Abstract

Nobody wants unnatural kinds. Just as we prefer all natural ingredients in our food, so also we prefer natural kinds in our ontology and epistemology. Philosophers contrast natural with merely “conventional” kinds, and scientists advocate for natural rather than artificial classification systems. A central plank of the desired naturalness is “mind independence”—the property of existing independent of human interests and desires. Natural kinds are discovered, not made. They reflect the structure of the world (“nature’s joints”) and for this reason justify the practice of inductive inference. Conventional kinds, by contrast, are dependent on human classificatory activities. They are created with an end in view and therefore lack “a real existence in nature” (J. S. Mill, A System of Logic, Ratiocinative and Inductive [London: Longmans, 1843], 1:165). Since their existence is dependent on human activities, nominal kinds need not track nature’s joints. Because scientists are interested in groupings that really exist in nature—not those fabricated for human purposes—their classificatory practices aim to achieve natural-kind classifications. Achieving these classifications is crucial to the success of science.