Parental speech to typical and atypical populations: a study on linguistic partial repetition

Parents often use partial self-repetitions with variation in successive utterances (e.g., Want to get your ball? Get your ball? Do you want to get your ball?). Such ‘variation sets’ contain latent distributional information about the building blocks of language and are predictive of children's lexical and grammatical structures. Because these properties in parents of atypically developing children are virtually unknown, we compared for the first time variation sets in parental speech directed to toddlers with Autistic Spectrum Disorders (ASD), Down Syndrome (DS), and a baseline group of Typically Developing toddlers (TD). In Study 1, we analyzed transcripts of mothers' child-directed utterances during naturalistic dyadic play interactions. While children's mean developmental age was the same across the three groups, we found that measures of partial repetitions in child-directed speech were larger in the ASD than in the DS and typical groups. In Study 2 we also found that these larger measures in the ASD group were mainly driven by the mother, as opposed to the father. Because partial repetitions decrease with chronological age of the child in typical groups, and the atypical children were older than the TD group, our findings suggest compensating modes of communication in parental speech to atypical populations, especially the ASD group. The study validates the extension of structural/statistical analyses of language to compare parental communication to typical and atypical populations.Accepted versionThis work was supported by the NAP Start-up Grant M4081597 (G.E.) from Nanyang Technological University Singapore. The founder agencies had no role in the conceptualization, design, data collection, analysis, decision to publish, or preparation of the manuscript