Laughing Across Languages: AI-Assisted Translation of Cultural Humor
November 7, 2025 (Friday) · 15:20-16:35 · Room 205
This study examines how AI-assisted translation helps learners develop paraphrasing strategies and metalinguistic awareness in Translation in Language Teaching (TILT). Previous research has focused on subtitle translation’s technical aspects (Kumar, 2023) and AI in language learning (Warschauer et al., 2023), but this study uniquely investigates how beginner learners engage with culturally specific humor through AI-supported paraphrasing. Using the Japanese TV show Tetsuko’s Room, we analyzed how 38 non-English major students (CEFR A2) worked with AI-generated translations, focusing on their paraphrasing strategies and language awareness development.
The study applies three theoretical frameworks: domestication and foreignization (Venuti, 1995), Skopos theory (Vermeer, 1989), and relevance theory (Carston, 2002). These frameworks guided our analysis of how learners actively engaged with culturally embedded humor under subtitle constraints. We observed how learners used AI-generated paraphrases as starting points, critically evaluated them, and adjusted their choices to develop effective paraphrasing strategies, balancing linguistic appropriateness and cultural nuances.
The findings highlight two key developments. First, learners improved their paraphrasing strategies by critically comparing options and selecting contextually appropriate expressions. Second, they showed increased metalinguistic awareness, becoming more sensitive to the relationship between language forms and cultural meanings. Critically revising AI-generated paraphrases helped learners handle complex expressions they might have otherwise struggled with, demonstrating the value of combining AI support with learners’ active involvement.
These findings suggest that AI-assisted translation tasks, through active learner engagement, can support A2-level learners in handling complex content and enhancing language awareness. This study contributes to ELT and translation studies by showing how AI-supported paraphrasing fosters linguistic sensitivity and cross-cultural competence at early stages of language learning, while underscoring the need for proper guidance in using AI outputs effectively.