Charles-Luce & Luce (1990) found smaller phonological similarity neighbourhoods in five- and seven-year-old children's expressive lexicons than in an adult receptive lexicon, a finding they interpreted as evidence that children need not employ fine-grained auditory perceptual analyses in lexical processing. In the present investigation, neighbourhood sizes were calculated for an expressive lexicon derived from two vocabulary lists representative of children aged 1;0 to 3;0 (Rescorla, 1989; Reznick & Goldsmith, 1989). Over 80% of the words in these early lexicons had at least one phonological neighbour; nearly 20% had six or more phonological neighbours. Very young children must have access to reasonably detailed phonological information in order to create and distinguish among such phonologically similar lexical entries.

Although adult-based words co-occur in the period of transition to speech with a variety of non-word vocalizations, little attention has been given to the formidable problem of identifying these earliest words. This paper specifies explicit, maximally ‘inclusive’ identification procedures, with criteria based on both phonetic and contextual parameters. A formal system for evaluating phonetic match is suggested, as well as a set of child-derived functional categories reflecting use in context. Analysis of word use across two samples of 10 children each, followed from 0;9 to 1;4, provides evidence to suggest that context-bound words can be ‘trained’ by focusing on eliciting language, but that the timing of context-flexible word use remains independent of such training.

The acquisition of language in the preschool years by multiple-birth children is often reported to be atypical. Some researchers have claimed that they use ‘twin language’, i.e. an autonomous language specific to a multiple-birth set. This claim was investigated by describing and comparing the phonological characteristics of the speech of 19 sets of two- to four-year-old multiple-birth children, and by measuring multiple-birth children's understanding of their twins' or triplets' context-free speech. The results indicated that multiple-birth children are prone to phonological disorder and consequently their speech is often unintelligible. Siblings' phonologies evidenced some similarities, although they were not identical. Multiple-birth children were better able to understand their siblings' mispronunciations than were other children of the same age, although that understanding was dependent on how closely the error form resembled the adult form. While these three factors conspire to give the impression that ‘twin language’ is common, none of these findings provided support for the claim that multiple-birth children use an autonomous language.

The past and present state of the empiricism–nativism issue is analysed. Empiricist philosophical doctrine (‘no innate ideas’) distinguished idea from structure or mechanism. However, Chomsky's conception of innate linguistic universals erased this distinction. The elimination left would-be empiricists without a coherent and defensible position. I argue that the issue remains alive primarily because of tension between two scientific tasks that face students of development. One is to discover what is cognitively and linguistically primitive, a task that encourages nativism. However, nativism is ultimately unsatisfactory because it systematically neglects the other task, which is to account for development, including the emergence of postulated innate primitives. To account for such primitives, it is necessary to relate them to particular central nervous System structures in such a way as to explain how the structure has the particular cognitive effects that define the primitive. That is likely to be difficult, and I show how the study of learning – much neglected in recent years – can help by reducing the number and type of innate primitives whose origin must be explained in that way.

Interaction sequences are explored which are initiated by either child requests or adult offers of objects. The focus is on those sequences in which the child does not want an object that is passed to her, and on how the child manages such an interactional contingency. Throughout the age range in question, 1;0 to 1;8, the child uses re-requests where this contingency occurs in request sequences. The analysis traces the development of these re-requests and compares them with other forms of re-request. In addition, differences are uncovered as between request and offer sequences concerning the child's ways of dealing with an unwanted object that is passed to her. Linkages are made between these themes and work on third position repair within conversation analysis. Home based video-recordings of one child between the ages of 1;0 and 1;8 constitute the data base for the study.

The phonemic inventories of 30 children (aged 3;4–5;7) with phonological delays were examined in terms of featural distinctions in order to address universal vs. individual accounts of acquisition. Phonetic inventories of these same children were also identified for comparison purposes. Across children, four hierarchical and implicationally related types of phonemic inventory were identified. The typology uniquely captured common distinctions maintained by all children, and at the same time, allowed for individual differences in the specific phonemic composition of each system. These cross-sectional results have theoretical implications for the longitudinal course of phonemic acquisition. In particular, children appear to have a number of linguistic choices that relate to the course, the specifics, and the mechanism of change in acquisition.

