Issues and international communication strategies for the intelligent translation and mediation of Chinese children’s science fiction: A case study of The Wandering Earth
国产儿童科幻作品智能译介问题与国际传播对策研究——以《流浪地球》为例

摘要

生成式人工智能降低文学海外译介成本，为国产儿童科幻拓展出海渠道，但低成本智能译介难以保障本土文化内核有效传递。本文以《流浪地球》为案例，运用跨文化传播“编码——转码——解码”分析框架，比对多版机器译文、结合海外馆藏与读者评论数据，剖析叙事编码、机器转码、海外解码各环节产生的文化损耗风险。研究发现大语言模型优先追求译文流畅，对文化负载要素识别不足，进一步放大原作自带的跨文化隔阂；据此构建译前文化分级标注、译中人机协同校对、译后多媒介反馈的全链条方案，兼顾翻译效率与文化保真，为国产儿童科幻深度文化出海提供实践参考。

Abstract

Generative artificial intelligence has cut down the cost for overseas translation of literary works and broadened export channels for Chinese children’s science-fiction. Nevertheless, cost‑effective intelligent translation fails to guarantee effective transmission of indigenous cultural connotations. Taking The Wandering Earth as the research object, this paper adopts the encoding-transcoding-decoding framework of cross-cultural communication. By comparing multiple machine-translated versions and analyzing overseas library collections as well as reader reviews, it identifies cultural-loss risks arising respectively from narrative encoding, machine-aided transcoding and overseas decoding links. The study reveals that large language models prioritize reading fluency yet lack sufficient recognition of culturally-loaded elements, which magnifies cross-cultural frictions embedded in source texts. Accordingly, this paper puts forward a full‑chain solution covering pre-translation cultural-hierarchy annotation, human-machine collaborative proofreading and post-translation multi-media feedback. Balancing translation efficiency and cultural fidelity, this strategy offers practical reference for in-depth cultural export of domestic children’s science-fiction.