Foundations of Intervention Consistency

Preprint: This work is a preprint, has not undergone peer review, and is made publicly available to establish a public scientific record. This manuscript establishes a minimal and audit-stable conceptual framework for reasoning about intervention consistency in experimental and protocol-level inference. It introduces precise definitions of interventions, audit-fixed evaluation rules, descriptions, consistency, and neutrality, without presupposing any physical theory, geometric structure, dimensionality, dynamics, or reconstructive ambition. The framework is intentionally non-constructive and does not introduce tests, protocols, experimental realizations, or empirical claims. Its sole purpose is to fix the logical language required to discuss consistency under interventions and to prevent hidden structural or geometric assumptions from entering subsequent analyses. In particular, the work clarifies the distinction between consistency and neutrality, and provides a terminological fixing of the term “geometric” as used in later Geometric Inconsistency Test (GIT) papers, explicitly excluding spatial, metric, or reconstructive interpretations. This paper serves as the foundational entry point (GIT 1) of the GIT series and defines the conceptual ground on which subsequent intervention-based inconsistency tests are formulated. Correspondence regarding this work may be directed to:[email protected]