The Individuation Problem: Why System-Based Frameworks Require a Prior Individuation Condition

Abstract

Frameworks across consciousness studies, cognitive science, systems theory, and engineering share a structural dependency: each requires a bounded, persistence-capable unit at a specific point in its formalism and supplies none. This paper identifies that dependency as a category error common to all post-system frameworks — frameworks that operate on systems without deriving the conditions under which those systems exist as individuated units. Six frameworks spanning information theory, inference, cognition, learning, engineering, and biological control are evaluated: Integrated Information Theory, the Free Energy Principle, Global Workspace Theory, reinforcement learning, classical control theory, and Consciousness Mechanics. Each presupposes its unit of analysis in a framework-specific way: a system for Φ, a Markov blanket, a workspace, an agent, a plant, an organism. The dependency is then exposed through a class of boundary-indeterminate configurations — tumor, fetus, split-brain, and eusocial colony — in which the system boundary is contested, transitional, divided, or constitutively unstable. No framework in the class can resolve these configurations from within its own formalism. The paper establishes that a prior individuation condition is not an optional refinement but a precondition for the coherence of any system-dependent account, identifies closure under maintenance as the minimal condition that resolves the problem, and demonstrates that the result relocates rather than replaces post-system frameworks within a dependency hierarchy they have left implicit.