김은성 Kim Eun-sun

DOI:10.21716/TKFL.82.1 Vol.82(No.0) 1-30, 2026

This study offers a bibliographical analysis of 『Beginning Korean』(1976), an elementary Korean language textbook compiled by the late Professor Kang Yoon-ho for the International Summer School at Ewha Womans University. Produced as a mimeographed edition without formal publication, this short-term instructional material from the 1970s has remained largely unexplored in the history of Korean language education. Using a documentary research method, the study examines the textbook’s structure and cultural features. Organized into Hangul acquisition, sentence pattern practice, speaking activities, and vocabulary learning, the textbook may reflect the influence of structuralist approaches and the audio-lingual method, along with an early attempt at functional syllabus design. The text units appear to emphasize repetition of sentence patterns, while the speaking units seem oriented toward everyday situational communication. Cultural content appears through linguistic, traditional, geographical, sociocultural, and intercultural elements, suggesting a relatively systematic approach for its time―one that can tentatively be understood as an early integrated model combining grammar, communicative function, and cultural education, though further comparative research is needed to substantiate this.

한국어 교재, Korean Language Textbook, 단기 교재, Short-Term Textbook, 국제하기대학, International Summer School, 한국어, Beginning Korean, 문헌 해제, Annotated Study

김지원 Kim Jiwon , 김영주 Kim Youngjoo

DOI:10.21716/TKFL.82.2 Vol.82(No.0) 31-58, 2026

This study examined how Korean learners’ uptake and target error correction differed according to the explicitness of GPT-generated written corrective feedback. Forty-one intermediate-to-advanced learners of Korean for academic purposes participated in explicit GPT feedback (n=15), guided GPT feedback (n=16), and a self-checklist (n=10). The explicit condition provided error locations, correct forms, and rationales, whereas the guided condition provided error locations, metalinguistic explanations, and hints without correct forms. Descriptively, all three groups showed lower error rates after revision, with the largest reduction in the explicit feedback group. The explicit group also showed significantly higher total uptake and successful uptake rates than the guided group whereas partial, unsuccessful, and non-uptake responses were relatively more common in the guided group. Case analyses showed that GPT occasionally identified acceptable expressions as errors or provided inappropriate revision directions.

GPT 피드백, GPT Feedback, 피드백 명시성, Feedback Explicitness, 서면 오류수정 피드백, Written Corrective Feedback, 학습자 수용, Learner Uptake, 표적 오류 수정, Target Error Correction

박지순 Park Jisoon

DOI:10.21716/TKFL.82.3 Vol.82(No.0) 59-83, 2026

This study examines how DEI (Diversity, Equity, and Inclusion) is realized in Korean language textbooks published in Japan. Four volumes from three textbook series were analyzed, focusing on visual materials and dialogues, along the three dimensions of diversity, equity, and inclusion. The findings show that nationality is largely confined to Korea and Japan, reducing the context of Korean language use to scenes of Korea-Japan exchange, and that characters of East Asian descent are predominant while characters of African descent are rendered invisible. In occupations and lifestyles, a gendered division of labor recurs: institutional and professional authority is assigned to men, whereas educational and care roles are assigned to women. Dialogue structures, by contrast, vary across the textbooks. In two of the three series, Korean men initiate conversations and lead in explaining and questioning, whereas in the remaining series the Japanese woman, positioned as a learner, performs face-threatening acts more frequently. Even so, initiating talk and providing information tend to be allocated to men and the work of maintaining relationships to women, so that conversational authority remains concentrated in particular genders and nationalities. As for inclusion, no overtly derogatory or discriminatory expressions were found; however, the repeated interactional pairing of Korean men with Japanese women represents Korean language learners and members of the Korean-speaking community through a narrow range of nationalities and genders. These findings suggest that, even without explicitly reproducing discrimination, such textbooks may convey particular social norms and power relations to learners at the level of the hidden curriculum.

한국어 교재, Korean Language Textbooks, 일본 한국어 교육, Korean Language Education in Japan, DEI, 한국어 교재 분석, Korean Language Textbook Analysis, 잠재적 교육과정, Hidden Curriculum

송한태 Song Hantae

DOI:10.21716/TKFL.82.4 Vol.82(No.0) 85-115, 2026

This study analyzes Thai beginner learners’ awareness and use of Korean particles from the perspectives of language transfer and interlanguage. Korean particles mark grammatical and semantic relations in a sentence, but Thai does not have a case particle system equivalent to that of Korean. Therefore, Thai learners may experience difficulty understanding the necessity of particles, distinguishing their functions, and using them appropriately in communication. The participants were ten beginner-level Korean learners majoring in Korean at a university in Thailand. Data were collected through particle-use tasks and semi-structured interviews. The tasks were used as supplementary data to examine learners’ basic awareness of Korean particles, while the interviews were used as the main data to analyze their perceptions and difficulties. The interview data were analyzed with reference to Braun and Clarke’s (2006) thematic analysis. The results showed that learners recognized the basic functions of ‘은 /는’ (-eun/-neun), ‘이/가’ (-i/-ga), ‘을/를’ (-eul/-reul), ‘에’ (-e), and ‘에서’ (-eseo) relatively well in controlled tasks. However, they reported difficulties in actual use, including delayed particle selection, confusion between similar particles, omission, and avoidance. These patterns were interpreted in relation to language transfer from Thai and interlanguage development during Korean particle acquisition. The study suggests that particle instruction should provide repeated practice in real communicative contexts.

