A study of idiom comprehension in children with semantic-pragmatic difficulties. Part II: Between-groups results and discussion.

Using a play-based methodology and a symptom checklist, this study investigated idiom comprehension in 26 children aged between 6-11 who were considered to have semantic-pragmatic difficulties. This group was compared with two groups of mainstream children and a group of children with (other) language disorders not primarily of a semantic or pragmatic nature. The results indicate that the children with semantic-pragmatic difficulties did, as a group, demonstrate significantly fewer appropriate idiomatic interpretations and significantly more inappropriate interpretations than did any of the other three groups. However, the higher level of inappropriate scores in the semantic-pragmatic difficulties group reflected a large number of 'fuzzy' actions rather than significantly higher rates of literality. This may indicate an awareness among these children that the literal meaning is inappropriate in the absence of adequate idiom vocabulary. It may also reflect difficulty in retrieving known idioms from memory and/or in selecting the most appropriate meaning from several possibilities in context. Despite relative weakness, the children with semantic-pragmatic difficulties displayed appropriate interpretations considerably more often than they evidenced inappropriate ones. Within-group analysis reveals that the children diagnosed with Asperger syndrome or high-functioning autism performed less well than did those diagnosed with 'semantic-pragmatic disorder'. Nevertheless, both of these subgroups encompassed a considerable range of comprehension ability with regard to the 12 common idioms tested. Analysis of the play task performance and symptom checklist suggests that this variation probably reflects differences in the critical semantic and pragmatic skills underpinning idiom comprehension. These include flexibility of thought, theory of mind, attention to context, prosody and overall coherence, as well as the ability to integrate world knowledge and current contextual information to guide inferencing. In combination with definition task data and broader knowledge of symptomatology, the play task may be used to identify the sites and sources of idiom comprehension breakdown in individual children. It is probable that idiom comprehension in this group of children with semantic-pragmatic difficulties was facilitated by the inclusion of only concrete idioms and by the visual support provided by the play set. Nevertheless, the moderate to age-appropriate ability displayed by this group in this context suggests that the characterization of children with semantic-pragmatic difficulties as predominantly literal needs to be revised. Rather than seeking a blanket characterization of an essentially heterogeneous group, it may be more useful to consider idiom comprehension a secondary manifestation of semantic and/or pragmatic difficulties. Since the children who exhibit these difficulties vary in symptomatology and, probably, in aetiology, it follows that they will also vary with respect to idiom comprehension and the reasons for its breakdown.