《红楼梦》回目英译：人工译本与AI译本的对比研究

The English Translation of Chapter Titles in A Dream of Red Mansions: A Comparative Study of Human and AI Translations

摘要: 在人工智能技术深度融入翻译领域的时代背景下，古典文学外译迎来传统人工译本与AI译文并存发展的新局面。《红楼梦》回目兼具叙事功能、文化内涵与对偶诗学特质，是检验文学翻译质量的典型文本。本文立足House翻译质量评估模式，选取四大类共8组经典回目，对比杨宪益–戴乃迭、霍克斯–闵福德人工译本与Kimi、DeepSeek的AI译本的翻译表现，从语场、语旨、语式三维度剖析优劣差异。研究发现，人工译本可平衡文化传递、叙事指引与诗学再现，AI译本存在形式缺失、文化内涵传递不到位、审美艺术表现力不足等短板；古典文学回目英译需兼顾语义忠实与文化审美，AI文学翻译仍需针对性优化。

Abstract: Against the backdrop of in-depth integration of artificial intelligence technology into the translation field, the external translation of classical literature has ushered in a new development landscape where traditional human translations coexist with AI translations. The chapter titles of A Dream of Red Mansions integrate narrative function, cultural connotation and antithetical poetic characteristics, serving as typical texts for testing the quality of literary translation. Based on House’s translation quality assessment model, this paper selects 8 groups of classic chapter titles divided into four major categories, and compares the translation performance of human translations by Yang Xianyi & Gladys Yang, Hawkes & Minford with AI translations generated by Kimi and DeepSeek. It analyzes their merits and demerits from the three dimensions of field, tenor and mode. The research finds that human translations can strike a balance between cultural transmission, narrative guidance and poetic representation, while AI translations suffer from formal deficiency, inadequate transmission of cultural connotations, and insufficient aesthetic and artistic expressiveness. The English translation of classical literature chapter titles needs to balance semantic faithfulness with cultural aesthetics, and AI literary translation still requires targeted optimization.