태국인 한국어 학습자, Thai Learners of Korean, 한국어 조사, Korean Particles, 조사 인식 및 사용 양상, Particle Awareness and Use, 모국어 전이, Language Transfer, 중간언어, Interlanguage

오솔미 Oh Solmi , 민정호 Min Jungho

DOI:10.21716/TKFL.82.5 Vol.82(No.0) 117-143, 2026

This study aimed to analyze the patterns of writing tasks presented in academic writing textbooks for international students from the perspective of heuristics. To this end, six textbooks independently compiled for liberal arts writing courses offered to undergraduate international students at Korean universities were selected, and the writing tasks in each textbook were examined based on an analytical framework consisting of textual, rhetorical, and discursive dimensions derived from McComiskey’s (2000) discussion of heuristics. The analysis showed that these textbooks generally reflected the formal structure and argumentative organization of academic writing, but relatively few devices were found that encouraged students to consider audience awareness and reflect on social context. This suggests that, although the textbooks are relatively effective in presenting the form and procedure of completed academic texts, they do not sufficiently provide heuristics-based devices that help learners construct meaning through questioning and exploration. Therefore, the design of writing tasks for international students should more evenly reflect textual, rhetorical, and discursive dimensions.

발견 학습, Heuristics, 학술적 글쓰기, Academic Writing, 글쓰기 교재, Writing Textbooks, 교재 분석, Textbook Analysis, 쓰기 과제, Writing Tasks

유홍빈 Yu Hungpin

DOI:10.21716/TKFL.82.6 Vol.82(No.0) 145-179, 2026

This study examines research trends in Korean language education in Taiwan by applying BERTopic topic modeling to a corpus of 214 theses and journal articles collected from academic databases in Korea (RISS) and Taiwan (NDLTD and NCL Taiwan Periodical Literature). After preprocessing of the Korean, Traditional Chinese, and English texts, BERTopic identified 12 topics, which were interpreted from three complementary perspectives: linguistic, educational, and learner-related. All 12 topics contained both Korean- and Chinese-language documents, suggesting that the multilingual embedding model captured cross-lingual semantic similarity; this alignment held after controlling for shared surface vocabulary and under document-level permutation tests. Contrastive linguistic studies constituted the largest research stream. In the primary corpus, topic proportions by period suggested a shift toward learner-related variables after 2022; however, this pattern was not reproduced when one supplementary document was added. Comparison by source revealed different emphases between materials published in Korea (cultural education and learner motivation) and those published in Taiwan (contrastive linguistics). Sensitivity analyses of corpus and language composition, together with truncation diagnostics showing that 94.9% of documents exceeded the model’s 128-token embedding limit, indicated that the topic structure is condition-dependent and should be interpreted as exploratory. This study contributes to Korean language education research by incorporating academic data from Taiwan and demonstrating both the applicability and the boundary conditions of BERTopic for multilingual research trend analysis.

대만 한국어교육, Korean Language Education in Taiwan, 연구 동향, Research Trends, BERTopic, 토픽 모델링, Topic Modeling, 한국어교육, Korean Language Education

이양양 Li Yangyang , 조건예 Zhao Jianyi , 왕팡링 Wang Fangling , 한재민 Han Jaemin

DOI:10.21716/TKFL.82.7 Vol.82(No.0) 181-211, 2026

This study aims to analyze the research trends in Korean language education in China from 2016 to 2025 using BERTopic, an embedding-based topic modeling approach. The dataset consists of 813 journal articles related to Korean language education collected from the CNKI database. The titles, keywords, and abstracts of the studies were combined and used as units of analysis. The analysis identified a total of 17 topics, which were categorized into five core domains: talent cultivation and educational policies, Korean language teaching and learning models, language and translation education, ICT-integrated Korean language education, and value and cultural education. Among these topics, applied and interdisciplinary talent cultivation(Topic 0) was identified as the largest research cluster, followed by speaking education based on situational language teaching(Topic 1), Korean pronunciation education(Topic 2), Korean language curricula in higher vocational colleges(Topic 3), flipped learning and microlearning-based blended teaching models(Topic 4), and Korean pragmatics and discourse education (Topic 5). Temporal analysis revealed that research from 2016 to 2018 mainly focused on language knowledge and skill acquisition, while studies after 2020 increasingly emphasized practical talent cultivation, digital-based teaching and learning, ideological and political education, and intercultural communicative competence. This study provides an objective and systematic analysis of research trends in Korean language education in China by applying BERTopic and offers implications for future research directions and the development of Korean language education in China.

