Manifold Traversability Dissociates Conscious Level from Neural Entropy: An Empirical Evaluation of Noetic Diffusion Theory

This paper evaluates whether conscious level is better captured by dynamical traversability than by neural entropy, explicitly framed against Churchland’s criteria for a mature neurocomputational theory. Using EEG and fMRI across anesthesia, sleep, epilepsy, and task engagement, we show that manifold geometry and second-order dynamics dissociate conscious level from entropy-like measures. Meta-Noetic Phase Space and Jacobian-based indices track loss and recovery of consciousness where entropy fails, supporting a dynamical–geometric account aligned with biological and computational constraints.