An exploration of pragmatics in autistic spectrum disorder and normal adults and children

This thesis investigates pragmatic abilities in the language of people with autistic spectrum disorders (‘Autistic Disorder’ as per APA, 2010) and of typically developing children. The thesis introduces a framework describing the levels of pragmatic abilities, and uses this framework to discuss the existing literature, before using it to discuss a corpus of clinical-pragmatic language. The literature review and naturalistic study reveal mixed evidence for pragmatic impairment, in particular revealing an absence of evidence for impairment to high-level pragmatics inferences known as particularised conversational implicature. The thesis then reports a set of experimental studies, using a method adapted from Surian, Baron- Cohen and Van der Lely 1996), with adults with autistic spectrum disorders, typical adults, and typically developing children. These studies showed a significant difference between groups, suggesting an autistic deficit in implicature that is compatible with the notion of a developmental delay. These results support existing studies that posit a pragmatic deficit in autistic spectrum disorders but conflict with existing research on the scalar implicature in autism; the reasons for the differences in results are discussed. The results also support existing studies that show a slow developmental trajectory for implicature in typically developing children; possible methodological reasons for this apparent slow trajectory (i.e. task demands) are discussed. Although the studies tested both IQ and cognitive models of utterance processing as predictors of performance, these explanations only proved useful with typical adults. The thesis discusses alternative, linguistic theories that may be useful in future experimentation: in particular, linguistic-pragmatic Relevance Theory (Sperber & Wilson, 1986/1995)