Does interspeaker variation in habitual speaking rate predict the size of the acoustic vowel space?

Abstract

In a previous study [Tsao and Weismer, J. Speech Lang. Hear. Res. (1997)], it was demonstrated that interspeaker variation in habitual speaking rate is predictive of interspeaker variation in maximal speaking rate. Thus speakers who are habitually slow speakers do not seem to be able to produce as fast a maximum rate as speakers who are habitually fast. This finding was interpreted to mean that some component of interspeaker variation in habitual speaking rate reflects neurological predispositions, rather than (for example) voluntary ‘‘choices’’ of different rates. The present study explores the hypothesis that habitually slow and fast speakers may have differently sized acoustic vowel spaces. The hypothesis is derived from previous observations that within speakers, the size of the acoustic vowel space fluctuates with speaking rate. Results suggest that there is substantial overlap between the vowel spaces of habitually slow and fast speakers, but that the slow speakers do tend to have larger vowel spaces. [Work supported by NIH.]