UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The early lexical acquisition of a child with autism spectrum disorder Gibson, Deborah 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2011_fall_gibson_deborah.pdf [ 13.65MB ]
Metadata
JSON: 24-1.0078370.json
JSON-LD: 24-1.0078370-ld.json
RDF/XML (Pretty): 24-1.0078370-rdf.xml
RDF/JSON: 24-1.0078370-rdf.json
Turtle: 24-1.0078370-turtle.txt
N-Triples: 24-1.0078370-rdf-ntriples.txt
Original Record: 24-1.0078370-source.json
Full Text
24-1.0078370-fulltext.txt
Citation
24-1.0078370.ris

Full Text

THE EARLY LEXICAL ACQUISITION OF A CHILD WITH AUTISM SPECTRUM DISORDER by DEBORAH GIBSON  A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Language Education)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  June, 2011  © Deborah Gibson, 2011  Abstract The case study examines two questions on the lexical acquisition of a language-delayed child with Autism Spectrum Disorder. The language development of the child, Graeme, from his birth to the age of three years, one month, is primarily sourced from naturalistic data collected in a daily diary by the mother/researcher. The first research question concerns the relationship between the child’s social/cognitive development and the major transitions in his earliest acquisition of comprehension and production. The major transitions in Graeme’s early word learning were the onset of comprehension, the onset of production, the onset and growth of intentional communication, the production of his initial idiosyncratic word/signs, the word spurt in comprehension, the semantic re-organisations during that period, and, at the end point, learning words spontaneously in comprehension. The milestones in Graeme’s social/cognitive development that were found to correlate with the transitions in his lexical acquisition were the naming insight, symbolic representation, exhaustive categorisation, and the joint attention abilities of gaze and point following, pointing, intersubjectivity, and initiating joint attention. These joint attention skills provided Graeme, a nonverbal child, with a means to ask for the names of things (his name question). The second research question examines the current definitions of the earliest words in production. A case is made for an extended definition of early words to include the acquisition of the earliest meaningful idiosyncratic productions of vocalisations and gestures of children with the language delay characteristic of autism.  	
    ii  Preface The renewal number for the Behavioural Research Ethics Board certificate of “approvalminimal risk” is H08-00542-A003. The Annual Renewal for Study for research involving human subjects was approved on December 12, 2010.  iii  Table of Contents Abstract .........................................................................................................ii Preface ..........................................................................................................iii Table of Contents......................................................................................... iv List of Tables ................................................................................................. x List of Figures .............................................................................................. xi List of Abbreviations ..................................................................................xii Acknowledgements ....................................................................................xiii Dedication................................................................................................... xiv Chapter 1: Introduction to the Study ......................................................... 1 The scope and purpose of the study and the research questions ................................. 1 Theoretical issues........................................................................................................ 2 Rationale for the study ................................................................................................ 3 Outline of the chapters ................................................................................................ 4  Chapter 2: Research and Theory on Early Lexical Acquisition .............. 6 Introduction................................................................................................................. 6 Theories of word learning ............................................................................................ 8 Introduction................................................................................................................. 8 Social-interactionist theories....................................................................................... 8 Cognitivist theories ..................................................................................................... 9 Associationist theories .............................................................................................. 10 Hybrid theories.......................................................................................................... 11 Developmental theories............................................................................................. 11 Summary ................................................................................................................... 14 Defining early words................................................................................................... 14 Introduction............................................................................................................... 14 Form, meaning, and function criteria for identifying early words............................ 16 Not words.................................................................................................................. 19 Lexemes .................................................................................................................... 20 Summary ................................................................................................................... 21 Prerequisites and conditions for learning words ..................................................... 21 Introduction............................................................................................................... 21 Social interaction and shared attention ..................................................................... 22 The child’s motivation to learn and caregiver input ................................................. 24 Perceptual cues.......................................................................................................... 25 Categorisation ........................................................................................................... 25 A priori constraints on word meaning....................................................................... 26 Perceptual salience and motor maturation ................................................................ 28 Summary ................................................................................................................... 28 iv  Acquiring the lexicon in comprehension .................................................................. 29 Introduction............................................................................................................... 29 Speech perception ..................................................................................................... 30 Research on acquiring lexical comprehension.......................................................... 31 Rates and timing of comprehension acquisition ....................................................... 34 The word spurt in comprehension............................................................................. 35 Summary ................................................................................................................... 36 Acquiring the early lexicon in production ................................................................ 37 Introduction............................................................................................................... 37 Mechanisms for learning early words in production ................................................ 37 Intentional communication ....................................................................................... 38 Gesture and sign........................................................................................................ 39 Babble ....................................................................................................................... 41 The phonetic form of early words............................................................................. 43 Onomatopoeic words ................................................................................................ 45 Idiosyncratic words ................................................................................................... 46 Context bound words ................................................................................................ 47 Acquiring meaning.................................................................................................... 48 Early semantic classes............................................................................................... 50 Summary ................................................................................................................... 52 The transition to reference......................................................................................... 53 Introduction............................................................................................................... 53 Symbolic representation............................................................................................ 53 Referential words ...................................................................................................... 54 Exhaustive categorisation ......................................................................................... 55 Social causes of referential development.................................................................. 56 Summary ................................................................................................................... 57 Ages, rates, and the entry point of acquisition ......................................................... 57 Determining the point of entry into the lexicon ........................................................ 57 Age, timing, and rate in the production lexicon........................................................ 58 Summary ................................................................................................................... 60 Chapter summary ....................................................................................................... 60  Chapter 3: Research on Lexical Acquisition in Autistic Spectrum Disorders...................................................................................................... 67 Introduction............................................................................................................... 67 Defining autism spectrum disorders ......................................................................... 67 Characteristics of language development in ASD.................................................... 68 Introduction............................................................................................................... 68 Heterogeneous language abilities in ASD ................................................................ 68 Early atypical language development within ASD ................................................... 70 Social communication in ASD ................................................................................... 71 Introduction............................................................................................................... 71 Joint attention impairments....................................................................................... 71 The effect of attentional impairments on fast mapping ............................................ 74 Gestural intentional communication ......................................................................... 75 Vocalised intentional communication....................................................................... 77 v  Social-communicative acts ....................................................................................... 78 Comprehension in ASD .............................................................................................. 79 Introduction............................................................................................................... 79 Comprehension delays .............................................................................................. 79 Predicted outcomes of comprehension delay............................................................ 80 Comparisons with comprehension in TD children ................................................... 82 Production in ASD ...................................................................................................... 83 Introduction............................................................................................................... 83 Speech perception and processing ............................................................................ 83 Phonological development........................................................................................ 83 The lexicon in production ......................................................................................... 84 Semantic development .............................................................................................. 86 Rates and variability of lexical acquisition in production......................................... 87 Predictors of language outcomes .............................................................................. 88 Introduction............................................................................................................... 88 Cognitive ability........................................................................................................ 89 Imitation .................................................................................................................... 90 Play skills .................................................................................................................. 91 Social-affective behaviours....................................................................................... 93 Environmental variables: input and therapeutic intervention ................................... 94 Theories accounting for language outcomes............................................................. 95 Summary...................................................................................................................... 97  Chapter 4: Diary Collection Method and Data Analysis ...................... 100 Part 1: Diary and case study research methods..................................................... 100 Introduction to diary and case study methodology ................................................. 