The Phonetic And Phonological Properties Of The Neutral Tone In Changsha Chinese

This thesis reports the acoustic experiment on the neutral tone in Changsha Chinese and gives a phonological interpretation to the acoustic and auditory properties of the neutral tone, makes an introduction to the design of my phonetic experiment on the tone system in Changsha Chinese, presents the findings of this experiment, and also gives a systematic analysis to the findings within the framework of autosegmental phonology. Up till now, the investigation into the tone in Changsha Chinese is perceptually-based. No researcher makes a phonetic experiment on the tone system in Changsha Chinese.The experimental items include the fundamental frequency (Fo), the rimal duration, the amplitude and the formants. The findings show that the F₀ values of yinping, yangqu and shang are level and the F₀ values of yinqu, yangping and ru are rising; in the comparative pairs of disyllabic expressions, the duration of neutral-toned second syllable is shorter than that of the full-toned second syllable, 60-70% to be exact; the energy of the neutral-toned syllable is as 60-70% as that of the full-toned syllable due to the shortness in duration, while the amplitude can be the same, weaker or stronger; the surface value of the neutral tone is a mid level and it is not affected by its preceding full tone; the change in formants of neutral-toned syllable is not obvious.Based on the findings of the phonetic experiment, I conclude that in Changsha Chinese, besides the neutral tone, the tone system consists of six full tones, four even tones and two rising tones: yin ping [33], yin qu [55], yang ping[13], yang qu [11], shang[44], ru [24]. The actual tone value of yin qu is [45], but if it is analyzed to be a level tone [55], the tone system is more typologically reasonable. The six full tones underlyingly are three tones: H, L and LH, composed of two tone elements [H] and [L]. There are two registers—[+upper] and [-upper] and each register contains these three tones. When the register makes an effect on the three tones, the six tones appear on the surface representation.There is no spreading in the tone-obtaining process of the neutral tone and the neutral tone obtains a mid level tone by default rule in Changsha Chinese. Because ofthe difference in duration, I assume that mora is the tone-bearing unit in Changsha Chinese. Consequently, although the pitch value mid of the neutral tone seems to be identical to the full tone yang ping 33, they are different tones. The rime of the syllable carrying the yang ping links two morae, while that of the syllable carrying the neutral tone links only one mora.The neutral tone being always shorter than the full tone in duration can be the proof that the shortness in duration is an essential property of the neutral tone. Because the amplitude of the neutral tone can be weaker, identical or stronger than that of the full tone, the weakness in amplitude is not an essential property of the neutral tone. The difference between the neutral tone and the full tone in vowel quality, which is shown by the formants, is not obvious. Therefore formants is not an essential property of the neutral tone either.Both low and mid have the possibility to be the default tone value, and the essential feature of default tone is [-upper]. In a two-level tone system, low is the default tone, while in the three-level or four-level tone system, as Changsha Chinese, the feature [H]/[L] should be taken into consideration, and mid, which has the feature [-upper, H], is default.This research is significant in these aspects: the structure of contour tone, the TBU in Chinese, the default tone in Chinese and the tone operative mechnism in Chinese.