“Tobby”, “swabwee”, and “strawberry”: Speech productions by toddlers in a linguistically-diverse community
Abstract
While there is growing awareness of the need to account for bilingualism in clinical assessments, there is a dearth of work on speech development adequately considering how input diversity, including bilingualism and accent variation, might affect toddlers’ productions. We recorded 2- and 3-year-olds with varied profiles of language input: (1) monolingual single-variety(Canadian English), (2) monolingual multiple-variety (additional English dialects/accents), and (3) bilingual. Target words were picture-elicitedand spontaneously-produced. In ongoing complementary studies, we are investigating whether input diversity helps account for extensive variation across children in early productive patterns and in their degree of intelligibility to adult listeners. Thus far, phonetic transcription demonstrates speech sound distortions, substitutions, and omissions common to young children, and interestingly, several reflections of patterns uncommonly-reported in more linguistically-homogeneous samples (e.g., backing, e.g. [kh] for /t/; frication, [v] for /b/; substantial distortions of vowels, [o] for /ɑ/). In an open-set transcription task presenting this corpus to Canadian English-speaking adults, we evaluate the hypothesis that deviations from canonical forms produced by monolingual single-variety toddlers are more predictable and interpretable to listeners than those by toddlers with varied input. Results will yield novel insights into speech development in diverse populations, with both theoretical and clinical implications.