歐盟議會英語語篇之詞串研究──口譯、筆譯及演說之語域差異

Lexical Bundles in English EU Parliamentary Discourse-Variation Across Interpreted, Translated, and Spoken Registers

套語（formulas）為語言的理解與產出帶來認知處理上的優勢，亦顯示使用者所歸屬的專業社群。本研究比較歐洲議會口譯員、筆譯員與各國議員英語套語的使用差異。研究方法分為兩步驟。首先，英語套語操作化為四字詞串（lexical bundles），採用語料庫驅動（corpus-driven）的方式辨認。語料庫有三種：一、歐洲議會全體大會的英語口譯逐字稿；二、與之對應的英語筆譯文本；三、大會的英語演說逐字稿。三種語料庫分別代表三種語域（registers）。第二步驟為詞串的文法結構及語用功能分析。研究發現，在詞串的實際使用上，歐洲議會口譯員、筆譯員與各國議員的差異極大，僅10%的詞串類別為兩者或三者共用。然而，三種語域的文法結構呈現類似模式，皆以名詞及介系詞片語的詞串為主，顯示議會語篇（parliamentary discourse）整體而言以資訊傳達為目的。功能分析顯示筆譯與演說兩語域較為類似，皆以「主題類詞串」為主，顯示資訊功能為重。口譯語域較為特別，其「立場類詞串」、「指涉類詞串」及「主題類詞串」比例相近，顯示資訊與溝通兩種功能並重。詞串功能分類有助於口、筆譯培訓，文末討論提升詞串教學價值的建議。

Formulas offer processing advantages in both language comprehension and production and signal one's identity in a professional community. This study attempts to explore the similarities and differences among European Parliament (EP) interpreters, translators, and Members of the EP (MEPs) in their use of English formulas, operationalized as four-word lexical bundles (LBs) identified through corpus-driven approach. English interpretations, translations, and speeches from the EP plenary sessions represent three registers, constituting the three corpora of this study. The LBs identified in each of the three corpora were then categorized based on their grammatical structures and pragmatic functions. Results showed that differences among EP interpreters, translators, and MEPs far outweigh similarities in actual bundle use, with only 10% of the total number of bundle types present in two or all three of the corpora. However, similar structural patterns of the three registers with noun phrase and prepositional phrase bundles playing a dominant role indicate informational purposes of the parliamentary discourse in general. Functional analysis revealed that the translated and spoken registers are more alike with the dominance of subjectspecific bundles, indicating informational priorities. On the other hand, the interpreted register stands out with its nearly equal proportions of stance bundles, referential bundles, and subjectspecific bundles, indicating a combination of informational and communicative priorities. Typology featuring LBs' functional characteristics may have pedagogical implications for translation and interpretation (T&I) training, and suggestions for enhancing LBs' pedagogical value are provided at the end of this paper.