<![CDATA[Augmented Reality has shifted from being a novelty technology in English Language Learning to an established pedagogical infrastructure in fewer than two decades. This position paper This position paper argues that the field is now entering a third generation in which AR's value is no longer derived from the technology itself but from its integration with artificial intelligence, mobile ubiquity, and learner-centered task design. Using a structured narrative review combined with a Scopus bibliometric mapping of 398 documents published between 2007 and 2026, the study traces the evolution of AR-supported English learning across three periods: a foundational period (2007-2016) dominated by marker-based and barcode prototypes, a mainstreaming period (2017-2021) shaped by mobile AR and learning theory, and a convergent period (2022-2026) characterized by an explosive growth of empirical work, with two thirds of all indexed publications appearing in this most recent five-year window. The findings indicate that AR consistently produces moderate-to-large positive effects on linguistic and affective outcomes, but that effects are uneven across skills, age groups, and contexts. The paper takes the position that further research should move beyond demonstrating that AR works, and instead address how AR-enhanced English instruction should be designed, governed, and integrated with generative AI in ways that protect equity, teacher agency, and pedagogical coherence. Implications for curriculum designers, teacher educators, and policymakers in English as a Foreign Language contexts are discussed, with particular attention to low-resource settings such as Indonesia, where mobile-first AR deployment offers a credible pathway to scale immersive English instruction.]]>