This work reports longitudinal evaluation of the speech-like vocal development of infants born at risk due to prematurity or low socio-economic status (SES) and infants not subject to such risk. Twenty infants were preterm (10 of low SES) and 33 were full term (16 of low SES), and all were studied from 0;4 through 1;6. The study provides the indication that at-risk infants are not generally delayed in the ability to produce well-formed speech-like sounds as indicated in taperecorded vocal samples. At the same time, premature infants show a tendency to produce well-formed syllables less consistently than full terms after the point at which parents and laboratory personnel note the onset of the canonical babbling stage (the point after which well-formed syllables are well established in the infant vocal repertoires). Further, even though low SES infants produce well-formed speech-like structures on schedule, they show a reliably lower tendency to vocalize in general, as reflected by fewer utterances per minute in recorded samples.

Young children learning English as well as many other languages frequently omit weakly stressed syllables from multisyllabic words. In particular, they are more likely to omit weak syllables from word-initial positions than from word-internal or -final positions. For example, the weak syllable of a weak-strong (WS) word like giraffe is much more likely to be omitted than the weak syllable of a SW word like tiger. Three hypotheses for this omission pattern have been offered. In two, children's weak syllable omissions reflect innate perceptual biases either to ignore initial weak syllables or to encode word-final syllables. In contrast, the SW Production Template Hypothesis states that children have a template for producing a strong syllable followed by an optional weak syllable. When they apply a series of SW templates to their intended utterances, weak syllables that do not fit the templates are more likely to be omitted than those that do. To compare the three hypotheses, young two-year-olds were asked to say four-syllable SWWS and WSWS nonsense words. Children's pattern of weak syllable preservations was highly consistent with the SW production template hypothesis, but not with the perception-based hypotheses. Implications of this research for children's function morpheme omissions and for the relation of metrical and segmental production templates are discussed.

In this study it is argued that the omission of closed class morphemes and of unstressed syllables within words is related to their common characteristic, viz. that they are unstressed, rhythmically weak parts of utterances. Several strands of evidence indicate that it is unlikely that children are unable to perceive these elements in the input speech. The pattern of (non)realization of unstressed syllables within content words and the class of determiners, was analysed in two Dutch children from 1;6 to 2;11. It appeared that polysyllabic words were quite generally truncated in such a way that they fitted a trochaic (strong-weak) pattern, particularly in the early samples. Some observations with respect to the (non)realization of determiners are suggestive of an influence of an SW-constraint on the realization of noun phrases. These findings support the hypothesis that in the course of utterance preparation, words and phrases are mapped onto S(W) templates. Anecdotal evidence suggests that the dissolution of the SW-constraint coincides with the acquisition of specific aspects of stress assignment in Dutch, such as quantity sensitivity.

A longitudinal diary-and-videotape study of the production of phonological segments by a Spanish–English bilingual child, age 1;1–3;9, revealed four stages in consonantal acquisition: presystematic variation, formation of a single system, separation into two systems corresponding to the two languages, and achievement of adult target values with later interference of one language in the other. Vocalic acquisition proceeded with widespread variation, followed by stabilization at target adult values, without any apparent intermediate unitary-system stage. Relevance of the data adduced here to the issue of whether bilingual children first function with a single system before acquiring two discrete languages is discussed. The issue of what it may mean to have one or two systems of phonology is addressed.

A blind child and her sighted twin sister were recorded at home once a month from 2;5 to 3;5. Their imitations, repetitions and routines (IRR speech) were analysed with respect to three dimensions (type, faithfulness to the model, and function). MLU for IRR speech was greater than that for productive utterances, which supports the idea that these productions facilitate language development. The blind girl used routines and modified imitations and repetitions more frequently than her sighted sister, and her use of modified repetitions increased during the study. She also seemed to use IRR speech for a longer period in her development. Both children frequently used these forms as a means of maintaining conversation and to fulfil their intentions. By introducing modifications (varying some element, or expanding the model) both children, especially the blind girl, analysed language and learned to use these utterances in the appropriate circumstances with a clear pragmatic function. Such behaviour is typical of holistic language processing and acquisition.

