Best Before Date Necessity: A Reply to Psillos

Abstract

This discussion paper is a reply to Stathis Psillos' paper "Induction and Natural Necessities" (J Gen Philos Sci 48(3):327–340, 2017. https://ift.tt/2OaCrSJ), published in this journal. In that paper, he attempts to refute David Armstrong's solution to the problem of induction. To accomplish this desideratum, he proposes that the best explanation for our observed regularities is a sort of "best before date" necessity. That is, necessary connections may break down and are not by default timeless. He develops arguments against my (Castro, Teorema 33(3):67–82, 2014) defence of the necessitarian solution regarding a previous paper by Beebee (Noûs 45(3):504–527, 2011. https://ift.tt/2vfiTFH). He alleges that (a) best before date necessity is no worse than timeless necessity; (b) his proposal does not imply any further inductive generalisation to timeless necessity; and (c) inductive inferences are justified. In this discussion paper, I provide arguments against these three claims.