This repository accompanies the paper “Observer Design and the Engineering of Epistemic Inevitability.” The paper addresses a complementary question to epistemic failure in scientific inference: given a universe and a target law, what observational access is sufficient or necessary to make that law epistemically inevitable? Working within fully specified, finite toy universes and analytically ideal but structurally constrained observers, the paper introduces a precise definition of epistemic inevitability. A law is inevitable if no alternative law remains observationally compatible, even given infinite data and ideal inference. Inevitability is treated as a structural property of the interaction between dynamics and observation, rather than as a feature of learning, optimisation, or statistical confidence. Two explicit constructions are developed: Universe C demonstrates that inevitability can be engineered. By appropriately designing observer interfaces, specific laws can be made unavoidable without revealing full microstates. Universe D demonstrates a sharp obstruction. It provides a concrete example of a universe in which no nontrivial law can be made inevitable without essentially complete access, regardless of observer sophistication. The paper develops a general framework based on observational equivalence and quotients, identifies inevitability thresholds across observer hierarchies, and introduces novel concepts including law interference and law capacity. Together, these results clarify what can and cannot be guaranteed in principle under restricted observation. This work is intended as a direct companion to: Jefferson, B. (2026). Observer-Limited Laws: Inference Under Restricted Observation. Zenodo.DOI: 10.5281/zenodo.18301565 The two papers together provide a structural account of law discovery as a property of (dynamics, observer) pairs, analysing both epistemic failure modes and epistemic guarantees.
Bob Jefferson (Tue,) studied this question.