the pro foundly deaf are born to deaf parents, and only these children are likely to be exposed to ASL from birth. Hearing parents may not realize that their child cannot hear because the child uses other senses to interact in an apparently normal way. Thus, the early childhood period may be normal in most ways but devoid of language that is accessible to the child. These children's later experience in learning sign language has been the subject of sorne important research related to the CPH. Like oral and written languages, American Sign Language (ASL) makes use of grammatical markers to indicate such things as time (for example, past tense) and number. These markers are expressed through specific hand or body movements. Elissa Newport (1990) and her colleagues studied the ability of deaf users of ASL to produce and comprehend grammatical markers. They compared Native signers (who were exposed to ASL from birth), Early signers (who began using ASL between four and six years of age), and Late signers (who began learning ASL after age 12). They found no difference between the groups in sorne aspects of their use of ASL, for example in vocabulary knowl edge. However, on tests focusing on grammatical markers, the Native group used the markers more consistently than the Early group who, in turn, used them more consistently than the Late group. The researchers concluded that 24 Language learning in early childhood their study supports the hypothesis that there is a critical period for first lan guage acquisition, whether that language is oral or gestural. Another line of research that has given new insight into the importance of early language experience comes from studies of 'international adoptees.' These are children who were adopted at an early age by families who did not speak the language the child had heard during infancy. In their review of studies of international adoptees, Johanne Paradis, Fred Genesee, and Martha Crago (2011) concluded that cognitive and linguistic outcomes were generally very positive. Sorne comparisons of their language with that of children the same age who had always heard the same language showed that subtle differences persist even after several years, but these are not the kinds of differences that most people would notice. Here again, of course, one cannot know whether something other than a late exposure to the language spoken in the adoptive environment also contributed to differences between these children and others who did not experience an abrupt change in their language environment. Nevertheless, with continuing research on children's linguistic behaviours and intuitions, as well as the neurological studies of infants' speech perception that we saw above, it is becoming clearer that language acquisition begins at birth, and possibly even befare, as the child's brain is shaped by exposure to the language(s) in the environment. The innatist perspective is thus partly based on evidence that there is a criti cal period for language acquisition. It is also seen as an explanation for 'the logical problem of language acquisition', that is, the question of how adult speakers come to know the complex structure of their first language on the basis of the limited samples oflanguage to which they are exposed. lnteractionistldevelopmental perspectives Developmental and cognitive psychologists have focused on the interplay between the innate learning ability of children and the environment in which they develop. They argue that the innatists place too much emphasis on the 'final state' (the competence of adult native speakers) and not enough on the developmental aspects oflanguage acquisition. In their view, language acquisi tion is but one example of the human child's ability to learn from experience, and they see no need to assume that there are specific brain structures devoted to language acquisition. They hypothesize that what children need to know is essentially available in the language they are exposed to as they hear it used in thousands of hours of interactions with the people and objects around them. Psychologists attribute considerably more importance to the environment than the innatists do even though they also recognize a powerful learning mechanism in the human brain. They see language acquisition as similar to and infl.uenced by the acquisition of other kinds of skill and knowledge, rather than as something that is different from and largely independent of  Language learning in earlychildhood 25 the child's experience and cognitive development. Indeed, researchers such as Dan Slobin (1973) have long emphasized the close relationship between children's cognitive development and their acquisition oflanguage. Piaget and Vygotsky One of the earliest proponents of the view that children's language is built on their cognitive development was the Swiss psychologist/epistemologist, Jean Piaget (1951). In the early decades of the 20th century, Piaget observed infants and children in their play and in their interaction with objects and people. He was able to trace the development of their cognitive understand ing of such things as object permanence (knowing that things hidden from sight are still there), the stability of quantities regardless of changes in their appearance (knowing that 1O pennies spread out to form a long line are not more numerous than 10 pennies in a tightly squeezed line), and logical inferencing (figuring out which properties of a set of rods (their size, weight, material, etc.) cause sorne rods to sink and others to float on water). lt is easy to see how children's cognitive development would partly deter mine how they acquire language. For example, the use of certain terms such as 'bigger' or 'more' depends on the children's understanding of the con cepts they represent. The developing cognitive understanding is built on the interaction between the child and the things that can be observed or manipu lated. For Piaget, language was one of a number of symbol systems that are developed in childhood. Language can be used to represent knowledge that children have acquired through physical interaction with the environment. Another influential student of child development was the psychologist Lev Vygotsky (1978). He observed interactions among children and also between children and adults in schools in the Soviet Union in the 1920s and 1930s. He concluded that language develops primarily from social interaction. He argued that in a supportive interactive environment, children are able to advance to higher levels of knowledge and performance. Vygotsky referred to a metaphorical place in which children could do more than they would be capable of doing independently as the zone of proximal development (ZPD). Vygotsky observed the importance of conversations that children have with adults and with other children and saw in these conversations the origins of both language and thought. The conversations provide the child with scaf folding, that is, a kind of supportive structure that helps them make the most of the knowledge they have and also to acquire new knowledge. Vygotsky's view differs from Piaget's. Piaget saw language as a symbol system that could be used to express knowledge acquired through interaction with the physical world. For Vygotsky, thought was essentially internalized speech, and speech emerged in social interaction. Vygotsky's views have become increasingly central in research on second language development, as we will see in Chapter 4. 