Beyond Prompt Engineering: Prompting (L)iteracy, Linguistic Capital, and Educational Inequality

This article advances the debate on artificial intelligence (AI) use in education by moving beyond the mechanistic notion of prompt engineering and the human-centered focus of prevailing AI-literacy frameworks. We introduce prompting (l)iteracy, a sociotechnical capacity that is simultaneously iterative (built through cycles of prompt revision and reflection) and critical (attuned to the economic logics, linguistic hierarchies, and distributed agencies that shape AI dialogue). Four theoretical lenses scaffold the concept: (1) post-Fordist political economy explains how each prompt exploits commodified linguistic labor embedded in large language models; (2) Bourdieu’s linguistic-capital thesis illuminates the social inequities likely reproduced by differential prompting proficiency; (3) Latour’s actor-network theory highlights non-human agency, as humans do not only prompt AI, but they are also prompted by AI; and (4) de Certeau’s tactics foreground users’ creative appropriations and resistances. Vignettes of AI interactions illustrate these dimensions and expose the twin pitfalls of techno-solutionism and techno-pessimism. We argue that future research must abandon self-report surveys in favor of authentic, trace-centered mixed methods that record authentic prompt–response logs and machine telemetry. Prompting (l)iteracy thus offers a richer analytic lens for understanding, and ultimately mitigating, AI-mediated educational inequality.