Abstract This thesis investigates the tonal variation of bisyllabic words in Hong Kong Cantonese, Macao Cantonese, and Zhuhai Cantonese. It focuses on the description of the direction and speed of tone merger and the analysis of the linguistically and socially constrained factors related to the tone merger. Finally, the prediction is also made on the variation trend of Cantonese tones in the three regions. The specific findings are as follows: The acoustic data were collected from 150 native Cantonese speakers from the IDADE DE three investigated regions. The results of phonetic transcription, growth curve analysis, and mixed-effect model of the acoustic data provide evidences for three tone mergers in Hong Kong Cantonese and Zhuhai Cantonese, namely, the high rising tone and mid- low rising tone (T2-T5), the mid-level tone and mid-low level tone (T3-T6), and the mid stopped tone and mid-low stopped tone (T8-T9). Meanwhile, the data reveals that the Macao Cantonese shows a different merger pattern. Apart from the three merger pairs found in Hong Kong and Zhuhai, the low-falling tone and low-level tone (T4-T6) Macao Cantonese, but onl are also found merge in Macao Cantonese, but only at the second syllable. Additionally, econd syllab the thesis also makes an overall comparison of the three regions and finds that the three tone merger pairs have merged in Zhuhai, but they are still in the merging process in Hong Kong and Macao. Macao Cantonese is undergoing more rapid merging process than Hong Kong Cantonese. The quantative method has been used to analyze the direction and speed of tone merger. As for the direction of tone merger, T2 merges into T5, T6 merges into T3 and T9 merges into T8 in Hong Kong Cantonese and Macao Cantonese. T4 merges into T6 in Macao Cantonese. As for the speed of tone merger, the ranking of merging speed at the first syllable in Hong Kong Cantonese is T2-T5>T3-T6>T8-T9; the rankingof merging speed at the second syllable is T3-T6>T8-T9>T2-T5. The ranking of merging speed at the first syllable in Macao Cantonese is T2-T5>T3-T6>T8-T9; the ranking of merging speed at the second syllable is T2-T5>T3-T6>T8-T9>T4-T6. The thesis also discusses the effect of tonal coarticulation on tone merger. Specifically, by analyzing the influence (i.e., carry-over effects and anticipatory effects) and magnitude of tonal coarticulation on target tones, the merger of T2-T5 of the first syllable in Hong Kong Cantonese and Macao Cantonese is found facilitated by tones with high onset of the second syllable, but at the second syllable, T2-T5 is found DADE facilitated by the tones with high offset of the first syllable. The merger of T8-T9 of the first syllable is found facilitated by tones with the middle onset tones of the second syllable. As for the T4-T6 merger only in Macao Cantonese, the merger is found facilitated when following the first syllable that carrying tones with low offset. This thesis uses the mixed-effect model to investigate the influential factors on tonal variation of Hong Kong Cantonese and Macao Cantonese. The data collected in this thesis suggests that the tone duration affects tone mergers except for T8-T9. The syllable position affects tone mergers except for T4-T6 in Macao Cantonese and T2- T5 in Hong ] g Kong Cantonese. The tone of first syllable affects tone mergers of T2-T5 in Hong Kong Cantonese and T4-T6 in Macao Cantonese. The interaction between nasal coda and vowel height affects the mergers of T3-T6 in Hong Kong and Macao Cantonese. Vowel height affects the merger of T8-T9 in Macao Cantonese. The interaction between stop coda and vowel height affect the merger of T8-T9 in Hong Kong Cantonese. Age is an important social factor affecting the mergers of Cantonese tones in the two regions. The interaction between age and speech style affects the mergers of T3-T6 and T8-T9 in Hong Kong Cantonese as well as T8-T9 in Macao Cantonese. The interaction between speech style and education level is only significant in the merger of T2-T5 in Hong Kong Cantonese. The factors of gender and social class show no significant effect on the tone mergers in Hong Kong and Macao Cantonese. This thesis also finds that identity category affects the merger of T4-T6 in Macao Cantonese. In terms of the merger mechanism, this thesis holds that the mechanism of tone merger in Hong Kong Cantonese and Macao Cantonese should be merger by transfer mechanism which belongs to one of the three merger mechanisms proposed by Labov. It is widely acknowledged that variationist sociolinguistics mainly focuses on the variation of consonants and vowels, with limited attention being paid to the tonal variation study. The main research object of tonal variation is citation tone, and rarely involves the tonal variation of bisyllabic words. This thesis discusses the tonal variation of bisyllabic words in Hong Kong Cantonese, Macao Cantonese and Zhuhai Cantoncse from synchronic level in apparent time. Based on age, this thesis uses statistical methods to investigate the tone targets of different tonal contours and clarifies the direction of tone merging in Hong Kong Cantonese and Macao Cantonese. Moreover, Using the method of vowel and consonant variation speed, this thesis NIVERSIDADE DE MA investigates the speed of tone merging in Hong Kong Cantonese and Macao Cantonese so as to capture the stage of tonal variation more clearly. The empirical vini materials of this thesis also contribute to theoretical research. In explaining the mechanism of tonal variation, it is found that tone merging is affected by tonal coarticulation. Aiming at the problem of tone merging in dialectology, this thesis uses the methods of sociolinguistic and the theory of sociophonetic to study, updates the results of tone merging in dialectology, interprets the constraint theory of language variation and change with sociophonetic, and provides a case of tone research for sociophonetic.

