한일 헤드라인 번역의 포스트에디팅 가이드라인 고찰

Abstract

We use Google Translate when reading foreign newspapers. Given that it is MT, we expect some degree of awkwardness. However, it still gives us easy and convenient access to the content and information. Today, many Korean media outlets provide articles in foreign languages, by selecting and instantly translating the many articles that are produced every day. The time pressure is extremely high in this kind of translation, and there can be occasional verbatim, hence unnatural, renderings. Yet, overall, the quality of work done by skilled translators tends to be high. If MT continues to advance, however, it seems that MT may be able to replace human translators in the near future, in terms of speed and accuracy. This paper looks into the outcome of MT on news headlines. In this age of overflowing information, headlines can be important indicators for selecting articles. With fewer words, they convey compact information, and it needs to be tested whether MT can serve the same function. While there is existing literature that covers the features of headlines in Korean and Japanese newspapers, and suggests translation strategies assuming human translation, the present study attempts to identify issues in headline translation done by machines. Further, it aims to discuss the potential of MT and implications for post-editing. Headline translation requires the understanding of the entire content of the article to avoid mistranslation, as well as the knowledge of the characteristics of headlines in both working languages. This paper examines headlines from different newspapers and how they are translated, in comparison to their MT versions.