First Language

Children with specific language impairment acquiring German and English were compared in terms of their use of articles, third person singular verb inflections, and copula forms. The German- speaking children showed higher degrees of appropriate use of third singular inflections and copula forms, but not articles. Verb errors in the form of bare stems outnumbered other types of errors even in the data for German. Of the four accounts of grammatical deficits in specific language impairment considered, one based on the morphological richness of the target language appears most consistent with the findings.

Previous research suggests that Estonian mothers of 2-year-olds are more interested in the control of their children's behaviour than in prompting conversational participation. The present study aims to investigate whether the finding holds true also for Estonian mothers of 4-year-old children. In addition, the amount and type of regulatory language used by mothers with 2-year-olds and 4-year-olds in two activity settings are compared: during meals and in puzzle-solving activities. Maternal utterances aimed at directing children's attention, physical activity and verbal activity were coded (imperatives, declaratives, questions). To establish the level of children's linguistic abilities, MLU and the longest utterance were calculated. The results showed that, although mothers directed more regulatory language to younger children than to older children, the pattern of using regulative language was the same with both age groups. Mothers were concerned chiefly with regulating physical activity and elicited little conversation. The preferred sentence type was imperative. In the younger age group, significantly more regulation of children's attention and physical activity occurred during puzzle solving than at meals. Some pecularities of Estonian language and culture are discussed that might determine the style of regulatory language used by Estonian mothers.

The present longitudinal study examined the use of verbs at the early stages of expressive language development by children with specific language impairment (SLI). Three children with SLI and their younger normal language learning siblings were videotaped during mother-child interactions in the home over a 2-year period. At the beginning of the study both the child with SLI and the younger sibling appeared to be in the same stage of expressive language development as measured by mean length of utterance (MLU). Thus, the younger sibling acted as a control for the child with SLI. Results indicate on the one hand no consistent significant difference in the proportion of verb types used by children with SLI and their younger siblings across the developmental period studied. On the other hand, the younger siblings all used significantly greater proportions of verb tokens than their impaired SLI brothers after a particular point in development. Comparison of particular verbs used suggest that children with SLI have difficulties applying the verbs in their lexicon across contexts. The normal younger siblings used a wider variety of verb forms in their constructions containing verbs and they also used similar forms for different verbs which the children with SLI did not do or did so to a lesser extent. The findings are discussed and, in particular, two hypotheses are proposed: an `SLI critical mass' hypothesis following the work of Marchman & Bates (1994) and an 'Extended Verb-island' hypothesis following the work of Tomasello (1992).

Data from several languages, including Swedish, have shown that children with specific language impairment (SLI) have particular difficulties with verb morphology. In the present study, the verb morphology of six Swedish children with SLI was further explored in longitudinal and cross-sectional data. It was found that the children with SLI had significant problems with the use of auxiliary and modals, compared with younger MLU-matched controls. Although the children with SLI omit auxiliary and modals less frequently with increasing age, their levels of omission are not compatible with their MLU levels. The results are discussed with reference to recent accounts of the underlying nature of SLI.