Cross-language speech perception in adults: Discrimination of Korean voiceless stops by English speakers

Abstract

This study examines the ways in which three classes of alveolar stops in Korean (voiceless 'tense' unaspirated /t*/, voiceless 'lax' slightly aspirated /t/, and voiceless heavily aspirated /th/) present different degrees of perceptual difficulty to adult English and Korean listeners. Results show that the /t/-/th/ contrast presented the greatest difficulty in perceptual discriminability for the American listeners (61% error rate) while the /t*/-/t/ and the /t*/-/th/ contrasts presented relatively easy discriminability with 9% and 3% error rates respectively. Painvise t-test results show that English listeners discriminated the /t*/-/th/ contrast significantly better than the /t*/-/t/ contrast, suggesting that a larger difference in VOT between stimulus items increases discriminability. These and other results suggest that English listeners' perception of Korean voicing contrasts is largely determined by phonemic status and to a lesser extent by the magnitude of acoustic difference.