AI-Augmented World Chat and AI Quest Narration as an Affordance for L2 Learning of English Slang and Pragmatics in WOW

Digital game worlds such as World of Warcraft (WoW) function simultaneously as social spaces and action systems, generating dense streams of ephemeral language that are rich in slang, netspeak, and informal pragmatic routines. For English-as-a-second-language (ESL) learners, these “digital wilds” offer valuable exposure beyond textbooks but are difficult to access because chat is fast, fragmentary, and governed by opaque community norms. Prior MMORPG research has documented vocabulary gains and emergent pragmatic routines, yet routine pragmatic acts and community slang remain underrepresented in instruction; ambient observation is underexplored as a learning mode; slang is rarely treated as an explicit target; affective factors such as anxiety and willingness to communicate are seldom linked to linguistic outcomes; AI is typically designed as a tutor rather than as an environmental affordance; and few studies triangulate quantitative measures with discourse-level evidence.To address these gaps, this article presents the design of a customized WoW server that embeds large language model (LLM)–driven “world chat” bots and text-to-speech (TTS) quest narration as configurable affordances for L2 learning of English slang and pragmatics. AI chatbots generate context-tuned, safety-filtered ambient messages that model high-frequency routines, while multimodal quest narration converts static text into synchronized audio–visual input. The article also outlines a mixed-methods, two-group pretest–posttest quasi-experimental design comparing this AI-augmented server with a standard WoW environment. Planned measures include slang comprehension, frequency and accuracy of target usage in production, complexity–accuracy–fluency metrics, affective scales, and discourse-level analyses of in-game chat. By conceptualizing AI as part of the game environment rather than a separate tutor, the study proposes a scalable framework for using AI-augmented MMORPGs as researchable L2 digital wilds for slang and pragmatic development.