Young children readily acquire new words with consonants and syllable structures already used accurately (IN words). They have more difficulty acquiring new words with consonants or syllable structures never before produced or attempted (OUT words). In the present study, we examined children's acquisition of a third type of word, containing consonants the children had attempted in the past but never produced accurately (ATTEMPTED words). IN, OUT and ATTEMPTED words and their object referents were presented to 11 young children in a series of play sessions. The children's production and comprehension of the words were then assessed. No comprehension differences among the three types of words were observed. However, ATTEMPTED words as well as OUT words were less likely to be acquired in production than IN words. Some revisions in models of child phonology are proposed to accommodate these findings.

Type/Token Ratios have been extensively used in child language research as an index of lexical diversity. This paper shows that the measure has frequently failed to discriminate between children at widely different stages of language development, and that the ratio may in fact fall as children get older. It is suggested here that such effects are caused by a negative, though non-linear, relationship between sample size (i.e. number of tokens) and Type/Token Ratio. Effects of open and closed class items are considered and an alternative Verbal Diversity measure is examined. Standardization of the number of tokens before computing Type/Token Ratios is recommended.

The study examined the transition to purposive use of intonation with single words for two children. Contrary to Bloom's (1973) claim of no systematic use of intonation in the one-word period, purposive use of rising tones was demonstrated in the context of indicating interest and naming objects. One child, by 1;3, clearly used rising tones to ask ‘legitimate’ questions in which the content of the replies mattered. If an informative response was not given he repeated his question accentuating the rising tone. The other child also sought informative replies, although this could only be established at 1;6. In addition, by 1;5 both children demonstrated widespread grammatical use of intonation in which a word was combined with distinct intonations to indicate a meaning distinction equivalent to one made by the grammar of the language. There were, however, earlier developments that presaged the way, so that the transition to grammatical use of intonation was gradual.

Two groups of 20 infants aged 0; 3 experienced either conversational turn taking or random responsiveness of an adult. All infant vocalizations were counted and then each was categorized as a speech-like (syllabic) sound or a nonspeech-like (vocalic) sound. The results of this experiment indicated that turn taking caused changes in the quality of infant vocal sounds. When the adult maintained a give-and-take pattern, the infant produced a higher ratio of syllabic/vocalic sounds. The effect of turn taking on infant vocalizations was discussed in terms of its possible adaptive value for adult responsiveness.

Stoel-Gammon & Cooper (1984) have recently presented phonological analyses in this journal of three children acquiring English. Their general conclusion is that the three children are each following their own unique pattern of acquisition. This paper is a response to that conclusion. We first point out that such research needs to distinguish different possible causes of variation. We then go on to criticize the method of analysis they used, arguing that it disposes the results towards individual variation. Next, we present an alternative analysis, using a different methodology, and show that the children's patterns of acquisition are actually relatively similar. Lastly, we conclude that both their and our results provide evidence for early phonological processing in opposition to the biological model presented in Locke (1983).

A new method to assess language comprehension in infants and young children is introduced in three experiments which test separately for the comprehension of nouns, verbs, and word order. This method requires a minimum of motor movement, no speech production, and relies on the differential visual fixation of two simultaneously presented video events accompanied by a single linguistic stimulus. The linguistic stimulus matches only one of the video events. In all three experiments patterns of visual fixation favour the screen which matches the linguistic stimulus. This new method may provide insight into the child's emerging linguistic capabilities and help resolve longstanding controversies concerning language production versus language comprehension.

This study evaluated the age-relationship, intra-sample reliability and grammatical validity of MLU-like measures of utterance length in the free speech of 21 normally developing children aged 1;8 to 2;8 with MLU between 1·05 and 3·06. Contrary to a previous study by Klee & Fitzgerald (1985). it was found that measures of utterance length hold a good relation with age, are reliable and predict grammatical development well in the age and MLU range considered. It is argued that there are identifiable points in the developmental evolution of MLU and MLU-like indices beyond which their reliability and validity have serious problems. Below these points, they constitute sound measures of syntactic development.

