Affective Control under Uncertainty: A Two-Level Theory of Consciousness

Major theories of consciousness explain important aspects of reportability, integration, or inference, yet none jointly accounts for three persistent facts: wakeful experience feels continuously present; effortful agency waxes and wanes with uncertainty rather than raw computational load; and most cognition proceeds without reportable awareness. Affective Control under Uncertainty (ACU) proposes a two-level architecture that separates consciousness from selfhood while explaining their functional coupling. Level 1 consists of affective control awareness: intrinsically valenced monitoring of regulatory viability under uncertainty, sustained as a standing readiness condition during wakefulness. Level 2 consists of self-modeling under uncertainty: transient, metabolically costly integration recruited when unresolved policy competition under differential affective stakes and temporal constraint forces categorical commitment. This recruitment is formalized as self-model recruitment intensity (SMRI), a graded control variable predicted to scale with policy entropy multiplied by affective variance under temporal forcing. ACU explains continuous presence without constant self-reference, effort scaling with uncertainty rather than prediction error magnitude, and the dissociation between vivid awareness and attenuated self-experience in flow, meditation, sleep, anesthesia and depersonalization. It advances falsifiable predictions concerning neural dynamics, pharmacology, development, comparative cognition, and artificial systems, and argues that phenomenality is identical with globally integrated, intrinsically valenced viability control rather than an additional property layered onto computation. By distinguishing baseline affective monitoring from transient self-model recruitment, ACU offers a unified control-theoretic account of presence, agency and their graded deployment under uncertainty.