Scaffolding with Generative AI in Language Classrooms: A Critical Thinking Framework

This paper proposes a critical thinking framework for integrating generative AI into English language instruction. Drawing on the tradition of Socratic dialogue, Vygotsky’s Zone of Proximal Development (ZPD) and collaborative learning, this paper argues that AI should function not as a substitute for instruction but as a scaffold supporting learner autonomy and reflective language growth. The study outlines three applications: AI-assisted scaffolding for ZPD-based writing improvement, collaborative prompting tasks grounded in the CoPIR model, and task designs fostering structural awareness and critical revision. Examples show how instructors can adjust feedback and prompt engagement through both peer and AI mediation. Rather than relying on AI for content creation, students are guided to revise, evaluate, and expand AI output, enhancing linguistic accuracy and self-monitoring. The framework positions AI as a tool for peer interaction and critical language use in the classroom, encouraging learners to think independently while engaging critically with AI-generated language.