Using an infant speech identification (ISI) procedure, English language environment infants, two- and six-year-old children, and adults were tested for their identification of sounds on a native (voiced/voiceless bilabial stop) and a non-native (prevoiced/voiced bilabial stop) speech continuum. Categorical perception of the two contrasts diverged as a function of age, increasing for the native contrast and decreasing for the non-native between two and six years. In Experiment 2, subjects of the same four ages were tested for their identification of a continuum of harmonic tones varying in pitch. Comparison of the results of Experiment i with the essentially continuous perception of this pitch continuum supports the view that the perception of the native contrast becomes more categorical with age, while perception of the non-native contrast becomes less categorical, especially at six years. Experiment 3, in which adults were tested on the three continua with a multi-trial open set procedure, demonstrated that results with the ISI procedure in Experiments 1 and 2 are comparable to results with more traditional methods. The results of the three experiments are discussed in terms of the role of specific linguistic experience in the development of categorical speech perception.

This study investigates the production of vocalization in adult–infant–toy interactions from 0;4 to 0;11. The hypothesis is that vocalizations are selectively uttered in relationship to their production context. Five infants (two girls, three boys) were intensively studied. Non-segmental acoustic features of vocalizations in four communicative contexts were analysed in relation to the individual infants, in order to reveal individual differences. The data were submitted to discriminant function analysis. Results show that (a) different patterns of non-segmental features characterize sounds produced in different contexts; (b) both inter-subject differences and intra-subject consistency are observed; (c) ‘selective production’ disappears after 0;9. These results are discussed in relationship to sound–meaning development.

The two- and three-syllable reduplicative babbling of five French-learning and five English-learning infants (0;5 to 1; 1) was examined in two ways for intonational differences. The first measure was a categorization into one of five categories (RISING, FALLING, RISE-FALL, FALL-RISE, LEVEL) by expert listeners. The second was the fundamental frequency (F0) from the early, middle and late portion of each syllable. Both measures showed significant differences between the two language groups. 65% of the utterances from both groups were classified as either rising of falling. For the French children, these were divided equally into the rising and the falling categories, while 75% of those utterances for the English children were judged to have falling intonation. Proportions of the other three categories were not significantly different by language environment. In both languages, though, three-syllable utterances were more likely to have a complex contour than two-syllable ones. Analysis of the F0 patterns confirmed the perceptual assessment. Several aspects of the target languages help explain these intonational differences in prelinguistic babbling.

Three children were observed interacting with their mothers at three different times: before the onset of single words, when vocabulary consisted of 10 words, and when vocabulary consisted of 50 words. Relations between communicative functions and acoustic analyses of prosodie variables (i.e. rise vs. nonrise of terminal contours) were studied. Considerable variability was found among the children in the number of rises produced overall and those produced for any function. Each child's use of rise was fairly constant over time and rises were produced RELATIVELY more frequently than nonrises with functions requiring a response from the listener. Factors affecting similarities and differences are discussed.

An unusual phonological pattern exhibited by a child aged 2;9 is described. The pattern involved the production of word-final strident continuants in words whose adult forms contain these features in initial, rather than final position (e.g. [ops] for soap). The data are interpreted as compatible with models of child phonology that permit direct mapping of features to final position.

The view that grammatical relations have substantial essence, designated as ‘subject’ or ‘object’ has difficulty in accounting for the variety of naturally acquirable grammatical relations. The acquisition of grammatical relations is examined from a theoretical framework, ROLE AND REFERENCE GRAMMAR, in which grammatical relations are decomposed into two separate types of structure: logical (semantic) structure and information (pragmatic) structure. The acquisition of grammatical relations from four languages is compared: (1) the definite accusative suffix and pragmatically motivated word order of Turkish; (2) Kaluli verb agreement, case and focus marking postpositions, and pragmatically motivated word order; (3) Hungarian definite and indefinite verb conjunction; and (4) Italian participial agreement and anaphoric, accusative case pronouns. Two conditions on structures are found to cause difficulty: the neutralization of a semantic or pragmatic distinction by interfering structures (e.g. Kaluli and Italian), and global case marking which forces the child to discover relevant semantic characteristics of both the actor and the undergoer (e.g. Hungarian and Kaluli). Structures that encode semantic or pragmatic distinctions independently are more easily acquired (e.g. Turkish). Piecing together discrete structures in a mosaic fashion, the child can acquire the great variety of grammatical relations that exist in human languages.

In Harris, Barrett, Jones & Brookes (1988), we reported the results of a detailed analysis of the initial uses of the first 10 words which were produced by four children. The present paper reports the results of an analysis of the subsequent uses of these 40 words. This analysis reveals that seven qualitatively different patterns of change occurred between the children's initial and subsequent uses of these words; the particular patterns of change which occurred support Barrett's (1986) model of early lexical development. In addition, it was found that, although there was a strong relationship between maternal speech and the children's initial word uses, the relationship between maternal speech and the children's subsequent word uses was very much weaker. These findings indicate that the role of linguistic input in early lexical development may decline quite sharply once the child has established initial uses for words.

