节点文献

变译理论指导下American Government(节选)的译后编辑策略

A Report of Post-Editing Techniques on American Government(Excerpts) Guided by Variational Translation Theory

【作者】 黄文

【导师】 郭威; 刘永俊;

【作者基本信息】 山西大学 , 翻译, 2024, 硕士

【摘要】 本实践报告以变译理论为指导,主要探索关于政治类文本机译译后编辑的有效策略,从而提高翻译质量。除了漏译、错译等明显的错误需要纠正以外,笔者发现仍然有许多语义不能被正确理解,因此进行人工译后编辑非常有必要:一是可以保证文本内容的完整性和简洁性;二是能增强作品的传播性。本次翻译素材节选自《美国政府》,其是由美国作家詹姆斯·Q·威尔逊所作的一本综合性政治类著作。《美国政府》主要介绍了美国政府的简要信息,包括美国的宪法、公民自由、民权、联邦立宪制、政治权利等一系列的具体内容,会存在一些科普类的知识。本文选取了第五、六、七章的文本内容,利用人工智能Chat GPT进行了汉英机器翻译。经研究发现,漏译、误译、语义繁冗和语序问题是这次翻译任务中最普遍的类型。基于此,本报告选取了适合解决上述错误类型的六种变通手段(增、减、缩、并、改、编)对该任务进行译后编辑,然后通过案例分析来证实译后编辑技巧的效果。由于变通手段之一——“述”指的是用译语语言将原文内容复述或改写,而原著中多为短小句子,理解起来较为简单,且机译的错误类型中不存在需要改写的内容,因此“述”这一变通手段不在本报告的讨论范围之内。本实践报告认为,使用变译理论指导下的变通手段进行译后编辑有助于解决机器翻译中所遇到的错误类型,从而提高机器译出的质量。具体而言,即采取增译的手段对原文中不对等、缺失的信息进行补充说明,在翻译特定文化表达时添加必要的解释或背景信息;采用减译的手段对原文中重复累赘的词组进行删除,但不改变语义;采用缩译的手段对原文中笔者认为译语读者不需要了解的内容进行浓缩、提取精华、简化句子,以减少读者的阅读负担;采用并译的手段将对相似内容进行整合或调整译文结构;采用改译的手段对句子的语序进行调整,使句子的逻辑关联更强;采用编译的手段对机译内容中生硬、直译的内容进行编写,使原文本信息更易被读者理解,争取提高政治类作品英汉机译的翻译质量。本报告尝试总结出不同机译问题的变通手段,从而为政治类作品的译后编辑提供更直接、有效的思路。

【Abstract】 Guided by the variational translation theory,this practice report discusses effective post-editing techniques to improve the quality of machine translation of political texts.In addition to the obvious errors,such as missing translation and mistranslations need to be corrected,the translator finds that there are still many semantics that cannot be correctly understood.As an effective means of translating political texts,post-editing has certain advantages in ensuring the integrity,conciseness,and improving the readability of the work.The translation material is an excerpt from American Government,a comprehensive political book by American writer James Q.Wilson,mainly introduces the American government,including a series of information such as the American Constitution,civil liberties and rights,federal constitutional system and political rights,with a certain character of scientific popularization.To this end,the report conducts an error analysis of the machine Chat GPT output of chapters 5,6 and 7 from American Government,identifying common errors in MT,including omissions,mistranslations,semantic complexity and incorrect sentence order.Six adaptation techniques—adding,deleting,condensing,integrating,altering and editing—are employed to resolve the above types of errors and post-edit the political works,and their effectiveness is elaborated through case analysis.As one of the workarounds,"narrating",refers to the retelling or rewriting of the original content in the target language.While the original text is mostly short sentences,which is relatively simple to understand,and there is no content that needs to be rewritten in the error types of machine translation.Hence,this workaround of "narrating" is not within the scope of discussion in this report.In this report,it is argued that the use of variational workarounds guided by the variational translation theory can help to resolve the types of errors encountered in machine translation and improve the quality of machine translation output.Specifically,it means using the methods of adding to supplement and explain the unequal or missing information in the original text,and adding necessary explanations or background information when translating specific cultural expressions.Using the method of deleting translation to remove redundant phrases in the original text without changing their semantics.Using the method of condensing to remove the content of the original text that the reader does not need to know,extracting the essence and simplifying the sentence to reduce the reader’s reading burden.Integration of similar content or adjustment of the translation structure is undertaken.Alteration in sentence order enhances the logical coherence.Furthermore,to facilitate better reader comprehension,the report emphasizes the importance of editing overly literal and blunt translations in the MT output.The overarching aim of the report is to consolidate effective adaptation techniques for various machine translation errors,providing concise and impactful guidance for the post-editing of political works.

  • 【网络出版投稿人】 山西大学
  • 【网络出版年期】2025年 07期
  • 【分类号】H315.9
节点文献中: 

本文链接的文献网络图示:

本文的引文网络