“求真”与“务实”的算法化:周领顺译者行为批评理论视阈下AI翻译行为分析——以DeepSeek英译中国古诗词为例
The Algorithmization of “Truth-Seeking” and “Utility-Attaining”: An Analysis of AI Translation Behavior from the Perspective of Zhou Lingshun’s Translator Behavior Criticism—A Case Study of DeepSeek’s English Translation of Classical Chinese Poetry
摘要: 随着生成式人工智能在翻译实践中的广泛应用,如何评价其翻译行为成为翻译研究关注的重要问题。本文以周领顺译者行为批评理论为框架,将DeepSeek古诗词英译活动视为具有译者行为特征的“拟译者行为”,选取经典中国古诗词英译案例,采用文本细读、译本对比和案例分析相结合的方法,考察其在“求真–务实连续统”上的行为表现。研究发现,DeepSeek的翻译行为整体呈现出较强的务实倾向:在词汇层面倾向于对文化意象进行规范化处理,在句法层面优先遵循目标语表达习惯,在意境传递方面则存在审美再现不足的问题。总体而言,DeepSeek更注重译文的流畅性、可接受性与表达稳定性,而在文化意蕴和诗学价值传递方面仍存在局限。本文认为,译者行为批评理论为分析人工智能翻译行为提供了新的研究视角,有助于进一步认识AI翻译的行为特征及其与人类译者之间的协作关系。
Abstract: With the widespread application of generative artificial intelligence in translation practice, how to evaluate its translation behavior has become an important issue in translation studies. Based on Zhou Lingshun’s Translator Behavior Criticism Theory, this study regards DeepSeek’s English translation of classical Chinese poetry as a form of pseudo-translator behavior characterized by translator-like features. By selecting representative cases of English translations of classical Chinese poetry, this study employs textual analysis, comparative analysis of translations, and case analysis to examine DeepSeek’s behavioral characteristics along the Truth-seeking - Utility-Attaining Continuum. The findings indicate that DeepSeek’s translation behavior generally demonstrates a strong pragmatic tendency. At the lexical level, it tends to normalize the translation of culturally loaded images; at the syntactic level, it prioritizes conformity to target-language conventions; and at the level of poetic conception, it shows limitations in reproducing aesthetic effects. Overall, DeepSeek places greater emphasis on fluency, acceptability, and stability of expression in translated texts, while facing certain limitations in conveying cultural connotations and poetic values. This study argues that Translator Behavior Criticism Theory provides a new perspective for analyzing artificial intelligence translation behavior and contributes to a deeper understanding of AI translation characteristics and its collaborative relationship with human translators.
参考文献
|
[1]
|
周领顺. 译者行为评价系统的学理性[J]. 外语导刊, 2024, 47(1): 106-114, 160.
|
|
[2]
|
周领顺. 再论译者行为研究中的“翻译外”研究[J]. 北京第二外国语学院学报, 2025, 47(2): 31-42.
|
|
[3]
|
马冬梅. 周领顺译者行为研究思想的系统性发展[J]. 北京第二外国语学院学报, 2024, 46(6): 64-86.
|
|
[4]
|
Brown, T.B., Mann, B., Ryder, N., et al. (2020) Language Models Are Few-Shot Learners. Proceedings of the 34th International Conference on Neural Information Processing Systems, Vancouver, 6-12 December 2020, 1877-1901.
|
|
[5]
|
Guo, D.Y., et al. (2025) DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning. arXiv: 2501.12948.
|
|
[6]
|
许渊冲. 翻译的艺术[M]. 北京: 中国对外翻译出版公司, 1984.
|
|
[7]
|
许渊冲. 文学与翻译[M]. 北京: 北京大学出版社, 2003.
|
|
[8]
|
Jiao, W., Wang, W., Huang, J.T., et al. (2023) Is ChatGPT a Good Translator? A Preliminary Study. arXiv: 2301.08745.
|
|
[9]
|
Newmark, P. (1988) A Textbook of Translation. Prentice Hall.
|
|
[10]
|
刘宓庆. 文体与翻译[M]. 北京: 中国对外翻译出版公司, 1998.
|
|
[11]
|
Nida, E.A. (1964) Toward a Science of Translating. Brill. https://doi.org/10.1163/9789004495746
|
|
[12]
|
Hendy, A., Abdelrehim, M., Sharaf, A., et al. (2023) How Good Are GPT Models at Machine Translation? A Comprehensive Evaluation. arXiv: 2302.09210.
|
|
[13]
|
Kocmi, T. and Federmann, C. (2023) Large Language Models Are State-of-the-Art Evaluators of Translation Quality. Proceedings of the 24th Annual Conference of the European Association for Machine Translation, Tampere, 12-15 June 2023, 193-203.
|