This dissertation aims at constructing a description of reduplication in Old Chinese, developing a generative theory of morpho-phonological interaction to account for the formation of the reduplication patterns, and re-examining general reduplication theories and issues of other linguistic components by drawing lessons from Old Chinese reduplication. The investigation of the source data reveals that Old Chinese reduplication has four basic patterns: progressive reduplication with either "smallness" or "vividness", retrogressive pattern with "repetition", fission reduplication with "specialization", and total reduplication with a vivid impression (a parasitic sense). The formation of the reduplication patterns results from the interaction between morphology and phonology. With motivation from semantics, the monosyllabic base is reduplicated as two identical syllables, which undergo further modification. 1) Since the reduplicative form with "diminutive" or "vividness" is semantically undecomposable, OOP (One Syllable One Meaning Principle) forces the two syllables to sound like one, which is achieved by raising the sonority of the onset of the second syllable. As such, the progressive pattern arises. 2) For the same reason, the reduplicative form with "specialization" has the same shape as the progressive at one stage. Pressure from the system thus compels it to undergo further modification, eventually producing the fission pattern. 3) The reduplicative form with a vivid impression is not under the control of OOP; thus it can keep its two identical syllables intact, yielding total reduplication pattern. 4) Reduplicative verbs are semantically decomposable; thus OOP does not come into effect. That the form is actually modified stems from the pressure of an already-existent total reduplication pattern, while this modification of the first rhyme is determined by quasi-iambic stress. This interaction produces a retrogressive pattern. This study sheds light on reduplication processes in general and other linguistic issues. During reduplication, full reduplication occurs first; then the reduplicant is modified. That reduplication operates on the interface between morphology and phonology is a universal phenomenon, but how this operation proceeds is language-specific. The consistent distinction between Type A syllables and Type B syllables seen in Old Chinese reduplication patterns indicates the unreasonableness of reconstructing a "medial" yod for Old Chinese.

This dissertation aims at constructing a description of reduplication in Old Chinese, developing a generative theory of morpho-phonological interaction to account for the formation of the reduplication patterns, and re-examining general reduplication theories and issues of other linguistic components by drawing lessons from Old Chinese reduplication. The investigation of the source data reveals that Old Chinese reduplication has four basic patterns: progressive reduplication with either "smallness" or "vividness", retrogressive pattern with "repetition", fission reduplication with "specialization", and total reduplication with a vivid impression (a parasitic sense). The formation of the reduplication patterns results from the interaction between morphology and phonology. With motivation from semantics, the monosyllabic base is reduplicated as two identical syllables, which undergo further modification. 1) Since the reduplicative form with "diminutive" or "vividness" is semantically undecomposable, OOP (One Syllable One Meaning Principle) forces the two syllables to sound like one, which is achieved by raising the sonority of the onset of the second syllable. As such, the progressive pattern arises. 2) For the same reason, the reduplicative form with "specialization" has the same shape as the progressive at one stage. Pressure from the system thus compels it to undergo further modification, eventually producing the fission pattern. 3) The reduplicative form with a vivid impression is not under the control of OOP; thus it can keep its two identical syllables intact, yielding total reduplication pattern. 4) Reduplicative verbs are semantically decomposable; thus OOP does not come into effect. That the form is actually modified stems from the pressure of an already-existent total reduplication pattern, while this modification of the first rhyme is determined by quasi-iambic stress. This interaction produces a retrogressive pattern. This study sheds light on reduplication processes in general and other linguistic issues. During reduplication, full reduplication occurs first; then the reduplicant is modified. That reduplication operates on the interface between morphology and phonology is a universal phenomenon, but how this operation proceeds is language-specific. The consistent distinction between Type A syllables and Type B syllables seen in Old Chinese reduplication patterns indicates the unreasonableness of reconstructing a "medial" yod for Old Chinese. / Arts, Faculty of / Asian Studies, Department of / Graduate

