專業翻譯資源建置：以貨幣政策決議新聞稿為例

Abstract

隨著電腦與資訊科技越來越發達，譯者的工作型態也有所轉變。過去譯者幾乎只能仰賴紙本參考資料和專家意見，反觀現在，翻譯資源百花齊放，更以各種不同形式存在。其中，能夠輔助譯者改善譯文精準度，產出道地語言的「語料庫」廣泛受到青睞，語料庫對翻譯及語言學習的益處也備受肯定。事實上，除了現成語料庫外，網路上也存在許多能幫助譯者根據自身需求，建置個人語料庫的免費工具。因此，筆者希望能依循各步驟，示範如何以語料庫為基礎，建置出專業財經翻譯資源，將方法提供給譯者參考。本研究結合平行語料庫與三個可比語料庫，前者的資源來自臺灣央行每季發布的中文貨幣政策決議新聞稿及其英譯，而後者則分別包含美國、英國及歐盟中央銀行系統以英文寫成的貨幣政策決議聲明。四組語料以一系列語料庫分析工具進一步處理，使用工具包含AntConc、kfNgram、BFSU Collocator、LF Aligner、Notepad++和CUC_ParaConc，而分析整理後所得到的結果以「全面性專業翻譯資源」的形式呈現，此資源包含經段對應及句對應的中英雙語語料，也涵蓋政府財經文稿中常出現的詞串、搭配詞和專門術語，且可以再搭配上述之多種語料處理工具一起使用，讓譯者能從單詞觀察到語篇，由不同層次檢視語言特徵，進而提升翻譯效率及產出品質；同時，筆者也盼望本研究能實際啟發譯者依據個人需求，建置有益其工作的翻譯資源。

Contemporary developments in science and technology have transformed translators’ working styles. Translation resources today are no longer limited to printed reference materials or subject field experts but available in a wide array of forms. Among numerous resources, corpora have become one of the top choices for translators who wish to produce high-quality and idiomatic texts, with the benefits of corpus use widely acknowledged. In fact, apart from ready-made corpora, there also exist a number of free tools that translators can employ to establish ad hoc corpora according to their needs. This study, therefore, aims to provide translators with a method of resource compilation through a step-by-step demonstration using a corpus-based approach. The research combines a parallel corpus, which contains the Chinese monetary policy releases issued by the central bank of Taiwan as well as correspondent English translations, and three comparable corpora, which consist respectively of the English-written monetary policy releases by the U.S. Federal Reserve System, the Bank of England, and the European Central Bank. These language materials are processed with a series of corpus analysis programs, including AntConc, kfNgram, BFSU Collocator, LF Aligner, Notepad++, and CUC_ParaConc. Results obtained through the analysis of the four corpora are presented as a comprehensive translation resource which encompasses paragraph and sentence-aligned bilingual texts as well as lexical bundles, collocations, and specialized terms occurring with a high frequency in financial and economic texts that belong to the text type of government publications. It is hoped that such a resource can not only facilitate the translation process by allowing translators to observe language features from the context to the vocabulary level but actually inspire them to build and use corpora to their advantage as well.