- 學位論文
音韻機率與漢語介音
Phonotactic Probability And The Prenuclear Glide In Mandarin Chinese
並列摘要
The thesis investigates whether the prenuclear glide in Mandarin Chinese tends to cluster with the onset or with the rime and the effect of phonotactic probability on the status of prenuclear glide in Mandarin Chinese syllables. The phonological status of the prenuclear glides in Mandarin Chinese is a controversial issue. Traditionally, the prenuclear glide in Mandarin Chinese was classified as belonging to the Final (Chao 1968, Cheng 1973). Other studies pointed out that the prenuclear glide was a cluster with the onset (Bao 1990) or a part of the onset as a secondary articulation (Duanmu 1990). Wan (1997) proposed that the prenuclear glides should be considered as ‘indeterminate’ and their status can be part of the onset or part of the rime depending on their preceding consonant. It has been found that spoken language processing is influenced by phonotactic probability (Jusczyk, Luce, & Charles-Luce 1994, Vitevitch, Luce, Charles-Luce, & Kemmerer 1997, Vitevitch & Luce 1999, Bailey & Hahn 2001). That is, the higher phonotactic probability of two segments is, the more easily the segments clustered. Thus, it is possible that the probability of co-occurrence between the prenuclear glide and its neighboring units affects the status of the prenuclear glide in Mandarin Chinese. The results of the experiments showed that the subjects preferred to classify the glide as part of the rime. However, the rime responses were also affected by phonotactic probability, especially where the glide was in the first input syllable. That is, the higher the probability between the prenuclear glide and the onset was, the fewer rime responses were produced. In addition, participants tended to prefer to put the glide with the rime in the second input syllable more than in the first input syllable. The Maximality principle (Prince 1985) results in the glide preserving tendency in the word-blending task. The subjects preferred to choose the largest possible outputs when they participated in the word-blending task (e.g. production experiment, forced-choose experiment, regression experiment).