In everyday conversation, people naturally coordinate or 'align' their language at a variety of different levels.
The interactive alignment account of dialogue assumes that coordination is achieved automatically and unconsciously through simple, low-level mechanisms of repetition, imitation and priming, rather than through high-level processes of negotiation and modelling of mental states. This raises the possibility that people with conditions characterised by poor conversational ability may be able to develop their skills through relatively simple low-level mechanisms.
Children with autism have been shown to align their language patterns (syntax) with a conversational partner on a highly structured task; I am currently investigating whether alignment is also noticeable in their natural conversation. Relatedly, I am also exploring the possibility of creating a system which provides alignment with its interlocutor, and which could therefore serve as a therapeutic tool to increase coordination in the conversations of children with autism. I enjoy hearing from anyone who is interested in my research, so please drop me a line if you have any questions about this, or if you'd like to take part in a study (z.hopkins@sussex.ac.uk).