The Acceptance Gate: Three Necessary Conditions for Acting Against One's Strongest Schema

Accounts of self-regulation converge on a two-part story about when a person acts against their strongest habit: a conflict must be registered, and control must be worth its cost. We argue that this story omits a necessary condition. Between noticing a truth about oneself and acting on it lies the acceptance of that truth, and acceptance can fail while detection and control capacity remain fully intact. We propose an architecture of situational autonomy in which three conditions are jointly necessary: an error signal must reach metacognition (detection), the truth it carries must be accepted (the acceptance gate), and the resulting intention must outweigh habit and the cost of control (override). Their relation is conjunctive at the margins: each condition caps autonomy, so as any one approaches zero, autonomy is driven toward the floor. The gate resolves as a logistic function of a global, non-contingent self-worth foundation set against the threat a given truth poses to contingent self-worth in a given domain. This is why the same surgeon concedes every error in the operating theatre and none at home. We disclaim novelty at the level of constructs; the novelty is the wiring: no current framework carries acceptance as a first-class condition on which the override depends. The architecture therefore forbids something additive accounts permit, and this is its test: where acceptance is low, the effect of forecasting depth on autonomy should vanish, not merely shrink. We develop the architecture, its formal core, and a measurement design that renders the prediction preregistrable. The empirical test is the subject of a companion paper.