This study compares the prosodie modifications in mothers' and fathers' speech to preverbal infants in French, Italian, German, Japanese, British English, and American English. At every stage of data collection and analysis, standardized procedures were used to enhance the comparability across data sets that is essential for valid cross-language comparison of the prosodie features of parental speech. In each of the six language groups, five mothers and five fathers were recorded in semi-structured home observations while speaking to their infant aged 0; 10–1;2 and to an adult. Speech samples were instrumentally analysed to measure seven prosodic parameters: mean fundamental frequency (f0), f0-minimum, f0-maximum, f0-range, f0-variability, utterance duration, and pause duration. Results showed cross-language consistency in the patterns of prosodic modification used in parental speech to infants. Across languages, both mothers and fathers used higher mean-f0, f0-minimum, and f0-maximum, greater f0-variability, shorter utterances, and longer pauses in infant-directed speech than in adult-directed speech. Mothers, but not fathers, used a wider f0-range in speech to infants. American English parents showed the most extreme prosodic modifications, differing from the other language groups in the extent of intonational exaggeration in Speech to infants. These results reveal common patterns in caretaker's use of intonation across languages, which may function developmentally to regulate infant arousal and attention, to communicate affect, and to facilitate speech perception and language comprehension. In addition to providing evidence for possibly universal prosodic features of speech to infants, these results suggest that language-specific variations are also important, and that the findings of the numerous studies of early language input based on American English are not necessarily generalisable to other cultures.

Speech samples from eight British female adults addressing young children and another adult were analysed in terms of vocal pitch. Increases in vocal pitch when addressing young children were observed but were smaller than the increases noted by North American researchers, and variability across speakers was marked. Some adults only slightly raised their vocal pitch to young children. Pitch increases were generally more marked for mean as against modal pitch measures, and for free speech as against reading aloud conditions.

We propose (contra Hyams, 1986) that the optional subject phenomenon in early child language arises because children have not yet acquired the morphological elements (primarily modals and tense) necessary to distinguish Subject-Taking (ST) verbs (e.g. finite verbs) from their non-Subject-Taking (NST) counterparts (e.g. infinitives). Unaware of this distinction, children are able only to observe that verbs sometimes occur with subjects and sometimes without. We show that our proposal makes a number of developmental predictions which we then test with the help of longitudinal data from three children. We conclude that: (i) There is no systematic morphological distinction between ST and NST verbs during the optional subject stage (OSS). (2) The emergence of the distinction between ST and NST verbs is gradual rather than sudden. (3) There is variation from child to child in terms of which morphologically-defined subclass of verbs is first recognized as subject-taking. (4) There is no link between the emergence of modals or contracted copulas and the end of the OSS.

Hirsh-Pasek, Treiman & Schneiderman (1984).and Demetras, Post & Snow (1986).have recently suggested that certain types of parental repetitions and clarification questions may provide children with subtle cues to their grammatical errors. We further investigated this possibility by examining parental responses to inflectional over-regularizations and wh-question auxiliary-verb omission errors in the sets of transcripts from Adam, Eve and Sarah (Brown 1973). These errors were chosen because they are exemplars of overgeneralization, the type of mistake for which negative information is, in theory, most critically needed. Expansions and Clarification Questions occurred more often following ill-formed utterances in Adam's and Eve's input, but not in Sarah's. However, these corrective responses formed only a small proportion of all adult responses following Adam's and Eve's grammatical errors. Moreover, corrective responses appear to drop out of children's input while they continue to make overgeneralization errors. Whereas negative feedback may occasionally be available, in the light of these findings the contention that language input generally incorporates negative information appears to be unfounded.

Recent controversy about the innateness of language has awakened interest in the genetic basis of linguistic development. In this study, we analysed more extensively data from Mather & Black (1984).in order to test the hypothesis that the speech articulation of monozygotic (MZ) twins would be qualitatively more similar than that of age- and sex-matched dizygotic (DZ) twins. Analyses revealed that 4–year-old MZ twin pairs were significantly more likely to misproduce the same sounds on an articulation test than were DZ twin pairs, and that DZ twins were no more likely to share errors than were children who were both genetically and environmentally unrelated. There was no evidence that MZ twins made more similar errors than DZ twins, and indeed it was difficult to make this determination since only broad categories of error (substitution, distortion, omission) were available for analysis. The greater amount of genetic material shared by MZ twins, and the presumably more similar morphology of their speech mechanisms, may have caused certain sound patterns to be of more nearly equal difficulty for both members of a MZ pair. However, these findings need to be confirmed with phonetically more detailed analyses.

The evolution of young children's categories, as measured by category name production, was studied. The focus was on categories whose names initially were overextended. Four children (initially aged 1; 1 to 1; 3) were visited at home twice a week over a four- to six-month period, until they were approximately 1;7. The play sessions included eli-citations of the names of specially chosen objects. Results indicated that, as predicted, four sequences of category evolution were found, formed by the intersection of two factors: overlap vs. mutual exclusivity and first re-assignment separate vs. first re-assignment joint. As expected, most sequences involved initial overlap, and the choice of first reassignment option varied as a function of the relationship (hierarchical or non-hierarchical) between the initial child-basic category and the new category.