26 Language learning in earlychildhood Cross-cultural research Since the 1970s, researchers have studied children's language learning envi ronments in a great many different cultural communities. The research has focused not only on the development oflanguage itself, but also on the ways in which the environment provides what children need for language acquisi tion. Between 1985 and 1997, Dan Slobin edited five volumes devoted to research on the acquisition of 28 languages, providing examples and analyses of child language and the language-learning environment from communities around the world. One of the most remarkable resources for child language researchers is the Child Language Data Exchange System (CHILDES), where researchers have contributed child language data in dozens of lan guages in recorded and transcribed forms that are available as electronic files from the CHILDES website (MacWhinney 2000). One feature of cross-cultural research is the description of child-rearing pat terns. Catherine Snow (1995) and others have studied the apparent effects on language acquisition of the ways in which adults talk to and interact with young children. In middle-class North American homes, researchers observed that adults often modify the way they speak when talking to little children. This child-directed speech may be characterized by a slower rate of delivery, higher pitch, more varied intonation, shorter, simpler sentence pat terns, stress on key words, frequent repetition, and paraphrase. Furthermore, tapies of conversation emphasize the child's immediate environment, picture books, or experiences that the adult knows the child has had. Adults often repeat the content of a child's utterance, but they expand or recast it into a grammatically correct sentence. For example, when Peter says, 'Dump truck! Dump truck! Fall! Fall!', Lois responds, 'Yes, the dump truck fell down.'  Language learning in early childhood 27 Researchers working in a 'language socialization' framework have found that the kind of child-directed speech observed in middle-class American homes is by no means universal. In sorne societies, adults do not engage in conversation or verbal play with very young children. For example, Bambi Schieffelin (1990) found that Kaluli mothers in Papua New Guinea did not consider their children t o be appropriate conversational partners. Martha Crago (1992) observed that in traditional lnuit society, children are expected t o watch and listen t o adults. They are not expected or encouraged t o par ticipate in conversations with adults until they are older and have more developed language skills. Other researchers have observed that in sorne societies, young children inter act primarily with older siblings who serve as their caregivers. Even within the United States, Shirley Brice Heath (1983) and others have documented substantial differences in the ways parents in different socioeconomic and ethnic groups interact with their children. Nevertheless, in every society, children are in situations in which they hear language that is meaningful to them in their environment. And they acquire the community language. Thus, it is difficult t o judge the long-term effect of the modifications that sorne adults make in speech addressed t o children. Tu.e importance of interaction The role of interaction between a language-learning child and an interlocu tor who responds t o the child is illuminated by cases where such interaction is missing. Jacqueline Sachs and her colleagues (1981) studied the language development of a child they called Jim. Hewas a hearing child of deaf parents, and his only contact with oral language was through television, which he watched frequently. The family was unusual in that the parents did not use sign language with Jim. Thus, although in other respects he was well cared for, Jim did not begin his linguistic development in a normal environment in which a parent communicated with him in either oral or sign language. A language assessment at three years and nine months indicated that he was well below age level in all aspects of language. Although he attempted to express ideas appropriate to his age, he used unusual, ungrammatical word order. When Jim began conversational sessions with an adult, his expressive abili ties began to improve. By the age of four years and two months most of the unusual speech patterns had disappeared, replaced by language more typical of his age. Jim's younger brother Glenn did not display the same type of language delay. Glenn's linguistic environment was different from Jim's: he had his older brother-not only as a model, but, more importantly as a con versational partner whose interaction allowed Glenn t o develop language in a more typical way. 28 Language learning in earlychildhood Jim showed very rapid acquisition of English once he began to interact with an adult on a one-to-one basis. The fact that he had failed to acquire language nor mally prior to this experience suggests that impersonal sources oflanguage such as television or radio alone are not sufficient. One-to-one interaction gives chil dren access to language that is adjusted to their level of comprehension. When a child does not understand, the adult may repeat or paraphrase. The response of the adult may also allow children to find out when their own utterances are understood. Television, for obvious reasons, does not provide such interaction. Even in children's programmes, where simpler language is used and topics are relevant to younger viewers, no immediate adjustment is made