韻素是韻律學在分析音節輕重理論系統中的最小單位。在韻素音步系統的語言中，韻素數量的多少決定音節的輕重。韻素在韻律系統中的作用，與音節中韻素數量的多少，元音響度的大小有直接關係。 / 自清代段玉裁發現"吾輕、我重"以來，這方面的探討一直未曾間斷，之前的研究雖提供了寶貴的材料，但卻不能很好地解決諸如輕重表現形式、語音與句法位置的關係和古今演變等方面的問題。而將韻素理論引入會使很多難題迎刃而解。本文在此基礎上，以上古漢語第一人稱代詞"吾"、"我"為研究對象，研究二者在先秦及兩漢以後的分佈特點及韻律輕重的表現，指出： / "我"是先秦漢語第一人稱代詞的標準形式，"吾"是在特定語音環境要求下出現的弱化形式。當句中位置具備語音弱化條件時，才使用弱化形式"吾"，其他位置使用標準形式"我"。因此，"吾"、"我"實為一詞的兩個不同形式。 / 本文主要包括以下幾個內容： / 第一章，緒論。本章分為三小節。第一小節，韻素理論的研究概況。概述韻素理論的基本觀點及理論體系。第二小節，漢語"吾輕、我重"研究綜述。本節對之前學者研究"吾輕、我重"的歷史進行總結和評價。第三小節，提出本文的研究目的、主要內容及方法。 / 第二章，第一人稱代詞"吾"、"我"的分佈現象及數字統計。本章著重描述先秦及兩漢以後重音位置上:"吾"、"我"出現的規律，分別考察以下幾個位置的分佈：語音空拍前位置、對比結構焦點位置及"前輕式"結構。 / 第三章，"吾"、"我"互補分佈特徵的韻律分析。本章分五部份。一、引入韻律相關理論，說明韻素輕重形式實現的語音條件。二、指出上古漢語"吾"、"我"的韻素特徵及二者的差異。三、"對"、"吾"、"我"互補分佈現象及統計結果進行韻律分析和解釋。四、闡釋"吾"、"我"對立現象的古今演變機制。五、重新審視"吾"、"我"對應屬性反映出的語音機制。 / 第四章，結論。我們概括了全文的觀點，強調：在韻素理論的框架下，本文提供了一系列發掘新材料的操作方法和驗證手段。同時，本文發現的新材料支持了上古韻素敏感的語音系統的理論和假說，同時支持了漢語韻律系統轉變的觀點，進一步論證了單音節音步是如何在上古漢語韻律系統中發揮功能的，同時也為漢語語音史的討論開闢了一個嶄新的領域。 / Mora is the smallest unit in the prosodic analysis. In quantity-sensitive languages, the weight of a syllable depends on the number of morae. The realization of the morae is relevant to the quantity and sonority of the vowel in the syllable. / Since Duan Yucai 段玉裁 proposed that "Wu 吾 is weak, while Wo 我 is strong" in the Qing Dynasty, the difference between the two pronouns has been extensively discussed. Although previous research has provided us with valuable materials, they cannot answer some crucial questions such as: what is the representation of weak form and strong form, what is the correlation between the stress and the syntactic positions in classical Chinese, and why the phonological contrast disappeared in the Han Dynasty. Within the framework of prosodic phonology, this thesis discusses the rule of the distribution of the first person pronouns Wu and Wo in archaic Chinese and proposes that: / Wo is the default form of the first person pronoun in archaic Chinese, while Wu exists as its reduced form in the specified phonological environments. In the language system, Wo is chosen as the basic form. Wu only occurs when the syllable is light, Wo occurs elsewhere. Wu and Wo are variants of the same pronoun. / The first chapter briefly introduces the relevant theoretical research principles on morae in section one, and then reviews some previous accounts for the contrasts of Wu and Wo in section two. The research purpose and methodology will be proposed in section three. / The second chapter shows statistics of the complementary distribution of Wu and Wo in archaic Chinese, especially focusing on the positions before the pause, contrastive stress and iambic structure. / The third chapter mainly presents the phonological and prosodic condition analysis to the phenomena listed in the chapter two. On the basis of the prosodic theories such as Nuclear Stress Rule (NSR), Invisible Condition (IC) and Default Rule (DR), I argue that the regularities of distribution result from the phonological distinction between Wu and Wo. Wo should be analysed as the underlying form, and the Wu represents the variable form. To give an explanation for the phonological process and mechanism in the evolution of these two pronouns, I gave a preliminary description of their changes after Han dynasty. This chapter also explores several properties of Wu and Wo to rethink how their contrast reflects some basic phonologic mechanisms / The fourth chapter summarizes the prosodic study of the two first person pronouns in this dissertation and emphasizes that: a set of evidence is provided to support the prosodical explanation, as well as to validate the hypothesis of Weight-sensitive System‘ in archaic Chinese. It is also suggested that aside from its significant role in prosodic system, monomoraic syllable foot could lead to great diachronic changes of the language. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / 趙璞嵩. / Parallel title from added title page. / Thesis (Ph.D.) Chinese University of Hong Kong, 2014. / Includes bibliographical references (leaves 122-129). / Abstracts in Chinese and English. / Zhao Pusong.