Translating the “untranslatable”: English-Mandarin machine translation of figurative language in literary texts

Summarise with VIRAT
Authors: Xin Yi Seah
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This thesis investigates the effectiveness of machine translation technologies—specifically ChatGPT, DeepL Translator, and Google Translate—in translating figurative language in literary texts between Mandarin and English, compared with human translations. Evaluators rated system and human outputs on a scale from ‘Excellent’ to ‘Bad’. Results varied across metrics and contexts, but several insights emerged: (i) DeepL and ChatGPT generally outperformed the human translation, whereas Google Translate lagged, though the difference between human and system performance is not statistically significant; (ii) providing more context in the input did not consistently improve translation quality; and (iii) evaluators with translation backgrounds tended to rate outputs more strictly than those without. Together, these findings illuminate the complexity around evaluating machine translation for literary texts.

Fields of Study & Subjects

Translation Studies and Practices Natural Language Processing Techniques Topic Modeling

Publication details

Journal / source
DR-NTU (Nanyang Technological University)
Publication date
1 January 2026
Language
EN