Hanxiao Ma
School of Foreign Languages, Gannan Normal University, Ganzhou 341000, Jiangxi, China.
*Corresponding author: Hanxiao Ma
Published: September 26, 2025
Article history: Updated on June 11, 2026
Abstract
This paper sets out to compare and evaluate the performance of the Qwen large language model (LLM) against human translation in the specific context of literary translation, using the Chinese versions of The Old Man and the Sea as a case study. To do so, the study fed two distinct instruction prompts into Qwen to generate its translations, and then compared these outputs with Sun Zhili’s 2012 Chinese translation. All three translations were subsequently analyzed using two software tools—Wordless, which measured lexical density, richness, and diversity, and AlphaReadabilityChinese, which assessed syntactic complexity. These di-mensions were examined in detail, with particular attention to how each translation handled semantic load and sentence structure. The results show that Qwen performs quite well in terms of lexical density, effectively conveying the original text’s core semantic information. However, when it comes to lexical richness and diversity—qualities that give literary translation its texture and nuance—Qwen still lags noticeably behind the human translator. This finding suggests that while large language models have made impressive strides in translation technology, human translation remains irreplaceable in the literary domain, where context sensitivity and expressive artistry are paramount. Consequently, rather than viewing AI as a substitute for human literary translators, future research and practice should focus on how the two might complement each other, integrating their respective strengths to elevate the overall quality of literary translation.
References
Hemingway, E. (2012). The old man and the sea (Sun, Z. L., Trans.). People’s Literature Publishing House. (Original work published 1952)
Huang, Z. L., & Zhang, T. (2021). Literary translation: To see through, but to translate through? —A comparative analysis of Li Xiyin and Sun Zhili’s Chinese versions of The old man and the sea. Journal of Foreign Languages, 44(4), 76-84.
Liu, C. X. (2023). A comparative analysis of Chinese translations of The old man and the seafrom the perspective of ecotranslatology. Masterpieces Review, (29), 133-135.
Luo, J. L. (2022). A comparison of Chinese translations of The old man and the sea from the perspective of intersubjectivity in translation. Literature Education, (4), 157-159. https://doi.org/10.16692/j.cnki.wxjys.2022.04.036
Wang, L. L. (2009). A corpus-assisted study on the stylistic translation of The old man and the sea (Master’s thesis, Jilin University).
Wang, M. H., & Lei, C. (2023). A cognitive-comparative analysis of two Chinese translations of The old man and the sea. Data of Culture and Education, (10), 6-9.
Wang, M. Y. (2018). A corpus-based comparative study on the style of The old man and the sea and its six Chinese translations (Master’s thesis, Henan University).
Wu, X. J. (2021). A comparative study on sentence processing in Yu Guangzhong’s translation of The old man and the sea. Modern English, (19), 56-58.
Xiang, R. (2011). A corpus-assisted study on the translator’s style of two Chinese versions of The old man and the sea. Journal of Hubei Normal University (Philosophy and Social Science), 31(3), 57-60.
How to cite this paper
Corpus-based Comparative Analysis of Large Language Model and Human Translation: A Case Study of the Chinese Version of The Old Man and the Sea
How to cite this paper: Hanxiao Ma. (2025). Corpus-based Comparative Analysis of Large Language Model and Human Translation: A Case Study of the Chinese Version of The Old Man and the Sea. Translation and Foreign Language Learning, 1(1), 61-65.
DOI: http://dx.doi.org/10.26855/tfll.2025.08.010