UC San Diego

Bilingual children’s language abilities cannot be fully understood by examining one language in isolation, because their linguistic knowledge is distributed across two interacting languages. One way to examine these cross-linguistic interactions is through cognates, or words that share meaning and form across languages. Prior research shows that bilingual children often process cognates more accurately or efficiently than noncognates, but the magnitude of this facilitation varies by child-level factors, word characteristics, language pair, task demands, and cognate identification methods. This dissertation examines cognate facilitation across increasingly complex assessment-relevant tasks: single-word vocabulary, sentence repetition, and story retell. Chapter 1 investigates whether Spanish-English bilingual preschoolers with and without parent-reported language concerns show a cognate advantage on a receptive English vocabulary task, and whether effects differ depending on subjective versus objective cognate classification. Chapter 2 uses a researcher-designed sentence repetition task to test whether cognate facilitation extends to target word accuracy and broader sentence-level performance in Spanish-English and Hebrew-English bilingual children. Chapter 3 examines whether cognate-rich story retell tasks support target word production, macrostructure, and language sample measures. Together, the chapters use cognates as a window into how bilingual children’s languages interact during clinically relevant language tasks.