100 Case study methods................................................................................................. 100 Diary study method................................................................................................. 101 Some limitations of diary methods ......................................................................... 103 The design of the diary for this study ..................................................................... 103 Diary updating: modern technologies ..................................................................... 105 The history of the diary on Graeme’s language acquisition ................................... 106 Being a parent observer........................................................................................... 107 Data collection in the diary ...................................................................................... 108 Concentration on comprehension data .................................................................... 108 Scope of the diary ................................................................................................... 109 Using diary data for research in Graeme’s lexical acquisition ............................... 114 Summary ................................................................................................................. 116 Part 2: Methods of Data Analysis............................................................................ 117 Introduction............................................................................................................. 117 Assembling the data from primary sources............................................................. 118 Developing the analytical tools............................................................................... 121 The Word Event list.................................................................................................. 121 Introduction............................................................................................................. 121 Reliability of observation and recording................................................................. 122 The criteria for lexical acquisition status in comprehension................................... 123 The criteria for lexical acquisition status in production.......................................... 125 vi  Age dates and the age of acquisition....................................................................... 127 The design and contents of the Word Event list ..................................................... 128 Word Event list organisation................................................................................... 132 Daily chronological organisation ............................................................................ 133 Grouped data organisation ...................................................................................... 135 Additional organisations of the WE data ............................................................... 139 Short versions of the WE list .................................................................................. 139 Organisation of the WE production data................................................................. 140 The semantic categories organisation ..................................................................... 141 The babble data organisation .................................................................................. 142 Summary ................................................................................................................. 143 The Social/Cognitive Development list.................................................................... 144 Introduction............................................................................................................. 144 The organisation of the SCD list............................................................................. 145 Emerging and continuing behaviours ..................................................................... 145 The variables in the SCD taxonomy ....................................................................... 147 Summary.................................................................................................................... 149  Chapter 5: Graeme’s Lexical Acquisition from Birth to 02.09.28 during the Period of Slow Word Learning............................................. 153 Introduction to the findings presented in Chapters 5 and 6 .................................... 153 Part 1: The prelinguistic period from birth to 02.06 ............................................. 154 Introduction............................................................................................................. 154 Social/cognitive growth in the prelinguistic period ................................................ 155 Intentional communication in the prelinguistic period ........................................... 161 Language development before 02.02.00 ................................................................. 163 Play behaviours in the prelinguistic period............................................................. 164 Assessments ............................................................................................................ 166 Diagnosis................................................................................................................. 167 Summary ................................................................................................................. 168 Part 2: The onset of Graeme’s lexical acquisition.................................................. 169 Introduction............................................................................................................. 169 The social/cognitive developmental spurt at the onset of word learning................ 169 Graeme’s comprehension lexicon at the onset of word learning ............................ 172 The earliest word/signs in Graeme’s productive lexicon........................................ 174 Summary ................................................................................................................. 177 Part 3: The period of slow lexical acquisition from 02.06.00 to 02.09.28............. 178 Introduction............................................................................................................. 178 Graeme’s comprehension in the period of slow word learning ............................ 179 Introduction............................................................................................................. 179 Context bound early lexical comprehension acquisition ........................................ 179 Rates of word learning per month........................................................................... 180 Comprehension of relational and nominal words ................................................... 181 Comprehension of directives .................................................................................. 182 Comprehension of proper names ............................................................................ 183 Language input to Graeme ...................................................................................... 184 Comprehension of phrases ...................................................................................... 185 vii  Partially acquired words in comprehension ............................................................ 186 Comprehension of early semantic categories.......................................................... 186 Semantic extension in comprehension .................................................................... 187 Summary ................................................................................................................. 189 The onset of production............................................................................................ 190 Introduction............................................................................................................. 190 Babbling during the period of slow word learning ................................................. 190 Intentional communication in the period of slow word learning ............................ 191 The extension of word/signs in production............................................................. 195 The acquisition of new word/signs in production................................................... 196 Summary ................................................................................................................. 200 Social/cognitive growth and play behaviour 02.06 to 02.09.28 ............................. 201 Introduction............................................................................................................. 201 Activities in solitary play ........................................................................................ 201 Imitation .................................................................................................................. 204 Manipulating objects............................................................................................... 205 Categorisation ......................................................................................................... 206 Music....................................................................................................................... 207 Symbolic play and representational ability ............................................................. 208 Social abilities in the period of slow word learning................................................ 210 Physical Abilities .................................................................................................... 212 Summary ................................................................................................................. 215 Chapter Summary .................................................................................................... 215  Chapter 6: Graeme’s Lexical acquisition from 02.10.00 to 03.00.28 during the Word Spurt in Comprehension ............................................ 218 Introduction............................................................................................................... 218 Part 1: The word spurt in comprehension.............................................................. 218 Quantitative changes in acquisition in the comprehension word spurt................... 219 The naming insight, attentional engagement, and the name question sign............. 220 Representational ability and picture books ............................................................. 222 The name question vocalisation .............................................................................. 227 Input, MLU, and verbal processing in the comprehension word spurt................... 231 Lexical misunderstandings during the word spurt in comprehension..................... 232 Comparative rates of acquisition in comprehension ............................................... 233 Social and cognitive growth from 02.10.00 to 03.00.28......................................... 234 Summary ................................................................................................................. 239 Part 2: Semantic development in the word spurt in comprehension ................... 241 Semantic categories in comprehension ................................................................... 241 Context dependence in the word spurt in comprehension ...................................... 242 Anomalies in Graeme’s comprehension lexicon .................................................... 245 Semantic extensions in word categories ................................................................. 245 Graeme’s strategies for determining semantic extensions and intensions in comprehension ........................................................................................................ 246 Lexical classes in the word spurt in comprehension............................................... 248 Words for animals and their sounds in comprehension .......................................... 250 Synonyms in comprehension .................................................................................. 251 viii  Colour names .......................................................................................................... 252 Summary ................................................................................................................. 256 Part 3: Graeme’s production during the comprehension word spurt ................. 259 Introduction............................................................................................................. 259 Babbling between 02.10.00-3.01.00 ....................................................................... 259 Intentional communication during 02.10.00 to 03.00.28 ........................................ 262 Word/signs acquired from 02.10.00 to 03.00.28..................................................... 264 Production of animal sounds................................................................................... 266 Phonological contrasts ............................................................................................ 268 Assessments ............................................................................................................ 270 Summary ................................................................................................................. 271  Chapter 7: Conclusions and Implications for Further Research........ 274 Dissertation questions............................................................................................... 274 Introduction............................................................................................................. 274 First research question ............................................................................................ 274 The transition to lexical onset ................................................................................. 275 The acquisition of comprehension and production during the slow learning period ...................................................................................................................... 280 The acquisition of the name question ..................................................................... 283 The semantic reorganisation and the acquisition of categorical words in comprehension ........................................................................................................ 285 Graeme’s lexical acquisition compared with that of TD children .......................... 289 Models of language acquisition supported by the findings..................................... 290 Second research question ........................................................................................ 291 The rationale for extending the definition of early words ...................................... 291 The rationale for including Graeme’s early word/signs in his productive lexicon..................................................................................................................... 