Speech addressed to children is supposed to be helpfully redundant, but redundant or predictable words addressed to adults tend to lose intelligibility. Word tokens extracted from the spontaneous speech of the parents of 12 children aged 1;10 to 3;0 and presented in isolation to adult listeners showed loss of intelligibility when the words were redundant because they had occurred in repetitions of an utterance (Experiment 1) or referred to an entity which was physically present when named (Experiment 2). Though children (N = 64; mean age 3;5, S. D. 6·1 months) recognized fewer excerpted object names than adults (N = 40), less intelligible tokens appeared to induce child listeners to rely on the word's extra-linguistic context during the recognition process (Experiment 3), much as such tokens normally induce adults to rely on discourse context. It is proposed that interpreting parental utterances with reference to non-verbal context furthers linguistic development.

Several recent studies have demonstrated strong relationships between lexical acquisition and subsequent developments within the domain of morphosyntax. A connectionist model of the acquisition of a morphological System analogous to that of the English past tense (Plunkett & Marchman, 1993) suggests that growth in vocabulary size may relate to the onset of over-regularization errors. However, this model suggests that the relationships between vocabulary size and morphosyntactic development are non-linear. Incremental increases in the number of verbs to be learned result in qualitative shifts in the treatment of both previously learned and novel forms, but only after the size of the lexicon exceeds a particular level (i.e. reaches a ‘critical mass’). In this paper we present parental report data from an extensive study of English-speaking children aged 1;4 to 2;6 using the MacArthur Communicative Development Inventory: Toddler form (N = 1130). These data corroborate several findings from previous studies, including the early usage of unmarked verb stems and the correct production of irregular past tense forms. Further, we demonstrate support for the ‘critical mass’ view of the onset of over-regularization errors, focusing on continuity among lexical and morphological developments. In our view, these data suggest that these linguistic milestones may be paced by similar, if not identical mechanisms.

Results are reported for stylistic and developmental aspects of vocabulary composition for 1, 803 children and families who participated in the tri-city norming of a new parental report instrument, the MacArthur Communicative Development Inventories. We replicate previous studies with small samples showing extensive variation in use of common nouns between age o;8 and 1;4 (i.e. ‘referential style’), and in the proportion of vocabulary made up of closed-class words between 1;4 and 2;6 (i.e. ‘analytic’ vs. ‘holistic’ style). However, both style dimensions are confounded with developmental changes in the composition of the lexicon, including three ‘waves’ of reorganization: (1) an initial increase in percentage of common nouns from 0 to 100 words, followed by a proportional decrease; (2) a slow linear increase in verbs and other predicates, with the greatest gains taking place between 100 and 400 words; (3) no proportional development at all in the use of closed-class vocabulary between 0 and 400 words, followed by a sharp increase from 400 to 680 words. When developmental changes in noun use are controlled, referential-style measures do not show the association with developmental precocity reported in previous studies, although these scores are related to maternal education. By contrast, when developmental changes in grammatical function word use are controlled, high closed-class scores are associated with a slower rate of development. We suggest that younger children may have less perceptual acuity and/or shorter memory spans than older children with the same vocabulary size. As a result, the younger children may ignore unstressed function words until a later point in development while the older children tend to reproduce perceptual details that they do not yet understand. Longitudinal data show that early use of function words (under 400 words) is not related to grammatical levels after the 4OO-word point, confirming our ‘stylistic’ interpretation of early closed-class usage. We close with recommendations for the unconfounding of stylistic and developmental variance in research on individual differences in language development, and provide look-up tables that will permit other investigators to pull these aspects apart.

Word learning by young children is viewed as a problem of deriving meaning from the use of forms in discourse contexts. Uses of causal and temporal terms in private speech by a child studied longitudinally from 1;9 to 3;0 are analysed from this perspective. Evidence is presented that words are first constrained to uses in specific discourse contexts, and later used more flexibly and with greater control over the semantics of the terms. Derivation of meaning from discourse is described as a dialectical process, and as such it is claimed to be more consistent with the full range of observational data, and with theories of word learning applicable to older children and adults, than other current theories of lexical acquisition in early childhood.

