An investigation of challenges in machine translation of literary texts : the case of the English-Chinese language pair

Abstract

In the absence of a focus on literary text translation in studies of machine translation (MT), this study aims at investigating some challenges of this application of the technology. First, the most commonly used types of MT are reviewed in chronological order of their development, and, for the purpose of identifying challenges for MT in literary text translation, the challenges human translators face in literary text translation are linked to corresponding aspects of MT. In investigating the research questions of the challenges that MT systems face in literary text translation, and whether equivalence can be established by MT in literary text translation, a qualitative method is used. Areas such as the challenges for MT in the establishment of corpora, achieving equivalence, and realisation of creativity in literary texts are examined in order to reveal some of the potential contributing factors to the difficulties faced in literary text translation by MT. Through text analysis on chosen sample literary texts on three online MT platforms (Google Translate, DeepL and Youdao Translate), all based on highly advanced neural machine translation engines, this study offers a pragmatic view on some challenging areas in literary text translation using these widely acclaimed online platforms, and offers insights on potential research opportunities in studies of literary text translation using MT.

Date of Award	2021
Original language	English