The use of context in pragmatic comprehension by specifically language-impaired and control children
The purpose of this paper is to explore how a theory of pragmatic comprehension, relevance theory (Sperber and Wilson 1995), can be used to examine contextual processing in pragmatic comprehension. It will present data from a task that explores pragmatic comprehension skills, within this theoretical framework, of 17 SLI children (5; 11-10; 10), 11 older controls (7;11-9;11) and 30 younger controls (4;0-5;11). The children were asked questions based on a storybook, that required different degrees of contextual processing, as defined by the theory. Reference assignment questions required the children to provide a referent that is verbally/pictorially given, hence the least demanding in terms of contextual processing. Questions that require the resolution of semantic underdeterminations via the process of enrichment are contextually more demanding than reference-assignment question. Questions requiring the resolution of an implicature require the greatest contextual processing, hence these are deemed pragmatically the most demanding questions. A study by Ryder and Leinonen (i.p.) of three-to five-year-old normally developing children has shown that the theory and the characterization of questions in terms of pragmatic complexity have developmental validity. This finding is further supported by the current study. It was found that the pattern of performance across the three question types was very similar for the SLI group and the younger controls and that this reflected the contextual complexity of the question. The older controls had the greatest number of correct answers to all question types when compared to the younger controls and the SLI group, and this difference was particularly evident in relation to the pragmatically more demanding question types (enrichment and implicature questions). When the SLI group was divided into "younger" (5;11-7;11) and "older" (8;0-10;0) groups, it became evident that the older SLI children performed worse than their chronological age peers on all three questions types, most significantly in relation to the enrichment and implicature questions. This shows that the older SLI children had difficulty utilizing the context in the two question types that required the greatest contextual processing. This trend was also evident for the younger SLI children when compared to the younger controls. However, here the differences were less pronounced for the enrichment and implicature questions, showing that the younger children, irrespective of their language functioning, had difficulty with the more complex contextual processing associated with these question types. It was also shown that the SLI children's grammatical language ability was not an indicator of success in answering the questions in this study.