The effect of talker accent on the perception of voicing in syllable-initial plosives by American English monolinguals

Abstract

Listen

Translate article

A listener's perception of speech is influenced by the talker's characteristics, including their accent. This syllable perception study, which uses a binary forced choice identification task, investigates the effect of talker accent at the level of speech sounds. A single ten-step /ba/-/pa/ continuum varying in voice onset time (VOT) was spliced into identical frame sentences spoken by a native speaker of American English and a non-native speaker whose native language is Mandarin. These frame sentences were modified so that the VOT for the bilabial plosives for both talker conditions is identical. This modification was done to control listeners from using frame sentence VOT as a cue. As the steps in the continuum differed only in terms of VOT for both talker conditions, listeners are only primed by the frame sentence for each item in the experiment. The participants' responses revealed that monolingual American English listeners are more likely to perceive syllables in the native talker condition as /pa/. In particular, the syllable at the 25ms step shows the largest variability between conditions. The results of this study show that listeners can perceive an acoustically identical syllable as being phonemically different when the syllables are spliced into sentences spoken by talkers with different accents.A listener's perception of speech is influenced by the talker's characteristics, including their accent. This syllable perception study, which uses a binary forced choice identification task, investigates the effect of talker accent at the level of speech sounds. A single ten-step /ba/-/pa/ continuum varying in voice onset time (VOT) was spliced into identical frame sentences spoken by a native speaker of American English and a non-native speaker whose native language is Mandarin. These frame sentences were modified so that the VOT for the bilabial plosives for both talker conditions is identical. This modification was done to control listeners from using frame sentence VOT as a cue. As the steps in the continuum differed only in terms of VOT for both talker conditions, listeners are only primed by the frame sentence for each item in the experiment. The participants' responses revealed that monolingual American English listeners are more likely to perceive syllables in the native talker condition as /pa/....