Description
Abstract: Artificial intelligence (AI) has changed translation from one-step conversion into an interactive process of explanation, comparison, and revision. This concept-driven integrative synthesis examines how machine translation (MT), neural machine translation (NMT), and generative artificial intelligence (GenAI) may support translation as learning in university English as a foreign language (EFL) classrooms. The analysis differentiated the roles of 32 empirical, review, conceptual, methodological, and policy sources. The literature indicates that AI-supported translation may promote noticing, lexical development, audience awareness, and comparative evaluation, but fluent output can conceal semantic shifts, pragmatic mismatch, cultural loss, and weak learner engagement. Productive use appears to require translation literacy, evaluative judgment, learner control, explicit criteria, independent verification, and proportionate process visibility. Drawing on task-based language teaching (TBLT), the paper proposes a five-stage Human-AI Translation Task Cycle: Contextualize, Translate, Interact with AI, Evaluate and Revise, and Reflect and Transfer. The framework offers a testable classroom design rather than a validated intervention.
Thông tin các tác giả
Full name: Vu Van Chinh
Academic degree and status: Master’s degree; PhD Candidate
Affiliation: FPT University
Address: 2411, Nguyen Van Loc, Hadong, Hanoi
Phone: 0902 092 091
Email: chinhvv@fe.edu.vn
Từ khóa
Artificial intelligence, EFL classrooms, human-AI translation, machine translation literacy, task-based language teaching
