This study explores the spontaneous acquisition of Italian inflectional morphology by three children (age range 1;4–3;0). Longitudinal, free speech samples are examined, focusing on the development of the morphological paradigms of Italian verbs, pronouns and articles. Data analysis is conducted using criteria appropriate to allow reliable cross-linguistic comparisons with data from English. By this means we evaluate the plausibility of a nativist, parameter-setting account of language development in Italian and English, as recently proposed for these two languages. Results show that the general developmental patterns observed in Italian are not significantly different from those found in English. These findings are not consistent with current interpretations of parameter-setting accounts of language development. Alternative explanatory models are discussed.

This study explores the role that gesture plays in the earliest stages of language learning. We describe how one-word speakers use gesture in combination with speech in their spontaneous communications, and interpret gesture presented in combination with speech in an experimental situation. Forty one-word speakers (ages 1;2.22 to 2;4.6) were videotaped in a free-play session which provided data on the child's spontaneous gesture and speech production. The children were also given a comprehension task in which the presence and absence of gesture were systematically varied in relation to speech. We found that (1) all of the children spontaneously produced gestures in combination with speech, and (2) all of the children were able to understand gesture when it was presented in combination with speech, not only when the gesture was redundant with speech but also when the gesture substituted for speech. These data suggest that, even at this young age, gesture naturally forms an integrated system with speech in both production and comprehension.

English has several classes of transitive verbs which can optionally appear without an undergoer. Despite their similar syntactic sub-categorization, there are at least three different semantic subclasses that allow undergoer omission. Information sources based on surface structure, for example, syntactic bootstrapping, cannot inform the child of the semantic representation of these verbs. The focus of this paper is the acquisition of a single English verb, eat. The transcripts of 40 children, who were audiotaped monthly from 1;0 to 3;0, showed that eat was the first member of this verb class to be acquired. Some 1276 eat sentences were analysed for the presence of overt undergoer arguments across levels of cumulative verb lexicon (CVL) size, and two discourse conditions: (1) UNDERGOER ACCESSIBLE and (2) OPEN (to undergoer omission). Results indicate that undergoer omission became associated with discourse conditions when CVL size rose above 75 types, at MLU approximately 2·4 and age approximately 2;3. This suggests that two-year-old children are sensitive to a relationship between undergoer omission and discourse context.

To examine the possible persistence of phonological selectional constraints on young children's lexical choices, the words attempted in the conversational speech of a longitudinal sample of 12 normally-developing preschoolers from age 2;0 to 5;0 were scored for syllabic length, presence of consonant clusters, and distribution of constituent phonemes. Except at the youngest ages, few developmental changes in target word characteristics were seen, and the observed differences were largely accounted for by syntactic, lexical, and pragmatic factors. The results suggest that selectional constraints persist only briefly in the course of language acquisition.

In this study, mental terms in mothers' and their children's speech at two and three years of age were studied in order to examine the relationships between maternal and child use. Nineteen mother and child dyads were videotaped for one hour on each of two days when the children were 2;0 and again for two one-hour sessions on separate days when they were 3;0, and mental terms were noted. The utterances in which mental terms were used were coded for function. Results supported the existing picture of children's mental term use. Few terms appeared at 2;0, but many were used at 3;0 with think and know predominating. Mental terms occurred more commonly in utterances used to regulate the interaction between the participants than in utterances referring to mental states. Children's mental term use mirrored that of their mothers. Further, mothers' use of mental terms for particular functions when their children were 2;0 predicted their children's use at 3;0. While allowing no conclusions about causation, our findings suggest that the development of mental state language, and thus presumably a theory of mind, is fostered by the linguistic environment. Specifically, it is argued that the tendency of mothers to focus their children's attention on mental processes by talking about them and, more importantly, by using utterance types which conceivably direct the children to reflect on their own mental states, is associated with children's use of mental terms.

Sociolinguists have investigated language mixing as code-switching in the speech of bilingual children three years old and older. Language mixing by bilingual two-year-olds, however, has generally been interpreted in the child language literature as a sign of the child's lack of language differentiation. The present study applies perspectives from sociolinguistics to investigate the language mixing of a bilingual two-year-old acquiring Norwegian and English simultaneously in Norway. Monthly recordings of the child's spontaneous speech in interactions with her parents were made from the age of 2;0 to 2;7. An investigation into the formal aspects of the child's mixing and the context of the mixing reveals that she does differentiate her language use in con-textually sensitive ways, hence that she can code-switch. This investigation stresses the need to examine more carefully the roles of dominance and context in the language mixing of young bilingual children.

Seventeen preschool (age range 2;10–3;6), 26 kindergarten (age range 5;5–6;7), and 26 fourth-grade (age range 9;5–10;5) children's knowledge structures were examined with a word association task and a match-to-sample picture task to determine whether or not children used slot-filler categories as a mediating structure between event-based and taxonomic knowledge structures, as proposed by Nelson (1985, 1986). In general, preschool children were able to provide event-based, but not slot-filler or taxonomic, relations; kindergarten children were able to provide event-based or slot-filler relations, but very few taxonomic relations; and fourth-grade children were able to provide all three relations. These findings support Nelson's hypothesis that taxonomic knowledge structures are derived from event-based knowledge structures with the aid of slot-filler categories.

This study had two purposes. The first was to examine age differences in the extent to which children infer and use a speaker's interpersonal goal to understand speech acts. To this end, the subjects responded to Do you have…? requests in a role-playing task. The goal behind such a request is implicit in the speaker's choice of noun phrase. We sought to determine whether children, like adults, use the noun phrase to infer the speaker's goal and thereby decide whether the request is a yes-no question or a directive. The second purpose was to examine age differences in the extent to which children select responses that carry implications appropriate to the speaker's interpersonal goal. To do this, Do you have… ? requests containing general category labels were addressed to the children when they possessed all the category exemplars and when they possessed only a few exemplars. A simple yes implies that the listener has nearly all exemplars and, therefore, is inappropriate for the speaker's goal in the latter situation. The subjects were six-, seven-, nine-, and eleven-year-olds and adults, with 12 subjects per age. Only the eleven-year-olds and the adults used the speaker's noun phrase and considered the implications of yes.

Pizzuto & Caselli (1992) claim that data from the morphosyntactic development of Italian-speaking children are inconsistent with nativist, parameter-setting models of language development. In the present Note it is argued that much of the data which Pizzuto & Caselli adduce is irrelevant to the specific hypotheses they are evaluating and that those data which are relevant fully support parameter-setting and linguistictheoretic models, contrary to their claims.

The HCRC dialogue database consists of over 700 transcribed and coded dialogues from pairs of speakers aged from seven to fourteen. The speakers are recorded while tackling co-operative problem-solving tasks and the same pairs of speakers are recorded over two years tackling 10 different versions of our two tasks. In addition there are over 200 dialogues recorded between pairs of undergraduate speakers engaged on versions of the same tasks. Access to the database, and to its accompanying custom-built search software, is available electronically over the JANET system by contacting liz@psy.glasgow.ac.uk, from whom further information about the database and a user's guide to the database can be obtained.