The effects of prosody on pitch and voice quality of White Hmong tones

Abstract

White Hmong contrasts two high-falling tones (one breathy, the other modal) and two low tones (one modal level-tone, the other creaky low-falling). Perceptual studies [Garellek et al. (2013)] have shown that listeners rely on breathy voice to distinguish between the high-falling tones, but ignore creaky voice when distinguishing between the low tones. We test whether such differences stem from prosodic variation, by examining tokens from stories read by native speakers. Vowels were annotated for phrasal position and neighbouring tones. We obtained f0 and voice quality measures. Results support and help elucidate previous perceptual research: (1) the breathy high-falling tone is breathy in all prosodic positions, (2) the breathy high-falling tone has a prosodically-variable f0, (3) the creaky low-falling tone has a prosodically-stable f0. The creaky low-falling tone is creakier than the modal low tone in all positions. Therefore, listeners may ignore f0 in the identification of the breathy high-falling ton...