Idioms As a Measure of Pragmatic Language Abilities in Adolescents with Autism Spectrum Disorders

Background:  

Individuals with autism spectrum disorders (ASD) have well-documented deficits in pragmatic language, such as narrative production and use of idiosyncratic language; however, few studies have examined comprehension of idiomatic expressions in ASD. (e.g., spilled the beans). While clinicians and parents of children with ASD often cite idiomatic expressions as an area of weakness, studies of idiom comprehension have found that, in the absence of significant structural language deficits, children with ASD tend to perform similarly to their peers with typical development (TD) peers in defining idioms.

Objectives:  

Our objective was to study the comprehension of idioms in ASD as a means of understanding the pragmatic language deficits associated with the disorder. We compared idiom definitions produced by adolescents with TD and ASD using standardized testing, quantitative methods (i.e., coding) and through the use of quality ratings from naïve readers.

Methods:  

Participants were adolescents with ASD (diagnoses confirmed by ADOS, n = 13) with a history of language delay but functioning in the average cognitive range and TD (n = 12). Groups did not differ on chronological age, gender, and full scale IQ. Participants completed the “Figurative Language” subtest of the Test of Language Competence (TLC), where they were asked to interpret an idiom used in context (i.e., “Situation: A boy talking about his girlfriend. Expression: She is easily crushed.”). Participants also read several common American idioms in isolation (turn a corner, swim against the tide, etc.) and were asked to define their meaning. Their responses were transcribed and coded for: 1) accuracy, 2) focus on literal definition, and 3) whether an example was included. The transcribed definitions were rated by 10-12 college students, naïve to diagnosis and study hypotheses, for goodness and accuracy. 

Results:  

The ASD group scored lower than the TD group on the TLC Figurative Language subtest (i.e., idioms used in context; F(1, 23) = 6.44, p < 0.02, h² =  0.22). Scores on the experimental idiom task did not differ (p’s > 0.22) on any of the coding measures (i.e., how accurate or literal the definition was, whether an example was used in the definition), indicating that the groups were similarly able to define idioms when presented in isolation, rather than in context. Furthermore, naïve readers were unable to detect differences between the ASD and TD definitions (p’s > 0.38), a notable result given that naïve readers have been shown to detect differences between ASD and TD in narrative quality. 

Conclusions:  

For idioms presented and defined in isolation, adolescents with ASD and TD produced similar idiom definitions, according to quantitative coding and quality ratings from untrained raters. However, when asked to interpret the meaning of idioms in the context of a social situation, adolescents with ASD performed significantly worse than their peers. This indicates that deficits in idiom comprehension likely reflect social deficits rather than language or word knowledge deficits. Interventions that teach idiomatic expressions should emphasize the use of idioms in context, rather than in isolation, to most effectively impact pragmatic language use.