Macrostructure, microstructure, and mental state terms in the narratives of English-Hebrew bilingual preschool children with and without specific language impairment

Children's bilingual status is important because the interest here is in narrative performance in both languages of bilingual children, in particular the within-subject, cross language comparisons. As Paradis (2010) has argued, there are some structures where performance differences will point to a temporary lack of opportunity for mastery, whereas other structures will be markers of underlying difficulties. We expect the discriminators to be language specific, depending on attested vulnerabilities for each of the languages involved. Narratives were examined for macrostructure (goals, attempts, and outcomes), microstructure (e.g., length, lexis, and morphosyntax), and mental state terms (MSTs). Thirty-one preschool children (TLD = 19, SLI = 12) retold stories accompanied by six pictures that were matched across content (Baby Birds/Baby Goats) and to the extent possible across languages (first language/second language) for macrostructure, microstructure, and MSTs in the framework of the Working Group on Narrative and Discourse Abilities in COST Action 0804 Language Impairment in a Multilingual Society: Linguistic Patterns and the Road to Assessment. The macrostructure results confirmed previous findings showing similar performance in both languages for children with TLD and those diagnosed with SLI. Consistent with previous findings on narrative abilities among bilingual children, microstructure analysis of verbal productivity, length of communication units, and lexical diversity distinguished children with TLD from those with SLI. An analysis of MSTs yielded more MSTs in children's second language, in particular more mental verbs. The most prevalent MSTs used in all narratives were early acquired perceptual and motivational verbs (see and want). Overall, distinctions between narratives of children with TLD and SLI were found primarily for microstructure features, where error analysis was particularly important in uncovering possible markers, especially in second languages.