Abstract
The effect of speaker ethnicity on speech perception remains unclear. Proponents of the bias hypothesis maintain that presenting an Asian or Mexican face to American participants triggers a certain kind of bias that could result in worse comprehension and even hearing a non-existent ‘foreign accent.’ Exemplar-based studies, on the other hand, have proposed that these findings merely reflect a mismatch between listeners’ expectations and the actual speech signal. While previous studies all used post-perceptual, offline tasks to examine the effect of speaker ethnicity on speech perception, this study made use of an online task instead. Thirty-two native English participants completed a speeded audio-visual sentence verification task, for which they had to classify statements as true or false. The utterances were paired with a photograph of an Asian face, a White face, or a fixation cross, and were presented in a mixed design. Both correctness scores and response times for all the different face-voice pairings were recorded. Results suggest that online processing was not affected by speaker ethnicity, as response times did not differ as a function of the various face-voice pairings. Additional findings showed that the foreign-accented voices took significantly longer to process than the native voices, and that false statements took longer to answer than true statements.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call