조영우 Young Woo Cho

DOI:10.15702/mall.2026.29.1.1 Vol.29(No.1) 1-31, 2026

As generative AI is increasingly being used for real-time interaction, existing SLA theories―notably Complex Dynamic Systems Theory (CDST)―face limitations in explaining the micro-mechanisms of how meaning is constructed and negotiated through language in real-time human-AI interactions. Adopting an abductive approach, this study proposes the Quantum Model of Language (QMoL) to reconceptualize language as probabilistic meaning potentials that collapse into specific form-meaning combinations through contextualized interaction. Grounded in research findings in quantum cognition, QMoL defines language use and acquisition as a cyclical process involving meaning-wave generation, contextual collapse, particle-like realization, and reintegration into the learner’s semantic field. This framework provides a parsimonious explanation for linguistic complexity, variability, creativity, and context sensitivity through the unified quantum concepts of superposition, collapse, decoherence, and entanglement. Pedagogically, the study proposes a paradigm shift from knowledge transmission to AI-enhanced prompt engineering and coaching. In this model, AI functions as a dynamic resonance field while teachers act as coaches guiding learners to exercise their agency in orchestrating their own meaning-construction processes during interaction. Implications are discussed for establishing QMoL as a coherent theoretical model for fostering learner agency, adaptive fluency, and human-centered communication in AI-mediated second language education.

Quantum Model of Language, generative AI, second language education, quantum cognition, learner agency

한수미 Sumi Han , 김민지 Minji Kim

DOI:10.15702/mall.2026.29.1.32 Vol.29(No.1) 32-53, 2026

This study examines how media discourse on early childhood English education in South Korea has evolved over a 36-year period (1990-2025). A total of 3,671 news articles were retrieved from BigKinds, a news big data platform operated by the Korea Press Foundation. Keywords were extracted using the platform’s TextRank-based automatic extraction tool and further refined through morphological preprocessing in Python. Keyword frequency analysis and co-occurrence network analysis were conducted using the Louvain algorithm, with centrality measures employed to trace structural shifts across three periods. The findings reveal that ‘parents’ and ‘kindergarten’ were the most frequent keywords, functioning as core hubs that bridged multiple discourse clusters. Media coverage was organized around four major axes: parental agency, institutional spaces, market dynamics, and policy intervention. In Period 1 (1990-2011), discourse centered on English, young children, and kindergarten, reflecting the institutionalization of early English learning. In Period 2 (2012-2019), policy-related keywords, including Ministry of Education, public education, and private education costs, gained prominence as regulatory debates intensified following the 2014 Act on Normalization of Public Education. In Period 3 (2020-2025), keywords such as online, contactless, and COVID-19 emerged alongside debates over entrance exams for young children, signaling a shift toward digital learning and concerns about educational equity. By quantitatively mapping long-term trends in media framing, this study demonstrates that discourse on early childhood English education has been continuously reconstructed at the intersection of education, policy, market forces, and technology, suggesting that the media functions as an active constructor of educational discourse.

early childhood English education, media discourse, news big data, network analysis

임관혁 Gwan-hyeok Im , 고경민 Kyoungmin Ko

DOI:10.15702/mall.2026.29.1.54 Vol.29(No.1) 54-72, 2026

This study explores the educational convergence of generative artificial intelligence (AI) and the metaverse in a universitylevel English course, with the aim of enhancing students’ understanding of English-speaking cultures and their digital literacy. A 16-week course was designed and implemented, in which 30 undergraduate students collaboratively developed English cultural content using generative AI tools (e.g., Vrew, Napkin AI, Hedra AI) and constructed immersive cultural spaces within a metaverse platform. Employing a case study approach, data were collected through a post-course survey and open-ended reflections. Quantitative data were analyzed using descriptive statistics, while qualitative responses were examined through thematic analysis. The results indicate that the course significantly promoted students’ intercultural understanding, particularly in recognizing cultural diversity and developing empathetic perspectives toward people from different cultural backgrounds. In addition, students reported high levels of confidence in using digital technologies, including real-time collaboration in the metaverse and creative multimedia content production with AI tools. Qualitative findings further revealed that learners perceived generative AI not as a substitute for learning, but as a collaborative tool that supported planning, visualization, and creative expression. Overall course satisfaction was also notably high, reflecting strong learner engagement and ownership. Based on these findings, the study highlights the pedagogical value of integrating generative AI and the metaverse in English education and suggests the importance of critical digital literacy to support effective and ethical technology use in higher education contexts.

