Production and perception of Yang tone breathiness in today’s Shanghai Chinese Jiayin Gao ([email protected]), Laboratoire de Phonétique et Phonologie – Paris 3/CNRS In Shanghai Chinese and other northern Wu dialects, Yang tone syllables are said to be accompanied with some breathy quality, known as 清音浊流 ‘clear sound followed by muddy airflow’, whereas Yin tone syllables are not. This breathiness may be viewed as a redundant feature of Yang tone syllables, which are otherwise characterized by a low tonal register. Several acoustic studies have examined this issue and most of them confirmed the presence of the breathy quality (e.g., Cao & Maddieson, 1992). Yet, physiological and perceptual studies remain relatively rare. Besides, less is known about the variation and the evolution of this feature. However, Shanghai Chinese, has undergone the most rapid changes among Wu dialects. In a recent study, we examined the production of Yang tone syllables by 11 speakers of two age groups, using physiological and acoustic measurements. The results showed large cross-age and cross-gender variation. Our female speakers, elderly or young, didn’t produce any differential breathiness in Yin vs. Yang syllables in terms of spectral configuration (H1-H2) and glottal open quotient (OQ). Clear breathiness in Yang syllables was found for our elderly male speakers. The young male speakers’ productions were more variable. The results suggested a general trend toward loss of breathiness in Yang tone syllables. In this study, we focused on the perceptual aspect of breathiness, in order to determine whether breathiness plays a secondary role in tone perception. We asked whether the loss of a redundant feature in production was preceded by the loss of its role in perception, as suggested by the listener-oriented explanation for sound change (Ohala 1993)? We therefore conducted two identification tests with 16 young native speakers of Shanghai Chinese (mean age: 22; from 18 to 26). In both tests, subjects were presented with tone continua between two “base” syllables in tone 2 (T2: Yin tone) and in tone 3 (T3: Yang tone). For each continuum, the base syllables were synthesized (test A) or naturally produced (test B) with either modal or breathy phonation. All the stimuli shared the /ɛ/ rime. Their onsets were /Ø, p, t, f, s, m/. On each trial, subjects had to identify between a Yin-Yang minimal pair (as illustrated by a Chinese character) the syllable that corresponded best to the stimulus presented. We recorded Yang response rates as well as Yin and Yang response times (RT) along the continua, according to the “base” syllable phonation type, in each test. The data of one subject in test A and of two subjects in test B were discarded. Our results showed that Yang response rate was higher overall for breathy than modal syllables in both tests A and B, as shown by the Yang identification functions’ shift toward the T2 endpoint for breathy syllables (Figure 1). RTs were consistent with the tendency to identify breathy syllables as Yang: For the breathy syllables, Yin response RTs were slower, and Yang response RTs faster (test B).

(A)

(B)

Figure 1. Identification curves of Yin responses of (A) synthesized stimuli and (B) natural stimuli.

Thus, breathy quality facilitates Yang responses. Even though young speakers are found to produce less breathiness than elderly speakers, our young listeners seem to perceive breathiness as a secondary cue to Yang tone. Our two studies suggested that, in the course of sound change, the loss of a redundant feature is observed in production before perception, contrary to Ohala’s (1993) listener-oriented explanation. Selected references Cao, J. F., & Maddieson, I. (1992). An exploration of phonation types in Wu dialects of Chinese. Journal of Phonetics, 20(1), 77-92. Ohala, J. J. (1993). Sound change as nature’s speech perception experiment. Speech Communication, 13(1), 155-161.