✶ The Evidential History of a Posterior: Update Trails, Path Integrity, and Doxastic Accountability

Credencing evaluates how confidence answers to evidence. The existing Credencing papers distinguish Objective Evidence, E0, Perceived Evidence, EP, Assigned Credence, CA, and Deep Rationality, SD; separate evidence-processing gaps from belief-integrity gaps; defend warranted uncertainty; diagnose rationalization; extend the model to institutions; analyze AI-mediated inquiry; and specify when raw gaps become warranted diagnoses. This paper adds the missing temporal layer. A posterior credence is not fully assessable as a final number. It is answerable to its evidential history: the prior it inherited, the evidence it registered, the alternatives it considered, the updates it made or refused, the uncertainty it preserved or erased, and the standards it applied across time. Two agents may end with the same assigned credence while having very different epistemic paths. One may have updated honestly from new evidence; the other may have protected a fixed conclusion through selective registration, asymmetric likelihood treatment, standard drift, hindsight laundering, or confidence ratcheting. The paper formalizes update trails, warranted posterior paths, dynamic Core Irrationality, path integrity, and audit-trail adequacy. It argues that credence evaluation must become record-sensitive whenever stakes, public claims, institutional authority, or AI-mediated evidence processing make hindsight reconstruction likely. The practical result is a temporal protocol for individuals, institutions, and human-AI systems: record prior confidence, live alternatives, evidence events, perceived evidential support, warranted uncertainty, assigned credence, update conditions, later revisions, and reasons for movement or nonmovement. Credencing then becomes not merely a snapshot diagnosis, but a disciplined account of how confidence earned its present location.