The achievement of labelling was investigated in a longitudinal study of one mother–infant dyad, using video-recordings of their free play in a period between 0; 8 and 1; 6. Analysis of joint picture-book reading revealed that this activity had very early on the structure of a dialogue. The child's lexical labels might be regarded as more adult-like substitutes for earlier communicative forms that he had utilized in the dialogue. These were smiling, reaching, pointing and babbling vocalizations, all of which were consistently interpreted by the mother as expressing the child's intention of requesting a label or providing one. Participating in a ritualized dialogue, rather than imitation, was found to be the major mechanism through which labelling was achieved.

Twelve one-word-stage children (aged 1; 4 to 2; 0) were given minimally contrasting two-word instructions, e.g. kiss duck, kiss car, bang duck and bang car. Since non-linguistic cues were eliminated, comprehension involved making non-syntactic inferences from the word combinations. The children could respond correctly to some of the instructions, and even carried out some unfamiliar activities such as tickle book. Observations of mother–child interactions revealed play routines that may help the child learn to understand more than one word at a time.

Children's imitations were analysed as a function of parental speech acts for six children in early Stage I of language acquisition. The relative frequency with which children imitated mothers reflected the relative frequency with which mothers imitated children (Spearman rank correlation = 0·77). Although parents' imitative expansions could all be categorized as having primary speech act functions (e.g. request for information) from the parents' point of view, expansions constituted a separate class of speech events in terms of children's responses. The children imitated imitations far more frequently than non-imitative speech acts in the same category. These findings suggest that individual differences in children's propensity to imitate may arise from the degree to which parents provide a model of imitation as a speech act.

Much of adult communication is carried on at the subtle level of indirect utterance meaning, where the speaker's intent is not literally expressed in his or her utterance. This work investigates the young child's ability to respond appropriately to the intended, as opposed to the literal, meaning of one class of such utterances, requests for action. The data were the responses of five children between 1; 7 and 2; 4 to the direct and indirect requests for action their mothers produced during a natural play session. All the children responded with action to requests for it regardless of how subtly the requests were expressed. The apparent ability of two-year-olds to deal with indirect speech acts is surprising in that the understanding of such utterances by adults is presumed to depend on a combination of sophisticated rule systems. Rather than grant the young child such knowledge, the alternative of an action-based response heuristic is proposed.

This study investigated 3-, 5- and 7-year-old Tagalog-speaking children's mastery of agent-focus (active) and patient-focus (passive) sentence structures. In contrast to the usual English developmental pattern with actives and passives, the Tagalog children generally showed better mastery of patient-focus sentences than agent-focus sentences. These results were attributed to the children's strategy of interpreting the first noun of a sentence to be the agent of the action. The structure and usage of these sentence types in Tagalog do not appear to provide linguistic pressures favouring such a strategy. Thus these results are consistent with the view that such a strategy might be non-linguistic in origin.

Spontaneous full passives and related constructions from 234 children aged 2; 0 to 13; 11 and elicited passives from 262 college students were analysed. Full passives were classified as REVERSIBLE (The dog was chased by the girl), INSTRUMENTAL NON-REVERSIBLE (The lamp was broken by (or with) the ball), or AGENTIVE NON-REVERSIBLE (The lamp was broken by the girl). The Agentive non-reversible did not appear until after age 9; and until age 11 no child produced both Reversible and Non-reversible passives. All the children used the passive in a semantically restricted way (but not in the same way). The possible developmental course of the full passive was traced for children who initially used only Reversible passives versus those who initially used only Instrumental non-reversible passives.

To explore the relationship of language use and speech adaptation to role and context the spontaneous speech of a boy aged 5; 6 was recorded in three different situations: playing alone, playing with a friend of the same age, and playing with his mother. Several analyses were made of the speech transcripts to examine the relationship between context and language use. The nature of play episodes was also investigated. The findings suggest that role play has an important function in the development of a child's ability to communicate effectively, and that a child's awareness of his own role and the expectations he has concerning social interactions lead to modification in language use.

This paper reports the results of a preliminary study dealing with the ways in which children between six and seven years of age organize spoken language. In particular it deals with aspects of the temporal and segmental structure of polysyllabic English words containing the syllable C/jul/ (e.g. pediculous). On the basis of data presented, it is suggested (a) that in order to meet certain phonological conditions it is the relations between syllable (and not segment) durations which must undergo modification towards the adult duration values, and (b) that the achievement of such values is accompanied by some quasi-systematic alterations in phonological rules.

Three groups of children, aged 5, 7 and 11 years, were tested in a clause-memory task, in order to investigate the role of syntactic and semantic factors in children's recall and processing of spoken continuous prose. The children listened to a story, which was interrupted at intervals during which they were asked to repeat verbatim the last sentence they heard. The test-sentences always consisted of two clauses: either of the main–main type, or of the main–subordinate or subordinate–main types. The prose material was also either normal prose, or semantically uninterpretable but syntactically normal prose. The results suggest a developmental sequence in which the youngest group's performance is dominated by semantic factors. They do not show the adult clause-memory effects, their recall is dominated by the main clause, and their recall of normal prose is not clausally segmented. The two older groups' performance more closely approximates that of adults.

In order to examine the extent to which adult judgments of first words depend on visual and auditory cues, spontaneous utterances were collected for boys and girls of 1; 5, 1; 6 and 1; 10 as well as adults naming the same toys. Vocalizations clearly appropriate to the original visual context were dubbed onto new contexts varying in the number of phonetic features (0-2) by which the first phoneme of the utterance differed from the initial phoneme labelling the visual context. College students' perceptual accuracy declined with degree of mismatch, although the identical speech occurred in each context. The older the speaker, the less perception was affected by visual context. Whereas the meaning of child speech was judged by visual cues, adult meanings were judged mostly by auditory input. Biases in perceiving early speech may prompt reconsideration of current methodology and theory in early language acquisition.