Examining the Efficiency of Machine Translation in Translating English Idioms used in American Media

Abstract

The use of idioms permeates the American media, be it movies, series, TV sitcoms, chat shows, radio chats, documentaries and so forth. By definition, idioms are fixed expressions whose meaning cannot be inferred merely from the meanings of the words that compose it. With this in mind, translators face a tremendous challenge when attempting to translate American media to another language, whether that was done through dubbing or via the use of subtitles. Nevertheless, today’s available resources can provide an appreciable assistance regarding such challenges. Among these resources are online machine translation services, particularly the ones that are based on machine learning and artificial intelligence. The central aim of this study was to explore the efficiency of online machine translation services in translating idioms used in American media. To attain this aim, three of the most reliable online machine translation services were examined to determine their level of accuracy concerning idioms translation: Google Translate, DeepL Translator, and Bing Microsoft Translator. The results have indicated that machine translation is capable of translating idioms with an average accuracy rate of 68.7%, with Bing Microsoft Translator being the most accurate. In most cases, the translation is done by paraphrasing. However, Bing Microsoft Translator sometimes opts for using idioms in the target language that are similar in terms of meaning to the ones in the source language, which reveals new horizons for the development of high-tech online machine translation services that would be able to accurately and meaningfully translate cultural-specific lexis such as idioms.