Search Abstracts | Symposia | Slide Sessions | Poster Sessions
“So much to say”: Investigating the receptive language abilities of autistic non- and minimal-speakers using electroencephalography
Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 3 and 4
Hannah Rapaport1, Selene Petit1, Alexandra Woolgar1; 1MRC Cognition and Brain Sciences Unit, University of Cambridge
Introduction Despite the exponential growth in global funding for autism research since 2001, approximately 30% of the autistic population who have no or minimal spoken language (henceforth, ‘non-speakers’) remain chronically under-researched. This empirical neglect is in part because standardised tools and experimental paradigms rarely accommodate autistic non-speakers’ communication, motor and sensory challenges, making it difficult to conduct reliable and valid assessments of their cognition. In the near-absence of empirical data, it has been widely assumed that autistic non-speakers have a limited understanding of language. However, this assumption is incompatible with the growing community of autistic non-speakers who have learnt to communicate by pointing to letters on a stencil or typing on a computer keyboard – suggesting that at least some non-speakers have excellent receptive and expressive language. Anecdotal reports of non-speakers’ language being underestimated by educators, family members and the public lead to concerns for non-speakers’ education, career prospects and wellbeing. It is therefore vital that we develop novel methods to empirically assess the language abilities of autistic non-speakers. To meet this need, the aim of the current study was to determine whether non-invasive brain recording could be used to detect neural signatures of language comprehension in individual autistic non-speakers and age-matched neurotypical counterparts. Methods Following our pre-registered plan (https://osf.io/3wfbx), we used a wireless, gel-free ElectroEncephaloGraphy system to record the brain responses of 28 autistic non-speakers (Mage = 16.7 years; range: 5-28; 6 females) and 37 neurotypical participants (Mage = 14.6 years; range: 5-37; 15 females) as they listened to a series of spoken sentences. Half of the sentences ended with a word that was congruent with the sentence (e.g., “In spring, flowers BLOOM.”), and half, incongruent (e.g., “In spring, flowers CRY.”). A difference in the brain response between the more-predictable congruent sentences and the more-surprising incongruent sentences (known as the ‘N400’ effect) would offer neural evidence of language comprehension. Notably, our neural measure of language comprehension bypasses the need for a behavioural response. Preliminary Results Using univariate analyses, we found significant, cluster-corrected N400 effects for the neurotypical group in central (352 to 680 ms) and right-temporal (617 to 781 ms) regions-of-interest, but no significant effects for the autistic group. At the individual-level, we found significant cluster-corrected N400 effects in a subset of neurotypical and autistic individuals. We are currently running multivariate pattern analyses to determine whether we can discriminate between patterns of neural activation associated with the congruent and incongruent sentences. Full analyses will be completed in mid-2024. Conclusion Our preliminary results provide an early proof of concept, demonstrating that it is possible to detect neural signatures of language comprehension in individual autistic non-speakers. This work has the potential to advance and potentially challenge scientific understandings of non-speakers’ language abilities, and could have important implications for how non-speaking autistic people are perceived and treated by family, practitioners, employers and the public.
Topic Areas: Disorders: Developmental, Speech Perception