Most arguments about misinformation assume that disagreements are primarily about truth. This paper argues that a large class of contemporary disputes arise earlier: many claims circulating in public discourse are not false but unevaluable. The framework introduced here provides three diagnostic instruments for analysing the evaluability of claims: • The Φ-gate – determines whether a claim specifies a measurement pathway connecting it to observable evidence.• Z-depth – measures the number of reconstruction steps required before a claim becomes evaluable.• Access Collapse – describes structural conditions under which predicates lose meaning when observational access approaches full system resolution. Together these instruments form an epistemic type system: a pre-evaluative filter determining whether a claim is structurally ready to enter truth-evaluative discourse. This paper synthesises a four-paper research series: I – Access Collapse: A Typed Structural Boundary for Operational AccessII – Epistemic Hygiene in PracticeIII – Reconstructability, Depth, and ValidationIV – Before Truth (this paper) Applications discussed include AI evaluation, cryptographic protocols, scientific discourse, and public deliberation.