From Thermodynamic Boundary Conditions to Bayesian Residue: A Constraint-Based Derivation of Consciousness

The persistent lack of convergence in consciousness research may reflect not a shortage of empirical data but the absence of a principled rule for identifying the appropriate level of explanation. This paper presents a constraint-based derivation of the minimal biological architecture required for a falsifiable theory of consciousness. Rather than beginning from axioms about experience, information, or specific neural mechanisms, the derivation begins from empirically unavoidable conditions for the persistence of living systems and proceeds stepwise: from thermodynamic boundary maintenance and metabolic closure, through multicellular chemical and electrical coupling, to the incorporation of organism–environment history via activity-dependent plasticity, and finally to the metabolically sustained integration of large-scale neural processes into a temporally extended dynamical regime. Each level is introduced only when the preceding one becomes insufficient to account for an identifiable capacity, and each transition defines explicit conditions for falsification. Within this architecture, conscious experience is characterised as a reversible, energy-dependent regime sustained in cortical association networks coupled to sensory, interoceptive, autonomic, and metabolic systems that encode organism-level viability constraints. The framework's central testable commitment is conjunctive: reportable conscious states are predicted to occupy a restricted region of a two-dimensional state space defined jointly by sufficient cortical metabolic support (CMRGlu ≈ 44–46% of normal) and preserved large-scale integration (PCI ≈ 0.31). States violating either constraint, including the off-diagonal case of NREM sleep, are predicted to be unconscious. The apparent persistence of a subject is interpreted as the probabilistic stabilisation of viability-weighted system history — a Bayesian residue.