A shared mechanism for perceiving and producing voice-onset times in speech

Abstract

A reaction-time method was employed to assess the relationship between speech perception and production mechanisms at the level of phonetic features. Subjects memorized a set of CV syllable pairs so that they could respond to the stimulus syllable of a pair by saying the corresponding response syllable. With acoustic presentation of the stimuli, response latencies depended on the phonetic relationship between the stimuli and the responses. Shared voice-onset time facilitated the response but shared place of articulation had no effect. This facilitation did not occur with visual presentation of the stimuli, indicating that it was due to speech perception priming speech production and not due to the ease of memorizing different syllable pairs. Similar results were found for different consonant types (stops and fricatives) and for stimulus-response combinations containing different vowels. These findings partially confirm classic views of the relationship between speech perception and production, such as those assumed by the motor theory of speech perception. It appears that speech perception and production may involve a shared voice-onset time mechanism. [Supported by NSF grant BNS-8206809.]