Incidental Learning of Multi-Word Expressions Through Watching a Video in Different Captioning Modes

Although Multi-Word expressions (MWEs) are widely recognized as reliable indicators of language mastery (Dang et al, 2021; Majuddin et al., 2021; Puimège & Peters, 2019), empirical research examining both receptive and productive knowledge of MWEs across on-screen captions remains inconsistent. The present study aimed to clear this uncertainty by focusing on Korean-speaking novice English learners and their incidental acquisition of MWEs using form recall and form recognition under three caption conditions: no captions, whole captions, and MWE-enhanced captions. A total of 55 undergraduates from various disciplines in South Korea participated in a three-session design with intervals of three to four weeks. An analysis included data from fifty participants who completed form recall and form recognition tests in all sessions. Results indicated all groups achieved modest improvement, but there were no statistically significant differences between captioning types. Nonetheless, the MWE-enhanced captions showed slightly better performance than the others, suggesting a moderate pedagogical promise for this type of input. A slight gap in MWE gains from form recognition and form recall highlights the instructional relevance of these dimensions of MWE learning. Overall, more research is needed to explore how recognition and recall of MWEs develop across different proficiency levels and input modes.