A Bidirectional Criticality TAC Model of Conscious Emergence: A Phase-Transition Framework under Temporal Integration and Resource Constraints

Existing theories of consciousness remain largely static: they describe irreducibility, binary access, or steady-state convergence, but rarely how conscious states switch, collapse, and rebuild over time. Under three constraints, namely finite resources, non-equilibrium maintenance, and discrete input, this paper constructs a minimal dynamical model in which temporal span and reconstruction activity serve as two control parameters and closed-loop completeness as the order parameter. It argues that the emergence and regression of consciousness are bidirectional critical phase transitions in this two-parameter space: closed-loop completeness sustains a high plateau inside a feasible region and falls cliff-like once a boundary is crossed, while discontinuous jumps, hysteresis, and critical slowing are unified as signatures of a single bistable structure. The model derives a scaling law in which the upper bound of the temporal window grows with the square root of the available power quota, and from it generates testable predictions on age-related cliff-like cognitive decline, hysteresis between anesthetic induction and emergence, working-memory scaling, and supercritical EEG spectral changes, together with five falsifiable hypotheses. The model presupposes no specific neural substrate; all conclusions are convergence results under resource constraints.