Self-Concealing Information and Observer-Modifying Dynamics

This record contains the PDF and LaTeX source for Self-Concealing Information and Observer-Modifying Dynamics: A First-Principles Measurable-State Theory of Internal Blindness, External Anchors, Structural Insulation, and Delayed Audit. The paper develops a first-principles measurable-state framework for observer-modifying information: informational exposures that can change an observer’s future readout and action channels, thereby making diagnosability itself endogenous rather than fixed. The theory formalizes self-concealing and observer-modifying dynamics in hidden-state controlled systems using measurable kernels, comparison laws, and pathwise observable processes. It derives exact and approximate results on internal blindness, bridge theorems linking local concealment to cross-regime discrimination loss, conditions under which external anchors improve diagnosis through testing deficiency and weighted total variation, observational structural insulation under restricted interfaces, and delayed or recurring audit recovery, including finite-horizon and changepoint-style results. The work is theoretical and theorem-driven; it does not present an empirical benchmark suite. The LaTeX source includes the canonical machine-readable theorem and certificate manifest, and the PDF is tagged for improved machine accessibility. This record is intended for researchers working on statistical experiments, stochastic control, hidden-state systems, sequential detection, information-flow security, and formal models of observer modification and auditability.