Abstract
Traditionally, the prevocalic glides are considered as structurally part of the final. However, in Duanmu (1990) and Bao (1990), the prevocalic glides are analyzed as forming a constituent with the initial consonant. To clarify the issues regarding the status of prevocalic glides, Wan (1997) gathered evidence from naturalistic slips of the tongue, and concluded that a prevocalic glide could be part of the onset or part of the rime depending on its preceding consonant. Faced with these contradictory analyses, we tried to gather evidence to resolve the controversy by administering two experiments: a syllable-expansion experiment and a syllable-contraction experiment.In the syllable-expansion experiment, the subjects were asked to select from two alternatives the one form that was preferable when they had to expand the key syllable into two syllables. In the syllable-contraction experiment, the subjects were asked to fuse two syllables into one on the basis of onset/rime division.As a result, we found that, unlike Duanmu’s (1990) and Bao’s (1990) analysis, there existed a general tendency for subjects to select “rime” answer in these two experiments. The place of articulation of the preceding consonant was not found to be a significant factor in the categorization of the prevocalic glide, neither was the type of glides significant, contrary to Wan’s (1997) findings. Furthermore, three other interesting findings, the superiority of the test strategy, the merging of the glides, and the replacement of the glides, were discovered in the syllable-contraction task.