Speaker
Description
This study examines how EFL learners and GPT-4 tend to interpret opaque English idioms in balanced ambiguous contexts, where both literal and figurative readings may still remain possible. Idioms are important in everyday communication, yet they are often difficult for foreign language learners because of semantic opacity and cultural specificity. Although there is growing attention to AI-assisted language learning, research is still somewhat limited in comparing how human learners and large language models negotiate idiomatic ambiguity when no single interpretation is clearly supported. To respond to this gap, the study adopts a task-based survey design with Vietnamese undergraduate students majoring in English Language Teaching (second- and third-year students). Participants complete ten interpretation tasks and choose literal, figurative, or indeterminate meanings, with brief explanations. The same items are also submitted to GPT-4 through standardized prompts. Responses will be analyzed quantitatively through option distributions and qualitatively through thematic coding of reasoning patterns. The study is expected to contribute to understanding human - AI differences in idiom comprehension and epistemic stance under uncertainty.
Keywords: Idiomatic ambiguity; Opaque idioms; EFL learners; GPT-4; Interpretive behavior
Biography
Nguyen Dang Nguyen Phuong holds a Master’s degree in TESOL from the University of Huddersfield, UK. She is a lecturer at the Faculty of Foreign Language Teacher Education, University of Foreign Language Studies, The University of Danang, Vietnam, teaching English language skills courses. She is also a PhD candidate at the same institution with research interests in EFL pedagogy and AI-assisted language learning.
| Affiliate type | University |
|---|