A cross-linguistic comparison of cognate production in bilingual children with and without language impairment

Purpose: The current study examined if bilingual children (English-Spanish) with language impairment(LI) and children in the low typical(LT) range display a cognate advantage as their typically developing(TD) peers do. Given the literature we posed two hypotheses; on one hand, learning cognates may be easier for bilingual children with language impairment over typically developing children, as their shared representations lend to overlap in input. Conversely, it is possible the children with SLI would exhibit a cognate disadvantage given that in early language development children reject lexical units with high competition. Method: We examined whether 117 Spanish-English bilingual children (5;0 to 9;11) displayed a cognate advantage in oral production relative to their typically developing peers. The cognate and noncognate items were derived from the English and Spanish versions of the Expressive One-Word Picture Vocabulary Test. Children’s average proportion of cognate and noncognate responses were compared across ability groups. Results: TD bilingual children exhibited a cognate advantage, while the bilingual children with LI exhibited a cognate disadvantage. TD bilingual children produced a significantly higher proportion of cognates across their two languages, while LI children produced most of their cognates in Spanish only. The LT children performed similarly to the LI group in terms of overall proportion correct of cognate pairs over noncognate pairs, but performed similar to the TD group in terms of the language of response (only English, only Spanish, or both languages) of the cognate pairs. Conclusion: Consistent with our second hypothesis, children with LI show a cognate disadvantage, while TD bilingual children show an advantage for cognate production. As expected, LT children’s performance fell between the LI and TD groups. We discuss the theoretical and clinical implications of these findings.