Contemporary theories of consciousness—including Global Neuronal Workspace Theory (GNWT) and Integrated Information Theory (IIT)—primarily characterize conscious states in terms of large-scale integration or ignition occurring within temporally local neural configurations. Here we introduce the Temporal Recursive Auto-Integrative Theory (TRAIT), a formal framework that posits consciousness as a dynamically self-maintaining process requiring diachronic integration across endogenous temporal scales. TRAIT advances two jointly necessary conditions: (i) a minimal regime of global metastable coordination, indexed by phase coherence R(t) above a dynamical threshold, and (ii) non-zero short-range diachronic integration (ΦD,short), operationalized as conditional mutual information between present and past neural states given input history. Unlike state-based metrics, ΦD quantifies endogenous temporal continuity independent of exogenous drive. We provide an operational estimation pipeline based on time-delay embedding and k-nearest-neighbor mutual information estimators with surrogate testing. TRAIT predicts dissociations between instantaneous integration and diachronic continuity in perturbational paradigms (e.g., TMS-EEG), differentiates metastable coordination from hypersynchronous collapse, and generates testable hypotheses across wakefulness, anesthesia, NREM sleep, and cerebellar disruption. The framework further proposes that affective valence correlates with the temporal derivative of internally generated entropy (−dH/dt), formulated as a testable informational hypothesis rather than an ontological identity. Positioned in light of recent adversarial collaborations contrasting GNWT and IIT, TRAIT shifts the explanatory focus from spatial integration alone to temporally recursive self-maintenance, offering a mathematically explicit and empirically falsifiable account of diachronic continuity as a constitutive constraint on conscious processes.
Building similarity graph...
Analyzing shared references across papers
Loading...
Angelo Ferraiuolo
Building similarity graph...
Analyzing shared references across papers
Loading...
Angelo Ferraiuolo (Tue,) studied this question.
www.synapsesocial.com/papers/69b25b1996eeacc4fcec9849 — DOI: https://doi.org/10.5281/zenodo.18934238