臺語韻律對語音調擾動及音質的影響

Prosodic Effects on Tonal Perturbation and Voice Quality in Taiwanese

董鴻鈞(Hung-Chun Tung)

指導教授 ： 安可思

國立臺灣大學/文學院/語言學研究所/碩士(2008年)

https://doi.org/10.6342/NTU.2008.10049

摘要

根據過去研究指出，語音的發聲動作與聲學表現會在韻律顯著的地方增強。然而，針對韻律與子音產生的語調擾動以及音質的關係方面的研究甚少。本論文旨在探討台灣閩南語中語調擾動與韻律單位開頭的增強現象，以及韻律與音質之間的關係。
三位台灣閩南語發音人HZM, WHX, 與ZZJ參與本研究的錄音。研究音節包含台灣閩南語的高平調、高降調、低降調、升調，以及中平調，並以台語中的無聲送氣、無聲不送氣，以及有聲塞音作為研究音節的開頭子音。韻律位置包含句子 (utterance) 、語調詞組(intonational-phrase) 、小詞組(small-phrase) 、 詞(word) 、以及音節(syllable)的開頭，測量的項目包含目標音節長度5%與目標音節前一個音節長度95%上的基頻(fundamental frequency)、 第一泛音與第二泛音音強之差 (H1 - H2) 、 第一泛音與最靠近第一共振峰之泛音音強之差 (H1- A1) 、 第一泛音與最靠近第二共振峰之泛音音強之差(H1- A2) 、第一泛音與最靠近第三共振峰之泛音音強之差(H1 - A3)。
語調擾動方面的結果顯示韻律單位開頭的增強現象並不顯著，可能是因為台灣閩南語是音調語言，如果語調擾動在韻律單位開頭的位置有增強的現象，那麼聽者有可能在語音辨識上面有所困難。
另一方面，就音質而言，HZM與ZZJ在目標音節前一個音節長度95%上的H1- A1與H1- A2會隨著韻律強度增強而增加，也就是說，音質在韻律單位末端會隨著韻律強度增強而變得更為氣化(breathy)。至於目標音節長度5%的音質，三個發音者並沒有表現出隨著韻律強度而改變的結果。
本研究是第一個研究語調擾動是否有在韻律單位開頭位置增強的論文。 我們的結果與 Jun (1996)針對韓語、法語、與英語等非音調語言所做的研究相反。我們假設韻律單位開頭位置增強可能僅在特定語言或是特定發聲動作存在。就音質而言，我們發現H1- A1與H1- A2是台灣閩南語中音質改變的顯著指標，這與Esposito(2007)針對福州話的研究結果相似。

關鍵字

台語 ； 音調擾動 ； 韻律 ； 音質

並列摘要

Many studies have observed that several articulatory gestures in speech production and their acoustic output tended to be strengthened at prosodically prominent positions; however, the relationship of tonal perturbation of stop consonants and the voice quality with the prosodic hierarchy have not been explored yet. The current study aims to investigate the prosodic domain strengthening effect in the tonal perturbation of stop consonants and the relationship between prosody and voice quality in Taiwanese.

Three native Taiwanese speakers, HZM, WHX, and ZZJ participated in the recording of this study. The target syllables carried high-level, high-falling, low-falling, rising, and mid-level tones, with voiced, voiceless aspirated, and voiceless unaspirated stops as the onset consonant. The target syllables were located at the initial position of utterance, intonational-phrase, small-phrase, word, and syllable. The fundamental frequency (F0), the differences in the amplitude between the first harmonic (H1) and the second harmonic (H2) (H1 - H2), between the first harmonic and the harmonic closest to F1 (H1- A1), between the first harmonic and the harmonic closest to F2 (H1- A2), and between the first harmonic and the harmonic closest to F3 (H1 - A3) were measured at the 5% time point of target syllables, and at the 95% time point of the syllable immediately preceding target syllables.

Results on the tonal perturbation show that the prosodic domain strengthening effect on tonal perturbation was not significant, which might be attributed to the fact that Taiwanese is a tonal language. If the tonal perturbation were strengthened, it is possible for listener to process the linguistic input with difficulty.

On the other hand, H1 - A1 and H1- A2 at the 95% time point of the syllable immediately preceding target syllables increased as the prosodic boundaries became stronger, i.e., the voice quality became breathier at the final of prosodic domains when the prosodic domain was larger. As for the voice quality at the 5% time point of target syllables, the three speakers did not show any obvious change in voice quality in terms of the prosodic hierarchy.

The study is the first study to investigate whether tonal perturbations would be strengthened at the boundary of larger prosodic constituents. Our study shows a contrastive results with Jun (1996), where atonal languages such as Korean, French, and English. We might postulate that domain-initial strengthening effect is language-specific or specific to some speech articulatory gestures. Concerning the voice quality, we found that H1- A1 and H1- A2 were successful indicators for the voice quality change in Taiwanese, which is similar with the Esposito (2007)'s findings in Fuzhou dialect.

並列關鍵字

Taiwanese ； tonal perturbation ； prosody ； voice quality

參考文獻

Thurgood, G. (1993). Phan Rang Cham and Utsat: Tonogenetic themes and variants. In K. J. G. J. A. Edmondson (Ed.), Tonality in austronesian languages (pp. 91-106). Honolulu: University of Hawaii Press.

Blumstein, S. E., & Stevens, K. N. (1981). Phonetic features and acoustic invariance in speech. Cognition 10(1-3), 25-32

Byrd, D., & Saltzman, E. (1998). Intragestural dynamics of multiple prosodic boundaries. Journal of Phonetics, 26(2), 173-199.

Byrd, D. (2000). Articulatory vowel lengthening and coordination at phrasal junctures. Phonetica, 57(1), 3-16.

Caisse, M (1982). Cross-linguistic differences in fundamental frequency perturbation inducedvoiceless unaspirated stops. Unpublished master thesis, University of California, Berkeley.