No abstract

In some dialects of Chinese and Miao, the nasals in syllable-initial position have been described as being accompanied by a homorganic stop, which are often transcribed with superscripts: [mb], [nd], and [ŋg], as a deliberate attempt to characterize these segments as phonetically distinct from prenasalized stops, [mb], [nd], [ŋg]. In our study, the acoustic nature of these “post-stopped” nasals will be explored, since no instrumental study has been conducted on them. A preliminary investigation of data from two Zhongshan Chinese speakers confirm that these nasals are different both perceptually and acoustically from the prenasalized stops in other languages. It is found that the so-called “stop” component in Zhongshan syllable-initial nasals is not a stop, but a burst that occurs simultaneous with the oral release following the nasal. Such bursts occur sporadically in English, but are consistently produced in the Zhongshan nasals, and are perceived as homorganic stops accompanying the nasals. The waveforms also show a characteristic shape, with a sharp rise in amplitude at vowel onset. The results suggest the precise synchronization of velic closure with oral release.

The understanding of the dynamic nature of the syllable has been deepened by investigating the P center/stress beat and segmental target concentration (Ren, 1986). This study is concerned with another feature, the temporal location of syllabic tonal target in a tone language, Chinese. Chinese has four syllabic tones with the phonological forms H, LH, L, HL, respectively, realized by changes in F0. Our acoustic data from three speakers shows that (1) in contrast to the anticipatory cross-syllable coarticulation at the segmental level, the cross-syllable tonal transition is a carryover, (2) with a conflicting preceeding tonal context, the temporal location of the only tonal target in H and L and the first target in LH and HL depends on the rate of F0 transition starting from the syllable onset, (3) with a harmonic preceding tonal context, the first target in LH and HL is located very close to the consonant offset or the vowel onset. The relations between P center/stress beat and tonal target in case (3) call for further investigation.

Durations of (1) closure, (2) frication noise, (3) aspiration, and (4) voice in Chinese sounds under different linguistic conditions were measured on oscillomink displays of waveforms. A set of rules of percentage duration variation were formulated to reflect the effects of each of several linguistic factors such as distinctive features, syllable structures, word lengths, tone and stress patterns, etc. The notion of “incompressibility” [I. Lehiste, J. Acoust. Soc. Am. 51, 2018–2024 (1972); D. H. Klatt, J. Acoust. Soc. Am. 54, 1102–1104 (1973)], i.e., that durations resist being shortened under combined shortening rules, is discussed. Like English, Chinese vowel durations are incompressible under several shortening rules; unlike English, Chinese consonant durations do not show clear incompressibility. Two problems were found in applying the formula Dj = K*(Di − Dmin) + Dmin [D. H. Klatt, J. Acoust. Soc. Am. 59, 1208–1221 (1976)] in Chinese, one is that Chinese data show two values of “minimum duration” (one duration symptomatic of word length effects, and a shorter duration under neutral tone); another problem is that different combinations of shortening rules may result in different powers of incompressibility in duration.

Phonetic orthodoxy treats the acoustic realization of a CVn, syllable (N.B. V = vowel or glide) as the concatenation of consecutive transitions from one element to the following one in the syllable. This study proposes an alternative view of the acoustic structure of this type of syllable, based on LPC trackings of F2 in sets of selected Chinese and English CVn syllables. The basic hypothesis is that the underlying F2 transitions between any two phonologically adjacent elements in the same syllable (e.g., the C → V1, V1 → V2,…) all originate at the same temporal position—the syllable initiation. Each transition has a specified rate. The acoustic realization will be a programmed “truncation” process; that is, the phonologically preceding transition truncates the phonologically following transition (e.g., the C → V1 transition truncates the V1 → V2 transition; the truncated V1 → V2 in turn truncates the V2 → V3 transition). This model provides elegant accounts and quantitative predictions of the undershooting and overshooting phenomena in acoustic target realization. [Work supported by NSF.]