Can We Logically Tell Good from Bad Reasoning?

The possibility of distinguishing good from bad reasoning – the precondition of any normative role for logic – has been challenged by a cluster of skeptical arguments in recent philosophy of logic. While much attention has focused on the gap between logical knowledge and ordinary reasoning highlighted by Harman and Russell, this paper identifies three additional arguments against logical normativity that have received insufficient scrutiny: one concerning the inherent limitations of formal evaluation methods, one concerning the pragmatic indeterminacy of illogical behavior in natural language communication, and one rooted in the early-twentieth-century doctrine of the ‘feeling of logicality’ directly opposed by Charles S. Peirce. The paper pursues two sequential aims: first, to show that skepticism about the logical normativity of reasoning is deeper and more historically entrenched than contemporary discussions suggest; second, to reconstruct Peirce’s theoretical framework – centered on the distinction between logica utens and logica docens , the tripartite account of good reasoning, and the method-based criterion of validity – and to demonstrate that it provides a unified response to all four skeptical challenges.