Structural Drift Cannot Be Detected by the Layer It Affects: The Substrate-Governance Case

Misclassification and structural drift are widely treated as problems of signal quality, interpretive sensitivity, or decision-layer coherence. This framing has produced a class of proposals — signal-aware systems — that attempt to detect drift by instrumenting the interpretive pipeline with additional sensors, meta-models, and feedback loops. This paper argues that such approaches are structurally incapable of governing drift because they operate inside the very layer that drift affects. When the interpretive pipeline deforms, its capacity to detect deformation collapses with it. The paper formalizes this as the Substrate-Governance Claim: structural drift cannot be detected by the layer it affects. The claim rests on three pillars: (1) an epistemological anchor in the Never Why doctrine, which establishes the impossibility of governing a drifting system through its own explanations; (2) field evidence from the CVS/Caremark pharmacy prior-authorization case, in which misclassification emerged not from signal failure but from mandate deformation invisible to the interpretive layer and only detectable from the technician vantage point outside it; (3) implementation proof in the existence of substrate-layer governance architectures that separate identity, privilege, admissibility, and lineage from the interpretive pipeline. The paper concludes that signal-aware systems cannot solve misclassification or drift because they treat governance as an interpretive problem. Drift is not interpretive. Drift is structural. Structural problems require substrate-layer governance.