for the needs of an individual child. Once children have acquired sorne language, however, television can be a source oflanguage and cultural information. Usage-based learning f u more and more research has documented the ways in which children interact with the environment, developmental and cognitive psychologists find further evidence that language acquisition is 'usage-based'. Inthis view, language acquisition is possible because of children's general cognitive capac ities and the vast number of opportunities they have to make connections between the language they hear and what they experience in their environ ment. Sophisticated electronic recording devices have been used to track and count words and phrases children hear in their daily lives. Deb Roy documented his son's acquisition of words, showing the frequency and the contexts for the occurrence of language. Most remarkable, perhaps, is the demonstration of the power of interaction between the child and the adults and how adults focus on the language the child has begun to use (Roy 2009). The usage-based perspective on language acquisition differs from the behav iourist view in that the emphasis is more on the child's ability to create networks of associations rather than on processes of imitation and habit for mation. Referred to by various names, including cognitive linguistics, this view also differs sharply from the innatists' because language acquisition is not seen as requiring a separate 'module of the mind' but rather depends on the child's general leaming abilities and the contributions of the environ ment. As Elena Lieven and Michael Tomasello (2008) put it, 'Children leam language from their language experiences-there is no other way' (p.168). According to this view, what children need to know is essentially available to them in the language they are exposed to. Sorne of the early research in this framework was done in the context of con nectionism and involved computer simulations in which language samples were provided as input to a fairly simple program. The goal was to show that the computer could 'leam' certain things if exposed to enough examples. The program was found to be able to sort out the pattems from the input and even generalize beyond what it was actually exposed to. Iteven made the  Language learning in early childhood 29 same kinds of creative 'mistakes' that children make, such as putting a regular -ed ending on an irregular verb, for example, eated. In a usage-based model, language acquisition involves not only associating words with elements of externa! reality.ltis also a process of associating words and phrases with the other words and phrases that occur with them, or words with grammatical morphemes that occur with them. For example, children learning languages in which nouns have grammatical gender learn to associ ate the appropriate article and adjective forms with nouns. So if children are learning French, they learn that la and une go with chaise (chair) and le and un go with livre (book). Similarly, they learn to associate pronouns with the verb forms that mark person and number-il aime (he likes) and nous aimons (we like). They also learn which temporal adverbs go with which verb tenses. Of particular importance to this hypothesis is the fact that children are exposed to many thousands of opportunities to learn words and phrases. Lea.rning takes place gradually, as the number oflinks between language and meaning and among language forms are built up. For usage-based theorists, acquisition of language, while impressive, is not the only remarkable feat accomplished by the child. They compare it to other cognitive and percep rual learning, including learning to 'see'. That is, the visual abilities that we cake for granted, for example, focusing on and interpreting objects in our ,risual field, are actually learned through experience. Language disorders and delays Although most children progress through the stages of language develop ment without significant difficulty or delay, there are sorne children for whom this is not the case. A discussion of the various types of disabilities (including deafness, articulatory problems, autism, dyslexia, and so on) that sometimes affect language development is outside the scope of this book. ltis essential that parents and teachers be encouraged to seek professional advice if they feel that a child is not developing language normally, keeping in mind that the range for 'normal' is wide indeed. 'While most children produce recognizable first words by 12 months, sorne may not speak before the age of three years. In very young children, one way to determine whether delayed language reflects a problem or simply an individual difference within the normal range is to determine whether the child responds 'ID language and appears to understand even if he or she is not speaking. For older children, delays in learning to read that seem out of keeping with a child's overall cognitive functioning may suggest that there is a specific problem in that domain. Sorne children seem to begin reading almost by magic, discovering the mysteries of print with little direct instruction. For most children, instruction that includes sorne systematic attention to 30 Language learning in earlychildhood sound-letter correspondences allows them to unlock the treasure chest of reading. Both groups fall within a normal range. For sorne children, however, reading presents such great challenges that they need expert help beyond what is available in a typical classroom. Childhood bilingualism The language development of children who learn multiple languages during childhood is of enormous importance throughout the world. Indeed, the majority of the world's children are exposed to more than one language. Sorne children learn multiple languages from earliest childhood; others acquire additional languages when they go to school. The acquisition and maintenance of more than one language can open doors to many personal, social, and economic opportunities. Unfortunately, as Jim Cummins (2000) and others have pointed out, chil dren who already know one or more languages and who arrive at their first day of school without an age-appropriate knowledge of the language of the school have often been misdiagnosed as having language delays or disorders. This includes immigrant and minority language children who do not speak the school language at home and children who speak a different variety of the school language. These children's knowledge of a different