artificial intelligence, metaverse, digital literacy, intercultural understanding, case study Applicable level, tertiary education

조규희 Kyuhee Jo

DOI:10.15702/mall.2026.29.1.73 Vol.29(No.1) 73-94, 2026

This study explores the educational potential of Vibe Coding by developing ‘English Sound Explorer,’ a web-based phonemic awareness application for Korean elementary students using Google Antigravity. The development process, involving 73 natural language prompts, demonstrated a collaborative evolution wherein the pedagogical requirements of the developer (i.e., the researcher) conflicted with, compromised, and evolved alongside the technical implementations generated by the Large Language Model (i.e., Gemini 3 Pro). This dynamic interaction resulted in a gamified tool based on a five-stage developmental hierarchy―rhyming, oddity, blending, segmentation, and manipulation―targeting specific phonemic difficulties (/r/-/l/, /p/-/f/, /b/-/v/) caused by L1 Korean interference. A suitability analysis by four elementary English education experts, conducted through thematic analysis, confirmed the app’s effectiveness as individualized scaffolding, particularly for slow learners. The findings highlight the critical role of the developer’s Content Knowledge (CK) and pedagogic reasoning in guiding AI to refine educational logic and verify its outputs. Taken together, Vibe Coding can empower teachers to become ‘Educational Architects’ capable of addressing specific classroom micro-needs through domain expertise and prompt literacy.

vibe coding, phonemic awareness, AI-assisted app development, educational architect, Google Antigravity

Joohoon Kang

DOI:10.15702/mall.2026.29.1.95 Vol.29(No.1) 95-114, 2026

Situated within Technology-Enhanced Language Teacher Education (TELTE), this qualitative multiple-case study adopts video-mediated reflective practice as its primary theoretical lens to investigate microteaching as a site for teacher development. While Multimedia-Assisted Language Learning (MALL) is retained only as a positioning frame for the digital artifacts, the study examines how five undergraduate English majors in a Korean university negotiated their pedagogy and identity across two iterative cycles. Data from a multimedia corpus―including video recordings, stimulated recall interviews, and reflection logs―were analyzed using reflexive thematic analysis. Findings reveal three key trajectories: (a) video evidence shifted participants from procedural method fidelity to data-driven adaptation by making the disconnect between plans and interactional reality visible; (b) technological affordances (pausing, rewinding) acted as a prosthetic sense, extending teacher noticing to missed learner cues; and (c) the screen created an “objectified self,” enabling participants to critique and edit their emerging professional identities. Crucially, the study suggests that these outcomes were not automatic effects of technology but were conditionally achieved through structured pedagogical supports that mitigated evaluative anxiety. The paper argues that video-mediated microteaching functions effectively when designed not as a performance test, but as a scaffolded inquiry into practice.

Video-mediated reflective practice, Technology-Enhanced Language Teacher Education, TELTE, teacher noticing, teacher identity, microteaching, reflexive practice, Korean EFL

Eun-jou Oh , Jun Seo Song , Yun Ji Yi , Jeong Hyun Kim , Seo Yeon Lee , Ye Eun Sung

DOI:10.15702/mall.2026.29.1.115 Vol.29(No.1) 115-144, 2026

This study investigates the development of a web platform for AI-assisted Content and Language Integrated Learning (CLIL) focused on Global Citizenship Education (GCED). Adopting a participatory inquiry design rooted in the Students as Partners (SaP) framework, the research utilized an iterative design-based research (DBR) process involving five undergraduate co-researchers and a teacher-researcher, supported by data from 14 student interviewees and 54 survey respondents. Analysis was structured around Garrett’s Five Planes of User Experience, synthesizing data from team meetings, interviews, a course-wide survey, and learner autoethnographies. Findings revealed cumulative design tensions that shifted over time: while early feedback focused on linguistic wayfinding and instructional purpose, sustained engagement uncovered systemic frictions related to interaction flow, functional efficiency, and the necessity for richer multimodal scaffolding. Consequently, the resulting platform, "Glove," underwent critical revisions, including the integration of bilingual interface labels, multimodal resource prioritization, and expanded AI prompt functions. Beyond technical refinement, the findings demonstrate that the participatory process substantively enhanced learner agency, pedagogically grounded AI literacy, and global citizenship dispositions. "Glove," serves as a shared epistemic space for translanguaging and cognitive scaffolding, facilitating the co-construction of knowledge regarding complex global challenges.

AI translators, reading comprehension, grammar, students’ perception, reading tasks