In the current cognitive framework of language teaching methodology, any language that can be taught is primarily considered a communication tool. The term "teachability" implies the existence of a set of rules, referred to as grammar. Teachability is synonymous with learnability, and as a result, deep learning and generative AI are increasingly maturing. Consequently, language teaching approaches and practices that are rule-based, including all that follow the rules of lexicography, syntax, and narrative grammar, are being challenged or may be even replaced. For instance, ChatGPT can now generate a poem that adheres to syntactic, grammatical, and even prosodic conventions and discourse grammar. However, the creation of an original poem is sometimes precisely an offense against these rules and conventions. Language learning based on interpersonal communication, including immersive learning, can be replaced by human-machine interaction. From the perspective of language teaching, human-based teaching will no longer focus on learning how to produce a grammatically correct utterance but on learning how to judge whether AI-generated output is plausible to human reality. Language pedagogy towards the AI era does not build on the imparting and practice of language rules—a component that can be entirely replaced by machines—but instead, it begins at a higher cognitive level in Bloom's taxonomy, specifically critical competence and critical thinking. This is the adjustment that current language pedagogy must make, as AI-empowered virtual classrooms will become a "dangerous supplement" to human beings, the language animal. Meanwhile, generative AI will also make the concept of language acquisition applicable only to a person's first language. It will be imperative for a “cogito” being to master a truly meaningful mother tongue, one that defines the self and constitutes I-subjectivity, a tool with creative, aesthetic and critical empowerment, rather than just a communication tool sans human individuality. In other words, human-human language teaching aims at a language that can be called human-empowered, and this is what transformation the current language pedagogy needs to be prepared for in the Generative AI Era.

Keywords: language pedagogy; the Generative AI Era; rule-based teaching; human-empowered

Dr. Fan Jinghua (NUS, Singapore) is a lecturer at the Singapore Centre for Chinese Language, Nanyang Technological University. His research interests and expertise encompass language teaching, translation practice and studies, literary studies, and critical theory. He has authored over 40 academic papers and served as the principal investigator for more than 20 research projects on language teaching, commissioned or sponsored by the Ministry of Education, Singapore.