Locating Belief on the Spectrum of Existence

Abstract

DOI: https://doi.org/10.26333/sts.xxxix1.02

Krzysztof Posłajko’s insightful and incisive book, Unreal Beliefs, successfully reignites the dormant debate about belief eliminativism. His key insight, inspired by Lewis’s (1983) distinction between natural and non-natural properties, is that a category can exist without being real: “completely arbitrary collections of modal objects” (Posłaj­ko, 2024, p. 68) exist, but since they are not natural kinds, naturalists need not accept that they are real. This opens up the possibility that beliefs exist in a minimal sense while not being real, since they may fail to be natural kinds, thus forestalling the counter-intuitive and paradoxical implications of belief eliminativism, in favor of Posłajko’s favored “minimal non-realism”. However, as Posłajko realizes, the distinction between natural and non-natural properties is “graded” (2024, p. 68), e.g., categories of the special sciences, like money, are not as natural as the categories of our best physics, but they are not completely arbitrary collections of modal objects either. This paper concerns where on this ontological spectrum to locate the category of belief, a question about which Posłajko is not fully clear. Although he argues successfully that there is little evidence that belief constitutes a natural, neurocomputational kind, I argue that there are at least two other ways belief might qualify for natural kindhood: as a property of Dennettian intentional systems, inevitably produced by natural selection, or as a socially maintained deontic kind, i.e., Brandomian discursive commitment, inevitable in populations using natural language to coordinate. Posłajko’s arguments for minimal non-realism about belief fail to rule out these versions of belief realism.