When the Archeologist's Career Ended in Ruins: An fMRI Study of Pun Comprehension in Autism

Background:   Understanding intended and contextual meanings in language is among the most affected aspects of language comprehension in autism (Happe, 1994; Tager-Flusberg, 1981). This difficulty may lie in the fact that comprehension of figurative language, such as irony and pun, may be atypical in autism (Happe, 1995). While several studies have examined other types of figurative language in individuals with autism, no brain imaging study to date has examined pun comprehension in autism. A pun is a rhetorical technique in which the speaker deliberately invokes multiple meanings through a single word or phrase. Comprehending puns is a challenging cognitive task since it involves several subtasks, such as identifying multiple meanings of the word, decoding the pun sentence, and understanding the joke. People with autism might have difficulty in all these processes, and the current study examined the cortical bases of such processes.

Objectives: The primary aim of this study was to investigate the neural mechanisms involved in interpreting figurative language, specifically comprehending puns in autism.

Methods: Seven high-functioning adolescents and adults with autism and nine age and IQ matched controls participated in this fMRI study (data collection in progress). Sentences containing puns (e.g. To write with a broken pencil is pointless) and control sentences were presented visually in a blocked design format with four blocks in each experimental condition. The participants’ task was to silently read and understand one meaning (in the literal condition) or two meanings (in the pun condition) in each sentence and press a button to indicate that they had read and understood the sentence. After the scanning session, the participants were asked to point out the two meanings in each sentence.

Results: Participants with autism showed reliably lower levels of activation than controls in left inferior frontal gyrus (pars opercularis aspect) and left middle temporal gyrus while comprehending pun, suggesting difficulty with determining word meaning and deciphering syntactic cues from the pun sentence. In contrast, participants with autism activated more right fusiform gyrus (part of the visual word form area) and right middle temporal gyrus (RMTG), indicating the utilization of a different approach to understanding word meaning. There was also lower activation in the right orbitofrontal cortex (OFC), an area thought to be involved in humor comprehension.

Conclusions: Underactivation in autism in left hemisphere language regions (especially pars opercularis of the IFG) while comprehending puns suggests that participants with autism may not be deciphering the syntactic cues of the sentence or understanding the multiple meanings involved. Greater RMTG activation in autism may corroborate this view by suggesting increased difficulty in comprehending the meanings of the sentence. Moreover, greater activation in autism in the visual word form area while reading pun sentences may suggest their possible focus on graphemic level instead of lexical semantic level. Finally, if participants with autism do not understand the multiple meanings of pun, they may not be detecting the humor (indicated by less activation in OFC). Overall, these findings suggest atypical cortical recruitment in autism while comprehending pun.