Earticle

This study investigates Vietnamese pre-service EFL teachers' readiness for ChatGPT integration through the AI-enhanced TPACK framework. The research addresses two questions: (RQ1) How do Vietnamese pre-service EFL teachers' perceptions and readiness align with AI-enhanced TPACK requirements? (RQ2) What culturally responsive implementation strategies do they plan for ChatGPT integration? Employing an explanatory sequential mixed-methods design, data were collected from 247 pre-service teachers via questionnaires, followed by semi-structured interviews with 12 participants selected through maximum variation sampling. Statistical analysis revealed significant disparities between Content Knowledge (M=5.82) and Pedagogical Knowledge (M=5.64) versus Technology Knowledge (M=4.93). Factor analysis identified four primary barriers: technical infrastructure, institutional support, professional development, and time constraints, explaining 72.4% of implementation challenges. While 83.7% intended to use ChatGPT for instructional planning, correlation analysis showed integration intentions were moderated by institutional support (ΔR²=.14, p<.001) and self-efficacy (β=.53, p<.001). Qualitative findings revealed hybrid implementation approaches emphasizing preparatory rather than real-time use, with cultural adaptation as a central concern. The study demonstrates that successful ChatGPT integration in Vietnamese EFL contexts requires addressing both technical competencies and culturally responsive pedagogical applications.

Nghiên cứu này điều tra sự sẵn sàng của giáo viên tiếng Anh tiền nghề nghiệp Việt Nam trong việc tích hợp ChatGPT thông qua khung AI-TPACK. Nghiên cứu giải quyết hai câu hỏi: (CH1) Nhận thức và sự sẵn sàng của giáo viên tiền nghề nghiệp EFL Việt Nam phù hợp như thế nào với yêu cầu của khung AI-TPACK? (CH2) Họ lập kế hoạch những chiến lược triển khai phù hợp văn hóa nào cho việc tích hợp ChatGPT? Sử dụng thiết kế phương pháp hỗn hợp tuần tự giải thích, dữ liệu được thu thập từ 247 giáo viên tiền nghề nghiệp qua bảng câu hỏi, sau đó là phỏng vấn bán cấu trúc với 12 người tham gia được chọn qua lấy mẫu biến thiên tối đa. Phân tích thống kê cho thấy sự chênh lệch đáng kể giữa Kiến thức Nội dung (M=5,82) và Kiến thức Sư phạm (M=5,64) so với Kiến thức Công nghệ (M=4,93). Phân tích nhân tố xác định bốn rào cản chính: cơ sở hạ tầng kỹ thuật, hỗ trợ thể chế, phát triển chuyên môn và ràng buộc thời gian, giải thích 72,4% thách thức triển khai. Trong khi 83,7% dự định sử dụng ChatGPT cho lập kế hoạch giảng dạy, phân tích tương quan cho thấy ý định tích hợp được điều tiết bởi hỗ trợ thể chế (ΔR²=0,14, p<0,001) và tự hiệu quả (β=0,53, p<0,001). Phát hiện định tính tiết lộ các phương pháp triển khai kết hợp nhấn mạnh việc sử dụng chuẩn bị hơn là thời gian thực, với thích ứng văn hóa là mối quan tâm trung tâm. Nghiên cứu chứng minh rằng tích hợp ChatGPT thành công trong bối cảnh EFL Việt Nam đòi hỏi giải quyết cả năng lực kỹ thuật và ứng dụng sư phạm phù hợp văn hóa.

This study investigates the impact of Online Interactive Learning Environments (OILE) integrated with AI technologies on students' English Language Learning (ELL) skills, specifically focusing on reading, writing, speaking, and listening. The research involved 50 English major students, divided into control and intervention groups, and utilized AI-assisted tools such as WhatsApp Groups, Google Documents, ELSA Speak, and Zoom to enhance interactive and personalized learning experiences. The results demonstrated significant improvements across all assessed areas, indicating the effectiveness of these interventions. The data revealed notable enhancements in reading comprehension, writing development, speaking proficiency, listening comprehension, vocabulary acquisition, and pronunciation accuracy. This study highlights the potential of AI-assisted OILE to create engaging, interactive, and personalized learning environments that foster active student participation and improve overall language proficiency. These findings contribute to the ongoing global discourse on the role of artificial intelligence in education, offering practical implications for educators, curriculum designers, and policymakers who aim to transform traditional EFL classrooms into dynamic interactive learning spaces. Future research should focus on optimizing these technologies to address specific skill deficits and to enhance the efficacy of AI-assisted language learning.

