[PS-1.25] 2.5-year-olds show no own-voice advantage: Implications for the nature of children's early lexical representations

Cooper, A. , Fecher, N. & Johnson, E.

University of Toronto

How do children mentally represent words? Toddler word recognition studies nearly always test children on unfamiliar adult voices. But we know that child and adult vocal tracts are quite distinct, and that children's pronunciations can differ dramatically in systematic ways from adult pronunciations. Would children recognize words better if they were tested on a child's voice? Or better yet, if words were presented in the child's own voice? In this study, 40 mother-child dyads recorded a set of words commonly known by 2.5-year-olds (e.g., strawberry). The 2.5-year-olds returned to the lab 1-4 weeks later and completed an eye-tracking task, where they viewed two familiar objects and heard a single word label. Each child was tested on labels produced in their own voice, their own mother's voice, an unfamiliar child's voice and an unfamiliar mother's voice. Results revealed better word recognition with adult than child voices, with performance best for productions from the child's own mother. No own-voice advantage was found; indeed, children demonstrated better word recognition for an unfamiliar adult talker than their own voice and comparable performance on their own and unfamiliar child's voice. These findings suggest that speech perception does not appear to be based on own-speech representations.