This study examines extramural L2 writing practices of third-age (60+) EFL learners in community-based lifelong learning contexts in Istanbul, Türkiye. Using interview data from a purposeful subsample of 30 learners drawn from a larger profiling study (N=535), it explores multilingual genre repertories, modalities, and literacy trajectories, advancing a lifespan and socially situated perspective on L2 learning and L2 writing.
This paper presentation argues that Cicero’s Five Canons equip L2 teachers to assess what AI cannot: rhetorical nuance. Drawing on over 1,400 peer feedback instances from 700+ students, we demonstrate how classical rhetoric validates teacher expertise and provides practical strategies for maintaining pedagogical authority in an increasingly automated world.
This study examined how annotation training affected Taiwanese undergrad students’ annotation quality and their perceptions toward such training. In English Composition II, students completed two self-annotated assignments for reviewer (instructor and peer) feedback. Preliminary findings showed that student held overall positive attitudes toward annotation training. Annotation quality improvement, however, varied individually.
This proposal explores L2 writing teachers’ identity dilemma and humanistic reconstruction in the AI era. It adopts a mixed-methods design, analyzes internal/external causes of role conflicts, and aims to propose a human-centered framework to balance AI application and humanistic teaching for teachers’ professional development.
This paper presentation situates generative AI in a conversational pedagogy for the L2 composition classroom in a way that de-emphasizes the mere linguistic production of the written text, but which preserves the dialogic practices of writing between writer, community, and discourse to encourage writerly development.
This qualitative study of 52 international graduate students examines how multilingual writers negotiate voice through AI collaboration. Drawing on Bakhtin’s dialogism and the APSE model (Wang & Wang, 2025), findings reveal a monologic-to-dialogic continuum shaped by positionality. Authenticity emerges not despite AI engagement, but through intentional, agentive negotiation with it.
