作为第三位肖像印在英镑钞票上的女性历史人物，简·奥斯汀在英国的地位不言而喻。耶鲁大学教授布鲁姆将其与莎士比亚相提并论，称之为反讽大师。奥斯汀的成名作Pride and Prejudice在英国小说家毛姆看来是她最完美的作品。这部小说描绘了几类婚姻观，对其中体现的人性弱点和不公正的男性继承制进行了反讽。反讽将批评隐藏于话语字面义之下，这既是奥斯汀的语言风格，又是其自卫的手段。正如哲学家克尔凯郭尔所言，没有反讽就没有人类生活。反讽的重要性可见一斑。 遗憾的是，国内外对反讽做了大量研究，但对反讽翻译研究关注不够。因此，本文选取大陆杨缤和王科一、香港东流、台湾陈玥菁及刘佩芳和邓盛铭译本为分析对象，从反讽的语言信号出发，基于范·路文兹瓦特翻译转换模型探讨Pride and Prejudice中言语反讽和结构反讽的汉译，以描述翻译转换对反讽效果的影响和揭示译者的翻译规范。本文关注词/短语等微观层面的反讽及翻译转换的同时，更关注宏观层面的反讽，即对某一主题的反讽，将宏观反讽分为对爱情婚姻观中男女主人公傲慢与偏见的反讽和对功利婚姻观的反讽，对于后者，采用概念隐喻理论将功利婚姻观与商业交易联系起来，考察大陆和港台译本宏观上的主题意义是否与原作对等。 本研究发现，五个译本存在诸多共同之处，大多数翻译转换出现了宽泛化、平实化、中性化、明晰化、语义窄化和情态/感量值低化等主观化解读，频次远高于客观化解读，折损了原文反讽语言信号的文体特征，减弱或丧失了反讽效果。主观化解读总体上存在一致性，是霍姆斯所论的结构性翻译方法。主观化解读表明未遵循源语规范，然而译者这么做并非为了适应目标语规范，译文既非图里所论的充分翻译，又非可接受性的翻译。五个译本的不同之处在于，杨译和王译未见漏译，是范·路文兹瓦特所论的整体性翻译，而其余三个译本为非整体性翻译。总体而言，大陆译本在主题意义方面比港台译本更贴近原作。 本文对反讽理论和翻译理论均有贡献。反讽理论上，揭示了结构反讽与言语反讽的相通之处，拓展了结构反讽的反讽对象，论证了反讽语言信号的重要性和对某些反讽话语的不可或缺性，理清了反讽信号和反讽源的关系，探讨了语义学、语用学和认知语用学在判断反讽话语方面的互补关系。翻译理论上，改进了范·路文兹瓦特翻译转换模型，提供了当原文译素存在至少两项符合语境的意义时确定核心译素、原文译素存在两种感情色彩时确定感情色彩、译文存在多项翻译转换时确定重要等级的方法，探讨了从翻译转换的类型和频次特征中提炼译者翻译规范的方法，并指出了反讽翻译应避免译文与原文所指相同但主题意义削减。此外，基于五个译本的得失，提出了反讽的翻译策略。

As the third historic female appearing on the pound note, Jane Austen’s importance in England is self-evident. Allan Bloom, professor with Yale University, compared her with Shakespeare and called her master of irony. Pride and Prejudice, the work that made Austen famous, is her best novel, according to English novelist William Somerset Maugham. The novel centers on marriages of different types and targets irony at human’s weaknesses and unfair inheritance law. Austen’s irony, with criticism hidden under its surface meaning, is her style and weapon for self-defense, just as what Kierkegaard remarked that without irony there was no human life. The importance of irony goes without saying. Worldwide much attention has been paid to the study of irony but far less to that of the translation of irony, which calls for more attention in this field. This dissertation, therefore, selects as texts for analysis Chinese versions of Pride and Prejudice translated by Yang Bin and Wang Keyi from mainland China, Dong Liu from Hong Kong, Chen Yuejing, Liu Peifang and Deng Shengming from Taiwan, approaches irony from its linguistic signals and uses van Leuven-Zwart’s translation shift model to study the translation of verbal irony and structural irony in Pride and Prejudice, unveil the changes or shifts in translation and their impacts on ironic effects, and unlock translation norms. Apart from concentration on linguistic signals of irony and its translation at microstructural level, this dissertation focuses on irony at macrostructural level and explores irony toward two themes, the pride and prejudice of the male and female protagonist and mercenary marriages. As regards mercenary marriages, this dissertation employs conceptual metaphors to link together marriages of this type with business deals and examines whether its five Chinese versions by translators from mainland China, Taiwan and Hong Kong have achieved equivalence in the thematic meaning of Pride and Prejudice. This dissertation has found many similarities among the five Chinese versions of Pride and Prejudice. Most translation shifts feature tendencies toward generalization, plainness, neutralization, explicitation, meaning deduction and modal/emotional value reduction. These subjective interpretations of the source text far outnumber objective ones and cut the stylistic features of linguistic signals of irony of the source text, thus reducing or losing the ironic effect. The subjective interpretations on the whole are coherent and consistent, and the translation method, according to Holmes’ definition, is structural translation. Subjective interpretations of the source text mean departures from the norm of the original language, but translators’ so doing does not aim at complying with the norm of the target language. The five Chinese versions, therefore, are neither, according to Toury’s terms, adequate translation, nor acceptable translation. The differences among the five Chinese versions are that Yang’s and Wang’s translations feature integral translation, a term put forward by van Leuven-Zwart, since nothing was omitted or deleted. The rest three are non-integral translation. Chinese versions by mainland China translators, on the whole, are closer to the source text in its thematic meaning than those by Taiwan and Hong Kong translators. This dissertation contributes to theories on irony. It reveals the connection between verbal irony and structural irony, expands the scope of victims of structural irony, proves the importance of linguistic signals of irony and their indispensability in certain circumstances, clarifies the relations between signals of irony and sources of irony and delves into the complementarities of semantics, pragmatics and cognitive pragmatics in identifying ironies. This dissertation also contributes to translation studies. It improves van Leuven-Zwart’s translation shift model, offers methods for determining architranseme when a source text transeme has at least two meanings compatible with context, ascertaining emotional coloring when a source text transeme has two, deciding scales of importance when several translation shifts occur in one translation and concluding translation norms on the basis of translation shifts, and points out that translation of irony should seek equivalence in the thematic meaning of a novel and avoid seemingly equivalence in what the word denotes. In addition, it puts forward strategies for translating ironies on the basis of the merits and demerits of the five Chinese versions of Pride and Prejudice.