Quiet But Not Stable

Deterrence is often inferred from observable non-events: no attack, no escalation, no visible crisis, no overt response. This article separates that observation from the underlying stability claim. Quiet is an observed output condition. Capacity is the counterfactual response potential implied by the current state. Under a single observation rule — quiet is recorded when a realized disturbance qt remains at or below directional capacity Ct — survival establishes only qt ≤ Ct. It does not reveal the distance to the boundary or whether that boundary is moving. The article develops this identification result, notes that an adversary who chooses the next test can endogenize qt, and specifies one state-transition mechanism through which potential can change beneath continued quiet: residual burden, amortization, and wear, defined as a decline in the ability to clear subsequent residual. A constructive recurrence with published parameters holds disturbance magnitude fixed and produces 45 consecutive quiet episodes on the wear path while capacity declines; visible failure appears only when the hidden boundary is crossed. The same parameterization does not fail under constant clearance, and a learning path restores potential rather than degrading it. The resulting framework treats perturbation, recovery, and inherited state as evidence about stability that non-occurrence alone cannot provide.