A Case Study on the Quality of Literary Translations Produced via MT+Post-Editing without Source-Language Competence

Machine translation (MT) is preferred by many translators because it is fast and convenient; it provides time savings especially in the production of high-volume technical texts. However, in contexts where cultural and linguistic nuance is decisive, particularly in literary translation, the importance of human intervention continues. In this study, based on the example of a published translation book, a translation practice produced through an MT+post-editing workflow without knowledge of the source language was examined in terms of translation quality. The analyzed work is Milan Kundera: Une Vie d’Écrivain by Jean-Dominique Brierre. This biography was translated into Turkish and published under the title Bir Yazarın Hayatı: Milan Kundera by Osman Akınhay, through his publishing house Angora Kitaplığı. Due to the scope of the article, the research was limited to direct quotations selected through purposive sampling; these quotations, selected from seven of Kundera’s works, were examined comparatively between published human translations and Akınhay’s translation. The evaluation was conducted on the basis of the criteria of fluency, semantic accuracy, and contextual appropriateness, using an approach that draws on the MQM (Multidimensional Quality Metrics) framework; the elements identified in each example were reported at Minor/Major/Critical levels. The findings showed that, in the examined examples, the translation is generally fluent and contextually appropriate; however, they also showed that greater caution is required regarding semantic accuracy, particularly due to nuance shifts that emerge in certain verbs and evaluative expressions.