The Entangled Agency Framework: Modeling Human–Machine Collaboration in the Writing Classroom

This concept preprint introduces the Entangled Agency Framework (EAF), a model for examining how rhetorical agency circulates within human–machine writing systems. Building on theoretical foundations from Kenneth Burke's notion of symbolic action, Bruno Latour's actor–network theory, and Karen Barad's concept of intra-action, the EAF reframes writing with AI not as a contest between human control and algorithmic automation but as a dynamic ecology of interpretation. Within this ecology, agency does not reside in any single actor but emerges from the interplay of multiple variables: system (S), alignment (A, denoting AI responsiveness and fit with human intent), reflection (R), prompt (P), human labor (H), and infrastructure (I). Transparency (T), defined as the degree to which interpretive processes remain visible, modulates them all. Expressed as E = f(S, A, R, P, H, I | T), the framework offers a rhetorical and computational heuristic for tracing how meaning and responsibility shift across hybrid writing events.