This longitudinal study of how preverbal infants communicate with their mothers utilized the situation in which the infant was seated in a highchair at lunchtime. This situation predisposed infants to use communication as a means, since they were often unable to achieve their goals without assistance. It was found that infants' communicative attempts were often unsuccessful; the present study focussed on how infants and mothers worked to establish the infants' intents after communication failures. In the preverbal negotiation of failed messages infants direct communicative behaviours to their mothers which their mothers fail to comprehend immediately, NEGOTIATIONS occur when mothers help infants make their intents clear. Negotiation episodes have four components: the infant's initial signal, the mother's comprehension failure, infant repairs and episode outcome. Changes in these components provide much information about how infants' communicative skills evolve during the transition to a linguistically based communication system. Negotiation episodes are contrasted with episodes called IMMEDIATE SUCCESSES in which the mother readily comprehends the intent behind the infant's signal, and MISSED ATTEMPTS in which the mother fails to pick up on the infant's signal. Taken together these three types of communicative episode reveal a degree of persistence and creativity on the part of the preverbal infant that is surprising in the light of prior research. Such episodes further reveal that the course of preverbal communication is NOT smooth.

One child's lexical development was followed from 1; 1 to 1; 7 to look for possible causes of early word overextensions. Lexical acquisition was observed for comprehension and production through the use of a parental diary and systematic language sampling. In addition, 16 nonsense words were taught in 10 experimental sessions over a two-month period to control for phonological effects. The results isolated six distinct factors behind early word overextensions. It is concluded that overextensions should not constitute the primary data for the construction of theories of semantic acquisition.

Observational/correlational research has established that children in stage 5 object permanence development use words that refer to the visible movement of objects, while it is not until stage 6 that they use words referring to the invisible movement of objects (e.g. gone). It is not clear, however, whether the cognitive competence evidenced by performance in the object permanence task is a prerequisite for these linguistic productions or whether they both emerge simultaneously from some underlying representation. The current study used a lexical training paradigm to teach object words, visible movement words, and invisible movement words to children at stage 5 and stage 6 object permanence development. Stage 6 children learned all three types of words equally well. Stage 5 children learned object and visible movement words, especially in comprehension, but did not learn the invisible movement word. These results indicate that the cognitive structures underlying performance in the object permanence task are genuine prerequisites for learning these types of words.

Mother–child (two-year-old) object labelling was studied in three natural discourse settings: (1) routine, caretaking; (2) free play; and (3) novel. Object labelling was found to be considerably more varied in these natural discourse settings than in experimental situations. While basic level tokens predominated in the free play context, they were significantly less prevalent in the routine and novel contexts. Additionally, subordinate level term usage was more common in the routine and novel contexts. The relation between mother and child labelling was also investigated and results indicated that context may be more important in determining level of labels than maternal modelling. Analyses of the discourse uses of non-basic level terms revealed that mothers were organizing the social-interactive context in ways that may facilitate child category formation.

This study examined topic discourse skills in 30 pre-operational level children interacting with their mothers in a free play situation. For each interaction, topic initiations/shadings were analysed for here-and-now, fantasy, and displacement topics. All turns (initiations, shadings and maintenance) involving these topic categories were also coded for communicative intent. Results indicated that children initiated/shaded more fantasy topics than their mothers, who initiated/shaded more displacement topics. Topic maintenance was significantly greater for fantasy and here-and-now topics than for displacement. Mothers initiated/shaded and maintained topics primarily by requests for all three topic categories. Children used more informatives to initiate/shade these topics, and maintained fantasy and displacement topics primarily by responses to mothers' questions. The results suggested a complex picture of children's management of discourse as it relates to topic and mothers' continuing role in facilitating their children's discourse skills.

This study assesses the role of frequency of input in the acquisition of the present perfect by Scottish and American children. Two questions were addressed: (1) Do adults speaking Scottish English use the present perfect more frequently in speech to children than those speaking American English? (2) If there is a difference in the frequency of input, how does this affect the course of development of this structure in the language of Scottish vs. American children? Cross-sectional data were collected from 12 Scottish and 12 American children aged 3; 0 to 6; 4 and from adults interacting with them in naturalistic settings. The data led to the conclusions that (1) Scottish adults use the present perfect construction in their speech to children much more frequently than American adults do; (2) Scottish children use the present perfect construction in their speech long before their American counterparts; and (3) frequency of input does play a major role in the timing and order of acquisition of the present perfect. However, its role appears to be an interactive one, in which it conspires with factors such as semantic, syntactic, and cognitive simplicity to make some forms easier to learn than others.

This paper reports on a cross-sectional acoustical investigation of the development of the voicing contrast in Thai word-initial stops. Voice onset time (VOT) productions associated with three bilabial stops (/b p ph/), three alveolar stops (/d t th/), and two velar stops (/k kh/) were measured in seven 3-year-old children, seven 5-year-old children, seven 7-year-old children, and seven adults. Within each age group, differences in mean VOT between homorganic stop consonants indicated that 3-year-olds have acquired all voicing contrasts except /b/ vs. /p/ and /d/ vs. /t/. By age five, children have acquired all voicing contrasts. However, as measured by differences in mean VOT across age groups, 5-year-olds still do not produce /b/ or/d/ in an adult-like manner. Issues concerning the order of acquisition and nature of development of stop consonants are discussed.

Two experiments are reported, where the development of pragmatic communication skills was assessed in a simple but realistic setting, so that even the youngest speakers could construct consistently accurate messages. It was shown that children of up to six years are specifically hampered by the speaker's use of left and right, but not by the use of named landmarks. Children were also allowed to create their own messages. The ability of the speaker to use language which would be valuable to a particular listener could then be measured. The results show a development between six and seven years in speakers' ability to select and use linguistic forms which are appropriate to listeners' receptive skills, though speakers could use objectively accurate information at an earlier age.

All uses of the connective but by narrating children 3; 6 to 9; 6 were classified as having primarily semantic function, pragmatic function, or as errors. The youngest children make the majority of errors, mistakenly using but when causal or precausal relationships exist. All children most commonly use but to signal semantic relationships such as semantic opposition or violation of expectation, although the oldest children use but proportionately more to encode more complex contrasts. However, children at all ages also use but for primarily pragmatic functions, to interrupt the flow of their narrative in order to insert relevant comments, monitor the listener's attention, or change topic. If discourse can be conceptualized as having multiple levels, but signals a change in level within a speaker's turn.