Logic has traditionally been construed as a normative discipline; it sets forth standards of correct reasoning. Explosion is a valid principle of classical logic. It states that an inconsistent set of propositions entails any proposition whatsoever. However, ordinary agents presumably do -- occasionally, at least -- have inconsistent belief sets. Yet it is false that such agents may, let alone ought to, believe any proposition they please. Therefore, our logic should not recognize explosion as a logical law. Call this the 'normative argument against explosion'. Arguments of this type play -- implicitly or explicitly -- a central role in motivating paraconsistent logics. Branden Fitelson (2008), in a throwaway remark, has conjectured that there is no plausible 'bridge principle' articulating the normative link between logic and reasoning capable of supporting such arguments. This paper offers a critical evaluation of Fitelson's conjecture, and hence of normative arguments for paraconsistency and the conceptions of logic's normative status on which they repose. It is argued that Fitelson’s conjecture turns out to be correct: normative arguments for paraconsistency probably fail.