本文考察了港澳珠三地粵語雙字組詞語的聲調合併現象,著重探討三地聲 調的變異方向和速度,從語言和社會兩個因素分析了聲調合併的原因,並預測 了三地粵語聲調的演變趨勢。具體的結論如下: 本文綜合運用聽音轉寫、增長曲線分析、混合效應模型的方法,對三地150 名粵語母語者語音數據分析,確定了香港粵語雙字組詞語出現了三個聲調合併 變項,分別是陰上——陽上(T2-T5)、陰去——陽去(T3-T6)和下陰入——陽入(T8- IVE DADE D T9)。這三組聲調在珠海粵語中已經完成合併。澳門粵語比香港粵語、珠海粵語 多了陽平——陽去(T4-T6)變項,但該變項僅出現在後字位置。從區域比較的視 角來看,珠海粵語已經完成了三組聲調合併;香港和澳門緊隨其後,聲調變項 均在合併中,澳門快,香港慢, 禮知信 學 門大 本文運用定量的方法分析了港澳兩地粵語聲調的合併方向和速度。就合併 文 骂骂 方向來說,兩地粵語 T2-T5的合併方向為 T2 向 T5 合併,T3-T6 的合併方向為 T6 向 T3 合併,T8-T9 的合併方向為 T9 向 T8 合併。澳門粵語 ·T4-T6位於雙字 組詞語後字位置時,T4 向 T6 合併。合併速度方面,香港粵語 T2-T5、T3-T6 和 T8-T9 位於前字位置時合併速度快慢排序為 T2-T5>T3-T6>T8-T9;位於後字 位置時合併速度快慢排序為T3-T6>T8-T9>T2-T5。澳門粵語 T2-T5、T3-T6和 T8-T9 位於前字和後字位置時合併速度快慢排序均為T2-T5>T3-T6>T8-T9, 澳門粵語 T4 和T6位於後字位置時,合併速度慢於 T8-T9。 本文探討了聲調協同發音對聲調變異的影響,發現排除連讀變調的因素, 雙字組詞語聲調合併會受到協同發音的影響。具體來說,當香港粵語和澳門粵 語的 T2-T5 位於雙字組詞語前字位置時,後字高起點聲調會促進二者的合併;位於後字位置時,前字高終點聲調會促進二者的合併。T8-T9 位於前字位置時, 後字中起點聲調會促進二者的合併。澳門粵語 T4-T6位於後字位置時,前字低 終點聲調會促進二者的合併。 ERSI 本文利用混合效應模型的方法探討影響港澳兩地粵語雙字組詞語聲調變異 的語言因素和社會因素。語言因素方面,除T8-T9,聲調時長影響了兩地粵語所 有聲調合併變項;除澳門粵語 T4-T6 和香港粵語 T2-T5,音節位置影響了兩地 粵語所有聲調合併變項;前字聲調影響了香港粵語後字 T2-T5和澳門粵語後字 T4-T6的合併;元音高度與鼻音的交互作用影響了兩地粵語 T3-T6的合併,元 FRRRR 音高度影響了澳門粵語 T8-T9 的合併,元音高度與塞音韻尾的交互作用影響了 香港粵語 T8-T9的合併。社會因素方面,年齡顯著制約兩地粵語的聲調合併; 年齡與語體的交互作用影響了香港粵語 T3-T6、T8-T9 和澳門粵語 T8-T9 的合 併;教育程度與語體的交互影響了香港粵語 T2-T5 的合併;身份類別認同影響 了澳門粵語後字 T4-T6的合併。本文未發現性別和社會階層因素對两地粵語聲 調合併的影響。本文認為兩地粵語聲調合併變項的機制均屬於拉波夫提出的三 NIVERSIDADE DE MACAU 種合併機制中的“遷移合併” 變異社會語言學的主要研究對象為輔音和元音的變異,聲調變異較少。聲 調變異主要研究對象為單字調,鮮少涉及雙字組詞語的聲調變異。本文以“顯 象時間”的數據,描寫了港澳珠三地粵語雙字組詞語聲調的變異情況。本文在 年齡分層的基礎上,對不同調形的聲調目標分別統計,釐清了港澳兩地粵語聲 調合併的方向。不僅如此,本文還借鑒元音、輔音變異速度的方法,考察了港澳 兩地粵語聲調變項的合併速度,從而清晰展現兩地粵語聲調變異所處的階段。 文章的實證材料還為理論研究做出貢獻。文章在解釋聲調變異的機制方面,發現聲調合併亦受到協同發音的影響。本文針對方言學中聲調合併的問題,採用 社會語言學的方法和社會語音學的理論加以研究,更新了方言學中聲調合併的 結果,用社會語音學詮釋語言變異的制約理論,為社會語音學提供一個聲調研 究的案例。