This work aims to analyse the specific contribution of social abilities (here considered as the capacity for attributing knowledge to others) in a particular communicative context. 74 normally developing children (aged 3;4 to 5;9, M=4·6) were given two Theory of Mind (ToM) tasks, which are considered to assess increasing complexity levels of epistemic state attribution: Attribution of knowledge-ignorance (Pillow, 1989; adapted by Welch-Ross, 1997) and Understanding of False-belief (Baron Cohen, Leslie & Frith, 1985). Subjects were paired according to their age and level of performance in ToM tasks. These dyads participated in a referential communication task specially designed for this research. The resulting communicative interchanges were analysed using a three-level category system (pragmatic functions, descriptive accuracy, and ambiguity of messages). The results showed significant differences among subjects with different levels of social comprehension regarding the type of communicative resources used by them in every category level. In particular, understanding of false belief seems to be the most powerful predictor of changes in the children’s development of communicative competence.

This paper explores the concurrent and predictive validity of the long and short forms of the Galician version of the MacArthur-Bates CDI (IDHC). Forty-two Galician-speaking children were longitudinally evaluated at age 1 ; 6, 2 ; 0 and 4 ; 0. On the first two occasions, the subjects' vocabulary and grammar skills were assessed through the IDHC. Simultaneously, lexical and grammatical measures were obtained from spontaneous speech samples. Standardized measures of general cognitive abilities (WPPSI-R) and receptive and expressive language (RDLS-III) were obtained at age 4 ; 0. Results showed high and significant levels of concurrent and short-term validity of the IDHC. Strong associations were found between lexical development at age 2 ; 0 and language scores two years later. These results coincide with those obtained with other CDI versions, and suggest that the IDHC is an effective and reliable tool.

Verbal interactions between three mothers and their young blind children, with special attention to the use of maternal directives, were examined. It was found that a simple analysis of maternal directives can be misleading. Thus, counting single occurrences of directives ignores the possible adaptive role that directive sequences may have in conversational interactions with blind children.

This study of the conversational interactions of three mothers and their infants—one who was sighted, one with low vision, and one who was congenitally blind—found similarities between the sighted and low vision dyads. However, the mother of the blind infant talked more and used more directives, and her directives contained more descriptions and were more likely to occur in clusters. The complex nature of directives and their adaptive role with infants who are blind are discussed.