Abstract
Figurative language, including metaphors and similes, is a crucial component of communication; yet, it presents significant challenges for individuals with autism spectrum disorder (ASD). A critical gap in existing research is the impact of bilingualism on the ability of children with ASD to understand and produce non-literal speech. This study addresses this gap by examining the comprehension and production of metaphors and similes in monolingual and bilingual Greek-speaking children with high-functioning ASD. To the best of our knowledge, this is the first study to investigate these abilities in bilingual children with ASD. Thirty-three monolingual and 18 bilingual children participated in tasks designed to assess comprehension, production, and error patterns for metaphors and similes. The study has also investigated the roles of non-verbal intelligence, language skills (expressive vocabulary), and executive functions (working memory) in the children's performance in the metaphor and simile tasks. Results showed that the two groups did not differ in metaphor comprehension; however, bilingual autistic children with higher non-verbal intelligence appeared to have superior performance in metaphor comprehension compared to their bilingual peers with lower non-verbal intelligence. The bilingual autistic children outperformed their monolingual peers in metaphor production, likely due to their higher non-verbal intelligence ability, despite the fact that the bilingual group had lower expressive vocabulary scores than the monolingual children. Simile comprehension, on the other hand, favored monolingual children, while no significant group differences were observed in simile production. Regarding errors, both groups exhibited similar error patterns, with literal interpretations being the dominant error type across both groups, suggesting that pragmatic language difficulty is a hallmark feature in ASD. The findings challenge the misconception that bilingualism hinders language development in children with ASD and highlight its potential to provide benefits in the realm of non-literal language processing.