Philosophy of Action · Predictive Processing · Active Inference
Began 2021, renewed 2026
Two Accounts of Agency
A working paper translating Bratman's planning theory into predictive processing.
Client
Ongoing · working paper
scope
Philosophy of Action · Predictive Processing · Active Inference
year
Began 2021, renewed 2026
duration
Ongoing
Overview
The question.
Predictive processing has reshaped how cognitive science understands the mind — not as a system that passively receives the world, but as one that actively predicts it. Separately, Michael Bratman's philosophy of planning agency has offered one of the most influential accounts of what it means to act with will, as a temporally extended self who plans, commits, and follows through. The two literatures have largely developed apart from one another.
The Challenge
Two literatures, developed apart. |
My research brings them together. Drawing on Andy Clark's active inference model, I translate Bratman's account of agency into the vocabulary of predictive processing, and argue that the two frameworks are structurally and functionally analogous — closely enough that predictive processing can supply Bratman's theory with the "mesoscale" grounding it currently lacks: a mechanistic account of how planning agency is actually implemented in a predictive brain. |
the solution
A modular argument, tested through aberrant precision. |
Structured as a modular series of chapters, the work uses aberrant precision — a concept drawn from computational psychiatry — as a key mechanism for testing the boundaries of free will and agency, generating novel, empirically testable hypotheses in the process. I see predictive processing as a genuinely fertile, interdisciplinary space for this kind of work: a framework capable of giving old questions about agency real mechanistic teeth. This project is still under construction — an active, ongoing piece of research rather than a completed one. |
