An acoustic comparison of dental and retroflex sibilants in Chinese Mandarin and Taiwan Mandarin

Abstract

Mandarin has both dental and retroflex sibilants. While the Mandarin varieties spoken in China and Taiwan are often considered the same, native speakers of Mandarin can tell the difference between the two. One obvious difference is that between the retroflex ([ʂ], [tʂ], [tʂh]) and dental sibilants ([s], [ts], [tsh]). This study investigates the acoustic properties of the sibilants of Chinese Mandarin and Taiwan Mandarin. Eight native speakers each of Chinese and Taiwan Mandarin produced the six target sibilants in word-initial position. A number of acoustic parameters, including spectral moments and duration, were analyzed to address two research questions: (a) which parameters distinguish the dental and retroflex in each type of Mandarin; (b) is there a difference between Chinese and Taiwan Mandarin? Results show that retroflex sibilants have a lower M1 and M2, and a higher M3 than dental sibilants in each language. Moreover, Chinese Mandarin has significantly larger M1, M2, and M3 differences than Taiwan Mandarin. This pattern suggests that, in contrast to Chinese Mandarin, Taiwan Mandarin is merging the retroflex sibilants in a dental direction.