Dilley, L., Shattuck-Hufnagel, S., & Ostendorf, M. (1996) Glottalization of word-initial vowels as a function of prosodic structure. Journal of Phonetics, 24, 423-444.
Summary: Angela
The phenomenon of glottalization of word-initial vowels is discussed in this paper by exploring its preference of occurrences at both prosodic and segmental levels. Prosodically, the authors were interested in whether prosodic boundary types would influence the frequency of glottalization in word-initial vowels. Specifically, the authors would like to see if intonational phrases and intermediate phrases had different manifestations in glottalization. In addition, the authors wanted to see whether accented vowels showed a larger proportion of glottalization compared with unaccented vowels. Segmentally, the preceding and the following context in which the word-initial vowels occurred were included to see if different segments or pauses would influence the rate of glottalization occurrence. FM radio news speech were collected and analyzed to answer the aforementioned questions. Glottalization was identified by both perceptual and acoustic means. Results of this study showed that word-initial vowels that occurred at larger prosodic boundaries such as an intonational phrase were more inclined to be glottalized compared with those that appeared at a smaller prosodic boundary such as an intermediate phrase. With regards to pitch accents, there is a tendency for glottalization to co-occur more with accented word-initial vowels than with unaccented word-initial vowels. At the segmental level, glottalization occurred the most when a pause preceded. Other segmental factors did not seem to influence the glottalization of word-initial vowels when the vowel occurred in the phrase-initial position. For non-phrasal initial position, it was found that the more sonorant the preceding segment was, the more possible a word-initial vowel was glottalized. According to the authors, the results of this study highlight the importance of taking prosodic structures more seriously in the analyzing of linguistic data.
No comments:
Post a Comment