Semester of Graduation

December 2022


Master of Arts (MA)


Communication Sciences and Disorders

Document Type



Background: Vocabulary composition and word-learning biases are closely interrelated in typical development. Such word-learning biases are influenced by perceptually and conceptually salient word features, including imageability, concreteness, iconicity, and attention to shape. Autistic children often have delayed language acquisition, but there is currently little research examining the underlying mechanisms autistic children use to acquire words. The current study aimed to examine the noun composition of autistic children across a range of vocabulary sizes by examining associations between expressive noun vocabulary size and imageability, iconicity, concreteness and evidence for the shape bias, and to examine whether these patterns differ from their non-autistic peers. Methods: Participants for the word features analyses included 246 autistic children who were reported to produce between 5-312 nouns, and 940 expressive-vocabulary non-autistic children. Participants in the shape bias analyses included 272 autistic children who were reported to produce between 1-312 nouns, and 1,021 expressive-vocabulary and non-autistic children. Expressive vocabulary knowledge was measured using the MacArthur-Bates Communicative Development Inventory (Fenson et al., 2007). Results: Each analysis indicated that imageability, iconicity, and concreteness were associated with noun vocabulary size for both groups. Both groups displayed nonlinear change across vocabulary size, and group differences were found for each perceptual feature except imageability The pattern of the associations differed slightly for concreteness and iconicity for the autistic group, primarily at the very earliest points in noun vocabulary learning. Across the three word features, both autistic and non-autistic children learn highly imageable, iconic, and concrete words during the earliest stages of noun learning. Both groups also demonstrated evidence for the shape bias system, but the trajectory of growth of evidence for the system differed between the two groups. Conclusions: Imageability, iconicity, and concreteness were identified as significant lexicosemantic features for describing expressive noun vocabulary size in autistic children. Although slight differences across word features were observed, autistic children seem to have broadly similar noun vocabularies to their non-autistic peers. Furthermore, autistic children with smaller vocabularies produced more words that are highly imageable, iconic, and concrete, whereas children with larger vocabularies produced words that are less perceptually salient, indicating a potential shift away from reliance on perceptual-based language processing. We report unique information about nonlinear growth patterns associated with each word feature, with evidence for the shape system, and distinctions in these growth patterns between groups. Future studies should explore word-learning and lexical growth patterns involving early verb acquisition in autistic children.



Committee Chair

Haebig, Eileen