Machine Translation and Legal Text: A Case Study around Terminology (Spanish-Chinese)
CARMEN VALERO GARCÉS,LIANGJIAN HU
DOI: https://doi.org/10.3726/jts012024.05
2024-01-01
Journal of Translation Studies
Abstract:Abstract This paper is a follow-up study to the research conducted by the authors in May 2023, in which they extracted some provisions from the Spanish Civil Code and used five popular free online translation tools, namely: GPT 3, Google Translate, DeepL, Baidu Translate and Youdao Translate, to translate these provisions into Chinese. The aim of the study was to record and analyse how those five free online translation tools translate seven selected legal terms contained in the provisions; to summarize the common errors observed; and to call for collaboration between translation professionals, on the one hand, and machine translation (MT) engineers, on the other, with a view to optimizing the performance of MT in legal translation (language pair Spanish-Chinese). In this follow-up research, the authors wish to determine whether any of the five tools has displayed any change in performance over the previous six months, and to repeat the experiment with seven legal terms, of which three are different from those of the previous study. This follow-up study shows that, overall, the types of translation errors found in the previous study persist, while, individually, GPT 4 has better rendition than GPT 3 in the translation of the limited group of terms chosen for the study.