294 Extending the definition for word as it applies to atypical lexical acquisition ....... 296 Other Findings......................................................................................................... 298 Limitations of the study .......................................................................................... 299 Implications of the findings for further research .................................................... 300  Afterword .................................................................................................. 302 Afterward! ................................................................................................................. 302  Appendix A Original Data Sources ...................................................... 367 Appendix B Taxonomies........................................................................ 387 Appendix C Documentation for Research Ethics................................ 412  ix  List of Tables Table 2.1 Table 4.1 Table 4.2 Table 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 4.9 Table 4.10 Table 4.11 Table 4.12 Table 4.13 Table 4.14 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 6.7 Table 6.8 Table 6.9 Table 6.10 Table 6.11 Table 6.12 Table 6.13 Table 7.1 Table A.1 Table B.1 Table B.2 Table B.3 Table B.4  Semantic classes and the most frequent first words ...................................... 52 Examples from the first four columns of the WE list .................................. 129 Age dates for the word event round/round about........................................ 130 Examples of production columns in the WE list ......................................... 131 Examples of context, age dates, and type from the WE list ........................ 132 Daily chronological word event entries in comprehension.......................... 134 Daily chronological word event entries in comprehension and production 134 Chronological sorting during rapid acquisition in comprehension.............. 135 Examples of grouped word events in comprehension ................................. 136 The word event knock knock........................................................................ 138 Short version with CDI comprehension percentages................................. 140 The short production data list .................................................................... 141 Babble data ................................................................................................ 142 Entries for emerging behaviours and on-going activities at 02.07.08 ....... 146 The SCD list taxonomy.............................................................................. 147 Types of crying (journal entry 02.06.09) ..................................................... 162 Meaningful utterances and gestures acquired at the onset of word learning 175 Graeme’s new comprehension acquisitions by month ................................ 180 Words partially acquired.............................................................................. 186 Contents of semantic categories in early comprehension ............................ 187 Word/signs in the period of slow word learning in comprehension ............ 196 The number of words per month Graeme acquired in comprehension........ 220 Words confused in comprehension during 02.11 and 03.00........................ 232 Graeme and Benedict’s subjects rate of comprehension acquisition........... 234 Semantic categories added in word spurt in comprehension. ...................... 241 Context of word learning in comprehension in 02.10 ................................. 243 Anomalies in Graeme’s comprehension vocabulary ................................... 245 Extensions of the semantic category of early words in comprehension ...... 246 Animals and/or their sounds in comprehension........................................... 251 Colour word comprehension during 03.00 .................................................. 256 Examples of semantic development .......................................................... 257 Babble between 02.10 and 03.01 ............................................................... 261 Acquisition of word/signs 02.10.00 – 03.00.28 ......................................... 264 Animal sounds in production ..................................................................... 267 Word status criteria of aspects of early words ............................................. 293 Types of data samples ................................................................................. 367 Word Event List: Comprehension............................................................... 388 Word Event List: Chronological Sort ......................................................... 398 SCD taxonomy for the prelinguistic period ................................................ 408 The SCD list taxonomy (Table 4.14 reprinted)........................................... 410  x  List of Figures Figure 4.1 The original handwritten journal (02.10.22) .............................................. 112 Figure 4.2 Using joint attention to prompt naming pictures (~03.00) ......................... 114 Figure 5.1 Playing his piano (01.04)............................................................................ 156 Figure 5.2 Helping F (01.06) ...................................................................................... 157 Figure 5.3 Stacking blocks (01.04) .............................................................................. 159 Figure 5.4 Lining up animal toys (03.01) .................................................................... 160 Figure 5.5 Stacking blocks at CHDC (02.05) .............................................................. 161 Figure 5.6 Working with G .......................................................................................... 167 Figure 5.7 Close listening to the speaker ..................................................................... 171 Figure 5.8 Happy pose (02.07.24) ................................................................................ 193 Figure 5.9 An attraction to all that spins (02.09) ......................................................... 198 Figure 5.10 Distracted from metal by being addressed (02.07.17)................................ 202 Figure 5.11 Temptation of garden string fence (02.07.16) ........................................... 204 Figure 5.12 Joint attention and pointing with Grandma (03.03).................................. 210 Figure 5.13 Immature run ............................................................................................. 213 Figure 6.1 Joint attention to picture books with R (~03.00) ....................................... 223 Figure 6.2 Engrossed in picture books (03.04) ............................................................ 224 Figure 6.3 Fell asleep over books (03.04).................................................................... 225 Figure 6.4 The affect of the name question on the rate of acquisition......................... 229 Figure 6.5 Tantruming over problems with the crayon lineup .................................... 237 Figure 6.6 Relaxing with a bottle and a bell rattle (02.11) .......................................... 237 Figure 6.7 Socially engaged (02.11) ............................................................................ 238 Figure 6.8 Crayon lineup, all points forward ............................................................... 255  xi  List of Abbreviations AoA  Age of Acquisition  ASD  Autism Syndrome Disorder  ASL  American Sign Language  CDI  Communicative Developmental Inventory  CHDC  Children’s Hospital Diagnostic Centre  CV  Consonant Vowel  F  Graeme’s Father  IDP  Infant Development Programme  IJA  Initiating Joint Attention  IPA  International Phonetic Alphabet  IQ  Intelligence Quotient  MLU  Mean Length of Utterance  SCD  Social/Cognitive Development (list)  SVO  Subject Verb Object  TD  Typically Developing  WE  Word Event (list) xii  Acknowledgements I gratefully acknowledgement all of the unfailing support and patient encouragement I have received from my dissertation advisor, Dr. Kenneth Reeder. His brilliant articulation of ideas and diplomatic guidance were extraordinarily valuable. My grateful thanks also to the members of my committee, Dr. Pat Mirenda and Dr. May Bernhardt, for their many contributions in ideas, content, organisation, and style, and for their commitment to research in Autism Spectrum Disorders and in language delay. My deep gratitude for the loving support of my husband (F for Father), in his days as caregiver to Graeme at the time of the diary and throughout the writing of this dissertation. My children, Graeme and Claire, deserve very special thanks, for having so long endured absences of attention on the part of a mother writing a dissertation. Grateful thanks also to: Dr. Margaret Early, my dear friend, who knows my deficiencies as a self-starter and propelled me into enrolling as a doctoral student for this project. Dr. David Ingram, my longtime friend, for first having identified Graeme's language delay, for his crucial and extraordinarily generous design and implementation of Graeme’s early language therapy, for his guidance in starting the diary, and for his continued encouragement of my dissertation. Dr. Patricia Schwartz, for her wisdom and her generosity. The many therapists, preschool teachers, and medical professionals who supported us, and who helped Graeme to acquire his words. Special thanks to the Infant Development Programme in Vancouver, the Children’s Hospital Diagnostic Centre in Vancouver, and Berwick Preschool at the University of British Columbia. The University of British Columbia, for the Graduate Student Fellowship.  xiii  Dedication  To my son, Graeme.  xiv  Chapter 1: Introduction to the Study The scope and purpose of the study and the research questions This dissertation is an examination of the early lexical acquisition of a toddler with Autism Spectrum Disorder (ASD), my son Graeme, whose word learning was delayed in both comprehension and production. It is based on a case study of his development from birth over a three-year period, primarily sourced from a diary collection of naturalistic data. There are two research questions addressed in this dissertation. 1. What is the relationship between the delayed and unusual early lexical acquisition, in both comprehension and production, of my son Graeme with ASD, and the major developmental milestones in his cognitive and social growth? 2. Can the meaning and function of his early idiosyncratic productions justify the inclusion of these in his lexicon as word/signs and, if so, how does this inclusion explicate the definition of what comprises a word in the lexical acquisition of children with ASD? In this study, a qualitative analysis of the linguistic and social-cognitive data is designed to reveal the temporal connections between Graeme’s developmental streams in language, cognition, and socialisation. A linguistic analysis of the phonological, pragmatic, and semantic aspects of his word learning in its initial phase covers three developmental periods. The first is the prelinguistic period from birth to the onset of comprehension at 02.06.001, the second is the period of slow word learning in both comprehension and production from 02.06.00 to 02.09.28, and the third is the period of the word spurt in comprehension. The analysis identifies the transitions that signalled reorganisations in Graeme's word learning. These include: the onset of comprehension, the re-emergence of babble, the onset and growth of intentional communication, the production of the early word/signs, and the word spurt in comprehension. Other structural 1  The shorthand format for ages is given as 02.06.13, meaning two years, six months, thirteen days of age. A variation is the format 02.02, meaning two years, two months of age.  1  changes identified by the analysis are two semantic reorganisations. The first is in extensions of meaning during the period of rapid word learning in the comprehension lexicon and the other is in lexical groupings of both the comprehension and production lexicons2. The analysis of Graeme’s emerging and continuing behaviour reveals his social/cognitive developments such as categorisation, symbolic representation, the naming insight, and joint attention abilities. Joint attention (attention to a shared referent) is a pivotal development in lexical acquisition in children with ASD. Descriptions of Graeme’s social/cognitive developments and his autistic characterististics are based on data drawn from the context of Graeme’s daily life. His lexical acquisition is examined in relationhip to his social/cognitive growth. Like that of typically developing (TD) children, Graeme’s comprehension evolved from slow, context bound, case-by-case word learning to spontaneous, efficient, and rapid acquisition. Graeme’s full and partial acquisition of form and meaning in the unusual early words of his small productive lexicon occurred within the continuum of his production of babble, intentional vocalisations and gestures, onomatopoeic and idiosyncratic words, and context bound word-object associations.  Theoretical issues The issue of what constitutes a word in a child’s early vocabulary is central to the discussion of lexical acquisition. My second research question addresses the definition of a word in child language acquisition research, and the criteria for the point of entry into the early lexicon in production. The various interpretations in previous research of when an infant speech production gains the status of being a word, and some of the theoretical questions that these interpretations raise, both in theory and in the practise of data-based studies of vocabulary acquisition, are investigated in the light of my findings. The  2  The two data sets of Graeme’s comprehension and production comprise his lexicons.  2  application of these interpretations to a child with ASD who may not acquire every criterion for word status is explored. Other theoretical issues discussed in this study are the differences, if any, between Graeme’s lexical acquisition and that of TD children.  Rationale for the study Although a primary diagnostic feature of ASD is the delayed acquisition of lexical comprehension and production (Lord, Risi, & Pickles, 2004), there have been few studies of early word learning of children with ASD (Bailey, 2006), few longitudinal studies of language acquisition of verbal children with ASD, and no detailed daily diary studies of lexical acquisition with concentrations on comprehension and on the earliest impressions of lexical emergence and early pre-speech intentional communication productions for this population. Tager-Flusberg writes, Little is known about the development of language in autism. There have been few longitudinal studies conducted during the critical period when children with autism are in the process of acquiring language .