Pan-indexicality and prompt: developing a teaching model for AI-mediated academic writing

Abstract AI-mediated academic writing calls for new pedagogical approaches to the application of prompt engineering for writing courses. Whereas previous studies mainly inform students of prompt engineering techniques, little is known about how prompt engineering functions from the perspective of meaning negotiation between the human and generative AI. This paper explores the integration of the Pan-indexical process of linguistic signs into a prompt-based teaching model (PBTM), emphasizing its potential to facilitate meaning negotiation in prompt engineering during the early stage of AI-mediated academic writing. The PBTM consists of four key components: encyclopedic knowledge, contextual information, evaluative critical thinking, and iterative design. The application of PBTM lies in the early stage of idea development of academic writing and is organized around four major steps: crafting the initial prompt; refining the prompt with contextual information; engaging in evaluative critical thinking; iterative progression toward a desired response. This paper suggests that the Pan-indexical process of linguistics signs can be employed to AI-mediated pedagogical approaches to enhance students’ ability in optimization of prompts through a deeper understanding of meaning negotiation between the students and generative AI to support their academic writing process.