大语言模型时代人工译者的翻译优势——基于语料库的人工译文与ChatGPT译文对比研究  

Human Translators’Translation Advantages in the Era of Large Language Models:A Corpus-Based Comparative Study of Human Translation and ChatGPT Translation

作　　者：侯雁慧[1] 朱玉 

机构地区：[1]辽宁对外经贸学院外国语学院,大连116052

出　　处：《翻译研究与教学》2024年第4期116-126,共11页Translation Research and Teaching

基　　金：2022年度辽宁省教育厅项目“应急语言服务语境下翻译能力的构建与培养研究”项目(项目编号:LJKMR20221882)的阶段性成果

摘　　要：本文结合语料库量化统计和实例分析,系统比较了人工译文和ChatGPT译文在词汇、句法和篇章三个维度下的文本特征差异,以探究大语言模型时代人工译者的翻译优势,进而明确大语言模型时代译者翻译能力培养的新需求和新导向。研究发现人工译者的翻译优势主要体现在三个方面:(1)聚焦译文受众,提高译文的易读性和可接受性;(2)深入分析原文,增强译文连贯性和衔接性;(3)把握文化语境差异,提升译文准确性和得体性。本研究能够为大语言模型时代的翻译教学提供参考,从而有针对性地培养符合时代需求的翻译人才。This study employs corpus-based analysis and case studies to systematically compare the differences between human translations and ChatGPTtranslations in terms of multidimensional linguistic features,including lexical,syntactic and discourse-related aspects.The aim is to identify the distinct translation advantages of human translators in the era of large language models,and thereby clarify the new requirements and directions of translation teaching in this context.The superiority of human translators is demonstrated primarily in three areas:(l)enhancing translation readability and acceptability by focusing on the target audience;(2)improving textual coherence and cohesion through thorough analysis of the original text;(3)boosting translation accuracy and propriety by accommodating cultural and context differences.The findings offer insights for translation teaching in the era of large language models,facilitating targeted training of translators to meet needs of the era.

关 键 词：ChatGPT译文 人工译文 语料库 大语言模型 

分 类 号：H315.9[语言文字—英语] H085