… Such longitudinal studies can encompass both an individual difference approach and a comparative approach to explore variation in the development of language among children with autism (including when children begin to speak, growth patterns within and across different components of language, developmental rates and end points) …. These kinds of studies are crucial for a more complete understanding of language and communication in autism that will inform basic research on underlying etiology and will ultimately lead to the development of new treatments that can improve the lives of children with autism, all of whom have significant deficits in this domain (Tager-Flusberg, Paul, & Lord, 2005 p.79). In addition, there have been few longitudinal studies using natural language data at this level of detail on the processes of lexical acquisition in the earliest word learning in comprehension and production of TD children.  3  Outline of the chapters The dissertation is organised into seven chapters and an afterword. The first chapter is this introduction to the research questions and the scope and purpose of the study. Chapter 2 reviews literature on early lexical acquisition research in TD children, including the prerequisites and conditions for the onset of word learning, and aspects of the acquisition of the lexicons in comprehension and production, such as the rates of learning. It also discusses theories and models of word learning and investigates definitions for early words. Chapter 3 reviews literature on early lexical acquisition research on children with ASD that examines characteristics of language development and impairments in this population, along with predictors of language outcomes. Chapter 4 contains a description of case study and diary study methodology, details the design and the methods of data collection in the present study, and lists other sources of data. This chapter defines the categories of analysis for Graeme’s linguistic and socialcognitive developmental streams, the criteria used for Graeme’s lexical acquisition in comprehension and production, and the design and rationale for the two taxonomies that are the analytical tools. In Chapters 5 and 6 the data are presented and interpreted with reference to the relevant literature. The chapters are organised chronologically, and include journal entries from the diary. Chapter 5 examines the correlation of Graeme’s linguistic growth, social behaviour, and cognitive abilities in three periods: the prelinguistic period from birth to 02.06, the period of slow word learning in comprehension until 02.09.28, and the onset of early word production in this period. Chapter 6 examines the correlation of Graeme’s lexical acquisition with his social-cognitive development during the period of the word spurt in comprehension. The discussion in Chapter 6 covers the pragmatic functions of his intentional communication, semantic reorganisations in comprehension during rapid word learning, the relationship of new areas of lexical knowledge in comprehension to symbolic representation, and the additions to Graeme’s productive vocabulary, including the phonological contrasts he employed.  4  Chapter 7 addresses the conclusions that can be drawn from the findings with regard to the two research questions, examines the limitations of the study, and summarises implications for current theories and for further research. The afterword is written and titled (Afterward!) by Graeme himself, and has not been edited. It demonstrates his current language ability while describing his involvement in this project and his current life. There are three Appendices. Appendix A contains the complete entries from the daily diary for the excerpts used in this study, a tape transcript, and an assessment from Graeme’s preschool. Appendix B contains samples of data from different periods in Graeme’s word learning from the taxonomies of the Word Event list and the Social/Cognitive Development list, as well as a list of Graeme’s complete lexicon in comprehension, in the order of the age of acquisition (AoA). Appendix C contains the letter of consent given to Graeme for his participation in this study. With a data set from a child with language delay and ASD, a focus on comprehension acquisition, and a fine-grained detailed set of the earliest lexical acquisitions analysed at varying levels of specifity with which to answer the research questions, this study hopes to contribute to the research on the lexical acquisition of children with ASD.  5  Chapter 2: Research and Theory on Early Lexical Acquisition Introduction Research and theory on early lexical acquisition reflects the complexity of both child language development and of the nature of words themselves. The various threads of social, conceptual, perceptual, cognitive, phonological, semantic, and communicative knowledge emerge and are woven together in word learning. This chapter will examine the major empirical research and theory on the onset of lexical acquisition in the comprehension and production modalities of TD infants and toddlers. The focus will be on research in the areas that apply to the two research questions in this dissertation; one, the relationship between early lexical acquisition and social/cognitive development, and two, the definition of what an early word is. Theoretical research in these areas attempts to explain the circumstances and processes in which infants first learn to perceive speech, perceive adult intentions, communicate their own intentions, establish joint attention, and attach meaning to adult words. Theoretical and empirical research also offers explanations for how infants achieve cognitive milestones, develop semantic categories, transition into rapid lexical comprehension, produce their first, idiosyncratic, non-systematic words and gestures, and acquire symbolic, referential words. The areas of research and theory covered in this chapter are the following: theories of word learning; defining early words; prerequisites for learning words; acquiring the comprehension lexicon; acquiring the early production lexicon; the transition to reference; and age, rate, and the entry point of acquisition. The first section looks at contemporary theories of lexical acquisition that provide models of word learning: the intentional models of the social-interactionists, attentional associationist models, the cognitivist, hybrid, and developmental models. The section on the prerequisites for learning early words discusses proposals put forward by various theoretical viewpoints: social interaction, categorisation, a priori constraints on word learning; perceptual 6  salience and motor maturation; and the child’s contents of mind, and caregiver’s input. Research on the acquisition of the comprehension lexicon includes an overview of major data-based studies, research on speech perception, the comprehension word spurt, and the rate of comprehension acquisition. The section on the acquisition of the early productive lexicon consists of studies on acquiring intentional communication, gesture and sign, babble, the phonetic form of early words, onomatopoeic words, context bound words, acquiring meaning, and early semantic categories. The section on the transition to reference examines research on acquiring referential words, exhaustive categorisation (spontaneously sorting objects into groups), and symbolic representation. The final section looks at age, timing, and rate of production acquisition, and determining the entry point of acquisition into the lexicon. Research in the lexical acquisition of TD children in both the comprehension and productive lexicons informs my first research question: What is the relationship between early lexical acquisition in both comprehension and production and the major developmental milestones in cognitive and social development? Definitions and terminology for the early meaningful productions of TD children are many and various, depending to some extent on the theoretical point of view. In this section, the criteria for word status of the earliest productions are considered in terms of form, meaning, and function. This involves research on phonetic form, variations in word learning, pragmatic functions of intentional communication and gestures, and the acquisition of meaning from context bound to symbolic words. Gestured and/or vocalised early forms with context dependent meanings traverse the fuzzy border of what defines a word in child language, whereas the transition to reference has an agreed-upon definition. Examining the criteria for word status applies to my second research question: Can the meaning and function of the initial idiosyncratic productions justify the inclusion of these in the early lexicon as word/signs and how does this inclusion explicate the definition of what comprises a word in the lexical acquisition of children with autism?  7  Theories of word learning Introduction In the field of study of early word acquisition, two contemporary theoretical issues emerge. One debate is over the primacy of social or cognitive factors in the ontogeny of word learning and in the development of the comprehension and production lexicons: which factor best prepares prelinguistic TD infants in their first year to identify words, and propels toddlers in their second year through the slow beginnings of acquisition to the transition to reference? The second debate concerns which explanation best accounts for the continuum of development in lexical acquisition. The relative importance of social and cognitive developments and the developmental aspects of word learning are differently weighted according to the theoretical views of researchers in the field. These views can be summarised as social-interactionist, associationist, cognitivist, hybrid, and developmental. While theories of word learning are mostly directed towards the acquisition of referential words in production, there are implications for comprehension and for acquiring early pre-referential words. The following sections look briefly at these theoretical models of lexical acquisition. Key theoretical claims will be discussed further in the sections on prerequisites for learning early words and acquiring the comprehension and production lexicons.  Social-interactionist theories Broadly, interactionist theorists regard linguistic, maturational, cognitive, and social variables as both affecting and being affected by language acquisition. For a comparison of theories in the process of lexical acquisition, the social-interactionists will be discussed separately from the cognitivists and developmental theorists. The social-interactionist theorists focus more on social aspects of word learning. Researchers Akhtar and Tomasello (2000) argue that the prime instruments of word learning are the socialpragmatic abilities of joint attention and knowledge of communicative intentions. This 8  theoretical position contends that words are learned in the context of interpersonal communication and attentional engagement (Estigarribia & Clark, 2007). Functionalist theorists similarly claim that word learning results from the pragmatic needs of the child (Halliday, 1975). Social-interactionists and functionalists propose descriptions of the onset and the developmental nature of lexical acquisition, as determined by the child’s motivation to function and communicate, the child’s contents of mind, social and pragmatic cues, and the input the child receives in social interaction (Adamson, Bakeman, & Deckner, 2005; Adamson, Deckner, & Bakeman, 2010; L. Bloom & Tinker, 2001; Diesendruck, Markson, Akhtar, & Reudor, 2004; Halliday, 1975; Markson, Diesendruck, & Bloom, 2008). These theories are sometimes termed “intentional” as they suggest that from infancy on, children access the intentions of others in social interactions through goal-directed action in order to determine word meaning (Woodward, 2004; Woodward & Needham, 2009). From this viewpoint, the entire process of word learning can be seen as inherently social.  Cognitivist theories The cognitivist position is that lexical acquisition is in itself a cognitive process. Many researchers in psycholinguistics claim that domain-general conceptual abilities (those tailored to process general knowledge such as attention, memory, and association) are precursors to, and necessary for, the onset of lexical acquisition and for rapid word learning in comprehension and production. Cognitive abilities specific to language that are necessary for lexical acquisition are categorisation (Clark, 2003; Gopnik & Nazzi, 2003), the naming insight (Dore, Franklin, Miller, & Ramer, 1976) and symbolic representation (Lifter & Bloom, 1989; Piaget, 1962). The acquisition of these abilities causes qualitative shifts in the child’s word learning. Some proponents of cognition-based lexical acquisition propose that children have innately determined linguistic constraints that narrow the search space in order to deduce word meaning (Booth, Waxman, & Huang, 2005; Cimpian & Markman, 2005; Dromi, 1999; Kuczaj & Barrett, 1999; Merriman & Evey, 2005). Waxman (2010) proposes a word learning model that considers the diverse and flexible developmental relationship between the lexical 9  linguistic system and conceptual organisation, with its roots in generative linguistics and cognitive psychology. She argues against a perceptually and attentionally based associationist view of the links between words and concepts, proposing that conceptual knowledge permits inductive references for noun, adjective, and verb labels (Waxman & Lidz, 2006; Waxman & Gelman, 2009).  Associationist theories Associationist learning theories from the field of cognitive psychology promote the associative basis of all attentional learning; namely, that associations between cues and outcomes are learned. The attentional model rejects the induction process and hypothesis testing as irrelevant to experiential development, instead accounting for word learning through attentional mechanisms such as perceptual saliency, association, prior learning experiences, and frequency (Kruschke, 2003; Plunkett, Karmiloff-Smith, Bates, & Elman, 1997; Samuelson & Smith, 1998; Samuelson & Smith, 2005; Sloutsky, 2003; L. B. Smith, Jones, Landau, Gershkoff-Stowe, & Samuelson, 2002). As babies learn which cues to attend to and become expert at word-object pattern recognition, they become able to generalise across exemplars to deduce the meanings of words (Tan & Schafer, 2005; Yoshida & Smith, 2005). L. Smith (2000) claims that the term association most precisely describes the attention babies pay to three sets of circumstances. These are: one, relations between forms and referents, and actions and intentions; two, triadic joint attention (the ability to maintain the interaction of two-way communication in eye contact, alternating gaze, and gestures of pointing and showing); and three, syntactic cues and meaning in social exchanges that emphasise gestures and ostensive object handling. These three associations develop through accrued learning experiences into biased learning mechanisms, enabling children to construct assumptions such as the shape bias (a predisposition to categorise same-shaped objects, generalising word meanings based on shape) (Houston-Price, Plunkett, & Harris, 2005). Proponents of the associationist theory claim that it offers a unified theory of internal mental states and processes of cognition though the mechanisms of prediction and forming expectations (Regier & Carlson, 2002; Samuelson & Smith, 2005; Woodward & Hoyne, 1999). Akhtar and Tomasello (2000) 10  criticise the associationist model for its failure to account for the social aspects of communication such as reading intentions and manipulating attention.  