Does LLM translation align with translation universals? A cross-genre simplification study on English-Chinese translation based on dependency grammar

Zeyuan Jiang

PLOS ONE, 2025, vol. 20, issue 6, 1-19

Abstract: This study examines the presence of simplification, a translation universal (TU), in English-to-Chinese translation by comparing the Mean Dependency Distance (MDD) and Mean Hierarchical Distance (MHD) of Crowdsourcing human translations, Large Language Model (LLM) translations, and original Chinese texts across fifteen genres. Through analysis of three balanced comparable corpora, the research found that: (i) Compared to original Chinese texts, both human-translated and LLM-translated Chinese texts demonstrated significant syntactic simplification across all genres. (ii) Human translations exhibited a more pronounced tendency toward syntactic simplification than LLM translations across all genres. These findings not only validate the simplification hypothesis at the syntactic level but also highlight the different cognitive and processing mechanisms underlying human and LLM translation processes. The research indicates that human translators possess an active ability to optimize complex syntax that current LLMs lack, providing valuable reference for future development of LLMs and methods for LLM-assisted translation. Additionally, by adopting MDD and MHD as holistic measures of syntactic complexity, this study offers new perspectives for TU research and provides empirical insights into the linguistic nature of crowdsourcing translations from an English-to-Chinese perspective.

Date: 2025
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0324830 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 24830&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pone00:0324830

DOI: 10.1371/journal.pone.0324830

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().