Sacrificing Natural Kinds

The article reevaluates Jerry Fodor’s key argument for the autonomy of the special sciences, which rests on the notion of multiple realization and the claim that special science predicates must be natural kinds. After outlining how Fodor’s view was shaped by the “syntactic” conception of scientific theories, it shows that the recent “semantic” approach in scientific theories challenges the idea that special science kinds must be natural and ontologically committing. On the semantic account, scientific models often invoke idealized or domain-specific predicates that do not have to be natural. We use Fodor’s example – Gresham’s law – to articulate a semantic perspective that preserves the unity of science: higher-level explanations can remain useful, real, and relatively autonomous without irreducible natural kinds. By “sacrificing” natural kinds, we retain the explanatory powers of the special sciences, create a simpler ontological picture of the world, and justify the modus operandi of the sciences, such as cognitive science, where knowledge from the different levels or disciplines that constitute it informs and refines our overall understanding of the world.