First International Conference on Spoken Language Processing (ICSLP 90)

Kobe, Japan
November 18-22, 1990

Vowel Quantity Contrast in Korean: Production and Perception

Minje Zhi, Yong-Ju Lee

Signal Processing Section, Electronics & Telecommunications Research Institute, Daejeon, Korea

In the present study, the acoustic properties of the vowel quantity contrast in the Seoul variety of Korean were investigated. Acoustic measurements included the duration and fundamental frequency of the vowels, and the frequencies of the first two formants of a-vowels of test words. Perception tests were also carried out with manipulated stimuli. The purpose was to ascertain to what extent listeners from Seoul and Busan, where the tone contrast is substituted for the quantity contrast, could distinguish between two words contrasted in the vowel quantity in a frame with neither semantic nor syntactic cues. In the acoustic analysis it was found that the vowel quantity contrast in the Seoul variety is manifested primarily by differences in vowel duration and sencondarily by differences in formant structure. The perception experiment revealed that a majority of the Seoul and Busan listeners could discriminate between the words contrasting minimally by the quantity of the vowel in the Seoul variety. The positive effect of vowel duration on the perception of the quantity contrast was observed for the listeners from both Seoul and Busan.

Full Paper

Bibliographic reference.  Zhi, Minje / Lee, Yong-Ju (1990): "Vowel quantity contrast in Korean: production and perception", In ICSLP-1990, 613-616.