중국의 한국어교육, Korean Language Education in China, 연구 동향분석, Research Trend Analysis, 버토픽, BERTopic, 토픽 모델링, Topic Modeling, 연도별 토픽 변화 양상, Temporal Topic Trends

이은주 Lee Eun-joo

DOI:10.21716/TKFL.82.8 Vol.82(No.0) 213-233, 2026

The purpose of this study is to propose a Korean teaching and learning plan for foreign learners of intermediate level or higher, focusing on the scenes related to fermented food and storage space in Kim Ae-ran’s 「Knife Scar」. In recent years, Korean language education has emphasized the improvement of language skills and the cultivation of cultural understanding skills, and food culture is attracting attention as a representative cultural education material that can understand Koreans’ lifestyles and values. Fermented food in the work has an important educational meaning as a cultural symbol that contains the accumulation and waiting of time, community values, and family labor and memories beyond simple dietary elements. In addition, food culture education using literary works helps learners to deeply understand Korean emotions and life culture beyond acquiring vocabulary and grammar. In particular, 「Knife Scar」 effectively reveals the cultural and emotional meaning inherent in Korean food culture by organically linking and showing food, human life, and family memories. Therefore, this study analyzed the meaning of fermented food and storage space in the work, and based on this, suggested a three-session teaching and learning plan for learners of intermediate level or higher. Furthermore, by exploring the possibility and educational significance of culture-oriented Korean language education using literary works, we tried to expand the scope of Korean culture education.

문학 작품 활용, Use of Literary Works, 발효 음식, Fermented Food, 상호문화적 성찰, Intercultural Reflection, 소설 「칼자국」, Novel 「Knife Scar」, 한국어 교육, Korean Language Education

주리리 Zhu Lili , 차윤정 Cha Yunjung

DOI:10.21716/TKFL.82.9 Vol.82(No.0) 235-275, 2026

Spoken discourse competence in Korean language education involves not only grammatical accuracy but also the ability to select appropriate expressions according to discourse situations and speaker-hearer relationships. With the growing use of generative artificial intelligence in language education, AI-generated dialogue has received attention as a potential educational resource. This study examines Korean spoken dialogues generated by ChatGPT and explores their potential and limitations as supplementary materials for spoken discourse education. Six dialogues representing requests, refusals, and suggestions in formal and informal relationships were analyzed as the primary dataset. Additional dialogues were used to supplement the initial analysis and to examine patterns under more specific relationship, request-burden, and prompt conditions. Because the datasets differed in generation time and purpose, they were analyzed qualitatively according to their respective roles. The analysis focused on grammatical accuracy, pragmatic appropriateness, and spoken discourse features. The results showed that the generated dialogues were generally grammatically stable and generally maintained contextually appropriate speech levels, basic discourse functions, and topic continuity. Basic pragmatic strategies, such as indirect requests, reason-giving, mitigation, gratitude, and alternative suggestions, were also observed. However, apologies, acknowledgment of responsibility, and compensatory strategies were not consistently realized in high-burden formal requests. In addition, hesitation, repetition, self-repair, and incomplete utterances were limited, while similar indirect expressions and relatively well-structured sentences appeared repeatedly. These findings suggest that ChatGPT-generated dialogues may be used as supplementary materials after teacher review and revision. They may support comparison and revision activities, but should not replace authentic spoken materials or existing textbooks. The findings are limited to the datasets and prompt conditions examined in this small-scale exploratory qualitative study.

한국어교육, Korean Language Education, 구어 담화 교육, Spoken Discourse Education, 생성형 인공지능, Generative Artificial Intelligence, ChatGPT, 교육 자료, Educational Materials

천민지 Chun Minji

DOI:10.21716/TKFL.82.10 Vol.82(No.0) 277-306, 2026

This study aimed to develop a generative AI-based Korean vocabulary learning application and explore its educational applicability through a pilot implementation with Korean language learners. The application, Vocab Talk-Talk, was designed according to four principles: contextualized learning, output-oriented learning, interaction, and staged learning. In Stage 1, learners inferred the meanings of target vocabulary from example sentences, while in Stage 2, they produced sentence-level output by answering questions using the target words. Stage 3 provided dialogue practice through a mobile chat interface, and Stage 4 involved practical speaking, writing, and role-play tasks using the target vocabulary in various communicative contexts. The pilot results showed generally positive learner evaluations, particularly of the dialogue practice in Stage 3. Future revisions should include pronunciation support, flexible stage selection, and improvements in technical stability and usability. This study is significant in that it proposes concrete design principles and presents an implementation case for a generative AI-based Korean vocabulary learning application while examining learners’ experiences and satisfaction through a preliminary application.

어휘 학습, Vocabulary Learning, 생성형 AI, Generative AI, 어휘 앱, Vocabulary Learning Application, 맥락 기반 학습, Context-based Learning, 생산적 어휘 지식, Productive Vocabulary Knowledge