Differences in glottal opening pattern between Tokyo and Osaka dialect speakers: Factors contributing to vowel devoicing

In Japanese, high vowels /i/ and /u/ are often devoiced when they are preceded and followed by voiceless consonants. The phenomenon is called vowel devoicing. The frequency of its occurrence is high in dialects of eastern Japan including standard (Tokyo) Japanese and low in dialects of western Japan including Osaka dialect. It has been claimed that consonants, but not vowels, are pronounced more carefully in eastern dialects compared to western dialects, which results in higher frequency of vowel devoicing. This paper examines whether the glottal opening gesture of voiceless consonants differs between those Tokyo speakers who consistently show vowel devoicing and Osaka speakers who do not show vowel devoicing. The results indicated that the duration of glottal opening for /k/ in voiced CV syllables was significantly longer for Tokyo speakers than for Osaka speakers. For Tokyo speakers, duration and maximum size of the glottal opening in devoiced /kite/ are significantly larger than /k/ in voiced syllables. This difference cannot be explained by the simple addition of the glottal gestures of [k] and [t]. Furthermore, the difference in vowel length between /i/ and /e/ was related, at least in part, to the difference in the amount of the airflow passing through the vocal tract about the time the stop closure was released.