Induction and Natural Kinds Revisited

Assoc. Prof. Howard Sankey (University of Melbourne)

Abstract:  In ‘Induction and Natural Kinds’ (Principia 1997), I argued that inductive inference is reliable to the extent that it is grounded in the natural kind structure of the world.  I drew upon some ideas of Hilary Kornblith and Brian Ellis in an attempt to resurrect the principle of uniformity of nature as part of the justification of our use of induction.  Basically, the best explanation of the reliability of induction is that there are natural kinds which underpin successful inductive inference.  Stathis Psillos has objected that this account is circular, since it employs ampliative inference to justify ampliative inference.  In this talk, I return to this topic, and present a response to Psillos:  inference to best explanation has nothing to do with it; it’s the world that makes induction reliable.