Metabolic Computation in a Closed Cognitive State Space: Pressure-Stability as a Criterion of Meaning

This work introduces a computational framework in which cognition is modeled as a metabolic process unfolding within a closed discrete state space defined by a 13×20 toroidal topology. The system operates through internal dynamics governed by the relation Φ = P·V = k·T, linking signal pressure, attention allocation, and processing intensity. Unlike conventional computational systems based on optimization, this framework does not rely on external objectives, loss functions, or predefined semantics. Meaning emerges operationally as the stability of structures that persist under repeated pressure cycles. The closed topology ensures continuous re-exposure of states, creating an intrinsic validation mechanism where only structurally resilient patterns endure. This work proposes a shift from goal-directed computation to constraint-driven stabilization, suggesting that meaning can be treated as a computable property of dynamical systems. The framework builds upon prior work in metabolic cognition, including Metabolic Weather, MAD (Metabolic Adaptive Dynamics), and MTOS (Metabolic Tzolkin Operating System).