Distributed · SSRN · Research paper

Human Agency Under Consequential State: Correction, Revalidation, and Recoverable Consequences in Persistent Computational Systems

Micheal Charles Preble · Independent Researcher · Manuscript dated September 21, 2026

Research paper · SSRN 7496758 · Distributed on SSRN

Summary

Research paper. This is the underlying research paper (SSRN 7496758). It is distinct from the policy paper, Human Agency Under Consequential State: A Policy Framework (SSRN 7509878), and is not an earlier version of it.

When a source record is corrected, the correction does not by itself settle what happens to consequences already produced from the earlier state. This paper develops Human Agency Under Consequential State (OD-HAS) as a cross-domain framework for that interval. It distinguishes state change from propagation, propagation from revalidation, a corrected same-ground value from a materially distinct replacement ground, present authority from historical dependency, and prospective disposition of a consequence from residual effects that remain recoverable. It defines four state-change types: withdrawal, replacement, annotation, and derived-state change.

The paper also reports two limited public-record pilots, one using 25 Consumer Financial Protection Bureau complaint records and one using 20 Merit Systems Protection Board procedural chains, followed by a frozen discovery search that closed without identifying a qualifying single public corpus. Across the record systems examined, public visibility of basis revision repeatedly separated from public visibility of interim authority and binding timing.

Evidence status: Conceptual framework with two bounded public-record pilots. The findings concern the structure of public records, not institutional reality, and the paper does not claim that no qualifying corpus exists. The framework is not validated.

Suggested citation
Preble, Micheal Charles. “Human Agency Under Consequential State: Correction, Revalidation, and Recoverable Consequences in Persistent Computational Systems.” Manuscript dated September 21, 2026. Available at SSRN 7496758, 2026.