Korean-as-a-Foreign-Language Learners’ Engagement with Machine Translation Output

Abstract

Researchers and practitioners in foreign language learning continue investigating evolving technologies like machine translators and other artificial intelligence-assisted programs in classrooms. However, less attention has been given to less commonly taught languages, which often lack relevant resources compared to widely taught foreign languages such as English, despite machine translators’ potential in facilitating foreign language writing. This exploratory case study aims to address this gap by examining how US university students learning a less commonly taught language, Korean, use machine translators to revise their writing. Students were trained in a modified version of the Guided Use of Machine Translation model, which requires active engagement from students in selecting, comparing, revising, and justifying their writing choices. By analyzing their cognitive and behavioral engagement, the study sheds light on students’ approaches to incorporating machine translators’ output as feedback into their writing process. The findings suggest that students primarily focus on differences in word choice between their original writing and machine translators’ output over other areas such as particle usage. Students generally demonstrated the ability to accurately identify and correct their errors. Various revision strategies were implemented but adopting exact machine translators’ output was more commonly used, while students also exercised autonomy in their language choices.