EUVIMEDEUVIMED entdecken
Crossref

Figurative Language Acquisition and Generative Artificial Intelligence in FLE: A Phraseodidactic Approach with Georgian Learners

Natia Abashidze, Natalia SURGULADZE

Journal of Narrative and Language Studies · 30.05.2026

DOI: 10.59045/nalans.2026.96 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Narrative and Language Studies, online erschienen am 30.05.2026. Verfasst von Natia Abashidze und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study explores the intersection of phraseodidactics and generative artificial intelligence (AI) within the context of teaching French as a Foreign Language (Français Langue Étrangère - FLE) to native Georgian speakers. Drawing on Conceptual Metaphor Theory (CMT) and cognitive approaches to second language (L2) vocabulary acquisition, this exploratory classroom case study investigates how intermediate university learners ($N = 15$) comprehend, interpret, and produce L2 metaphorical expressions when supported by large language models (LLMs). Over a two-week action research intervention, participants engaged with a structured pedagogical framework alternating between autonomous AI consultation and human teacher mediation. Written production artifacts, interaction logs, and learner reflection journals were systematically evaluated using a dual-coded analytical rubric across four primary dimensions: frequency of use, contextual adequacy, conceptual complexity, and originality. The empirical findings reveal a distinct performance gradient dictated by cross-linguistic transparency; learners demonstrated high operational mastery of transparent expressions sharing conceptual structures with Georgian or English (M = 74.3\%), but experienced sharp drops in accuracy when handling culturally unique, opaque idioms (M = 29.6%). Crucially, qualitative analysis demonstrates that while LLMs provide effective low-friction validation for transparent forms, their automated definitions fail to override literal image interference in opaque structures without explicit intercultural scaffolding. Furthermore, the study identifies a robust correlation between idiomatic opacity and the triggering of high-level metacognitive monitoring behaviors. Based on these insights, a three-phase instructional integration model is proposed to optimize technological scaffolding while safeguarding the irreplaceable role of human pedagogical mediation.

Zitieren

APA

Abashidze, N., & SURGULADZE, N. (2026). Figurative Language Acquisition and Generative Artificial Intelligence in FLE: A Phraseodidactic Approach with Georgian Learners. Journal of Narrative and Language Studies. https://doi.org/10.59045/nalans.2026.96

Vancouver

Abashidze N, SURGULADZE N. Figurative Language Acquisition and Generative Artificial Intelligence in FLE: A Phraseodidactic Approach with Georgian Learners. Journal of Narrative and Language Studies. 2026. doi:10.59045/nalans.2026.96

BibTeX
@article{Abashidze2026,
  author = {Abashidze, Natia and SURGULADZE, Natalia},
  title = {Figurative Language Acquisition and Generative Artificial Intelligence in FLE: A Phraseodidactic Approach with Georgian Learners},
  journal = {Journal of Narrative and Language Studies},
  year = {2026},
  doi = {10.59045/nalans.2026.96}
}
Kostenlos starten →