Hybrid theories The emergentist coalition model is a hybrid theory specific to word learning. These theorists propose that both intentional and attentional cues are necessary for word learning (Golinkoff et al., 2000; Hirsh-Pasek, Golinkoff, Hennon, & Maguire, 2004). The emergentist coalition model combines aspects of both the social developmental and cognitive positions, stating that neither cognitive developments nor social understanding alone can explain the nature of lexical acquisition over time (Golinkoff & Hirsh-Pasek, 2006). Hybrid theorists Hollich, Hirsh-Pasek, and Golinkoff disagree with the associationist claim that children’s word learning abilities derive solely from their prior contingencies of word learning leading to systematic expectations about form-referent relationships (Golinkoff et al., 2000). They argue that this claim does not account for knowledge of social-pragmatic cues that enable the child to determine which cues are relevant. Nor do these claims explain how the child decides which aspect of the word meaning to extend: similarity, function, or taxonomic membership. Golinkoff, Mervis, and Hirsh-Pacek (1994) propose that children both actively attend to input and experience. They employ emerging (not innate) principles and biases to construct better strategies for lexical acquisition. This theory regards lexical principles as the product rather than the means of word learning, changing in character as the baby develops into a mature language learner. Reilly (2007) suggests developmental hybrid models of lexical acquisition are best able to incorporate the multiple levels of processing in sound and meaning involved in lexical acquisition, rather than theories which lean towards either solely cognitive or social-interactionist explanations.  Developmental theories Theorists in developmental models of word learning claim that a model must account for the gradual, continuous nature of language learning. The emergentist coalition model proponents contend that word learning is embedded in concurrent social, emotional, 11  cognitive, and motor growth. Therefore, a word learning theory must explain the changes in developmental processes and behaviour from pre-speech origins through to the qualitatively different acquisition of words during the vocabulary spurt (Akhtar, 2005; Hirsh-Pasek et al., 2004; Hollich et al., 2000; Pruden, Hirsh-Pasek, & Golinkoff, 2006). Dynamic system theories, in which early words gradually enable the development of underlying concepts through the child’s input and experiences, are supported by several researchers (Baldwin & Tomasello, 1998; Dromi, 2008; L. B. Smith, 1999; Tomasello & Akhtar, 1995). The connectionist model is a contemporary developmental model which proposes that gradual, complex, continuous maturation in neurophysiological functions, such as articulatory and phonological ability, influence changes in lexical comprehension and production processes (Clark, 1993; E. &. K. Kaplan G., 1971; Kent, de Schonen, Jusczyk, McNeilage, & Morton, 1993; Mills, Coffey-Corina, & Neville, 1997). Kent (2001) presents a model describing children’s sensory-motor neural development, which, in combination with their language exposure, underlies speech. This development adapts as vocal tract motor control, and auditory and visual perception are established. The connectionists use a quantitative approach to show that if acquisition data are modelled by the acquisition of a small number of new words rather than by age, a more gradual curve of growth is apparent, rather than the shift to a more rapid acquisition, known as a word spurt3 (David Ingram, personal communication, March 13, 2006). Ingram suggests that cognitive shifts can be gradual, as can the resulting lexical development. Connectionist theorists using the competition model view cognition as united to the body, the brain, and the social context (Bates & Benigni, 1979; MacWhinney, 2005b; MacWhinney, 2006). The functionalist model that connectionist researchers espouse is the competition model, which focuses on the word learning mechanism rather than the outcome of the learning process. Multiple, probabilistic cues of varying strength compete to interpret word meaning according to the interrelations of the underlying conceptual structures of previous learning and processing (Bates & MacWhinney, 1987; MacWhinney, 2005a; MacWhinney, 2006; MacWhinney, 2008).  3  An extended discussion of word spurts will be offered below.  12  Supporters of theories in which each new development emerges out of prior learning argue against instance-based learning and the notion of discrete language learning stages (P. Bloom, 2004). Clark (2003) regards the onset of word production as a period of continuous development in which children learn to produce increasingly intelligible utterances. Earlier researchers, however, proposed two distinct phases in the single word stage of the lexicon in production. First, there is an initial slow emergence from onset of production of meaningful communication until around 50 words, and next, a later period when symbolic words appear just before the vocabulary spurt (L. Bloom, 1973; L. Bloom, 1993; Ingram, 1989). Differing definitions of what constitutes a word influence the 50-word total, as does individual variation in children. Nelson (1988) proposes a second phase after the first 30 words, when children acquire a sign/sound for what’s that?, which corresponds to the naming insight, and a third phase between 03.00 and 04.00, in which words are combined. The continuum of lexical acquisition is shown in data-based studies. M. Robb, Bauer, and Tyler (1994) show that symbolic words overlap with children’s early word vocalisations throughout the one-word period, just as reduplicated babbling overlaps with canonical babbling prior to word use (B. L. Smith, Brown-Sweeney, & Stoel-Gammon, 1989; Stoel-Gammon & Cooper, 1984). Bates (1979) proposes that context bound performatives evolve into context-flexible referential use. Data-based studies also provide evidence for discrete stages (Fenson, Dale, Reznick, & Bates, 1994; Nelson, 1988; Rescorla, 1984). Dromi’s case study revealed her daughter’s gradual restructuring of her earliest words in production to categorical, referential words (Dromi, 1987; Dromi, 1999; Dromi, 2008). Tager-Flusberg, Rogers, Cooper, Landa, Lord, Paul, Rice, Stoel-Gammon, Wetherby, and Yoder (2009) provide the criteria for overlapping and developmental in nature language acquisition stages in children with ASD. Children with ASD will often present a mixed profile, with proficiency in one domain such as vocabulary, and weakness in others; typically, pragmatics. Although theories and models of word learning have excited the greatest attention in the field of lexical acquisition recently, researchers including Benedict, Bloom, Clark, 13  Dromi, Goldin-Meadow, Halliday, Ingram, and Nelson, to name a few, have based their theoretical views on data-based studies of naturalistic child language. Ingram states that “in order to determine word meaning in young children we need to know both the specific words children acquire, and their contexts of usage.” (Ingram, 1989).  Summary As child language acquisition is a pre-paradigmatic field (Kuhn, 1996), many theories of word learning are proposed, including the social-interactionist, the cognitivists, and developmental theories. One current theoretical debate is over the primacy of social or cognitive factors to account for the onset of lexical acquisition in comprehension as well as in production, and for the transition from the slow learning of early words to referential symbolic words. Another theoretical debate is between instance-based models of word learning based on qualitative shifts in cognition and developments in social interaction, and the developmental models, which view lexical acquisition as a gradual process in a continuum of maturing physical and cognitive skills in a language context. Research into aspects of lexical development will be discussed in regard to these theories.  Defining early words Introduction Defining what a word is, in the child’s earliest lexicon, is a key methodological issue in child language studies (Vihman & McCune, 1994), and is a focus of research in this dissertation. Central to an examination of lexical acquisition in the one-word stage is the notion of when an infant vocalisation, gesture, or intentional communication gains word status, which determines its point of entry into the acquired lexicon. Researchers have often excluded the earliest meaningful utterances and gestures at the inchoate beginning of word learning from studies on lexical acquisition. While most agree that the initial meaningful productions in a child’s spoken or signed lexicon differ qualitatively from the words acquired later in the one-word stage, what to call and how to define the earliest utterances and gestures has had many different interpretations. The beginnings of a 14  child’s lexicon are so individual, the concepts of what defines an early word so various, and the terminologies from different theories of language acquisition so numerous, that the field of language acquisition has resorted to using scare quotes on ‘word’, an indication of a provisional and contested term. ‘Word’ is the name and form used for the earliest words and intentional communications, to differentiate them from later, ‘real’, symbolic referential words. The following are some questions that research in lexical acquisition has considered with regard to including the earliest forms of meaningful communications under the rubric word. What are the various criteria for distinguishing vocalisation from word on the continuum between the onset of intentional vocalisations and the word spurt? Is the definition of a word in child language acquisition determined by form or consistent meaning, or both? If by form, how close to adult pronunciation does it have to be to be a word? Can an unconventional or onomatopoeic non-speech vocalisation be a word, even if it falls outside the speech sounds of the native language or of any human language? Can a word consist of a gesture? Or must it be a recognisable approximation of adult pronunciation, albeit subject to the motor articulation skills and emerging phonological rules of the child? Can it be accepted as a word if it is comprehensible to only the child’s intimates, or must it be understood beyond the child’s immediate circle? If being a word depends on having a regular extension (the range over which a term applies) of the word’s meaning, will a non-speech sound or gesture with consistent meaning that is understood by the child’s intimates qualify? Will a context bound speech-sound vocalisation that possesses irregular and underextended extensions of the adult meaning meet the criteria? Or must the utterance have conventional adult extensions of meaning? The following sections will examine various answers to these questions with regard to the terminology, definitions, and criteria for word status for early meaningful productions.  15  Form, meaning, and function criteria for identifying early words The first meaningful productions are generally referred to by what they are not: not real or genuine, not systematic, non-categorical, non-symbolic, non-differentiated, nonarbitrary, non-referential, and pre-lexical or non-lexical. In contrast, later words acquired in the one-word stage are positively defined: they are symbolic, referential, categorical, representational, arbitrary, and systematic. The definition and properties of symbolic referential words that children acquire on their way to using conventional words are more or less agreed upon in the field. The criteria for what constitutes the earliest prereferential words is variously defined (or sometimes undefined) by form, function, and meaning. Dromi refers to the beginning of the one-word stage as the “emergence of systematic, repeated productions of comprehensible (by the caregivers at least) phonetically consistent forms produced in expected contexts.” (Dromi, 1999). Werner (1963) used the term vocables for these productions, while Bruner called them non-systematic phonetically underspecified early forms (Bruner, 1975). These early utterances, which precede recognisable attempts at single words such as mama, are recognised by a familiar listener as linguistic units that convey meanings. Throughout the one-word stage these productions are modified in their phonetic form to become more comprehensible. M. Robb, Bauer, and Tyler (1994) followed the trajectories of six infants’ word production using a strict criterion for what counted as a real word compared to a nonword, based on 100% agreement between the researcher and caregiver. In their study, the child’s word is defined by its phonetic closeness to the adult form, taking into account children’s common phonological reduction rules. A non-word in their definition is an intelligible, phonetically transcribable vocalisation such as babbling or phonetically consistent vocalisation. The definition of words (and possibly even of non-words) in this study excludes onomatopoeic utterances with non-speech sounds, on the grounds that these are not phonetically transcribable. Marginal (non-dictionary) forms are included as words; for example uh-oh, and beep-beep. However, since these are words in an adult lexicon they pose no difficulty in their acceptance as words by any standards. An 16  example of an early word whose form is far from an adult target is mm-mm to refer to a specific object (Deuchar & Quay, 1999), which would not be accepted as a word in M. Robb et al. (1994). Vihman and McCune’s (1994) definitive publication, “When is a word a word?” specifies that early words do not depend on a close relationship with either the adult form or meaning. Their definition is a balance between an inclusive definition that accepts nonadult words but runs the risk of false positives, and a more exclusive narrow definition that restricts words to those that sound like adult words but risks missing productions that are words in the child’s communication system (McCune, 2009). In their 1994 study, determining whether a vocalisation has word status depends on three factors: 1. Context criteria: situational consistency, multiple use in multiple episodes, and parental identification. 2. Vocal shape criteria: phonetic resemblance to the adult form by matching at least two segments of the adult form, having a prosodic match, and phonetic stability of invariant phonological shape. 