Studi ini menyelidiki dampak Lingkungan Pembelajaran Interaktif Daring (OILE) yang terintegrasi dengan teknologi AI terhadap keterampilan Pembelajaran Bahasa Inggris (ELL) siswa, khususnya berfokus pada membaca, menulis, berbicara, dan mendengarkan. Penelitian ini melibatkan 50 mahasiswa jurusan Bahasa Inggris, dibagi menjadi kelompok kontrol dan intervensi, dan memanfaatkan perangkat berbantuan AI seperti Grup WhatsApp, Google Documents, ELSA Speak, dan Zoom untuk meningkatkan pengalaman belajar yang interaktif dan personal. Hasilnya menunjukkan peningkatan yang signifikan di semua area yang dinilai, yang menunjukkan efektivitas intervensi ini. Data menunjukkan peningkatan yang signifikan dalam pemahaman membaca, pengembangan menulis, kemahiran berbicara, pemahaman mendengarkan, penguasaan kosakata, dan akurasi pengucapan. Studi ini menyoroti potensi OILE berbantuan AI untuk menciptakan lingkungan belajar yang menarik, interaktif, dan personal yang mendorong partisipasi aktif siswa dan meningkatkan kemahiran berbahasa secara keseluruhan. Temuan ini berkontribusi pada wacana global yang sedang berlangsung tentang peran kecerdasan buatan dalam pendidikan, menawarkan implikasi praktis bagi para pendidik, perancang kurikulum, dan pembuat kebijakan yang bertujuan untuk mengubah ruang kelas EFL tradisional menjadi ruang belajar interaktif yang dinamis. Penelitian di masa mendatang sebaiknya berfokus pada pengoptimalan teknologi-teknologi ini untuk mengatasi kekurangan keterampilan tertentu dan meningkatkan efektivitas pembelajaran bahasa yang dibantu AI.

본 연구는 영어를 외국어로 학습하는 (EFL) 대학생들이 글쓰기 과정에서 생기는 질문을 담당 교강사와 ChatGPT 에게 어떻게 나누어 묻는지 살펴보고 각각 맡은 역할이 어떻게 다르면서 서로 보완되는지 논의한다. 한국에서 영어교육을 전공하는 1 학년 학생 17 명이 15 주간의 작문 수업에 참여하여 초고를 작성하고, 두 대상(ChatGPT 와 교강사) 모두에게 질문을 했으며, 학습 일지와 인터뷰를 통해 성찰했다. 수집된 자료를 분석한 결과, 학생들은 주로 번역 지원, 어휘 대안, 문장 흐름, 아이디어 생성과 관련해서는 ChatGPT 를 활용한 반면, 뉘앙스, 문화적 적절성, 내용의 깊이, 평가와 관련된 질문은 교강사에게 더 많이 제기하는 것으로 나타났다. 이러한 선택은 편의성과 접근성 뿐 아니라 ChatGPT 의 한계에 대한 메타인지적 인식과 저자성에 대한 우려에 의해 좌우되었다. 학생들은 AI 를 탐색을 위한 자원으로 활용하면서도, 학문적 기대와 가독성에 대한 권위 있는 안내자로서 최종적으로는 교강사에게 의존한다고 설명했다. 본 연구 결과는 질문하기가 피드백을 구하는 핵심 실천임을 강조하며, 글쓰기 교육에서 ChatGPT 와 교사 피드백을 경쟁적 지원이 아니라 상호보완적 지원으로 위치시키는 접근이 유익할 수 있음을 시사한다.

