Corrigibility under Temporal Constraint: A Framework for Specifying and Evaluating Corrective Opportunities

A corrective intervention may remain nominally available after the conditions required for its effective implementation have disappeared. Evaluating this possibility requires distinguishing intervention latency, the closure of corrective opportunities, and the eventual outcome of an intervention. This article proposes a framework for specifying those distinctions without defining successful correction into a temporal metric. An assessment begins from a fixed time and state, a reference policy, a declared preservation criterion, and a class of admissible interventions. Intervention latency is measured until a modification becomes operational, independently of whether preservation is achieved. Corrective opportunity is represented by a set of activation times \(W_D(ω)\) within the declared evaluation horizon at which at least one admissible intervention can attain the preservation threshold given the observable history of a realisation. The framework distinguishes an opportunity set fixed by the information available at the assessment time, a set conditioned on the observable history of a realisation, and a later reassessment. Temporal feasibility is summarised by \(C_p = P_M(T_r ∈ W_D(ω) D)\), where \(D\) specifies the assessment. In the special case of a single initial window this can reduce to an inequality involving a closure horizon \(T_v\), with the boundary convention stated explicitly. \(C_p\) is not a new mechanism, a probability of successful correction, or a claim to information absent from a richer model of the same dynamics. Three minimal cases separate timely successful correction, an intervention completed outside the opportunity set, and a timely modification that fails the preservation criterion. The article then specifies an evaluation design, comparison requirements, and failure conditions. No empirical advantage is claimed. The contribution is a formal and methodological framework for testing whether temporal representations assist prospective assessment of corrective opportunities.