Cortical auditory evoked potential correlates of categorical perception of voice-onset time

Anu Sharma, Michael Dorman

    Research output: Contribution to journalArticlepeer-review

    161 Scopus citations


    The goal of this study was to examine the neural encoding of voice- onset time distinctions that indicate the phonetic categories /da/ and /ta/ for human listeners. Cortical Auditory Evoked Potentials (CAEP) were measured in conjunction with behavioral perception of a /da/-/ta/ continuum. Sixteen subjects participated in identification and discrimination experiments. A sharp category boundary was revealed between /da/ and /ta/ around the same location for all listeners. Subjects' discrimination of a VOT change of equal magnitude was significantly more accurate across the /da/-/ta/ categories than within the /ta/ category. Neurophysiologic correlates of VOT encoding were investigated using the N1 CAEP which reflects sensory encoding of stimulus features and the MMN CAEP which reflects sensory discrimination. The MMN elicited by the across-category pair was larger and more robust than the MMN which occurred in response to the within-category pair. Distinct changes in N1 morphology were related to VOT encoding. For stimuli that were behaviorally identified as /da/, a single negativity (N1) was apparent; however, for stimuli identified as /ta/, two distinct negativities (N1 and N1') were apparent. Thus the enhanced MMN responses and the morphological discontinuity in N1 morphology observed in the region of the /da/-/ta/ phonetic boundary appear to provide neurophysiologic correlates of categorical perception for VOT.

    Original languageEnglish (US)
    Pages (from-to)1078-1083
    Number of pages6
    JournalJournal of the Acoustical Society of America
    Issue number2
    StatePublished - 1999

    ASJC Scopus subject areas

    • Arts and Humanities (miscellaneous)
    • Acoustics and Ultrasonics


    Dive into the research topics of 'Cortical auditory evoked potential correlates of categorical perception of voice-onset time'. Together they form a unique fingerprint.

    Cite this