This study investigates how EFL undergraduates distribute their writing-related questions between ChatGPT and their instructor by highlighting the distinct yet complementary roles of each. Seventeen first-year students majoring in English education in Korea participated in a 15-week composition course where they produced drafts, asked questions to both sources, and reflected through journals and interviews. Analysis of these data revealed that students consulted ChatGPT primarily for translation support, vocabulary alternatives, sentence flow, and idea generation, while directing questions about nuance, cultural appropriateness, depth, and evaluation to their instructor. Decisions were shaped not only by convenience and access but also by metacognitive awareness of ChatGPT’s limitations and concerns about authorship. Students described using AI as a resource for exploration while ultimately relying on the instructor as the authoritative guide for academic expectations and readability. These findings underscore questioning as a central feedback-seeking practice and suggest that writing instruction can benefit from positioning ChatGPT and teacher feedback as complementary, rather than competing, supports.

본 연구는 언어 소실 양상을 측정하기 위해 고안된 심리언어학적 과제인 HALA(Hawai‘i Assessment of Language Access)의 자동 데이터 분석 도구 개발 과정을 제시한다. HALA는 언어 소실 연구에서 유용한 측면이 있으나, 반응시간 측정 과정에서 과도한 수작업이 필요하다는 점에서 연구의 확장 측면에서 한계가 있다. 본 연구는 음고(pitch) 탐지 기술을 활용하여, 시각 자극과 함께 제시되는 비프음과 화자의 발화 시작 사이의 시간 간격을 자동으로 산출하는 도구를 개발하였다. 총 146명의 참여자로부터 수집된 3,755개의 발화 데이터를 대상으로, 자동 분석과 연구자가 수작업으로 분석한 결과를 비교한 결과, 반응시간의 32.68%는 100ms 미만의 오차를 보였고, 33.9%는 100–500ms의 오차를, 나머지는 500ms 이상의 오차를 보였다. 질적 분석 결과, 참여자가 목표 단어를 말하기 전에 uh 또는 oh와 같은 채움구를 표현한 경우 자동 분석과 연구자의 수작업 분석 사이에서 반응시간 차이가 큰 것으로 나타났다. 이러한 결과는 자동 분석에 대한 가능성과 동시에 추가적인 개선을 시사한다. 본 연구는 연구자의 수작업 분석 결과에 대한 보조적인 정보로서 자동 코딩을 효과적으로 활용할 수 있는 방안을 논의한다.

This study presents the development of an automated data analysis tool for the Hawai‘i Assessment of Language Access (HALA), a psycholinguistic task designed to assess lexical access speed as an indicator of language attrition. While HALA has been instrumental in studies of language attrition, its manual response time (RT) measurement process is labor-intensive and limits the scale of studies. To address this issue, we developed an automated tool that determines the time interval between the beep that accompanies the visual stimulus and the onset of a respondent’s verbal response, leveraging pitch patterns. Validation against 3,755 human-coded trials from 146 participants showed that 32.68% of automated RTs deviated by less than 100 ms, 33.9% by 100–500 ms, and the remainder by more than 500 ms. Qualitative analysis revealed that deviations often occurred when participants produced filled pauses such as uh or oh prior to naming the target word. These results suggest that while automation is feasible, further refinement is needed to better align with human measurement. Based on these findings, we discuss the effective use of this tool as a supplement to, rather than a replacement for, human annotation.

