Dual Negative Transfer: Mother‑Tongue‑Driven Cognitive Interference and AI‑Triggered Linguistic Deviation in EFL Writing and Targeted Mitigation Strategies

Abstract

Digital language learning has prevailed among college students. They widely use generative AI for writing revisions. While convenient, AI triggers a new type of linguistic interference that coexists with conventional first-language (L1) negative transfer, creating two sources of writing errors. Researchers have spent years studying this cognitive interference, yet most earlier work barely touched on how digital tools mess up students’ language output. When students copy AI writing without thinking, they pick up stiff, repetitive phrases — this AI-caused transfer works totally differently from native language interference. Current studies tend to talk about these two issues separately; few compare them comprehensively, and there are not enough targeted teaching plans to fix both types of errors together. Based on a systematic literature review, this paper constructs a three-dimensional framework to compare the two negative transfers, analyzes their multiple causes, and proposes tiered strategies alongside a four-step closed-loop teaching model. It enriches negative transfer theory for digital English as a foreign language (EFL) education and provides practical guidance for university English instructors.