The Social Brain and Language Learning

Background: Research on typically developing infants during the initial stages of language learning suggest that social interaction is a necessary component for learning. Work has shown, for example, that infants exposed to a new language during an early sensitive period for phonetic learning, can learn phonemes from the new language if exposed socially during interaction with a live human being. Infants exposed to the same language material, in the identical setting and on the same schedule but via television, learned nothing. This stark contrast between learning under social and nonsocial conditions led to the “social gating hypothesis (SGH),” the notion that infants require social interaction to enact other computational processes (“statistical learning”) when exposed to natural language.

Objectives: I will describe the research stemming from the SGH, which includes research on children with autism spectrum disorder (ASD) using phonemes, words, and the motherese style of speaking to infants to probe possible early markers of risk in children with autism.

Methods: The studies described utilize both behavioral measures on typically developing infants and children with ASD as well as brain measures (EEG and MEG).

Results: The results suggest, in typically developing children, that early language learning depend on a social context and social interaction. In children with ASD, our results hold promise that our brain measures of language processing may some day contribute to the development of reliable “biomarkers” for children at risk for ASD.

Conclusions: I discuss the meaning and implications of the claim that language learning is “gated” by the social brain.