Temporal correspondences between the attainment of specific milestones in play and language were examined through a longitudinal study of four Japanese children (aged 0;7 to 0;11 at the beginning of the study). There were developmental correspondences between the onset of six language landmarks (the emergence of first words, naming words, vocabulary spurts, word-chains, nonproductive two-word utterances, productive two-word utterances) and the onset of subcategories of play. Language and play both reflected the development of underlying symbolic ability, and both developed in a parallel manner at the single-word stage. After the emergence of word-chains, language and play developed interdependently. All the children proceeded through the same sequence of stages, but the rate of development was different depending on their environment.

Four kinds of language learning-related behaviours (LLRBs) were examined in the home conversations of six English children studied for six months from age 2;0 to 2;6. The speech of the children was coded for the number of times they participated in language lessons, language practice, metalanguage and revisions of prior language. All the children were active and frequent producers of LLRBs, with revisions being observed most commonly. Further analysis revealed that the majority of the revisions that the children initiated involved grammatical changes, with the revised utterances tending to be more grammatical than their predecessors. An auditory monitor is proposed as a partial explanation for revision behaviour, and is evaluated relative to other possible accounts of the findings.

Maternal directiveness, assessed by the mother's use of prescriptives, is correlated with slow vocabulary development. As prescriptives are most often used to redirect a child's attention to a different object or activity, it is hypothesized that attentional regulation underlies this negative relationship. In the present study, twelve mothers were videotaped interacting with their children aged 1;1, and 100 maternal utterances were coded for pragmatic intent. Prescriptives were coded as either changing (LEADING) or FOLLOWING the child's focus of attention. Only the frequency of mothers' FOLLOW-prescriptives correlated significantly with a productive vocabulary measure taken at 1;10. This correlation was high and positive, indicating that, given joint focus, directing a 13-month-old's behaviour can have beneficial effects on subsequent vocabulary development.

One of the most widely used indices of language development is a measure of utterance length in morphemes (MLUm). This study examines the applicability of MLUm to the acquisition of Irish. MLUm was calculated for data from Cian, aged 1;11–3;0. Even when an attempt was made to ‘assume the maximum’ by counting all possible morphemes, the correlation between a morpheme MLU (MLUm) and a word count MLU (MLUw) was very high (0·99). This points to MLUw being as effective a measure of Irish development as MLUm, as well as being easier to apply and more reliable. MLUw was calculated for the two younger children in the study (Eibhlís 1;4–2;1 and Eoin 1;10–2;6). An examination of the relationship between the three children's MLUw values and their grammatical complexity as measured on ILARSP (the Irish adaptation of LARSP) indicates that MLUw is a useful preliminary index for early development in Irish. However, further data are necessary to check whether MLUw loses its predictive relationship with grammatical complexity after a certain point. The study emphasizes the caution necessary in applying MLU to languages whose acquisition has not hitherto been studied, and underlines the role of MLU as a preliminary measure, which must not be overinterpreted.

Various theories of learning for the categories COUNT NOUN and MASS NOUN are compared. It is argued that children assign words to these categories on the basis of intuitions arising from perception that are relevant to Macnamara's (1986) semantic definitions of the categories. These definitions rest on the centrality of identity in the meaning of nouns and the centrality of individuation in the meaning of count nouns but not mass nouns. Empirical evidence is presented that supports the hypothesis that young children classify words as count nouns or mass nouns on the basis of perceptual information about the extension of the words, that is, whether or not the extension consists exclusively of enduring individuals whose discreteness from one another is perceptually salient (count nouns) or not (mass nouns). In an experiment, 48 children with a mean age of 2;10 (S.D. = 0;5) were taught a word for a kind of object (i.e. a perceptually distinct individual) or for a kind of substance (i.e. a collection of small granules). For some children the word was syntactically COUNT and for others it was syntactically MASS. Half of the children received incongruous perceptual and syntactic cues. For most of these children, classification of the word was guided by the object- or substance-like appearance of the stimulus despite the presence of incongruent syntactic cues. Syntactic cues influenced classification of the word for a minority of subjects, most of whom were among the oldest in the sample. It is concluded that perceptual information is critical in early decisions about membership in the categories COUNT NOUN and MASS NOUN.