3. Relation to other vocalisations criteria: imitation by the child, reformulation by mother, and no inappropriate uses. This study presents an explicit procedure to recognise the earliest words through the degree and type of point-to-point phonetic match to an adult word. Vihman and McCune (1994) note that an awareness of a child’s phonetic patterns (vocal motor schemes) helps the researcher identify an utterance as a word, just as the context of use and function helps to ascertain meaning. Vihman and McCune’s (2001; 1994) definition of a word includes the predecessors to symbolic words. Non-dictionary items such as vroom, ow, whee, and yum have word status, as do onomatopoeic forms, including animal sounds; however, inappropriate uses, homonyms, favourite sound patterns, and global, loosely defined, or uninterpretable utterances in their data do not (McCune & Vihman, 2001). Their studies confer word  17  status on context dependent utterances. These are differentiated from symbolic words that are used spontaneously with multiple occurrences over a range of various contexts. In two studies, Hirsh-Pasek (1999) and Golinkoff (1999) have also proposed a useful definition for identifying a child’s first words, derived from commonly accepted criteria in different theoretical perspectives. According to their definition, children have acquired a word when it is used with the intention to communicate, has a consistent phonological shape, has a consistent meaning, and is extended to multiple exemplars. In this definition, context bound words therefore are not words, since they are underextended and do not have an underlying referential meaning (Golinkoff et al., 1994). The distinction between intentional communications and words can also be determined by function. Most researchers in the lexical acquisition of TD children have distinguished early gestured or vocalised intentional communications from symbolic words, though the latter of course also convey intentional communication. Halliday (Halliday, 1975) regarded his son Nigel’s first consistent vocalisations, which he termed protowords, as the foundation for lexical development. He hypothesised that protowords occur within a protolanguage, defined as a set of basic communicative functions derived from the internal ability of the infant to express pragmatic intentions, rather than from the adult language. Halliday regarded protowords as non-words with no syntax, only a direct mapping from sound to the meanings that Nigel had created himself, as opposed to those he had learned from his language environment. Nigel produced 29 protowords when he was 01.01.05. Bauer (1986), however, includes pragmatic functioning as an early word criterion. M. Robb et al. (1994) note that non-words in their study might have been pragmatically functioning as words because they were identified as such at later age periods. This illustrates a core problem with defining early words from testing samples of data rather than from systematic, longitudinal, naturalistic daily diary collection. That is, the dilemma associated with filtering symbolic words from the earliest words persists without a naturalistic context or the pragmatic and developmental history of the particular words, and with the additional difficulties inherent in the general unintelligible nature of young children’s vocalisations (Vihman & Miller, 1988). 18  Not words Terminology and theory used in psychology and psycholinguistics help to identify the limitations of the earliest words in terms of cognitive development. The earliest words are said to lack categorical properties, in that they do not represent an underlying concept. For example, the word car is not specific to one object; it represents a number of exemplars that make up its underlying concept. Carter (1979) employs the term sensorimotor communicative schema to describe how the infant maps its phonetics onto schemata (Piaget’s term for the infant’s cognitive structures), as opposed to having concepts and categories. Like Halliday (1975), Bates, Camaioni, and Volterra (1975) describe early words as performatives without underlying meanings because children do not have concepts at this stage. Mandler and McDonough (1993) likewise make the distinction between schematic and categorical structures of representation in cognition, describing early meaningful productions as phonetically consistent forms that display intentional meaning and have some phonetic variation in form. Nelson calls the first 30 productions of the child pre-lexical terms (Lucariello & Nelson, 1985), describing them as holistic labels only, inseparable from the child’s undifferentiated representations of reality, such as people, objects, and actions (Lucariello, 1987; Nelson, 1982; Nelson, 1988). Barrett (1989) also does not regard very early words as true communicative symbols but as “frozen linguistic strings” with the characteristics of sensori-motor action schemata; that is, an absence of stable object concepts and an inability to represent reality in symbolic terms. Early utterances, in his view, are “behaviorally conditioned mimicry”, unanalysed representations of habitual situations devoid of communicative intent. They are merely a response associated with specific external stimuli consisting of physical objects, settings, and ostensive behaviours. These early productions are elicited by a familiar context; sometimes even by a sound pattern of a similar word. Dromi (1993) terms the earliest words as situational words, which link a phonetically under-specified sound pattern (Bruner, 1975) to a specific object.  19  Lexemes Because word is such an ambiguous and difficult term to define, the field of morphology employs the term lexeme. Lexemes are the minimally distinctive unit of lexical meaning in word formation, an abstract unit that underlies the set of inflected forms taken by that word (Aronoff & Feldman, 2000; Crystal, 1987). The term lexeme, however, is not synonymous with an early word in a child’s lexicon, although there are similarities. A single lexeme may in fact contain more than one word, as in the idioms take off, and all gone. This parallels narrowly context bound expressions such as this little pig, no way, and put up with, which may each function as one early word in the child’s early lexicons in comprehension and production4. Another overlap between lexemes and early words is that a child also will use one form for word variations of a lexeme or related set of lexemes, such as using yum to mean eat, ate, eating, eaten, and eat up (Ammon & Slobin, 1979). A case can also be made that there is an even wider set of meanings for one lexeme in comprehension. The toddler might understand yum to include the meaning for feed, food, meal, dinner, good, open wide, and swallow as well. Although it may be argued that children’s earliest utterances are composed of single lexemes, lexeme theory does not account for the temporal variation in phonetic form and meaning of early words in child language. Nor do the inflectional and derivational lexeme functions apply usefully to the early lexicon. More centrally, the claim cannot be made that children at the onset of speech have an underlying abstract unit of fundamental reference for their earliest words. In this dissertation, considering the lack of agreement on the definition of an early word, and for the sake of simplicity, intentional productions at the start of children’s speech production that have a consistent meaning and form will be described as early words. Arguments to support this decision will be further elaborated in Chapter 4, in the section on the criteria for lexical acquisition status in production; in the findings Chapter 5, Part 2, in the section on the onset of Graeme’s lexicon in production; and in the discussion on the findings in Chapter 7. These arguments will propose an answer to my second research 4  Take, off, all, gone, this, little, and pig are also all single lexemes.  20  question: what is the definition of a word in the lexical acquisition of children with ASD, and can the meaning and function of Graeme’s early idiosyncratic productions justify the inclusion of these in his lexicon as word/signs?  Summary Early pre-symbolic words in the one word stage usually do not have the full adult form or meaning, nor are they understood by many outside the child’s intimate circle. Researchers’ definitions of early words agree that they must be consistent in meaning and shape. However, criteria for word status varies, including the stipulation that early words must be used in multiple situations, and resemble adult words phonetically in at least two segments though they may be underspecified phonetically (Vihman & McCune, 1994). Words that are context bound in meaning and do not extend to other exemplars are accepted by some researchers as words, and not by others. Less universally accepted as words are onomatopoeic utterances, non-dictionary child forms, homonyms, gestures by hearing children, and vocalisations that are unintelligible or made with non-speech sounds (though these may be intentional communications in that they have pragmatic functions and are interpretable). What to call these earliest of words is problematic. While many terms have been suggested, in the absence of agreed terminology the most commonly used is ‘word’, with scare quotes, which is how they are informally referred to in online discussions of the first meaningful productions of researchers’ own children. The sections below on intentional communication, gestures, the phonetic form of the earliest meaningful productions, and the acquisition of symbolic capacities will examine research on and theories about the precursors to formally designated referential, symbolic words.  Prerequisites and conditions for learning words Introduction Different theoretical models propose various prerequisites for early word learning in both comprehension and production. This section will examine these necessary prelinguistic 21  conditions for lexical acquisition. These include social, cognitive, and physical developments in social interaction, categorisation, a priori constraints on word meaning, perceptual salience, linguistic input, and articulatory maturation. These prerequisites for language readiness are of particular importance to the onset of lexical acquisition in a language-delayed child.  Social interaction and shared attention Social-interactionists emphasise the importance of prelinguistic infant/adult interactions in preparation for the beginnings of lexical acquisition (L. Bloom, 2000; Gopnik & Meltzoff, 1992; Gopnik & Meltzoff, 1997). They theorise first, that early word learning is a process that occurs in a natural social and cultural context where language is used relevant to shared activities (Nelson, 2007a; Snow, 1999); second, that increases in perceptual and conceptual knowledge, such as recognising faces and objects and identifying recurrences of prior sensual experiences, help the infant to perceive similarities and to sort like with like (Clark, 2003); and third, that infants begin to pay attention to people and objects, initially responding to the adult’s acts with both actions and with some utterance of sounds (Akhtar & Montague, 1999). Harris (1988) suggests that the non-linguistic setting of predictable, consistent contexts and the linguistic input of the caretakers’ frequent modelling enable the child to find a way into speech. The social foundations for word learning are joint attention and intersubjectivity (the child’s awareness that attention on an external entity is shared)5 (Baldwin, 1993b). Social-interactionist research on word learning has focused on attention, gaze, and gesture in order to predict rate of growth, the emergence of new linguistic abilities, and eventual outcomes. In infancy, children learn to communicate and collaborate through the social-cognitive skills of shared intentions and responsive joint attention in eye contact and gaze monitoring (Mundy, Sullivan, & Mastergeorge, 2009; Tomasello, 2007). Neural brain imaging studies show children using responsive joint attention by 00.05, and initiating joint attention prior to 00.09 (Mundy & Van Hecke, 2008; Mundy & Jarrold, 5  Research in joint attention, intentional communication, gesture and sign, and context bound words in children with ASD will be more fully examined in Chapter 3.  22  2010). Baldwin and Sabbagh (2005) suggest that babies actively pursue this social coordination in order to gain information about referential intentions. The shared attentional focus of gaze and point following enables the infant to begin to infer referential intentions, and, with the aid of parents’ frequent ostensive labelling (presenting an object or event in such a way that attention is drawn to it) of perceptually salient whole objects, to begin to identify word meaning (Baldwin, 1995; Baldwin & Tomasello, 1998; Houston-Price, Plunkett, & Duffy, 2006; Meltzoff & Brooks, 2009; Sabbagh, Henderson, & Baldwin, 2007). Dare (1993b) refers to responsive joint attention as a self-reference that reduces error. The more highly developed pre-verbal social-cognition skill of triadic joint attention (intersubjectivity) is the ability to initiate and maintain the interaction of two-way communication through the employment of the visual and auditory cues of eye contact, alternating gaze, gestures of protodeclarative pointing and showing, intonation, and emotional displays (Charman, 2003; McCathren, Yoder, & Warren, 1999; Tomasello, 1995). These abilities, also known as initiating joint attention (IJA), are precursors of language development, enabling the child to determine which object, action, or attribution is being labelled. IJA, therefore, has a central importance in word learning, producing a greater likelihood of learning a new word than responsive joint attention does (Mundy, Gwaltney, & Henderson, 2010). Another link between joint attention and word learning is in the acquisition of the social foundations of symbolic representation. Kaplan (2006) proposes that children overtly practise responsive joint attention for six to 18 months before it becomes internalised, enabling the development of symbolic representation. Symbols in the form of pictures elicit joint attention, and, conversely, joint attention, in improving the depth of processing, may improve a child’s memory for pictures (Hirotani, Stets, Striano, & Friederici, 2009).  