Jessie Barrot

Prompting with Purpose: Rethinking Generative AI for English Language Teaching through CLARIFIES+ Prompting Framework

Generative artificial intelligence (AI) is increasingly used in English language teaching (ELT) to assist with lesson preparation, communicative activities, feedback generation, and assessment support. Despite its growing role, classroom use frequently relies on improvised prompts that do not always produce pedagogically appropriate outputs. Language-learning tasks are shaped by multiple instructional variables, including learners’ proficiency levels, targeted skills, genre and register requirements, and sociocultural learning conditions. Consequently, ELT requires prompting approaches that function as structured instructional specifications rather than simple technical commands. This lecture introduces CLARIFIES+, a contextualized prompting framework designed to guide the pedagogical use of generative AI in teaching, learning, and assessment. The presentation begins with a review of commonly discussed prompting strategies and several widely cited prompting frameworks. It then highlights a critical gap between general technical guidance on prompting and the pedagogical needs of ELT, particularly regarding contextual variables, assessment alignment, and ethical considerations. To address this gap, the CLARIFIES+ framework organizes prompt design around nine essential elements: Context, Limitations, Audience, Role, Intent, Format, Inputs, End Product, and Style. An additional “+” layer introduces adaptive mechanisms, including process management, clarification prompts, tool integration, and safety safeguards. Through practical examples, the lecture demonstrates how CLARIFIES+ can support differentiated instruction, develop learners’ prompt literacy and self-regulated use of AI, and promote greater transparency and consistency in AI-assisted formative and summative assessment. The session concludes by discussing potential limitations of the framework, challenges associated with classroom implementation, and directions for future research.