以語料庫分析譯者翻譯風格:以愛倫坡短篇小說譯本為例

愛倫坡（Edgar Allan Poe）能成為中文世界美國文學的代表性作家，翻譯於其中扮演關鍵角色。基於此點，愛倫坡作品的譯本具有翻譯研究價值。綜觀目前愛倫坡譯本相關的翻譯研究，文本分析仍為主要研究方法。本研究則嘗試加入語料庫分析，將梁永安、陳福成與謝瑤玲三位譯者的譯本轉換為語料庫，運用AntConc、Sketch Engine和CRIE3.0等語料庫工具，分析三個不同版本的譯文在量性數據呈現何種特色；再結合文本分析，歸納出三位譯者的翻譯風格，並參照譯者寫下的譯序或譯後記，試圖證明譯者背景會影響其翻譯風格。結果發現譯者風格確實與其背景有關聯：三人之中僅有謝瑤玲研究英美文學，受此影響，其翻譯風格偏向異化，且透過謝瑤玲的譯序，可知其在翻譯過程追求形式對等與動態對等；梁永安與陳福成皆非英美文學學者，翻譯時不惜刪減、增譯或補充原文資訊，較偏向歸化。本研究所得結果支持翻譯學者勒菲弗爾（Lefevere）的觀點：「翻譯是一種重寫」；但也指出Bensimon和Antoine Berman重譯假說觀點過於片面，重譯本異化程度不僅取決於讀者熟悉異國文化與否，譯者的翻譯觀也決定譯本異化程度。

關鍵字

愛倫坡；重譯本；譯者風格；語料庫分析；異化與歸化

並列摘要

Over the years many studies have been devoted to examining translations of Edgar Allan Poe’s work into Chinese. Thus far, however, most scholars have followed a textual analysis approach when looking at these translations. This thesis employs a different method, that of corpus analysis. It uses three types of software (AntConc, Sketch Engine, and CRIE 3.0) to analyze the translations of three very different translators, the three translators being Liang Yung-an, Chen Fu-cheng, and Hsieh Yao-ling. Prior to conducting the corpus analysis, it was hypothesized that such an analysis could help to reveal the translation styles of the three translators. It was further hypothesized that their style has something to do with their educational background. After the corpus analysis was completed, it was found that the three translators do indeed have significant stylistic differences. Furthermore, it was found that the educational background of the translators does appear to have an influence on their stylistic choices. Among the three translators, Hsieh is the only one who adopts a foreignization translation strategy. This probably has to do with the fact that she was a professor who studied and taught Western literature, a job which places emphasis on the importance of the original text. As for Liang and Chen, they did not study Western literature in a formal setting, and both employ a domestication, rather than a foreignization, translation strategy. The findings of this thesis lend support to Lefevere’s claim that translation is a “form of rewriting.” At the same time, however, they also suggest that the “retranslation hypothesis” suffers from a number of defects.