23  The child’s motivation to learn and caregiver input Social-interactionists have investigated the relationship of adult input to children’s vocabulary acquisition. The social factors that assist children in discovering word meanings are the input the child receives, along with responsiveness to and understanding of the child (Baldwin & Meyer, 2007). Nelson (2007c) states that learning to talk is a process emerging from characteristics of the individual child, the linguistic and social context, and, most likely, also from unspoken adult expectations and the child’s opportunities for expression. Clark (2003) notes that children form hypotheses about word meanings in conversational settings. They identify and make inferences about conventional meanings from the reactions of other speakers to their own usage, in order to restrict, expand, and fine-tune the definitions and pronunciation of the new words they are mapping. In Bloom’s view (2001) the development of the child’s contents of mind (the child’s reality as an internal motivation for word learning), plus the agency and action of the child, are the central mechanisms driving lexical acquisition in the first two years. In her view, the cognitive developments of symbolic capacity and conceptual structures are products, rather than prerequisites, of word learning. Like Halliday, (1975), Bloom (Golinkoff et al., 2000) believes word learning originates within the context of the whole reality of the child, rather than from adult input (as Clark suggests) or from internal wordlearning principles. Bloom, Margulis, Tinker, and Fujita (1996) argue against the idea that it is the caregiver who directs and constrains conversation as the mechanism for word learning, proposing that children employ moment to moment thinking and intentional states when they are comprehending and producing language. Akhtar and Tomasello (Golinkoff et al., 2000) concur with Bloom, stating that since language learning takes place for the purpose of communicating, word learning occurs because of the child's increased motivation for social communication, to understand what their addressor is saying and attending to. From this point of view, children are active agents in word learning, as evidenced by their learning words in complex, non-ostensive interactions where they are not directly addressed (Akhtar & Tomasello, 1998). Evidence for the child’s contents of mind being central to word learning comes from children taking the 24  lead in vocal interactions with adults and in their play activities in turn-taking, when at least two-thirds of the time adults are responsive rather than directive in word-learning situations (L. Bloom et al., 1996). Joint attention assures that the adult’s responsiveness to the child is relevant to the child’s thinking (Harris, 1992). Bloom (1993; 1998; 2000) argues that theories solely dependent on either social context or on specific principles are both reductionist and isolated from the rest of language development. She claims that word acquisition is primarily driven by the child’s contents of mind and the development of cognition, emotion, and social connectedness.  Perceptual cues Many researchers concur that early words are learned in an attentional and associative manner that links perceptually salient objects with sounds (Akhtar, 2005; Dromi, 1999; Hirsh-Pasek et al., 2004; Nazzi & Bertoncini, 2003). In early word learning the child pays attention to the cues of temporal continuity, perceptual salience, and prosody (Hollich et al., 2000). The novice learner needs numerous overlapping and interactive perceptual cues like ostensive handling, bright colours, and frequent labelling. They use their own perspective to attend to what they find perceptually salient (Hirsh-Pasek & Golinkoff, 2008). Tan (2005) notes that the frequency of parents’ ostensive naming, accompanied by preferential looking, influences word learning. Associationists account for lexical acquisition through attentional mechanisms such as perceptual saliency, association, prior learning experiences, and frequency (Kruschke, 2003; Plunkett et al., 1997; Pruden, Hirsh-Pasek, Golinkoff, & Hennon, 2006; Samuelson & Smith, 1998; Samuelson & Smith, 2005; Sloutsky, 2003; L. B. Smith et al., 2002).  Categorisation Ingram (1989) states that there is no well-defined theory of linguistic readiness; rather, there is an assumption that the order of language acquisition in a TD child is a reflection of sequences of cognitive developments necessary for language, including categorisation. The ability to categorise objects and actions develops throughout infancy from being perceptually based at 00.04 (Quinn, Eimas, & Rosenkrantz, 1993). In Piagetian terms 25  (1962), the child realises that though two objects are distinct they can be sorted by whether they behave or look alike, or have a similar function. The ultimate development of perceptual categorisation is the child’s realisation that every object belongs to a category6. A dissenting view, however, claims object categories are primitive conceptual notions that need not be learned (Huttenlocher & Smiley, 1987). Hall (2009) proposes that the ability of young children at the onset of lexical acquisition to learn proper names as well as count nouns is evidence of their conceptual biases to perceive people as individuals and objects as instances of a category. Recent research in the relationship of words and object categories provides evidence that parents’ labelling aids infants in associating words with categories (Cohen & Brunt, 2009).  A priori constraints on word meaning A priori constraints are cognitive developments that cognitivists propose as assisting word learning after its onset. The theory is that innate, domain-specific constraints on word meaning assist the child in narrowing down the possible meanings of a word (Gelman & Greeno, 1989; Merriman & Evey, 2005; Newcombe, Huttenlocher, Drummey, & Wiley, 1998; Waxman, 1991; Woodward & Markman, 1998). Constraint theories for early word learning include the whole-object assumption and the shape bias. The whole-object assumption is a child’s predisposition towards interpreting new words as labelling objects as wholes rather than as their parts or properties (Golinkoff et al., 1994; Hall, Waxman, & Hurwitz, 1993; Imai & Gentner, 1997; Landau, Smith, & Jones, 1988; Markman, 1989). Gentner (1982) claims that this is the most perceptually apparent level of analysis for babies, providing rich conceptual information. In addition, cognitivists claim a bias towards words referring to objects rather than events or actions. Social-interactionists Akhtar and Tomasello (Golinkoff et al., 2000), however, assert that children are predisposed to actively search for cues to what is most relevant to a speaker’s referential intent rather than to whole objects.  6  The further development of exhaustive categorisation will be discussed in the section on the transition to reference.  26  Waxman (1991) proposes that children have a predisposition to categorise same-shaped objects and hence to generalise on shape. Booth (2005) documented findings that infants at 01.06 were influenced by whether a test object was described as a thing, in which case they extended their meanings on the basis of shape; or an animal, in which case they used both shape and texture. These findings show that both perceptual and conceptual information influences word learning at an early age. Dewar (2009) found that infants of 00.10 expect different labels to refer to different kinds of objects, rather than shapes of objects, as demonstrated by the fact that their expectations about the internal soundproducing properties of the objects were influenced by how they were labelled. Socialinteractionists (Markson et al., 2008) argue that because the shape bias emerges early and is related to categorisation, it originates in children’s understanding of language and the world, rather than from mechanistic associations between categories of objects and names. Another argument against the shape bias as the sole mechanism accounting for word learning is that it is specific only to object labels. The shape bias and the noun bias (Gentner, 1982) therefore do not account for the full range of words that are learned from the start of lexical acquisition, such as social greetings, event terms, spatial words, and question words. Akhtar and Tomasello (2000) query the failure of innate constraint theory to account for why language emerges when it does. They propose instead that words are learned only in interaction with other people and are constrained by children’s understanding of what is going on in the social situation in which they hear a new word (Akhtar & Montague, 1999). Ingram (personal communication, April 4, 2006) questions whether constraints are real and empirical, noting that constraint theories do not come out of data-based studies. Hirsh-Pasek, Golinkoff and Hollich (Golinkoff et al., 2000) also stress that it is essential to use data to validate models and demonstrate the evolution of principles for word learning, from early word use to the sophisticated and rapid word acquisition of later word learning.  27  Perceptual salience and motor maturation Researchers have identified perceptual salience and motor maturation as other factors necessary to lexical acquisition. The connectionist model proposes that the gradual maturation of the neurophysiological functions of articulatory and phonological ability, as the child develops control of the vocal tract interacts with auditory and visual perception and language input to underlie speech (Kent et al., 1993; Kent et al., 2001).  Summary Theoretical models of early word learning propose various social, cognitive, environmental, and physical prerequisites for both comprehension and production. These include social interaction, perceptual salience, categorisation, a priori constraints on word meaning, input, and articulatory maturation as necessary prelinguistic conditions for lexical acquisition. Social theories suggest that joint attention and intersubjectivity are the foundation for lexical acquisition. Gaze following reveals speakers’ referential intentions in order to determine which object, action, or attribution they are labelling. Children are often the leaders in word learning situations, motivated by their intentions and contents of mind. Parents’ ostensive handling and labelling of perceptually salient objects and events, presented repeatedly in familiar contexts, is widely regarded as necessary for early word learning, in order for a child to associate the sound of a word with an object or event. Cognition-based researchers regard categorisation abilities as fundamental to the onset of language. Prelinguistic perceptual categorisation develops into the realisation that every object belongs to a category. Social-interactionists suggest that increases in perceptual and conceptual knowledge such as recognising familiar faces and objects and identifying recurrences of previous experiences in familiar contexts assist the child in perceiving similarities and sorting similar relations. Cognitivist theorists propose that a priori constraints and probabilistic assumptions reduce the reasoning required for a child to interpret an unfamiliar word. These include the whole-object assumption and the shape 28  bias. Developmental theories propose models that account for the gradual nature of word learning. This dissertation will examine the applicability of this research to the data, in order to answer the first research question on the relationship between the delayed and unusual early lexical acquisition of my son Graeme and the major developmental milestones in his cognitive and social growth.  Acquiring the lexicon in comprehension Introduction Though there are differences of emphasis, most researchers in lexical acquisition agree upon a recognition of sequential milestones in word learning. The onset of language comprehension co-occurs with the child’s expressions of pragmatic intentions through vocalisation and gesture, and the earliest consistent meaningful productions emerge at the same time as rapid word learning occurs in the comprehension lexicon. This is followed by the production of symbolic words, the vocabulary spurt in the productive lexicon, and, shortly after, the first word combinations. This sequence unfolds in language-delayed children as it does in TD children, though the timing of word learning differs in children with language delay. Studies of infants’ first systematic responses to adult word sounds show that their lexical acquisition begins early in the first year of life. The comprehension lexicon emerges through the comprehension of the adult words and signs of the native language. The lexicon in production has its origin in the emergence of the earliest productions of intentional communication, with word production typically emerging in the second year of life. Used as an index of word learning, an examination of the acquisition of comprehension reveals what it is that babies know when they begin to produce words. How babies form hypotheses about the meanings of the many words they hear is a central issue in theories of word learning. While many theoretical models of word learning are directed towards 29  production, the child’s acquisition of word meaning is a comprehension-based process (Kuczaj & Barrett, 1999). The focus in this section will be on research in the onset of speech perception, the acquisition of the comprehension lexicon, including rates and timing, and the word spurt in comprehension in TD infants, in order to be able to compare these with Graeme’s lexical development.  Speech perception This section presents a brief overview of research on speech perception. Jusczyk (1995a) proposes that in the first year, speech perception develops from unspecified languagegeneral processes to processes specific to the infant’s native language. Lexical acquisition, a higher level of language learning, is a language-specific development. Infants show a preference for their mother’s voice at 00.01 to 00.01.15, and for childdirected speech by 00.04 (Fernald & Kuhl, 1987). Perceptual and conceptual developments in the first year prepare infants to recognise and identify some words in their linguistic input. The infants’ perceptual systems build from their sensitivity to speech to their perception of the language-specific acoustic properties of a word (Gervain & Werker, 2008; Hirsh-Pasek, Tucker, & Golinkoff, 1996; Werker, 2004). Infants develop their ability to segment the speech stream of their native language and to perceive speech sounds as distinct in order to attach preliminary meaning to recurrent phonetic forms (Ingram, 1989; Werker & Yeung, 2005). Infants’ pre-verbal knowledge in conceptual organisation and in the linguistic domain enables them to perceive distinct words and to map them to meanings (Waxman, 2009). Acoustic perception is necessary for phonemic perception, the complex cognitive ability to discriminate speech sounds as phonemic and to link sounds to meaning (Ingram, 1999; Jusczyk, 1995b; Jusczyk, 1995a; Werker & Yeung, 2005). Tincoff and Jusczyk (1999) contend that some of the phonological forms of adult words begin to have meaning for the child at 00.06. Yeung and Werker (2009) demonstrate that infants at 00.09 can learn the sound patterns of meaningful contrasts in phonetic categories in their native language before they have acquired many words in production, by learning to discriminate contrastive pairings through visual cues in social situations where words are labelled. 30  Children gradually develop the perceptual ability to identify single segment differences between language specific words at around 01.00 (Ingram, 1989). Infants are highly attentive to many levels of speech processing relevant to word recognition, such as positive and negative affect conveyed through intonation, exaggerated intonation contours, and prosodically specified sound patterns in word-size speech segments (Fernald, 1989; Jusczyk, Hirsh-Pasek, Kemler, & Kennedy, 1992; Nazzi & Bertoncini, 2003). Ferry (2010) provides evidence that infants at 00.03 to 00.04 categorised words comprised of labelling phrases more successfully than with tone sequences. Infants at 00.06 can utilise familiar names to segment fluent speech (Bortfeld, Morgan, Golinkoff, & Rathbun, 2005). Signs and a wider range of sounds, such as whistles, are accepted as names for objects by babies of 01.01, whereas by 01.06 they do not accept non-speech sounds as readily as speech sounds for labels (Namy & Waxman, 1998; Woodward & Hoyne, 1999). Developmental theories speculate that there is a continuum of learning between phonological processing and lexical acquisition as babies learn to identify words through perceptual mechanisms (Fisher, Church, & Chambers, 2004). Werker and Tees (2002) present the theory of a specialized perceptual-motor system of infant processing that has evolved to serve human speech, but which functions concurrently with other developing abilities in babbling and word learning. Kuhl (2009) notes that cognitive, linguistic, and social abilities contribute to early speech perception, with individual variance in infants' phonetic learning predicting both their first language development patterns and second language learning.  