The Acquisition Of Complement Phrasal Constructions In Korean Is Examined In Spontaneous Speech Data From Two Children, Who Were Observed From One And A Half To Three Years Of Age. In Spite Of Typological Differences Between English And Korean, Both Syntactic And Semantic Characteristics are found to be shared by children acquiring complement constructions in the two languages. however, certain language-specific features of korean complement structures make it possible to address theoretical points concerning the structure of infinitival complements which cannot be resolved with the acquisition data on English. The error pattern in the acquisition of certain ‘subject-equi’ verbs in Korean poses problems both for LEG and GB accounts of the constituent structure of infinitival complements and the acquisition of those constructions. On the basis of the Korean data, I propose that base-generated VP complements are acquired first, with semantically motivated reanalysis of previously acquired infinitival complement structures occurring at a later stage.

This study investigated the relationship between control and coreference using an act-out task involving 81 children ranging in age from 3;1 to 8;0 and eight adults ranging in age from 30 to 55. The results replicated previous findings in revealing five developmental stages in children's interpretation of PRO, an empty pronominal element. A significant relationship was observed in the patterns of children's interpretation of forwards sentences containing PRO and those containing overt pronouns. However, there was no relationship in the development of restrictions on control and restrictions on coreference.

We report an experiment designed to identify how contextual information can influence children's performance on an experimental task involving temporal terms. Grain (1982) reported improved performance on a comprehension task when subjects were provided with contextual information, and he suggested that the improvement was due to satisfaction of presuppositions. However, this contextual information might have served to simplify task demands by providing prior information concerning an important aspect of the task. The present study distinguishes these factors by incorporating contextual information into the subordinate clause of the test sentences in a comprehension experiment (to satisfy presupposions) or into the main clause (to provide comparable prior information without satisfying presuppositions). We conclude that contextual information results in a significant improvement only when such information can be used to satisfy presuppositions.

The distinctions between the mental terms, know, think and sure was examined in an experiment with 60 children between three and six years of age. The children were required to find an object hidden in one of two places. Their only clues were two statements involving contrasting mental terms, with each statement referring to one of the possible hiding places. Results showed a significant improvement with age for the know-think and sure-think contrasts, with think treated as a less reliable index of location than both know and sure by four to five years of age. No change with age was found for know-sure contrast. It is concluded that by four to five years of age, children recognize the function of mental terms to express degrees of certainty, and that this understanding is probably not based on the factive properties of the terms.

This study examines how children understand the argumentative function of the French connective même (‘even’). Two completion tasks, related to the argumentative properties of the morpheme, were used: (1) to infer the conclusion of an ‘even’ sentence, and (2) to infer the argument position. Two main factors likely to influence the comprehension were investigated: the semantic context, and the syntactic form – affirmative vs negative. The argumentative function of même was globally mastered at the age of eight, but comprehension depends on semantic context. Performances were better for negative sentences, which suggests that negation plays a facilitating role; the specific properties of the negative expression pas même/même pas (‘not even’) in discourse are discussed.

Elsewhere we have argued on the basis of cross linguistic studies of directionality effects on anaphora in child language, that there is no universal ‘forward directionality preference (FDP)’; rather such a preference is linked to languages with specific grammatical properties. Although such a preference has been attested in English acquisition, matched experimental designs in Japanese, Chinese and Sinhalese, for example, do not show this effect. In this paper we argue that current attempts to show that forward directionality effects can also be induced in Japanese acquisition do not succeed in supporting the FDP. Specifics of the design of stimulus sentences in these experiments vary previous cross-linguistic designs so as to favour forward directionality on independent grounds, and confound cross linguistic comparisons. They in fact support a universal structure dependence in children's hypotheses about directionality of anaphora’.

This article challenges the position of Clark (1988) with regard to Contrast. The evidence cited in support of this theory from adult languages is far from clear, and the evidence coming from child language admits of alternative explanations. In addition, the ‘logic’ of assuming Contrast leads to insurmountable difficulties in the acquisition of irregular forms and allomorphs. An alternative to Contrast is proposed. This draws on the drive towards the adult system, the development of non-linguistic concepts, the acquisition of language in context, and the use of the Co-operative Principle by participants in conversational exchanges to account for those features of language and language acquisition that Contrast was designed to explain.

Murray & Trevarthen (1986) experimentally manipulated the behaviours of mothers and infants in an attempt to determine who is responsible for the fine temporal patterning during vocal engagements. In the present note it is argued that the measures of content and style of maternal talk used by Murray and Trevarthen do not directly indicate the ways in which the behaviours of the partners are combined on a moment-to-moment basis.