The chaos–state space dream model: Integrating phenomenology and neurodynamics in dream consciousness.

We propose the Chaos-State Space Dream Model (CSSDM): a candidate neural correlate of dream consciousness. CSSDM treats dreaming as the brain's exploration of hierarchically organized attractor landscapes when sensory anchoring lifts and prefrontal control attenuates during REM sleep. The proposal combines two published frameworks. From chaos-theoretic accounts of dream phenomenology, we take the observation that dreams have dynamical structure: attractor-like coherence, bifurcation-like transitions, sensitivity to subtle influence. From an embedding-theoretic account of consciousness, we take a candidate substrate: delay coordinate embedding via generalized synchronization in recurrent cortical networks. The synthesis maps dream phenomenology to specific dynamical properties: narrative coherence to attractor occupation, scene transitions to bifurcations, emotional intensity to limbic modulation of control parameters, bizarreness to access to normally inaccessible state-space regions, immersion to attenuated metacognition. Three predictions follow. First, bifurcation signatures in EEG should precede reported scene transitions. Second, cross-regional REM coherence should predict post-sleep insight. Third, nightmare episodes should show attractor entrapment signatures that normalize with successful therapy. The model generates falsifiable predictions about the neural dynamics of dreaming.