Data from an experiment on gender acquisition with 160 Spanish children from four to eleven years of age are presented in this paper. In Spanish there are three possible clues (semantic, morphophonological and syntactic), that speakers can use to determine the gender of a noun and the agreement of other variable elements accompanying it. Items where only one of the clues was present, items where there was a combined effect of two of them in agreement (both were feminine or masculine), and items where clues were in conflict (one masculine and the other feminine) were introduced in the experiment. This experimental manipulation made it possible to test the relative strength of the different types of competing clues. In particular, the aim of the present study was to determine the relative importance of intralinguistic and extralinguistic clues, as evidenced by the ability of Spanish children to recognize the gender of a noun upon hearing it in a particular frame, and consequently, to establish the agreement of other variable elements accompanying it. A procedure similar to that used by Karmiloff-Smith (1979).was employed. The results (which are compared with those obtained in other languages) give support to the theoretical view that children pay far more attention to syntactic and morphophonological (intralinguistic) information than to semantic (extralinguistic) information.

The development of forms and functions of negation in parent–child interactions in the early stages of language acquisition is discussed with illustrative examples from field data relating to two Tamil-speaking children and their parents. An attempt is made to provide a scheme for analysing the negatives in children's speech, maintaining the three basic functions of Bloom (1970) unaltered. Subcategories of these functions are introduced, taking into consideration the interactional situations as they are perceived by the parents and children. Children, it is observed, acquire the forms and functions of negation in a clear developmental sequence. They acquire and use the free forms naanaa (veeṇḍaam) ‘No, I don't want’ first, followed by negatives involving bound forms. Certain forms like maaṭṭeen ‘I won't’, muḍiyaadi' I can't’ and even veeṇḍaam are used in isolation (as free forms) first. Only later are they used with verb infinitives. The development of functions of negation follows the sequence: rejection, non-existence, prohibition and denial. Thus the children of this study show preference for rejection over non-existence as observed by earlier scholars.

The purpose of this research was to evaluate the development of temporal location within a cross-linguistic experimental design. The research focused on the transition from a temporal system based on absolute temporal relations involving speech time and event time to a more complex system involving relative temporal relationships and reference time. A comprehension test was constructed with problems which were diagnostic of two salient distinctions within each of three temporal systems. The procedure was based on a two-choice sentence–picture matching task. The children who participated were from Poland, the USA and Finland, and there were 12 children at each of the following age levels: 2;6, 3;6, 4;6, 5;6, and 6;6. The Polish and American children solved problems requiring absolute temporal location at 2;6, and the older children eventually solved most of the problems requiring relative location. The Finnish children followed a different pattern taking longer to comprehend both types of problems. A second experiment confirmed the Finnish pattern of development. The results of the comprehension test were compared to observations of conversational and narrative discourse which were the product of two elicitation procedures. The research demonstrates the way in which conceptual development places a universal constraint on the developmental process and how the specific properties of individual languages also have an effect.

This study investigated the claim that very young children avoid backwards coreference in their interpretation of sentences containing pronouns. Eighty-one children ranging in age from 3;1 to 8;0 and eight adults acted out four types of pronominal sentences. Cross-sectional data and individual response patterns reveal that children initially, prefer internal coreference even when such a response is disallowed for structural reasons. Avoidance of backwards coreference appears to be a late developing phenomenon characteristic of six-year-olds. Adult response patterns, which are manifested by some very young children, emerge as the dominant pattern by age seven.

This on-line study investigates the processing of word order by 30 French children (6;6, 8;6, 10;6) and 10 adults. Its main objective is to show that the privileged status granted to ‘canonical SVO sentences’ is inadequate to account for the on-line processing of pronominal utterances in spoken French. Using a word monitoring task, we showed that word order (NVN vs NNV): (1) is a significant factor in sentences containing no clitic pronoun; (2) stops being significant when sentences contain either one or two clitic pronouns. These results suggest that processing complexity depends upon co-reference (‘linear’, ‘crossed’ or ‘embedding’) assignment constraints rather than upon word order per se. We conclude that, in French, word-order processing always interacts with acceptability considerations provided by cliticization.

An extensive fuzzy boundary was identified within the adult extension of two exemplary object-words, ‘dog’ and ‘ball’. The fuzzy boundaries were incorporated into the standard of adult extension against which the preschool child's use of ‘dog’ and ‘ball’ was compared, and the investigation provided a framework in which children's responses to fuzzy boundary referents could be accommodated. Responses to fuzzy boundary referents were considered as having precedents within adult extension and the data revealed the provision of precedents for a significant number of the young child's applications and non-applications of the words. The investigation concluded that, through adherence to an oversimplified concept of adult extension, previous research has underestimated the degree of correspondence that exists between child and adult extension.

An experimental elicitation task with children between the ages of 1;8 and 11;3 shows that children learning Thai numeral classifiers begin with purely distributional information: specifically, (1) that classifiers must appear in the post-numeral position, and (2) that classifiers comprise a conventional, closed set of words. Semantic organizing features, such as salient features of the head noun's referent, appear later than these syntagmatic organizing features. Use of such semantic information is not an immature ‘first guess’ at grammatical categories, but rather, a necessary component of adult linguistic competence, because the categories are productive both for older children and for adults.