Master of Arts (MA)
Faculty of Science
Jeffrey A. Jones
The purpose of these studies was to test the main assumptions outlined in the Motor Theory of speech perception that (1) speech perception is linked to speech production, (2) audiovisual integration of speech occurs automatically and after the motor commands are activated, and (3) we perceive the intended gestures, which are extracted by a specialized ‘phonetic module’ in the brain. In Experiment 1, we used a Stroop-like paradigm, where participants viewed and listened to a speaker producing speech syllables (/aba/ or /aga/) in three conditions: audio-only, visual-only, and audiovisual. Participants were asked to ignore irrelevant speech stimuli, and to identify vocally or manually the target letters (BA or GA) that appeared over the speakers’ face, as quickly and accurately as possible. If speech perception is closely tied to speech production, then we should find faster response times to vocally produce a syllable that matched the perceived syllable than mismatched. Indeed, we found that response times were quicker when the target and irrelevant speech were compatible, than when they were incompatible. This finding was consistent across all modality conditions for verbal, but not for manual responses, suggesting a close perception-production link for speech. The same stimulus-response interference was found when the irrelevant stimuli were static pictures portraying speech or non-speech gestures (Experiment 4), demonstrating that even implied speech gestures interfere with speech production. Furthermore, these compatibility effects were seen in Experiment 2 when we use conflicting auditory and visual irrelevant speech information (e.g., auditory /aba/ dubbed over a visual /aga/). Our results showed both modalities interfered with speech production to the same degree. However, once the modalities were integrated (i.e., eliciting the McGurk effect; Experiment 3), our data showed faster response times to produce the target that was compatible with the integrated percept, than those targets that were compatible with either modality alone. Although not statistically significant, the trends in Experiment 3 suggest that integration may occur before the response stage is reached, implying that another mechanism may be responsible for integration. Overall, our findings provide some support for the Motor Theory view of speech perception, demonstrating that speech perception effects speech production.
Jarick, Michelle A., "Audiovisual speech perception: A speech production approach" (2007). Theses and Dissertations (Comprehensive). 799.