英文摘要 |
As in other languages with 'voiced' stops, the phonetic realization of Taiwanese /b/ and /g/ is highly variable from one context to another. Zhang (1983) even claims that they are replaced by homorganic nasals [m] and [ŋ] when the syllable is closed by a nasal. The positional variants were examined using oral and nasal airflow recordings. Three types of syllable structures were chosen with initial voiced stops in CV, CVN, and CVC. The target syllables are placed in utterance initial position and in medial positions after nasals, vowels, and voiceless stops. Utterance initially, there is a voiceless prenasalization, although its occurrence is strongly speaker-dependent When preceded by nasals, the prenasalized voiced stops alternate with post-stopped nasals. When preceded by vowels, the prenasalization was found only in nine out of630 tokens. |