Natural Kinds and Scientific Classification

Core Idea

Are scientific categories like 'gold', 'acid', 'gene', or 'species' discovered in nature or imposed by human interests? Natural kind theory holds that some classifications carve nature at its joints: genuine categories reflecting real boundaries based on underlying essences or causal structures. Scientific progress often involves discovering what really distinguishes entities. The nature of natural kinds shapes our understanding of scientific taxonomy, reduction, and the objectivity of science.

Explainer

Your study of universals and particulars introduced the question of whether properties and categories exist independently of the minds that think about them. Natural kind theory extends this into philosophy of science: among all the ways we could classify the world, some correspond to genuine joints in nature and others are merely conventional. A natural kind is a category that picks out a real grouping — one that supports inductive generalization, figures in laws of nature, and reflects underlying causal structure rather than human convenience. The contrast is with nominal kinds: categories defined by stipulation or surface features that do not track anything deep in the world.

The classical example is gold. Before chemistry, gold was classified by observable properties: yellow, heavy, malleable, shiny. But these surface features are not what makes something gold; they are just reliable indicators. What really makes something gold is having atomic number 79. This is the microstructural essence, the hidden nature that Kripke and Putnam argued determines membership in the kind. Importantly, we might have been wrong about the surface features — fool's gold (iron pyrite) looks like gold — but we cannot be wrong about what gold *is* at the fundamental level. The kind is discovered, not defined. This connects to your study of essentialism: natural kinds have real essences, and scientific investigation reveals what those essences are.

Not every scientific category is so clean. Species is the famous trouble case. The biological species concept (organisms that can interbreed and produce fertile offspring) seems natural enough, but it does not apply to asexual organisms, hybridizing populations, or extinct species known only from fossils. Alternative concepts — phylogenetic, ecological, morphological — carve up the same organisms differently. Philosophers of biology debate whether species are natural kinds at all, or whether they are more like individual lineages — historical entities rather than types defined by intrinsic properties. If species are individuals rather than kinds, then generalizations like "robins lay blue eggs" are more like claims about a particular population than like laws instantiated by every member of a universal category.

The deepest question natural kind theory raises is whether the world has a unique correct classification at all, or whether multiple valid taxonomies can coexist for different scientific purposes. Promiscuous realism (Philip Kitcher) holds that there are many equally real ways to carve nature, depending on the explanatory and predictive tasks at hand. Genes might be natural kinds in molecular biology while the same DNA sequences are not natural kinds in ecology. This pluralism preserves the realist intuition that classification is constrained by the world while acknowledging that different sciences, with different explanatory goals, may legitimately use different categories. The debate matters enormously for reduction: if kinds in higher-level sciences (psychology, economics, ecology) do not map onto kinds in physics, the prospects for reduction are dim — and the autonomy of those sciences depends on their categories being naturalistically respectable in their own right.