Research on acquiring lexical comprehension Studies in word production greatly outnumber those in comprehension, which are limited by the difficulty in maintaining the interest and attention of infants. Word comprehension is commonly assessed by longitudinal data-based studies using both parental diaries and word checklists, or by daily observations of naturalistic data. In addition, scoring infants’ responses can be problematic in terms of determining whether and to what degree they 31  understand the word in the same way adults do. Commonly used methods to determine comprehension include: presenting the child with alternatives that are systematically varied; attempting to get responses in controlled circumstances; using several scorers; and a cautious interpretation of responses. In Benedict (1979) the criteria for determining a word’s point of entry into the lexicon in comprehension was that the child must show a correct and consistent response to the word, verified by repetition within the following days. A discussion of several longitudinal studies in comprehension acquisition follows. Nelson (1973) published an analysis of the content and semantic structure of the first 50 words of 18 TD children between 01.00 and 02.00 collected in a longitudinal study. In 1979, Benedict modified the classification in order to better analyse comprehension, in the first controlled study of the initial 50 words comprehended and produced in TD infants. Comparing the onset of comprehension and production, Benedict found that comprehension began in the pre-linguistic period at around 00.09 and reached the 50word level at around 01.01, preceding production by almost four months. The rate of acquisition for the first 50 words in the lexicon in comprehension was twice that of production, which began at 01.01.21 and reached 50 words at 01.09.15. Her subjects took around two weeks to acquire their first 10 words in comprehension, but four weeks in production. These are averaged over the eight children studied, but Benedict’s and subsequent studies show there can be great variation in the sizes of an individual child’s vocabulary in comprehension and production. Huttenlocher (1974) presented data from a longitudinal investigation of comprehension development, looking at the first stages in systematic responses to word sounds in three TD children aged 00.10 to 01.06. Her data also provide evidence that comprehension precedes production. Huttenlocher concludes that there is not a simple relationship between children's lexicons in comprehension and production, and that there is an increasingly complex progression within the child’s response to words in the prelinguistic period before speech. Oviatt (1980a; 1980b) investigated babies’ comprehension of recently taught items. She found that babies younger than 01.00 who were taught words learned very little; that 32  much word repetition was necessary, and that they retained only the highly salient items such as the most recently taught words. However, between 01.03 and 01.05, there was both immediate and longer-term comprehension of newly acquired terms for objects and actions, suggesting advances in cognition. Rescorla (1984), in a longitudinal study, investigated individual differences in lexical comprehension of six children from ages 01.00 to 01.10. She examined the size, rate, order, and timing, finding that individual differences in the rate of early word comprehension had emerged by 01.02. These were highly predictive of large differences in language skill, IQ score, and play development at 02.00 and 03.00. Early progress in comprehension acquisition was associated with a referential language style (acquiring primarily naming words), greater use of overextension as a production strategy, a more highly differentiated early category development, and greater reciprocity in mother-child interaction. The benchmark data-based large sample studies that provide definitive measurements of lexical acquisition are based on the MacArthur Communicative Developmental Inventories (CDI) (Fenson, 1989). In the CDI, parental checklists of the verbal production of 1,803 TD children from 08.00 to 02.06 were developed to provide data on the lexicons in comprehension and production of TD children. The infant scale was designed to collect data between 08.00 and 01.04 and the toddler scale between 01.04 and 02.06. This study measured the lag between acquisition of comprehension and production, finding that babies understand between 50 and 100 words at 01.00, when they produce fewer than 10 words, and that by 01.01 babies can acquire a new word in comprehension with relatively little exposure. In 1994, Fenson et al. analysed the CDI data in the infant scale to show the variability in early communicative development in comprehension, and in gestural and vocal production (Fenson et al., 1994). The high rate of variability was attributed to a combination of maturational and environmental factors such as the talkativeness of the mothers, but not to demographic factors of gender, social class, birth order, and parent’s education. A 1996 study used the CDI infant scale results to provide  33  monthly lexical developmental norms for both the comprehension and production of 396 words (Dale & Fenson, 1996). While most studies on context dependent words are directed towards production, words restricted to specific contexts are present in the comprehension lexicon. At the onset, comprehension is scaffolded by an adult’s ostensive labelling and frequent repetition of a word in familiar routines (Nelson, 1988; Rescorla, 1980). A longitudinal study of the development of lexical comprehension and production observed six TD children from 00.06 every two weeks for 18 months in order to determine whether their vocabulary items were contextually flexible or context bound (Harris, Yeeles, Chasin, & Oakley, 1995). The researchers found a close relationship between the two types of words in both early lexicons, with evidence of contextually flexible comprehension emerging very early. They observed that the contextual variance differed in the extent to which comprehension preceded production and in the rates of acquisition in both lexicons.  Rates and timing of comprehension acquisition Individual variation in TD children is documented in the ages, rates, and times of acquisition, but not in the sequence of emergence. Benedict’s (1979) eight subjects had no words in comprehension at a mean age of 00.10.14. At around 01.00, subject to individual variation, children consistently understand words of the adult language to have some meaning, though not necessarily the exact meaning of the adult definition (McCarthy, 1954). The rate of comprehension acquisition is the same as, or greater than, that of production (Benedict, 1979; Ingram, 1989). For most TD children, the onset of the 20th word produced precedes the 100th word comprehended, although both TD and language delayed children can have a lag between the word spurt in comprehension and the onset of language production (Benedict, 1979; Gibson & Ingram, 1983). Most TD children have at least 100 to 200 words in comprehension before they produce their 50th word, estimated to be around the same time as their first word combinations (Ingram, 1989).  34  The word spurt in comprehension Many researchers in lexical acquisition have described periods of accelerated change, both quantitative and qualitative. These periods of rapid word learning have been termed word spurts, when the slowly acquired repertoire in comprehension and production of the novice word learner suddenly increases as words are added to the child’s lexicon at the rate of a more expert word learner. The existence, conditions for, and causes of the vocabulary spurt are the subject of debate among researchers in lexical acquisition. Ganger (2004) reanalyses data from earlier studies to question whether a word spurt occurs in every or indeed in any child, proposing a more gradual acquisition process. Among researchers who espouse the notion of word spurts, the social-interactionists claim that development in the child’s social-pragmatic abilities, contents of mind, attentional engagement, joint attention abilities, and communicative needs are the underlying causes of the engine that drives rapid and efficient increases in understanding novel words. The driving forces of the word spurt, according to cognitivist arguments, are exhaustive categorisation, symbolic representation, and the emergence of cognitive constraints on the possibilities of word meaning. Studies of word spurts in lexical acquisition have been mainly limited to production and not all researchers agree that word spurts occur in comprehension. However, several studies show evidence of a word spurt in comprehension in many TD children, beginning between 01.03 and 01.04 after a few months of an initial slow growth from the first to around the 50th word understood and lasting for several months (D. Bauer, Goldfield, & Reznick, 2002; Benedict, 1979; Fenson et al., 1994; Gibson & Ingram, 1983). Bauer, Goldfield and Reznick (2002) measured vocabulary in comprehension in TD male and female children at two month intervals from 01.02 to 01.10, in order to analyse individual and gender differences in the rate of early vocabulary development. They showed that although the lexical development of girls outpaced that of boys, there were distinctive fast and slow trajectories for both comprehension and production that were not exclusively segregated by gender. The presence of a surge in the comprehension lexicon was associated with the presence of a production spurt occurring around 01.08 to 01.10. Children from 01.03 to 02.00 developed greater proficiency in the speed and accuracy of 35  their word recognition from continuous speech, making rapid gains in verbal processing and in extracting lexical information prior to and concurrent with the production spurt (Fernald, Pinto, Swingley, Weinberg, & McRoberts, 1998). A study of children at 01.08 found that those with large lexicons in production recognised novel words after two repetitions, while those with smaller lexicons needed five repetitions (Torkildsen et al., 2009). These results suggest a relationship between the rate of receptive familiarity and the onset of the word spurt in production. An earlier study of this group indicated that the efficiency of fast-mapping (the linguistic ability to rapidly and efficiently associate unfamiliar terms for objects with real-world concepts through the process of elimination) was substantially greater in children who have entered the word spurt than in those who had not (Torkildsen et al., 2006).  Summary The acquisition of lexical comprehension begins early in the first year of life with the perception of language-specific acoustic properties of a word, and the developing abilities to segment the speech stream and to pay attention to intonation and familiar names. These abilities lead to phonemic perception of meaningful contrasts which link sounds to meaning. Long term data-based studies confirm that comprehension generally precedes production and is acquired at a faster rate, although there is much individual variation in rate and timing of the onset and acquisition of both comprehension and production. The first words in comprehension are context dependent. They are learned slowly, with ostensive labelling, in interactive linguistic environments. After advances in perceptual mechanisms and attentional engagement, rapid word learning in comprehension cooccurs with several key areas of social-cognition related to lexical acquisition. These are social-pragmatic ability (Charman, 2003), contents of mind (L. Bloom & Tinker, 2001), attentional engagement (Meltzoff & Brooks, 2009; Sabbagh et al., 2007), joint attention abilities (Tomasello, 2007), symbolic representational ability (Lifter & Bloom, 1989; Piaget, 1962), categorisation (Gopnik & Meltzoff, 1987; Gopnik & Meltzoff, 1992; Mervis & Bertrand, 1994), and the naming insight (Dore et al., 1976). The connectionist model proposes that gradual, continuous maturation in children’s sensory-motor neural 36  development influences changes in lexical comprehension and production processes, and results in a more gradual curve of growth than a word spurt. Further discussion of research on delays in the acquisition of the comprehension lexicon common to children with autistic spectrum disorders is in Chapter 3. Additional research on comprehension using diary studies is in Chapter 4.  Acquiring the early lexicon in production Introduction This section will examine research on the form and content of the earliest productions. The form of the early productive lexicon includes initial utterances, such as babble, that lead to meaningful gestured and vocalised intentional communication; early idiosyncratic words; and onomatopoeic words. Theories on the content of early words include acquiring meaning through context bound words, and on semantic classes. These studies apply to my second research question regarding the inclusion of Graeme’s early idiosyncratic, gestured, onomatopoeic, and context bound productions in his lexicon as acquired word/signs. Theories of early word learning are also relevant to the definition of what constitutes a word in the lexical acquisition of children with autism.  Mechanisms for learning early words in production Many researchers regard the earliest and later words to be acquired through very different mechanisms (Dromi, 1993; Nazzi & Bertoncini, 2003). In this view, children learning their first words initially employ an associationist mode of learning, mapping phonetically underspecified sound patterns to a perceptually salient event or object in simultaneous repeated context bound situations (Hollich et al., 2000; Nazzi & Bertoncini, 2003; Werker, Cohen, Lloyd, Casasola, & Stager, 1998). Early words are acquired slowly, and may be child-generated and idiosyncratic. A few of the more familiar early words develop into referential words as these links are extended to exemplars of the target category on a case-by-case basis. Referential words, in contrast, evolve smoothly 37  into conventional adult words through phonetic modifications and semantic extensions (Dromi, 1993; Dromi, 1999; Dromi, 2008; Lucariello, 1985).  Intentional communication Prelinguistic intentional communication emerges concurrently with the acquisition of the vocabulary in comprehension and joint attention skills when TD children from 00.09 to 01.02 are finding a way into the system and becoming sensitised to the forms of language (Bates et al., 1975; Nelson, 1991)7. Children develop the ability to communicate their intenti