AI时代下儿童文学翻译中价值观的伦理准则构建研究——基于DeepSeek与马爱农《小书房》译文对比的分析

Research on the Construction of the Ethical Principles for Value Transmission in Children’s Literature Translation in the AI Era—Based on a Comparative Analysis of DeepSeek and Ma Ainong’s Translation of The Little Bookroom

摘要: 当下AI技术正在重塑翻译实践，儿童文学翻译面临价值观传递的伦理挑战。本研究将以《小书房》为例，对比DeepSeek与马爱农译本的价值观传递差异。研究发现，马爱农译本体现“价值介入”伦理，译者主动进行文化调适、情感保护和诗意升华；DeepSeek译本体现“价值中立”技术立场，在价值观传递层面存在明显局限。本研究构建向善性、可理解性、诗性守护三维伦理框架，揭示AI受训练语料历史沉积影响及自身认知缺失的局限，提出“语义转换 + 伦理润色”的人机协同工作流。该框架与工作流具有探索性和案例性，其普适性仍需要更广泛语料、多语种、多题材、多模型及多样化翻译场景检验，以期为AI时代儿童文学翻译实践提供理论参照与操作路径。

Abstract: As AI is reshaping translation practices, children’s literature translation faces ethical challenges in the transmission of values. Taking The Little Bookroom as a case study, this research compares the differences in value transmission between DeepSeek’s translation and Ma Ainong’s translation. The findings reveal that Ma Ainong’s translation embodies a “value-engaged” ethics, where the translator actively undertakes cultural adaptation, emotional protection, and poetic enhancement, while DeepSeek’s translation adopts a “value-neutral” technical stance, displaying notable limitations in value transmission. This study constructs a three-dimensional ethical framework comprising benevolence, comprehensibility, and poetic preservation, identifies AI’s limitations related to historical sedimentation in training corpora and lack of embodied cognition, and proposes a human-machine collaborative workflow of “semantic conversion plus humanistic review”. The framework and workflow are exploratory and case-based; their universality still requires verification through broader corpora, multiple languages, diverse genres, multiple models, and varied translation scenarios, offering both theoretical reference and operational pathways for children’s literature translation practice in the AI era.