Tutuba語の歯茎閉鎖音と〔n〕について : 音韻論的観点から

Abstract

This work sets out to examine how the Tutuba language phonemes /d/ and /t/ should be described phonologically compared with previous research. Tutuba language is one of the vernaculars of the Republic of Vanuatu and is now supposedly spoken by less than 500 people. This study answers the following three points. 1. Is the sound 〔nḍ〕 linguo-labial? 2. Although /t/ is phonologically described as 〔t〕 in the previous research, is it more appropriate to be shown as both 〔●〕 and 〔t〕? 3. Is 〔n〕 used in 〔nḍ〕 one segment or an additional sound to 〔ḍ〕? To analyze the native informant data, I measured Fl, F2, F2-F1 of /d/ and /t/ on the spectrograms as well as the length of 〔n〕 and 〔n〕. The empirical analysis leads to the following conclusions. (Each number corresponds to the previous list.) 1'. The sound 〔nḍ〕 is not linguo-labial. The spectrogram does not show the same transition as linguo-labial. 2'. It is more appropriate to use 〔●〕 and 〔t〕 to show the phoneme /t/. Spectrograms makes it clear that the points of articulation of /t/ in /tete/ (father) and /tasi/ (brother) are different. Although previous research describes 〔t〕 only, judging from the fact that /t/ in /tete/ is articulated slightly forward of the point of articulation in /tasi/, 〔●〕 should be added phonologically. 3'. In the Tutuba language, the consonant cluster is not uncommon and sometimes appears at the head of the word. Although 〔n〕 used in 〔nḍ〕 is revealed to have approximately the same time length as 〔n〕, 0.05 sec, considering that 〔n〕 always appears immediately in front of 〔ḍ〕 and the combination between an additional sound and a consonant is supposed to be constructed by a nasal and a stop, this implies the 〔n〕 is not a segment but an additional sound.