從句長與句中成份觀察譯者語言選擇模式－－基於中日雙向平行語料庫之描述性研究

Exploring Translation Behavior Regarding Sentence Length and Sentence Constituents: A Descriptive Study Based on Chinese-Japanese Bi-directional Parallel Corpus

本研究採取語料庫語言學的質性以及量化的研究手法，建置雙向、對比、特定文類的中日日中平行語料庫，描述翻譯文本與非翻譯文本在句子與句中成份的使用特徵，並探討中文與日文在語言使用習慣不同的情況下，譯者的語言選擇的傾向，藉此推演出翻譯行為的選擇模式。由語料分析結果得知，譯者傾向在不違背目標語的語言使用習慣之下，接受、沿用源文的語言表現，不會選擇複雜而特殊的語言表現，譯者所採取的策略因而累積、堆砌出翻譯文本的特徵，使得翻譯文本與非翻譯文本在語言使用上呈現不同的篇章構成的特徵。本研究透過分析結果建構出一套翻譯行為特質的理論解釋，希望在中日翻譯實踐以及翻譯教育的應用方面，提供審視譯文、檢討翻譯策略的參考依據。

This research investigates the features of translated text and non-translated texts through analyzing the sentence length and constituent format in Chinese to Japanese and Japanese to Chinese translation. It uses a corpus-based methodology and sets up a model to describe translators' linguistic choice patterns. The small-scaled bi-directional, comparable and parallel corpus is constructed to facilitate a detailed investigation into the relationship between language usage of SL and TL. This study concludes that translators tend to avoid complex and unique linguistic features in SL and TL and tend to accept features of source texts with the condition that the rules of TL will not be violated. This paper intends not only to contribute to the theoretical consideration of specific features of translation, but also offer insight into translation pedagogy and practicesbetween Japanese and Chinese.