Empirical Admissibility and the Conditions of Confirmational Reference

This paper analyzes empirical admissibility as the set of conditions under which empirical claims can meaningfully enter confirmational discourse at all. It shows that confirmation presupposes admissible reference: data can support a claim only if the reference relations connecting observation, description, and claim are standing-preserving under admissible redescription. Within a closed admissibility framework, empirical evidence does not confer legitimacy by accumulation, probability, or convergence. Instead, admissibility determines whether evidence is even eligible to count as confirmational. When reference is unstable—due to scope drift, illicit redescription, or representational mismatch—no amount of empirical data repairs the failure. Apparent confirmation in such cases is purely notional. The analysis is eliminative. No empirical methodology, testing protocol, or theory of evidence is proposed. The paper instead establishes a boundary condition: empirical confirmation is admissible only when reference is fixed in a way that remains coherent across all licensed descriptions. Where this condition fails, confirmation is not weak or inconclusive—it is undefined. By isolating admissibility as prior to confirmation, the paper dissolves disputes over realism, empiricism, and evidential support that presuppose evidence can operate independently of reference stability. It closes these disputes by showing that disagreements about confirmation typically arise from unexamined failures of empirical admissibility rather than from conflicting interpretations of data.