Generative artificial intelligence (GenAI) tools such as ChatGPT and Grammarly have rapidly entered English as a Foreign Language (EFL) education, offering new opportunities for feedback and learning autonomy while raising ethical concerns. This exploratory study examined Vietnamese translation stu-dents’ perceptions of GenAI through the combined lenses of the Technology Acceptance Model (TAM) and Self-Determination Theory (SDT). A survey of 44 undergraduates measured self-report ratings on perceived usefulness (PU), perceived ease of use (PEOU), autonomy, competence, and ethical use using a 32-item Likert-scale questionnaire. Results revealed uniformly high self-perceptions across all con-structs, indicating strong subjective acceptance and confidence. Correlational analyses showed signifi-cant positive relationships among all variables, while regression analyses demonstrated that PU signif-icantly predicted competence and autonomy, and competence alone predicted ethical use. These find-ings suggest that students’ ethical engagement with GenAI stems primarily from their perceived com-petence rather than ease or usefulness alone. The study highlights a developmental shift from mere AI acceptance toward ethical empowerment, emphasizing competence-based ethics and autonomy-sup-portive pedagogy. Integrating TAM and SDT provides a holistic framework to understand how cogni-tive beliefs and motivational needs jointly shape responsible AI literacy in EFL education. Pedagogical implications include designing guided AI literacy tasks that transform technological utility into ethical capability.

Các công cụ trí tuệ nhân tạo tạo sinh (GenAI) như ChatGPT và Grammarly đang nhanh chóng được tích hợp vào giảng dạy tiếng Anh như một ngoại ngữ (EFL), mang lại cơ hội cho phản hồi và tính tự chủ trong học tập, đồng thời cũng đặt ra những vấn đề đạo đức. Nghiên cứu này khảo sát nhận thức của sinh viên chuyên ngành ngôn ngữ Anh tại Việt Nam về GenAI dựa trên Mô hình Chấp nhận Công nghệ (TAM) và Thuyết Tự Quyết (SDT). Khảo sát với 44 sinh viên thông qua bảng hỏi Likert 32 mục để tự đánh giá tính hữu ích (PU), tính dễ sử dụng (PEOU), tính tự chủ, năng lực và việc sử dụng có đạo đức. Kết quả cho thấy mức độ tự đánh giá cao ở tất cả các biến, phản ánh sự chấp nhận và tự tin mạnh mẽ của người học. Các phân tích tương quan chỉ ra mối quan hệ tích cực có ý nghĩa giữa các biến. Phân tích hồi quy cho thấy tính hữu ích dự đoán đáng kể năng lực và tính tự chủ, trong khi năng lực là yếu tố duy nhất dự đoán việc sử dụng có đạo đức. Những kết quả này cho thấy việc sử dụng GenAI có đạo đức chủ yếu xuất phát từ năng lực cảm nhận của người học, thay vì chỉ từ tính hữu ích hay tính dễ sử dụng. Nghiên cứu nhấn mạnh sự chuyển dịch từ chấp nhận công nghệ sang trao quyền đạo đức, đồng thời gợi ý thiết kế các nhiệm vụ học tập nhằm phát triển năng lực sử dụng AI một cách có trách nhiệm.

While interpersonal communication has been widely examined in the language-learning area, the role of attachment anxiety in shaping foreign language anxiety (FLA) and learners’ second-language willingness to communicate (L2WTC) remains insufficiently understood. This study investigated the relationship among attachment anxiety, FLA, and L2WTC in 193 college students using structural equation modelling. The structural equation modelling results indicated that: (1) higher attachment anxiety predicted greater perceived FLA and lower L2WTC; and (2) FLA significantly mediated the relationship between attachment anxiety and L2WTC. These findings highlight attachment anxiety as an important individual-difference factor in L2 communication and learning. They also suggest that interventions to enhance L2 participation may be more effective when they explicitly target learners’ underlying emotional and psychological vulnerabilities.

許多研究已探討人際溝通在語言學習中的作用，但依戀焦慮對外語焦慮以及個體參與溝通意願的具體影響，仍相對缺乏深入研究。本研究旨在以193名大學生為樣本，探討依戀焦慮、外語焦慮與第二語言溝通意願之間的關係。結構方程模型結果顯示：（1）依戀焦慮程度越高，個體感受到的外語焦慮越強，且第二語言溝通意願越低；（2）外語焦慮在依戀焦慮與第二語言溝通意願之間具有中介作用。這些發現凸顯依戀焦慮在塑造語言習得與溝通行為中的關鍵角色。本研究亦有助於深化我們對「心理特質個別差異如何影響語言學習結果」的理解，並指出有必要發展更具針對性的介入策略，以處理第二語言有效溝通背後的情緒與心理障礙。

