Poster B115, Sunday, March 25, 8:00-10:00 am, Exhibit Hall C
The Perception and Cognition of Racialized Voices
Tedra James1, Maxime Bouvagnet1, Psyche Loui1; 1Wesleyan University
The human voice contains abundant information about its owner. Human listeners are able to determine racial and/or ethnic background from the voice based on acoustic and linguistic cues (Perrachione et al, 2010), but little is known about the neural mechanisms that underlie this determination. Here we report a combined behavioral and EEG study to investigate the perception and cognition of race identification from the human voice. We generated a database of short (<1s) audio samples of speech utterances by black and white speakers from natural speech downloaded from TED.com. A norming study was conducted online in which subjects (n = 100) listened to each sample and completed two-alternative forced choice (2AFC) tests to identify the race of the speaker. Samples that were racially identifiable above chance level were then used in a behavioral and EEG study. 64-channel EEG was recorded while participants listened to each sample and, in randomized trials in a 2AFC task, identified either the speaker’s race or the word spoken. ERPs showed a large early left anterior negativity around 100 ms, followed by another later negativity in left prefrontal sites at 600 ms, in response to black voices compared to white voices. Compared to word-identification trials, race-identification trials show an early frontocentral positivity (200 ms) followed by a late left-lateralized negativity (900 ms). Taken together, we observe early perceptual as well as late cognitive mechanisms of race identification from the human voice. This neural time-course of race identification has implications for understanding and reducing racial prejudice.
Topic Area: EMOTION & SOCIAL: Person perception