Taiwanese final stops and following initial voiced stops and nasals

Abstract

The places of articulation of unreleased Taiwanese final stops, e.g., /p,t,k/ and glottal stop, have been known to assimilate into that of initial voiceless stops in following a syllable, depending on the speed of articulation (Peng, 1997). Besides assimilation in place of articulation, nasalization can be observed across syllables between voiced stops and nasals (Pan, 1999). However, little is known about the assimilation in place of articulation, voicing, and nasalization between Taiwanese final stops, and following voiced stops and nasals across different prosodic boundaries. This study uses EPG, airflow, and acoustical data to investigate the coarticulation between Taiwanese final stops, and initial nasals, initial voiced stops, across syllable boundary, morpheme boundary, phrase boundary of narrow focus, and intonation boundary. Preliminary results showed that Taiwanese final stops are lost in certain contexts. There is assimilation in place of articulation between unreleased final stops and initial voiced stops and nasals. No progressive nasalization is observed between final stops and initial nasals. Taiwanese final stops are easily coarticulated with following segments. Though the presence of final stops and short syllable duration are both vital cues to Taiwanese entering tones syllables, duration is a more invariant cue than the presence of final stops. [Project supported by NSC 87-2411-H-009-008.]