This qualitative study explores the demotivating experiences of English as a Foreign Language (EFL) teachers in Korea, Myanmar, and Uzbekistan. The purpose of this study was to examine how these demotivating experiences vary across contexts and how teachers respond to them. The research was guided by three questions focusing on the nature of demotivators, coping responses, and factors supporting teacher motivation. Drawing on interviews with 15 female teachers and employing grounded theory analysis (Strauss & Corbin, 1998), the study identifies three core themes: (1) systemic and institutional demotivators, including heavy workloads, inadequate salaries, and credential-related pressures; (2) sociocultural and interpersonal stressors, such as student misbehavior, parental demands, and exam-oriented instruction, and (3) professional resilience and coping strategies, including emotional regulation, supportive networks, and intrinsic motivation. Demotivation appeared more frequently in Myanmar and Uzbekistan, where experienced teachers were more likely to demonstrate adaptive coping mechanisms. Overall, the findings underscore that EFL teacher demotivation results from a complex interplay of systemic, cultural, and individual-level factors. Rather than responding passively, teachers actively navigate and negotiate these challenges. This study advocates for a more nuanced understanding of teacher motivation and calls for comprehensive institutional reforms, empowering leadership, and contextually responsive resilience initiatives to enhance teacher well-being across diverse educational contexts.

