An acoustic and electroglottographic study of White Hmong phonation.

Abstract

This study examines the phonation of White Hmong, a language with seven tones (traditionally described as high, mid, low, high-falling, mid-rising, low-falling, and mid-low), five of which are associated with modal phonation, and two of which are associated with non-modal phonation; the low-falling tone is creaky and mid-low tone is breathy. Thirty-three speakers were recorded producing words with all seven tones; 12 also made electroglottographic (EGG) recordings. Acoustic measures were cepstral peak prominence (CPP) and harmonic amplitudes H1* and H2*, H1*-H2*, H1*-A1*, H1*-A2*, H1*-A3*, and H2*-H4*. EGG measures were closed quotient (CQ) and peak-closing velocity (PCV). Measures were made automatically using VOICESAUCE and PCQUIRERX. Results showed that none of the measures tested distinguished all three phonation types. However, several measures distinguished two categories: H1 distinguished creaky versus non-creaky, H1-H2 distinguished breathy from creaky, and the EGG measures CQ and PCV both distinguished breathy from non-breathy. H1*-A1*, H1*-A2*, H1*-A3*, and H2*-H4* did not distinguish any of the phonation types. This suggests that phonation contrasts are realized across several phonetic dimensions in White Hmong. In addition, there was a gender difference in the production of phonation, with females having significantly higher CPP than males, suggesting that female phonations are noisier. [Work supported by NSF.]