RESUMEN
BACKGROUND: Children who do not produce single words by the expected age have been described as 'late talkers' or as demonstrating 'late language emergence' (LLE). Although their short-term growth in vocabulary is often strong, longer-term consequences of LLE remain in dispute. It has been argued that the majority of school-age children who had LLE move into the average range for narrative production, though studies have not examined narrative comprehension. It has also been argued that school-age children with LLE score in the average range on standardized tests of syntax, though studies have not examined performance in conversational contexts. AIMS: This article compared school-age children with and without histories of LLE for performance on standardized narrative comprehension and production tasks, as well as the use of complex sentences and relative clauses in narration and conversation. Both complex syntax and relative clause use are reduced in children with specific language impairment (SLI), so these structures may be useful as indicators of linguistic weakness. METHODS & PROCEDURES: The participants were twenty-two 8-year-old children, divided into two groups. Eleven children who had been diagnosed with LLE at 30 months were compared with a control group of 11 children with typical development (TD). All participants completed a standardized test of narrative comprehension and production and a 10-min conversational sample. Both narrative and conversational samples were analysed for the number of complex sentences and relative clauses. OUTCOMES & RESULTS: Overall results indicated that children with a history of LLE did not have comprehension or production scores that were significantly different from the TD group on the standardized narrative test; nor did groups differ for production of complex sentences or relative clauses in narrative samples. However, a significant difference was found for the production of complex sentences in conversational samples, with the children diagnosed with LLE producing fewer complex sentences than the TD group. There was no difference between groups for relative clause use in conversation or in narratives. CONCLUSIONS & IMPLICATIONS: These data suggest that children with a history of LLE may exhibit age-appropriate performance on a standardized narrative test, but still lack the syntactic complexity of their TD peers in conversation. Assessments for school-age children with a history of language delay should include analysis of syntactic complexity in conversation to identify continuing weakness. Future research should examine use of other specific types of complex structures (e.g. infinitival and clausal complements) in this population, as well the feasibility of increasing complex sentence production through intervention. In addition, future studies should examine whether this decreased production of complex syntax in conversation is noted by naive listeners.