language or language variety is often incorrectly interpreted as a lack of normal language development and a lack of background knowledge for school subjects. They may be placed in remedia! or special education classes because schools are not equipped to provide an adequate assessment of children's ability to use their home language or of their general cognitive abilities or their knowledge of school subjects, learned through another language. Researchers have recently made important progress in providing guidelines that can help educators distinguish between disability and diversity (Paradis, Genesee, and Crago 2011), but much practica! work remains to be done so that children can make the most of their cognitive and linguistic abilities. Children who learn more than one language from earliest childhood are referred to as 'simultaneous bilinguals', whereas those who learn another lan guage later may be called 'sequential bilinguals'. We sometimes hear people express the opinion that it is too difficult for children to cope with two lan guages. They fear that the children will be confused or will not learn either language well. However, there is little support for the myth that learning more than one language in early childhood is a problem for children who have adequate opportunities to use each one. There is a considerable body of research on children's ability to learn more than one language in their earliest years. Although sorne studies show minor early delays in one or both languages for simultaneous bilinguals, there is no evidence that learning two  Language learning in early childhood 31 languages substantially slows down their linguistic development or interferes with cognitive development. Indeed, many children attain high levels of proficiency in both languages. Ellen Bialystok (2001) and other cognitive and developmental psychologists have found convincing evidence that achieving bilingual proficiency can have positive effects on abilities that are related to academic success, such as metalinguistic awareness. Limitations that may be observed in the language of bilingual individuals are more likely to be related to the circumstances in which each language is learned than to any limitation in the human capacity to learn more than one language. For example, if one language is heard much more often than the other or is more highly valued in the community, that language may eventually be used better than, or in preference to, the other. One aspect of bilingual language use is referred to as code switching-the use of words or phrases from more than one language within a conversation. For example, a child who speaks both French and English might say, 'I'm playing with le chateau'. Such switching between languages may sometimes reflect the absence of a particular vocabulary word or expression, but it can just as often be the intentional use of a word from the other language for a variety of interactional purposes. Highly proficient adult bilinguals also code switch when they speak to others who also know both languages. The use of both languages within a bilingual context is not evidence of a lack of proficiency.lt may have many different motivations, from expressing solidarity to making a joke. Psychologists have shown that speakers of more than one language are constantly making choices about how to express themselves and that code switching is patterned and often predictable. lndeed, this experience in making choices has been identified as contributing to cognitive flexibility throughout life (Bialystok 2009). As children learn a second language at school, they need to learn both the variety oflanguage that children use among themselves (and in informal set tings with familiar adults) and the variety that is used in academic settings. In his early research on childhood bilingualism, Jim Cummins called these two varieties BICS (basic interpersonal communication skills) and CALP (cognitive academic language proficiency). Characteristics of the two varie ties overlap to a certain extent, but there are important differences, not just in the range of vocabulary that each requires but also in the way information is expressed. Mary Schleppegrell (2004) and others have sought to discover what exactly it is that characterizes these varieties oflanguage and the interac tion patterns that tend to go with them, and sorne aspects of the distinction remain controversia!. l t is widely agreed, however, that the language needed for academic discourse is more difficult for children to acquire than the infor mal language of day-to-day interaction (Cummins 2000). 32 Language learning in earlychildhood Children entering school with little or no knowledge of the language spoken there may acquire BICS within a relatively short ti me - a s little as a year or two. They learn from watching and imitating interactions among their peers and between teachers and students. They make connections between fre quently heard words and phrases and the routines and recurring events of the dassroom, cafeteria, and playground. For this reason, students are sometimes perceived as 'fluent' in their second language. This can lead teachers to assume that any difficulties in academic tasks are not due to limited language skills but to other causes-from lack of motivation to learning disabilities. More careful observation shows that the students, while fluent in social settings, do not have the CALP skills needed for academic tasks such as understanding a problem in mathematics, defining a word, or writing a science report. Virginia Collier (1989) found that, for most students, acquiring age-appro priate CALP takes several years. As the second language learner tries to catch up, the children who carne to school already speaking the school language are continuing to learn hundreds of new words every year and to learn the concepts that these words represent. If second language learners have limited knowledge of the school language and do not have opportunities to continue learning academic content in a language they already know, it is not surpris ing that they fall behind in learning the academic subject matter that their peers have continued to develop. Children need time to develop their second language skills. Many people assume that this means that the best approach is to start learning as early as possible and to avoid the use of the child's previously learned languages. Certainly, it is important for children to begin learning and using the school language as early as possibl