Logos and episteme
Nicola D’ALFONSO
ABSTRACT: This paper argues that each of the three core steps in Goodman’s formulation of the ‘new riddle of induction’ is fallacious. First, it challenges the parallel between the justification of deduction and that of induction, highlighting how the latter’s ampliative nature renders Goodman’s purported dissolution of Hume’s problem untenable. Second, it rejects the a priori distinction between lawlike and accidental generalizations, maintaining that any observed regularity provides a prima facie legitimate basis for an inductive inference, thus reconceptualizing Goodman’s critique of Humean habit. Finally, it addresses the “grue” paradox, arguing that its origin lies not in an inescapable dependence on language, but in the illegitimate employment of beliefs that embed presuppositions about an as-yet-undetermined future within the very description of the evidence. As a solution, a normative criterion of exclusion is proposed: the validity of an inductive practice requires that the description of observational data first be purified of any future-dependent belief. This epistemological solution, which operates at a level logically prior to the inference, is then compared to the main alternatives in the literature (including those of Quine, Putnam, Kripke, and Davidson), and its capacity to resolve the riddle without recourse to problematic metaphysical or pragmatic commitments is demonstrated.
