Abstract
Large Language Models (LLMs) are transforming translation practices. The Human-Centred, Augmented Machine Translation (HCAMT) framework emphasises empowering users (Briva-Iglesias, 2024) in an era of increasing automation. This study investigates how LLMs may support cultural adaptation (CA), a process requiring creativity, cultural sensitivity and critical human judgement, from the perspective of translation students. A classroom study was conducted with 24 BA-level participants performing English-to-Chinese CA of culturally loaded proper nouns. All students completed a web-search-supported human adaptation task first and an LLM-supported human adaptation task second. Final translations were evaluated through a depth-of-cultural-adaptation rubric, while questionnaire responses and LLM interaction logs were analysed to examine learner perceptions, workflow patterns and human-LLM collaboration. The findings are interpreted as exploratory because the activity used a small sample, a fixed task order and different source items across conditions. Students overwhelmingly preferred LLM support because it was perceived to accelerate brainstorming and information retrieval and to reduce effort. However, the LLM-supported workflow did not consistently produce deeper cultural adaptation and sometimes may have encouraged cognitive offloading or offered shallow cultural transfer and misleading suggestions. The study therefore contributes to translation pedagogy by reframing LLMs not as replacements for cultural interpretation, but as tools whose value depends on human agency, verification and critical control.
| Original language | English |
|---|---|
| Number of pages | 27 |
| Journal | Linguistica Antverpiensia, New Series – Themes in Translation Studies |
| Publication status | Accepted/In press - 17 Jul 2026 |
UN SDGs
This output contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 4 Quality Education
User-Defined Keywords
- Cultural Adaptation
- LLM-supported Translation
- Human-centred Augmented Machine Translation
- English-Chinese Translation
- Student Perceptions
Fingerprint
Dive into the research topics of '(AI)dapting Cultures: Exploring LLM Support for Interlingual Cultural Adaptation from the Perspective of BA Students at a Hong Kong University'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver