The Scope and Effects of Preboundary Prosodic Lengthening in Japanese

要旨

The acoustic duration of phonemically short and long Japanese vowels with three degrees of removal from an intonation phrase boundary--immediately preboundary, one mora preboundary, and two moras preboundary--was measured. Immediately preboundary vowels showed significant acoustic lengthening relative to like controls, with short vowels showing proportionally greater lengthening than long vowels. However, vowels even one mora preboundary were generally not significantly different from controls. Interpreted in light of the π-gesture model of prosodic influence (Byrd and Saltzman 2003; Byrd et al. 2006), this suggests that the temporal scope of preboundary prosodic lengthening in Japanese is approximately one mora and, thus, only the latter portion of a preboundary long vowel overlaps the π-gesture. This difference in the degree of lengthening of short versus long vowels also significantly reduced the long-to-short ratio of immediately preboundary vowels relative to controls. Moreover, while there was a vowel-to-word duration ratio that reliably classified control vowels as either short or long, applying the same ratio to immediately preboundary vowels classified all of them as long. These findings challenge the claim that relative acoustic duration in Japanese is invariant across speaking rates (Hirata 2004).

主題