人工智能在中国古诗词翻译中的应用与挑战——以李商隐《无题》英译为例
Applications and Challenges of Artificial Intelligence in Translating Ancient Chinese Poetry—Based on the English Translation of Li Shangyin’s To One Unnamed
摘要: 随着神经机器翻译与大语言模型的发展,人工智能在文学翻译中的应用不断深化,尤其在中国古诗词英译语法重构以及意象直译方面等展现出显著优势,但也存在明显局限,难以精准捕捉深层审美意蕴和文化语境。在李商隐《无题》的英译中,人工智能模型(Gemini 3.1 Pro)展现出处理高度结构化文学文本的算法优势,能严格恪守五步抑扬格与高度平行的句法结构,在英语语境中成功重塑原诗的音韵与形式骨架。然而,其在深层文化意象与情感意境的传达上存在明显短板,难以跨越底层语言逻辑破译特有的谐音双关,且易破坏原诗情感主体的统一性与朦胧含蓄美。因此中国古诗词的英译应确立“人工智能辅助 + 人工审校优化”的协同模式,借助算法解析文本背后的共性与规律,依托人类译者的创造性与情感积淀为译文注入灵魂,从而实现技术效率与文学品质的平衡。
Abstract: With the development of neural machine translation and large language models, the application of artificial intelligence in literary translation has continuously deepened. AI demonstrates significant advantages, particularly in the syntactic reconstruction and literal translation of imagery when translating ancient Chinese poetry into English. However, distinct limitations persist, as it struggles to accurately capture profound aesthetic implications and cultural contexts. In the English translation of Li Shangyin’s “To One Unnamed”, the AI model (Gemini 3.1 Pro) exhibits algorithmic superiority in processing highly structured literary texts. It strictly adheres to iambic pentameter and highly parallel syntactic structures, successfully reconstructing the phonological and formal skeleton of the original poem within an English context. Nevertheless, it reveals noticeable shortcomings in conveying deep cultural imagery and emotional atmosphere. It struggles to transcend underlying linguistic logic to decipher unique homophonic puns, and it tends to disrupt the unity of the emotional subject as well as the hazy, implicit beauty of the original poem. Therefore, the English translation of ancient Chinese poetry should adopt a collaborative model of “AI assistance + human post-editing and optimization”. By leveraging algorithms to analyze the universalities and underlying patterns of the text, while relying on the creativity and emotional depth of human translators to breathe soul into the work, a balance between technical efficiency and literary quality can be achieved.
参考文献
|
[1]
|
葛颂, 王宁. 人工智能时代的文学翻译: 挑战与机遇[J]. 外语与外语教学, 2024(1): 94-101+149-150.
|
|
[2]
|
张可人, 郭国良. 智能时代文学翻译的“艺∙术∙心三维”——与郭国良教授的对谈[J]. 翻译研究, 2024(2): 15-27.
|
|
[3]
|
许渊冲. 翻译的艺术[M]. 北京: 五洲传播出版社, 2006.
|
|
[4]
|
王贇, 张政. ChatGPT人工智能翻译的隐忧与纾解[J]. 中国翻译, 2024, 45(2): 95-102.
|
|
[5]
|
吴勇立. 基于人工智能的文学翻译探赜——以《浮士德》和《大腕》德汉互译为例[J]. 翻译研究与教学, 2024(1): 32-39.
|
|
[6]
|
刘宓庆. 翻译美学导论[M]. 北京: 中译出版社, 2019.
|
|
[7]
|
万兵. 从文体学视角看格律诗的翻译——以许渊冲英译李商隐的《无题》为例[J]. 天津外国语学院学报, 2010, 17(6): 24-28.
|