ဤအရည်အသ ွေးဆ ိုငရ် ာ ိုသေ န ည် က ိုရ ွေးယာွေး၊ မြနြ် ာနငှ ် ဥဇဘက်ကစ္စေနန် င ို င် ြံ ာွေးရ ှ နင ို င် မံ ာွေးဘာ ာအင်္ဂလ ပ် (EFL) ဆရာ/ဆရာြြ ာွေး၏ စ္ ေ်အာွေးသလ ာ သစ္သ ာ အသေ ွေ့အကကံြုံြ ာွေးက ို စ္ ွေးစ္ြ်ွေးသလ လာ ည်။ ဤ ိုသေ န၏ ရည်ရ ယ် က်ြှာ စ္ ေ်အာွေးသလ ာ သစ္သ ာ အသေ ွေ့အကကံြုံြ ာွေး ည် အသမ အသနအြ ြုံွေးြ ြုံွေးေ င် ြည် ိုက ွဲမပာွေးပပ ွေး ဆရာြ ာွေးက ထ ိုအသမ အသနြ ာွေးအာွေး ြည် ို ေိုံမပ နက်က ည်က ို စ္ ွေးစ္ြ်ွေးသလလ ာမ ငွေး် မြစ္် ည်။ ိုသေ နက ို စ္ ေ်အာွေးသလ ာ သစ္သ ာ အသကကာငွေး် ရငွေး် ြ ာွေး၏ သဘာေရာွေး၊ က ိုငေ် ယ်သမြရှင်ွေးြှု ေိုံမပ န ် က်ြ ာွေးနငှ ် ဆရာြ ာွေး၏ စ္ ေ်အာွေးထက် နြ် ှုက ို သထာက်ပ ံသပွေး ည ် အသကကာငွေး် အရာြ ာွေးဆ ိုငရ် ာ သြွေး နွေး် ိုံွေး ိုမြင ် ဦွေးေည်ညွှနက်ကာွေး ွဲ ည်။ ဆရာြ ၁၅ ဦွေးအာွေး သေ ွေ့ဆိုံသြွေးမြနွေး်ကာ Strauss & Corbin (1998) ၏ grounded theory ိုံွေး ပ်နည်ွေးမြင ် ိုသေ နမပြုံ၍ အဓ က သ ေါငွေး် စ္ဉ် ိုံွေး ိုက ို သြာ်ထိုေ်ထာွေး ည်။ (၁) စ္နစ္်ဆ ိုငရ် ာနငှ ် အြွဲ ွေ့အစ္ည်ွေးဆ ိုငရ် ာ စ္ ေ်အာွေးသလ ာ သစ္သ ာ အသကကာငွေး် ရငွေး် ြ ာွေး၊ ဥပြာ အလိုပ်ောဝနြ် ာွေး ြ ာွေးမပာွေးမ င်ွေး၊ လစ္ာြလိုံသလာက်မ င်ွေးနငှ ် လက်ြှေ်/အ အြှေ်မပြုံလက်ြှေ်ဆ ိုငရ် ာ ြ အာွေးြ ာွေး; (၂) လ ြှု-ယဉ်သက ွေးြှုနငှ ် လ နငှ လ် ကကာွေး ြ အာွေးြ ာွေး၊ ဥပြာ သက ာင်ွေး ာွေးြ ာွေး၏ ြ ငသ် ော်သ ာ အမပြုံအြ ြ ာွေး၊ ြ ဘြ ာွေး၏ သောင်ွေးဆ ို က်ြ ာွေးနငှ ် စ္ာသြွေးပ ွဲအသမ မပြုံ ငက်ကာွေးြှုြ ာွေး၊ (၃) ပသရာ်ြက်ရှငန် ယ် ံနင ို ရ် ည်နငှ ် က ိုငေ် ယ်သမြရှငွေး် နည်ွေးြ ာွေး၊ ဥပြာ စ္ ေ် ံစ္ာွေးြှု ထ နွေး် ြုံပ်မ ငွေး် ၊ သထာက် ံပ ံပ ိုွေးြှုက နယ် က်ြ ာွေးနငှ ် အေ င်ွေးပ ိုငွေး် စ္ ေ်အာွေးထက် နြ် ှုေ ို ပေါဝင ် ည်။ စ္ ေ်အာွေးသလ ာ ြှုက ို မြနြ် ာနငှ ် ဥဇဘက်ကစ္စေနန် င ို င် ြံ ာွေးေ င ် ပ ိုြ ိုကကံြုံသေ ွေ့ရပပ ွေး၊ အသေ ွေ့အကကံြုံရ ှ ဆရာြ ာွေး ည ် က ိုက်ည စ္ ာ သမြရှငွေး် နင ို ် ည် နည်ွေးလြ်ွေးြ ာွေးက ို ပ ိုြ ိုမပ ွဲကက ည်။ စ္ိုစ္ိုသပေါငွေး် ရလဒြ် ာွေးအရ EFL ဆရာြ ာွေး၏ စ္ ေ်အာွေးသလ ာ ြှု ည ် စ္နစ္်ဆ ိုငရ် ာ၊ ယဉ်သက ွေးြှုဆ ိုငရ် ာနငှ ် ေစ္်က ိုယ်သရအဆငဆ် ိုငရ် ာ အသကကာင်ွေးအရာြ ာွေး၏ ရှုပ်သထ ွေးစ္ ာ သပေါင်ွေးစ္ပ်ြှုြှ မြစ္်သပေါ်သကကာငွေး် အထ ွေး မြင ် သြာ်မပထာွေး ည်။ ဆရာြ ာွေး ည် ေိုံမပ နမ် ငွေး် ြမပြုံဘ ွဲ ဤစ္ နသ် ေါ်ြှုြ ာွေးက ို ေက်ကကစ္ ာ ဦွေးသဆာငက် ာ ည နင ှု ွေး် သမြရှင်ွေးကက ည်။ ဤ ိုသေ န ည် ဆရာြ ာွေး၏ စ္ ေ်အာွေးေက်ကကြှုက ို ပ ိုြ ိုနက်ရှု ငွေး် စ္ ာ နာွေးလည်ရနန် ငှ ် အြွဲ ွေ့အစ္ည်ွေးအေ ငွေး် မပြုံမပငသ် မပာငွေး် လွဲြှုြ ာွေး မပြုံလိုပ်ကာ သ ေါငွေး် သဆာငြ် ာွေးအာွေး အငအ် ာွေးမြည်ေ ငွေး် ရန ်ေ ိုက်ေ နွေး် ည်။