The first 50 words of three blind children were collected and analysed using procedures used by Nelson (1973) on 18 sighted children. The early vocabulary of the blind children paralleled that of the sighted children in the age and speed of acquisition, and in the underlying characteristics of what the children chose to label. These reflect a sensorimotor organization in which self-action and perceptual change are the salient variables. The early words of the blind children differed from those of sighted children in the percentage of words in each of Nelson's grammatical categories. This suggests differences in how the children use language. These differences are discussed as a function of the children's lack of vision and their particular language learning context.

The hypothesis that prelinguistic vocalizations contain extensive and systematic sound-meaning correspondences was examined through an exhaustive analysis of the babbling of five infants in their second year. These infants were videotaped over a period of three to six months at home and at a day-care centre. Their babbled utterances were transcribed phonetically and categorized according to consonant-type and vowel-type. Contexts for each utterance were also categorized, primarily according to the infant's simultaneous action. A quantitative analysis of co-occurrences between phonetic and contextual categories determined that across infants between 14 and 40% of utterances recurred in particular contexts with a greater than expected frequency. These findings support Halliday's (1975a) notion that babbling is not entirely random but contains consistent sound-meaning relations that are not adult-modelled. They also extend the notions of continuity between prelinguistic and linguistic stages of development to the semantic domain.

Phonetic and phonological analyses were performed on spontaneous speech samples of six 2–year–old monolingual Puerto Rican Spanish-learning children. The analyses showed a number of patterns of sound usage similar to those found in English-learning children of the same age, as well as children from other linguistic backgrounds. These findings add support to the claim that certain universal patterns exist in phonological development. However, a number of patterns were also observed which seemed to be accounted for by the target language being acquired. Similarities and differences among the individual children are also discussed.

This paper seeks to account for a culturally patterned set of analogic renamings found in Huli baby talk, nursery rhymes and children's verbal games. These socialization activities evidence a marked concern with body motifs and appellations. In accordance with ethnographic paradigms of explanation, the frame of reference is broadened to include consideration of inter-adult behaviour involving ‘talk about the body’ to assess what is being learnt from such interactions as well as what communicative intents are encoded. The argument is developed that in addition to the significance of these play routines in sensitizing the child to cultural rules about speaking, these ludic forms also appear implicated in both an evolutionary and logico-operational sense in the conventional anatomical nomenclature.

The communicative functions of pitch direction and pitch range during the second year were investigated. The research focusses on initial plus repeated requests, either for objects or for co-operation, which were contrasted against giving, showing and labelling. Infants aged 1; 2 to 1;10 were observed in interaction with their mothers in a semistructured situation. Vocalizations were coded separately for communicative function and for prosodic features. Results show that pitch range is consistently higher (1) for repeated requests than for initial requests for objects and (2) for initial requests for objects and for co-operation than for labelling. Giving and showing have an intermediate rank between requests and labelling. Trends in prosodic contours were observed in both requests and labelling: rising tones were more frequently associated with requests whereas falling tones appeared more often in conjunction with labelling.

Utterances for which clarification is requested have not always been considered an integral part of the ensuing contingent query sequence. To examine this issue, characteristics of children's utterances which were targeted for requests for clarification were examined to determine if they influence the nature of responses. Any relationship here would (a) establish the role of the initial utterance as part of the contingent query discourse unit, and (b) begin the documentation of that role. Twenty-six children (nine aged between two and three, eleven between three and four, and six between four and five) were videotaped in a one-hour free-play session. Twenty times during the session an experimenter signalled non-comprehension of a comprehensible utterance by asking ‘What?’. The queried utterances were categorized as serving one of three functions (interpersonal, child activity or environmental) and as occurring in one of three social contexts (eye contact, other social or private). Responses were classified as being either a no-response, repetition or revision. Results showed that social context interacted with response type: more no-responses and fewer repetitions occurred in the private relative to the eye-contact context. This interaction demonstrated that the initial utterance plays a role in response determination, and is therefore an important part of the contingent query sequence. Findings also indicated that response type changed with development.

