A Dynamical Return Constraint for Consciousness Perturbational Closure as a Cross-Theoretical Invariant

We propose the Return Constraint Model (RCM), a perturbation-based dynamical framework for discriminating conscious from unconscious states (Massimini et al., 2005). Rather than defining consciousness as a static informational or metaphysical property, RCM characterises conscious states as occupying a dynamical regime in which perturbation evokes global propagation, directed causal return within a bounded temporal window, and transient stabilisation in a structurally selective attractor manifold (Friston, 2000; Kelso, 1995). The model is substrate-neutral but biologically scoped, explicitly falsifiable, and designed to resist trivial recurrence, arousal confounds, and superficial simulation (Chalmers, 1996). RCM is formulated as a necessary, not claimed sufficient, operational condition. It does not resolve metaphysical questions regarding subjective experience, but provides a falsifiable intervention-based framework for comparing classical, quantum, and artificial system implementations. The temporal window of conscious integration — W2 (80–250 ms) — is derived from two independent physical mechanisms grounded in directly measurable neural anatomy: Population B corticothalamo-cortical loop timing (Layer 6 axonal conduction times of 2–50 ms one-way, producing round-trips of 22–190 ms, Stoelzel et al., 2017) and TRN alpha gating physics (one to two alpha cycles at 8–12 Hz). This derivation is independently corroborated by five independent methodological branches comprising over 180 independent evidence items — alpha EEG oscillations, attentional blink, backward masking and recurrent processing, Pöppel temporal integration epochs, and sequential metacontrast mandatory windows — none constructed to support RCM, each converging on the same 80–250 ms range from a different experimental paradigm. The W2 derivation is not a free parameter. It is a boundary condition readable from published electrophysiology and independently confirmed by five independent methodological literatures. Applied to the thalamocortical dysrhythmia (TCD) family, RCM generates a four-category gating failure taxonomy applicable across nine neurological and psychiatric conditions. The cross-domain clinical programme constitutes an independent validation not achievable by existing frameworks from equivalent first principles. Explicit falsifiers, operational testing protocols, and anti-simulation criteria are specified. Version 2: Supplementary Appendices A–D added (Interface with Contemporary Neuroscience; Interface with Collapse and Quantum Models; Systems and Physics Interpretation; Artificial Systems Interface).