Hyperintensionalist Metaphysics as a Case Study in Overfitting

Abstract One main trend in metaphysics over the past thirty years has been hyperintensionalism, a view on which many of the most significant metaphysical distinctions distinguish between entities (such as propositions, states of affairs, properties, and relations) that are necessarily equivalent but not identical to each other. A paradigm of this approach is Kit Fine’s seminal work on essentialist distinctions, arguing that they cannot be reduced to modal terms. Various forms of hyperintensionalist semantics have been proposed for such constructions, involving impossible worlds, truthmakers, and Russellian propositions. The key evidence for hyperintensionalism is supposed to come from decisive counterexamples to the opposed view, intensionalism. The chapter argues that the alleged counterexamples are unreliable, since they are explicable as based on heuristics in the psychologists’ sense, efficient ways of answering questions quickly and easily that are less than fully reliable. The relevant heuristics assess statements about ‘essence’, ‘because’, and the like in terms of answers to ‘Why?’ questions that evaluate their truth in terms of the goodness of the corresponding explanations. The results are sensitive to differences in manner of presentation and the like, and so easily generate illusions of hyperintensionality. Such methods generate provably false results in some cases and so are merely heuristic. Thus hyperintensional metaphysics is an artefact of overfitting, a well-known pathology in the natural and social sciences by which theories become ever more complicated as they are fitted to error-prone data.