Language alignment in children with an autism spectrum disorder

This thesis examines language alignment in children with an autism spectrum disorder (ASD), a neurodevelopmental disorder characterised by impaired social understanding and poor communication skills. Alignment, the tendency for speakers to repeat one another’s linguistic choices in conversation, promotes better communication and more satisfying interactions (cf. e.g., Fusaroli et al., 2012). By corollary, deficits in alignment may adversely affect both communicative and affective aspects of conversation. Across three studies, I consider whether ASD children’s conversational deficits relate to disrupted patterns of alignment, and explore the mechanisms underlying this. In the first study, I adopt a corpus-based approach to show that syntactic alignment effects are observable in ASD children’s ‘real-life’ conversations, not just in an experimental context. The second study draws on research into the role of inhibitory control in communicative perspective-taking (Nilsen & Graham, 2009) to show that lexical alignment is not socially mediated in ASD. I develop this work in the third study, which highlights how, for ASD children, conversation can be compromised when lexical alignment is driven exclusively by priming mechanisms. Taken together, these studies advance our understanding of conversational deficits in ASD, and particularly how impaired social understanding affects ASD children’s language processing in dialogue. I conclude that, while ASD children have intact alignment, reduced social understanding may prevent them from ‘diverging’, which can be necessary to move a conversation forward (Healey, Purver, & Howes, 2014). More broadly, the thesis addresses questions of theoretical relevance to the study of alignment, by clarifying the contributions of unmediated (i.e., priming) and socially mediated (i.e., audience design) mechanisms to children’s alignment behaviour, both in ASD and typical development.