文本類型與譯者修改過程之探討

修改為翻譯過程中不可或缺的一部分，大多數譯者都會在交出譯文之前修改譯文，以確保譯文一來沒有二元性錯誤，二來翻得「恰當」。翻得恰當與否跟原文的文本類型有相當大的關係，因不同類型的文本具有不同的特色和需求。本研究便以放聲思考法，探討譯者修改不同文本類型之譯文時的行為，試圖了解文本類型是否影響譯者修改譯文的過程。受試者共七位，皆為受過兩年翻譯研究所訓練的譯者，且過去一年皆以筆譯為主要收入來源。本研究所選擇之文本類型乃根據德國學者 Reiss 於 1971 年所分之資訊類、表達類、和操作類文本，並分別以報導文章、短詩、廣告文案作為代表。研究結果顯示，受試者修改三類文本時，作出修訂的頻率差異不大。但在修改表達類文本時，受試者時常幾經考量後，決定不修改，原因是受試者認為表達類文本的解讀空間較大，不像資訊類文本有明確必須傳達的訊息。受試者修改資訊類文本時，著重於更正錯誤，較能明確地說出譯文的問題；修改表達類及操作類文本時，則花較多時間試圖將譯文修改得更有創意，文字讀起來更漂亮，有時譯文沒有必須修改之處，但受試者仍會將之改成自認為更好的版本。除了實際作出修改之處，研究中也觀察到，有時受試者針對某部分譯文，起初認為應作出修改，但考慮到文本特色，最後決定保留原譯文。此外，研究也觀察出譯者修改過程中的盲點，若受試者腦海中的字典內建有錯字，則受試者檢查譯文再多次，也不會發現錯誤。除錯字外，受試者對不熟悉的概念也容易以為自己懂了，但其實抓錯重點。本論文可能為國內首次以文本類型為變項，針對翻譯修改所作的放聲思考研究。結果顯示文本類型的確在翻譯修改過程中扮演重要角色。受試者在修改譯文的過程中，時常以文本目的和特色作為考量依據。

關鍵字

翻譯修改；文本類型；放聲思考法

並列摘要

This study utilizes think-aloud protocol (TAP) to probe into the influence that text types have on translators during their revision process. Seven participants were asked to translate three texts of different text types and then revise their translations using TAP. All participants have at least two years of training in translation and have indicated translation as their primary source of income over the past year. The texts chosen for this study include an excerpt from a news report, a short poem, and a cosmetics advertisement, representing the three text types (informative, expressive, and operative) proposed by German linguist and translation scholar Katherina Reiss in 1971. The results show that the average number of corrections per line made by each participant did not vary significantly with text type. When revising informative texts, their focus gravitated towards error checking the first draft of the translation. This group tended to be most proficient at identifying and explaining translation mistakes. For expressive and operative texts, rather than correcting the translation per se, most people in the study was inclined to refine the translation and make it sound more creative, to the point that edits were made even when the participants themselves thought the edits were not strictly speaking necessary. The study also shows that some edits were avoided specifically due to text type, and lastly, that some participants have certain blind spots when revising the translation, leaving binary errors still to be found in the finished translation. This investigation is perhaps the first in Taiwan into the role that text types play in the revision process for translations and it illustrates that it is indeed an important factor to keep in mind when revising a translation.