Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Health and Rehabilitation Sciences


Janis Oram Cardy, PhD


Language in individuals with nonverbal learning disabilities (NLD) and autism spectrum disorder (ASD) has been described as semantically empty and impoverished, despite apparently average word knowledge. Here, inter-related studies explored semantic representations in adults with these disorders of social perception. Studies highlighted semantic integration, a form of gestalt perception in which new concepts are developed by connecting familiar terms in novel ways. Semantic integration was compared to vocabulary breadth, and to nonverbal gestalt perception, comparing clinical groups to each other and to adults without a diagnosis. Because weaknesses in gestalt perception have been seen in NLD and ASD, it was expected that the clinical groups would show difficulty with semantic integration compared to controls, but that vocabulary would not differ between groups.

Chapter 2 presents results from two surveys administered to investigate autism symptoms in adults with NLD, ASD, or no diagnosis. Results corroborated social perception impairments in both clinical groups, and adults with NLD had survey scores above thresholds for potential ASD. Chapter 3 found no differences between groups in breadth of vocabulary, as hypothesized. The NLD group had lower scores for tests of semantic integration and gestalt perception than controls. The ASD group, however, had equal or better scores than the other groups for semantic integration, an unexpected result potentially related to formulaic language. In Chapter 4, adults with NLD provided fewer meanings for polysemous words than controls, and scores for this measure were predicted by nonverbal perceptual reasoning. Results supported clinical observations that individuals with NLD are less likely to form links between unlike but familiar words, and suggested a specific cognitive underpinning for this difficulty.

Overall, there was little to no difference between clinical groups for either quantitative or observational data concerning vocabulary breadth, but quantitative differences were seen for underlying cognitive measures. Outcomes suggested different cognitive paths by which these adults arrive at similar destinations in regard to their linguistic strengths and weaknesses.