EUVIMEDEUVIMED entdecken
Crossref

A Comparative Study of the Translation of Cohesion in English-Chinese Business Texts between Human Translators and ChatGPT

Na Tang, Mohamed Abdou Moindjie

Malaysian Journal of Social Sciences and Humanities (MJSSH) · 31.03.2026 · 11 (3) · e003859

DOI: 10.47405/mjssh.v11i3.3859 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Malaysian Journal of Social Sciences and Humanities (MJSSH), online erschienen am 31.03.2026. Verfasst von Na Tang und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study examines how cohesion is rendered in English-Chinese translation by human translators and ChatGPT, with a focus on business texts. The analysis of cohesion draws on Halliday and Hasan’s (1976) model of cohesive devices, while the classification of translation methods uses Vinay and Darbelnet’s (1995) taxonomy. A mixed-methods approach is adopted, combining qualitative exploration of the translation strategies with quantitative analysis of the frequency and similarity of translation methods for cohesive devices between human and ChatGPT translation. A self-compiled parallel corpus comprising eight English business annual report passages, along with their human translations and ChatGPT outputs, is employed, totaling 46,423 words. The results reveal that human and ChatGPT translations similarly employ six main translation methods, namely literal translation, borrowing, transposition, modulation, equivalence, and omission. Direct strategies (literal translation and borrowing) dominate, followed by omission, while oblique strategies (transposition, modulation, and equivalence) are least frequent. The similarity rate in the treatment of cohesive devices between the two versions exceeds 82%, suggesting a high degree of convergence. Nevertheless, differences also emerge. ChatGPT translation relies more on omission and direct strategies, whereas human translation demonstrates slightly stronger orientation toward the target language and greater use of oblique strategies. Translation errors are also identified in ChatGPT outputs, emphasizing the need for post-editing. This research contributes to translation studies by providing a systematic framework for cohesion analysis, offering corpus-based evidence on similarities and differences between human and ChatGPT translations.

Zitieren

APA

Tang, N., & Moindjie, M. A. (2026). A Comparative Study of the Translation of Cohesion in English-Chinese Business Texts between Human Translators and ChatGPT. Malaysian Journal of Social Sciences and Humanities (MJSSH), 11(3), e003859. https://doi.org/10.47405/mjssh.v11i3.3859

Vancouver

Tang N, Moindjie MA. A Comparative Study of the Translation of Cohesion in English-Chinese Business Texts between Human Translators and ChatGPT. Malaysian Journal of Social Sciences and Humanities (MJSSH). 2026;11(3):e003859. doi:10.47405/mjssh.v11i3.3859

BibTeX
@article{Tang2026,
  author = {Tang, Na and Moindjie, Mohamed Abdou},
  title = {A Comparative Study of the Translation of Cohesion in English-Chinese Business Texts between Human Translators and ChatGPT},
  journal = {Malaysian Journal of Social Sciences and Humanities (MJSSH)},
  year = {2026},
  volume = {11},
  number = {3},
  pages = {e003859},
  doi = {10.47405/mjssh.v11i3.3859}
}
Kostenlos starten →