The current study documented one child's earliest use of prepositions during her second year of life. The spatial oppositions up-down, on-off, in—out and over—under were first to be learned. These words were all used initially in non-prepositional senses (mostly as holophrastic, verb-like requests for activities) prior to prepositional usage. They were seldom omitted or misused. The prepositions with, by, to, for, at and of were learned later. Four of these were used to express at least two distinct case relationships, and some case relationships (instrumental and dative) were indicated by more than one of these words. These later learned prepositions were not used by adults or learned by the child as holophrases, but rather they were acquired as distinct lexical items by means of analytic learning strategies that employed some form of ‘extraction’ from adult phrases. These words were omitted and misused much more often than the spatial oppositions. Differences in the acquisition pattern of these two groups of prepositions were attributed to linguistic rather than to cognitive factors.

The language of mothers and fathers to their toddler sons and daughters at play with three different types of toy was analysed for both sentence types and structural elements. Two of the play contexts, dolls and vehicles, were highly sex-role stereotyped, and the third, shape sorters, was neutral. Few differences in speech were found as a function of either parent or child gender. Each of the three play contexts, however, elicited its own language pattern. With dolls, parental speech was characterized by a relatively large amount of language and a frequent use of questions and nouns, especially by contrast with vehicle play, which involved little language. Play with shape sorters elicited a high proportion of directives and attentionals and a relatively low variety in language. These results suggest that children who play frequently with dolls may receive more opportunities to learn and practise language than do children who select other toys for play. In this way early toy preferences may contribute to differential socialization by parents.

Infant temperament characteristics of activity level, task persistence and affect were measured in an eight-month longitudinal study of eight children. Discourse and pragmatic features of their mothers' speech, specifically total number of utterances and use of self-repetitions, directives and attention-getting devices, were also monitored. Pearson correlation analysis suggested complex interactions between maternal speech and infant temperament. In some cases a maternal speech pattern appears to maintain an infant temperament characteristic. In others, a child characteristic appears to maintain a maternal speech form. The results are used to argue that non-linguistic child behaviours may influence maternal speech and deserve further attention in investigations of the link between maternal speech and child-language development.

The phonological acquisition of a young girl whose parents spoke two dialects of Cantonese was examined. The father's dialect had a phonological distinction between initial /l/ and /n/ which was merged into /l/ in the mother's dialect. The child was followed bi-weekly for approximately one year. The results indicate that she acquired neither the mother's nor the father's dialect. Instead, she acquired [l] and [n] as freely varying allophones of a single phoneme. In the first months, [n] was the most frequent realization of the phoneme, with [l] becoming the most frequent one in later sessions. The results are interpreted as supporting the claim that children use all available input in acquiring language rather than limiting themselves to a primary language model.

This study investigated the use of stative, process, punctual, and non-punctual verbs by a child acquiring Japanese as a first language between the ages of 1;0 and 4;11 in an attempt to find evidence for two of Bickerton's (1981) proposed language acquisition universals, which form part of the language bioprogram hypothesis of language acquisition. As predicted by Bickerton's state-process hypothesis, it was found that all sampled present progressive verb forms occurred with process verbs while these forms were never used with stative verbs. Also, with only one exception, all omissions of present progressive forms occurred with the early use of ‘mixed’ verbs, i.e. verbs which behave syntactically as process verbs in Japanese but are nonetheless semantically stative. However, contrasting with Bickerton's hypothesis that children initially use the past tense to mark punctuality, no relationship between past tense use and punctuality was found.

Children's reported copying-without-deletion errors, like ‘Whose is that is?’, have often been interpreted as the result of a non-adult rule of subject–aux inversion. By contrast, this study presents a performance account of these errors and supports this account by investigating what factors impede children's performance. Using an elicitation task, 16 test sentences were evoked from 16 3- to 5-year-old children. In particular, errors appeared (1) when the subject NP contained a relative clause, (2) when the relative clause had an object gap, and (3) when the relative clause was long. Since errors occurred in response to some sentence types and not others, and the children who made copying-without-deletion errors produced at least four correct responses, these results were interpreted as support for the performance account.

The evidence for the Contrastive Hypothesis (Clark 1980, 1983a, b, 1987, Barrett 1978, 1982) is reviewed. An examination of data from the acquisition of object words, relational words and superordinate terms reveals little support for this hypothesis that young children automatically assume that every two words in their lexicons contrast. Further, theoretical problems with the positions that children assign words to semantic fields as they are acquiring them and that innovations are used to fill lexical gaps make these stances untenable.