A phonetic study of the nasals and vowel nasalization in standard Chinese

Abstract

This dissertation investigates the phonetic characteristics of the nasals and vowel nasalization in Standard Chinese by acoustic and aerodynamic experiments. The main purposes are to investigate (1) the timing relationship between a vowel and its adjacent nasal consonant, and the temporal extent of vowel nasalization; (2) the acoustic properties of vowel nasalization in terms of the parameters of A1-H1 (the amplitude difference between the first formant and the first harmonic) and A1-P1 (the amplitude difference between the first formant and the nasal peak at about 950Hz); (3) the spacial extent of vowel nasalization in terms of the ratio of nasal airflow volume relative to the total 'oral + nasal' airflow volume (NFR, or nasal flow ratio); and (4) the effects of various segmental (vowel quality, nasal type, and adjacent segment) and prosodic (syllable position, syllable boundary, and speech rate) factors on (1)-(3). Researchers have noticed that the syllable-final nasals in Standard Chinese are "weak nasals in that they are often produced with incomplete oral closure and are shortened or deleted in natural speech. Results of the present study show that the syllable-final nasals are different from the syllable-initial nasals in Standard Chinese in two aspects. On the one hand, the initial nasals have relatively short but constant duration and is not likely to be shortened or deleted. There is always a clear boundary between an initial nasal and the flanking vowels. While the syllable-final nasals have longer duration, they are subject to shortening or deletion when followed by a continuant (fricative or vowel), especially when speech rate is fast. On the other hand, anticipatory vowel nasalization is greater when a vowel is followed by a tautosyllabic nasal ending than a heterosyllabic nasal initial. The two nasal initials [m] and [n] are similar in terms of nasal duration and the degree of vowel nasalization. But the two syllable final nasals [n] and [ŋ] are different in that [ŋ] is longer and less likely to be deleted than [n]. The two nasal endings are also distinguished in terms of the formant values in the preceding vowel. The extent of vowel nasalization is quantified in terms of duration, A1-H1/A1-P1, as well as nasal airflow ratio. Vowel nasalization in Standard Chinese is found to have longer duration for lower and longer vowels than higher and shorter vowels. In terms of A1-H1/A1-P1, the acoustic extent of nasal coupling is smaller and slower in VN.V where the nasal ending is frequently deleted. In addition, the acoustic nasal coupling is smaller and slower when speech rate increases. In terms of nasal flow ratio, the spatial extent of nasalization tends to be smaller for the low vowel [a] than the mid vowel [ə] and the high vowels [i, u, y], due to the lower oral impedance for lower vowels. The spatial extent of nasalization is also smaller when the nasal ending is deleted in VN.F and VN.V contexts and the vowel of the rhyme is fully nasalized (Ṽ) or partially nasalized (VṼ). Finally, the type of nasal consonant does not show any significant effect on the extent of vowel nasalization, which implies that the size of velic opening is similar for the vowels preceded or followed by the different nasal consonants. This study provides detailed experimental data for the description of the nasals and vowel nasalization in Standard Chinese. These phonetic data also help us to gain a better understanding of the nasals and vowel nasalization in human languages. The timing relationship between a nasal and its adjacent vowels and the degree of vowel nasalization is affected by the phonetic properties of the vowel and nasal of the language per se, but is more importantly conditioned by the process of articulatory coordination between related gestures.

Date of Award	15 Jul 2014
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Wai Sum LEE (Supervisor) & Yun Yang ZEE (Co-supervisor)