| This study attempts at a systematic description of voice onset time (VOT) for word-initial stops in mono-syllabic words in 7 Xiang dialects(NXDs including NXD, YYD and XYD; OXDs including SFD, XXD, XND and CBD), and makes further verification and discussion on related hypotheses. The research questions include 1) What pattern do the word-initial stops display for each dialect in terms of VOT? 2) Can VOT serve as an effective acoustic parameter to distinguish stops in Xiang dialects? 3) Is there a trade-off between place of articulation (POA) and VOT in all the dialects?8 native speakers (4males and 4 females) were selected and prompted to read a wordlist of about 108-150 monosyllabic words in each dialect, whose speech were recorded by a ZOOM recorder and a Lenovo computer. The VOT values of a full set of stops for each dialect were measured and analyzed, and based on these statistics the pattern of the stops was displayed from the perspective of acoustics. Meanwhile, the effectiveness of VOT's distinguishing feature and the trade-off between POA and VOT are to be verified.The major findings of this study were as follows. First, voiced stops do exist in all the sampled Old Xiang dialects, and they are absent in all the sampled New Xiang dialects. Second, statistics show that typical VOT patterns, namely negative lead, short positive lag and long positive lag can differentiate the stops with either voicing or aspiration in Xiang dialects. For example, voiceless unaspirated and voiceless aspirated stops in New Xiang dialects have a mean VOT value of+15ms and +90ms respectively, in accordance with those for Mandarin Chinese and Cantonese, and also similar to those in English and Dutch in former studies. In Old Xiang dialects, voiced unaspirated, voiceless unaspirated and voiceless aspirated stops have a mean VOT value of -80ms,+15ms and +80ms respectively, very similar to those in Thai and Hindi, which also make a tripartite categorization of stops. Third, the trade-off effects between POA and VOT are uncertain. To be specific, no dialect completely conforms to such effects, and most demonstrate conformity in some but not all stop categories. Voiceless unaspirated stops in Xiangyin and Yiyang dialects, voiceless aspirated stops in Changsha dialect, voiceless stops in Chengbu and Shuangfeng dialects, voiceless unaspirated and voiced stops in Xinning dialect all show such trade-off. In some of the dialects, these effects are partially realized in that VOT values for velar stops may be significantly longer than those for alveolar and bilabial stops; however, such effects are missing between alveolar and bilabial stops. Shuangfeng is a case in point:in the voiceless unaspirated and voiceless aspirated stops, velars have the longest VOT among all the stops, but there is no sign of such trade-off between the alveolars and bilabials. There are also cases which even demonstrate the opposite effects. For example, in both Shuangfeng and Xiangxiang dialects, voiced velar stop has the smallest VOT values. In short, our analysis has unsystematic results concerning the POA and VOT trade-off effects, which might suggest such effects be at best interpreted as language-specific.This study has the following contributions. First, it carries out an acoustic analysis of the stop patterns for representative New and Old Xiang dialects for the first time. Second, it supplies acoustic evidence to confirm the pervasive existence of voiced stops in Old Xiang dialects, a point with only impressionistic data in former studies. Third, it draws on statistic data to confirm that VOT is operative in defining stops in Old and New Xiang dialects, which adds to the wealth of cross-dialectal VOT studies. Fourth, it employs data in Xiang dialects to dismiss the universality of the trade-off effects between POA and VOT. The current study can also furnish the description of Xiang dialects, and the results on VOT can be of value in the synthesis and recognition of these dialects in the future. |
PDF Full Text Request