We formalise a distinction between two epistemically independent knowledge channels: direct experiential knowledge K_d and formal symbolic knowledge K_f. We prove that K_d constitutes a valid epistemic ground independent of K_f, that K_f is a translation function rather than a validity operator, and that the demand for K_f as a precondition for accepting K_d constitutes a *category error*. We term this error the **Verification Fallacy**. Three corollaries are derived: - **Independence Corollary**: K_d and K_f are generated by structurally disjoint processes and admit independent validity conditions. - **Resistance Corollary**: Under conditions of sufficient cross-context stability, K_d cannot be defeated by the absence of K_f. - **Burden Corollary**: The burden of proof in K_d-vs-K_f disputes lies with the party asserting K_f as necessary, not with the holder of K_d. The theorem is illustrated via the Wagyu case — direct sensory knowledge of A5 wagyu beef quality that is (a) valid, (b) not reducible to Certified Angus Beef grading protocols, and (c) not defeated by the absence of formal certification. Formal connections are established to Polanyi's tacit knowledge framework, with explicit identification of what this theorem contributes beyond the existing literature. The paper introduces operationally real phenomena defined by inter-subjective consistency at the phenomenal level (not requiring institutional protocols), approximate surjectivity of the translation function τ (acknowledging Polanyi's limits on symbolic articulation), and a scope condition for resistance based on cross-context stability rather than an unverifiable universal claim.