The study investigates how basic communicative functions expressed by utterance modalities (declarative, exclamative, injunctive, interrogative) emerged in the early language of a French child, and examines whether and how morphosyntactic and prosodic devices were used to mark these contrasts. A longitudinal corpus of naturalistic productions was collected between the ages of 1;2 and 1;9, and 960 utterances were subjected to functional, prosodic and grammatical analyses. Declarative, exclamative and injunctive utterances were found from 1;2, and first interrogatives appeared at 1;6. Intonation contours varied as a function of utterance modality and were largely in accordance with the patterns in French: declaratives and exclamatives were falling, interrogatives rising and injunctives split between falling and rising contours depending on their specific functions. A quarter of the productions involved an elementary grammatical marking of utterance modality such as interjections, imperative or indicative verbal forms, or interrogative morphemes. These findings indicate an early and complementary use of prosodic and grammatical devices in the child's construction of the linguistic system.

Universally, object names make up the largest proportion of any word type found in children's early lexicons. Here we present and critically evaluate a set of six lexical principles (some previously proposed and some new) for making object label learning a manageable task. Overall, the principles have the effect of reducing the amount of information that language-learning children must consider for what a new word might mean. These principles are constructed by children in a two-tiered developmental sequence, as a function of their sensitivity to linguistic input, contextual information, and social-interactional cues. Thus, the process of lexical acquisition changes as a result of the particular principles a given child has at his or her disposal. For children who have only the principles of the first tier (REFERENCE, EXTENDIBILITY, and OBJECT SCOPE), word learning has a deliberate and laborious look. The principles of the second tier (CATEGORICAL SCOPE, NOVEL NAME – NAMELESS CATEGORY’ or N3C, and CONVENTIONALITY) enable the child to acquire many new labels rapidly. The present unified account is argued to have a number of advantages over treating such principles separately and non-developmentally. Further, the explicit recognition that the acquisition and operation of these principles is influenced by the child's interpretation of both linguistic and non-linguistic input is seen as an advance.

The purpose of the present study was to show that children under age 3;0 are capable of reformulating requests in different ways depending on how their addressee responds to the request. This adaptive ability is considered to be an indication of the use of both communicative and conversational skills. Children from French-speaking families were observed at two developmental thresholds: the end of the prelinguistic period (1;6) and the start of the linguistic period (2;6). The verbal and nonverbal outputs of the two groups of 12 children (1;6 and 2;6) were compared in three object request situations: the adult immediately complies with the request (satisfaction), the adult asks a clarification question (clarification), and the adult refuses to comply with the request (refusal). The ways in which the children adapted to each situation were found to be similar at the two ages considered. In the clarification situation, vocal productions and their repairs were more numerous, whereas in the refusal situation, non-reformulations and gazes to and from the requested object and addressee predominated. The discussion deals with the significance of these results in the development of communicative and conversational skills in children.

Recent research suggests that young children assume, as a default, that a word applied ostensively to an unfamiliar object (e.g. a man riding in a car) is a basic-level count noun (e.g. person) rather than a situation-restricted count noun (e.g. passenger). In this experiment, 14 mothers and their children (mean age 3;7; range 3;0 to 4;6) participated in a story-book reading session in which the mothers taught their children both a basic-level count noun and a situation-restricted count noun for a series of object drawings. An analysis of mothers' spontaneous teaching strategies revealed that, for a given object, they typically taught a basic-level count noun before a situation-restricted count noun. Furthermore, they tended to teach basic-level count nouns exclusively through estensive definition, especially when the objects were unfamiliar. Moreover, they were (1) more likely to use estensive definitions and (2) less likely to provide additional information concerning the application of the word, when teaching basic-level count nouns than when teaching situation-restricted count nouns. The results provide new evidence that word teachers, like young word learners, assume that a basic-level count noun is the psychologically-privileged candidate for an ostensive definition involving an (unfamiliar) object.