Production and perception of work‐initial stops by Korean adults

Abstract

In Korean there are three classes of stops: Voiceless unaspirated [t], voiceless slightly aspirated [tb], and voiceless heavily aspirated [thh]. For these classes of stops produced by male and female Seoul and Pusan speakers, voice‐onset time (VOT), amplitude of aspiration, and f0 of the first five glottal pulses were measured. VOT was longer for slightly aspirated than for unaspirated stops, and longer still for heavily aspirated stops. Unaspirated and heavily aspirated stops were both produced with significantly higher f0 (more than 30%) than slightly aspirated stops. Gender and dialect (Seoul versus Pusan) differences in these variables were also noted. Then, six 12‐step series of syllables differing in VOT (5–82 ms), fundamental frequency (100, 125, 150 Hz), and amplitude of aspiration noise (−33, −21 dB re: vocalic energy) were synthesized. These stops, varying perceptually from [ta] to [tba] to [thha], were presented to 12 native‐Korean listeners. Results indicate that all three stimulus variables were perceptually significant for the Korean listeners. [Work supported by NIDCD Grant No. DC‐00719 and NSF Grant No. DBS‐9258482.]