Anti-Phase Theta-Gamma Coupling: A Neurodynamic Hypothesis Derived from a Dynamic Self-Model (Accepted by CNS 2026)

The nature of the self remains a central enigma in philosophy and a key frontier in science. This paper proposes a novel “Dynamic Self-Model”—the “minimal self” emerges from the rapid alternation between first-order “awareness” (A) and second-order “awareness-of-awareness” (AoA)—inspired by phenomenological insights from Buddhist meditation. As a core component, we define AoA—typically not accessible to ordinary introspection—as a spontaneous, non-inferential “knowing” state distinct from effortful “meta-cognition” and attentional “meta-awareness”, thereby advancing it as a prime candidate for the Higher-Order Representation. This conceptual model places clear constraints on candidate neural implementations, thereby recasting the long-standing philosophical puzzle of selfhood as a testable question about neural processes. Among the many possible neural bases, we propose a parsimonious, falsifiable neurodynamic hypothesis as a leading candidate mechanism: an “anti-phase” theta-gamma coupling, where sensory gamma activity (A) and anterior-insular/fronto-insular gamma activity (AoA) are driven by opposite phases of the same theta rhythm. We then outline a proof-of-concept study comparing a simple task baseline with an induced flow state (reduced self-referential processing); we predict robust anti-phase coupling at baseline and selective attenuation during flow. Furthermore, regardless of whether the leading candidate is supported, a systematic testing framework—encompassing comparative, boundary, and causal tests—should be developed. Finally, we argue—within an inference-to-the-best-explanation (IBE) framework—that this parsimonious “Dynamic Self-Model” offers a unified explanatory account of three enduring puzzles in self-consciousness studies (Higher-Order vs. Reflexive theories, the classic infinite regress problem, and the mechanism for phenomenological “transparency”), which in turn provides abductive support for the model.