This study investigates the effect of using an AI-powered Interactive Response System (IRS), Pear Deck, on students’ achievement in an online English as a Foreign Language (EFL) writing course. While teleconferencing tools have been helpful, they might not be sufficient to facilitate a successful online EFL writing class. Pear Deck, which facilitates real-time writing and monitoring, was integrated to address this issue. A quasi-experimental design was used with 70 freshman participants divided into a treatment group (taught with Pear Deck slides) and a control group (taught with static PowerPoint slides). The two groups were taught by the same lecturer using identical lesson plans and assessment tasks for over 12 weeks. Results from a 2 x 2 mixed design ANOVA showed a significantly greater improvement in writing proficiency for the treatment group compared to the control group. Students who used the interactive platform had significantly higher achievement. Additionally, the platform’s features, including the real-time feedback function, interactive format, and semi anonymous participation, might have contributed to increased engagement and reduced anxiety.

This study investigates Korean university EFL students' first-hand experiences, perceptions, and attitudes toward the use of generative AI (GenAI) in English learning. A descriptive survey was administered to 73 undergraduates enrolled in compulsory English courses. Results showed that ChatGPT (65.8%) and Google Translate (42.5%) were the most frequently used tools, primarily for coursework and convenience. Although students generally perceived GenAI as moderately effective for improving English skills, many reported limited familiarity and only moderate technological readiness. Over half of the participants indicated they had never received formal training on using AI tools for language learning. Students rated real-time feedback, translation support, and access to diverse learning resources as key opportunities. These findings suggest that Korean universities should provide structured AI literacy instruction to support pedagogically meaningful and responsible GenAI use in EFL programs, particularly within large, mixed-proficiency general English courses.

이 연구는 영어 매개 수업(EMI) 교실에서 메타언어(metalanguage)의 역할과 그것이 제2언어 습득(SLA)에 미치는 영향을 분석한다. EMI는 주로 내용 학습에 초점을 두지만, 동시에 학습자들이 영어 능력을 향상시킬 수 있는 기회를 제공한다. 그러나 EMI 교사들은 언어 교수 전문성이 제한되어 있어 언어적 문제를 효과적으로 다루는 데 어려움을 겪는 경우가 많다. 본 연구는 한국의 고등학교와 대학교 수업을 대상으로 한 교실 관찰과 담화 분석을 통해 EMI 교사들이 사용하는 메타언어의 유형과 기능을 분류하였다. 연구 결과, 교사들은 주로 반기술적(semi-technical) 및 비기술적(non-technical) 메타언어를 사용하며, 이를 선제적으로(preemptively) 활용하기보다는 주로 사후적으로(reactively) 사용하는 경향이 있는 것으로 나타났다. 또한 학문 분야에 따라 메타언어 사용 방식에도 차이가 있는 것으로 확인되었다. 이러한 결과는 EMI 교사들이 학습자의 이해와 제2언어 습득을 효과적으로 촉진할 수 있도록 언어적 전략을 갖추게 하는 전문성 개발 프로그램의 필요성을 시사한다. 본 연구는 학문 내용 지도와 제2언어 발달 간의 상호작용을 강조함으로써 EMI 교수법 연구에 기여한다.

This study examines the role of metalanguage in English-medium instruction (EMI) classrooms and its implications for second language acquisition (SLA). While EMI primarily focuses on content learning, it also provides opportunities for students to develop their English proficiency. However, EMI teachers often struggle with addressing linguistic issues due to limited language-teaching expertise. Through classroom observations and discourse analysis in high school and university settings in South Korea, this study categorizes the types and functions of metalanguage used by EMI teachers. Findings reveal that teachers predominantly use semi-technical and non-technical metalinguistic terms, often reactively rather than preemptively. Disciplinary differences also shape how metalanguage is employed. These insights highlight the need for professional development programs that equip EMI teachers with effective linguistic strategies to enhance student comprehension and SLA. This study contributes to EMI pedagogy by emphasizing the intersection of disciplinary instruction and second language development.