DeepSeek-R1与传统译者对《在细雨中呼喊》创伤叙事的英译对比研究

DeepSeek-R1 vs. Human Translators: A Comparative Study of English Translations of Trauma Narrative in Cries in the Drizzle

摘要: 本研究基于诺德“功能加忠诚”理论，对比人工智能DeepSeek-R1与人工译者白亚仁对余华《在细雨中呼喊》的英译策略。重点分析二者在创伤意象、隐喻及文化专有项处理上如何平衡文本功能、作者风格忠诚与读者接受忠诚。研究发现，AI在文化词转换及语义准确性上优势明显，能高效生成符合英语习惯的译文；而人工译者在情感表达、隐喻创新及文化补偿方面更优，更符合创伤文学的伦理需求。研究为AI文学翻译优化及人机协同提供新视角。

Abstract: This study employs Nord’s “Function plus Loyalty” theory to compare the English translation strategies of AI (DeepSeek-R1) and human translator (Allan H. Barr) for Yu Hua’s Cries in the Drizzle. Focusing on their approaches to trauma imagery, metaphors, and culture-specific items, it examines how they balance textual functions, loyalty to authorial style, and loyalty to reader reception. Findings reveal that AI demonstrates superior efficiency in culturally-loaded term conversion and semantic accuracy, producing natural English output, while the human translator excels in emotional expression, metaphorical creativity, and cultural compensation, better addressing trauma literature’s ethical demands. The study offers new insights for optimizing AI literary translation and human-machine collaboration.