@

초록

This paper analyzes the Korean-Japanese translation patterns of Japanese passive voice and the “temorau-sentence” structure to elucidate the speaker-centered linguistic perception structure inherent in Japanese. The subjects of analysis were four groups: “1. General Learning Materials”, “2. Student Translations”, “3. Professional Translator Translations”, and “4. AI Translations”. Comparing and contrasting these materials to examine common patterns and differences revealed the following issues, regardless of the group: (1) frequent use of double passive constructions like ‘보여지다’, (2) unnatural expressions resulting from literal translations like ‘~해 받다’, (3) misunderstandings regarding the role of the passive recipient in V-temorau constructions, and (4) repeated arbitrary conversion of passive constructions into active voice. These tendencies stem from the fact that while Japanese emphasizes ‘from whose perspective the situation is described’ within a sentence, Korean tends to center on an agent-focused narrative. Particularly in “temorau-sentences,” the core elements are ‘who requests what, who responds by doing what, and who receives the result.’ However, numerous cases were found where the roles of the agent and passive subject were reversed during translation, altering the meaning. Therefore, this paper proposes that rather than reducing passive voice and the “temorau-sentence” to a simple substitution problem between individual sentence patterns, it is necessary to approach them within an analytical framework that clarifies the role relationship between the agent and the passive subject. This approach is useful for understanding the background of recurring misunderstandings and mistranslations. Furthermore, by observing the reproduction of the same tendency across diverse data types-including literary texts, everyday discourse, professional translation, and machine translation-it demonstrates that Japanese is a system that grammatically marks speaker-centered language perception.

키워드