Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Emotional Prosody Processing in Autistic Adults
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Azia Knox1, Jacob Burger1, Arild Hestvik; 1University of Delaware
Theory of Mind (ToM; Baron-Cohen, 2000), the ability to understand and interpret the mental states of others, is often cited as something autistic individuals inherently lack (Frith et al., 1991). One feature comprising ToM is successfully inferring the emotional states of others, often communicated by a speaker utilizing prosodic vocal cues like intonation, intensity, pitch, and the combination thereof (Frick, 1985). Prior research has examined neural responses to emotional and neutral prosodic contours (Mitchell et al., 2003) and how prosodic expectancy violations are elicited. Upon hearing a sentence beginning with neutral prosody, listeners predict that the sentence will continue with that same prosody. If this prediction is violated by splicing the sentence to switch the emotional valence, a Prosodic Expectancy Positivity (PEP) ERP occurs parietally ~470ms post-violation onset; a switch in linguistic valence elicits a PEP ~620ms post-violation (Paulmann et al., 2012; Wickens & Perry, 2015). This study implements a similar paradigm with autistic adults to test their success in identifying emotional and linguistic prosody, and whether the aforementioned linguistic and emotional prosody expectancy violations may be elicited. Our hypothesis is that since autistic adults can generally recognize when a speaker is upset, the same PEP should be elicited with high behavioral accuracy. We developed a modified version of prior PEP methodologies. A native English speaker recorded sentences in three prosodic conditions: neutral prosody (baseline condition), angry prosody (emotional condition), and question prosody (linguistic condition). All sentences began with “She had” followed by a simple noun phrase (e.g. “a quiet voice”). Splices were created by combining the neutral reading of “She had” with either an emotional or linguistic prosodically contoured noun phrase. Sentences are presented over two speakers while a fixation cross is at the center of a screen. Shortly after the sentence offset, subjects are tasked to press a button to categorize the sentence as neutral, emotional, or question. Continuous EEG is recorded with 65-channel EGI system. So far, eight native English speakers (2 male, 2 nonbinary, Average age=22.37) with self or official diagnosis of autism have participated, with a target N=40. We plan to analyze two difference waves: spliced emotional prosody-unspliced neutral, and spliced linguistic prosody-unspliced neutral. Preliminary analysis of the emotion difference wave shows an early anterior negativity peaking ~150ms post splice, followed by a right parietal positivity peaking at ~556ms, replicating prior studies with a slightly later positive peak. The linguistic difference wave shows a more muted early anterior negativity again peaking at ~150ms with a left-lateralized parietal positivity at ~624ms. If the preliminary results remain in the target sample size, the violation neural response would provide evidence of knowledge of what emotional or illocutionary states are inferrable from prosody in autistic adults. These findings will also provide pushback to ToM’s hypothesis that autistic individuals have an inherent deficit in interpreting emotional mental states. This EEG data is key as even if behavioral accuracy is near ceiling, the neural response will show that this ability is not wholly absent or deficient in autism.
Topic Areas: Prosody,