Natural Kind

Concept and Historical Background

A natural kind is commonly understood as a category or grouping that corresponds to real, objective divisions in nature, rather than to merely conventional or arbitrary classifications. When philosophers say a category is a natural kind, they typically mean that it “carves nature at its joints”: it tracks genuine similarities and structures in the world that are independent of our linguistic or conceptual schemes.

The idea has deep roots in ancient philosophy. In Plato, the metaphor of carving nature at its joints suggests that correct classification aligns with the metaphysical structure given by the Forms. Kinds are not just convenient groupings; they reflect an underlying order. Aristotle develops a more biological and taxonomic approach. His scheme of genera and species, and his talk of essences, provides an early model for thinking about natural kinds as groups characterized by essential features that determine their behavior and capacities.

In early modern philosophy, the notion becomes tied to debates about essences and scientific classification. John Locke distinguishes between real essences (the underlying constitution of a thing) and nominal essences (the set of observable qualities we use to define a term). Locke is skeptical that our classifications map neatly onto real essences, suggesting that many species terms are pragmatic groupings rather than strictly natural kinds. David Hume, emphasizing habit and resemblance, is often read as weakening the metaphysical weight of natural kinds, portraying grouping as based on psychological association rather than deep ontological joints.

Throughout the 19th and early 20th centuries, with the rise of modern science, talk of natural kinds became closely associated with scientific taxonomy, especially in chemistry and biology. The idea that elements, species, or chemical compounds form objectively privileged groupings inspired a renewed interest in what makes some categories “natural” and others merely conventional.

Natural Kinds in Modern Philosophy of Science

In the 20th century, the term “natural kind” became central in analytic metaphysics and the philosophy of science. Several interrelated themes emerged:

1. Natural kinds and laws of nature.
For logical empiricists and later philosophers such as W. V. O. Quine, natural kinds were invoked to explain how induction and scientific laws are possible. Kinds are those categories that appear in lawlike generalizations and support projectible predicates—predicates that can be reliably extended from observed to unobserved cases (e.g., “All copper conducts electricity”).

2. Essentialism and reference.
A major revival of essentialist thinking about natural kinds came with Saul Kripke and Hilary Putnam in the 1970s. They argued that terms like “water,” “gold,” and “tiger” are rigid designators: once their reference is fixed (often by paradigm samples), they designate the same kind in all possible worlds. The essence of a natural kind is typically understood in microstructural terms (e.g., H₂O for water). Putnam’s famous Twin Earth thought experiment suggests that surface similarities are insufficient: what makes something water is its underlying molecular structure, not just how it looks or behaves.

On this view, natural kinds are discovered rather than stipulated, and scientific investigation can reveal their essential natures. Scientific progress, especially in chemistry and molecular biology, appears to provide paradigms of such discoveries.

3. Property cluster and homeostatic accounts.
Critics of strict essentialism have developed more flexible accounts. Homeostatic Property Cluster (HPC) theories hold that a natural kind is a cluster of properties maintained by homeostatic mechanisms (e.g., biological reproduction, feedback systems, social practices). Instead of a single necessary and sufficient essence, there is a characteristic pattern of traits that tends to co-occur because of underlying stabilizing processes. Biological species and many kinds in the special sciences are often cited as fitting this model better than rigid essence-based accounts.

These approaches retain the link between natural kinds and explanatory success—they are the categories that support robust explanation, prediction, and induction—while allowing for variation, borderline cases, and historical change.

Contemporary Debates and Challenges

Current discussions of natural kinds are diverse and often interdisciplinary, involving metaphysics, philosophy of science, social philosophy, and scientific practice.

1. Realism vs. conventionalism and constructivism.
A central debate pits realists about natural kinds against more conventionalist or constructivist views.

• Realists claim that some categories genuinely reflect the structure of the world, and that this fact explains their scientific utility. On strong realist accounts, the world comes pre-packaged into kinds, which our best theories gradually discover.
• Conventionalists and constructivists argue that classification is always interest-relative: what counts as a kind depends on our goals (prediction, control, explanation, moral evaluation). They may still grant that the world contains patterns, but deny that it dictates a unique or privileged way of carving it.

Some philosophers adopt pluralist positions, suggesting that there may be multiple, equally legitimate systems of natural kinds, tailored to different explanatory or practical purposes.

2. Social and human kinds.
An important strand of recent work concerns whether there can be natural kinds in the social and human domains, such as gender, race, mental disorder, or social roles.

• Some argue that such categories are heavily shaped by social practices, norms, and power relations, making them paradigms of social kinds rather than natural kinds.
• Others propose that social kinds can still be “natural” in a broader sense if they track stable, causally significant patterns—sometimes modeled using HPC or similar cluster theories.
• A further complication is “interactive kinds” (Hacking): classifications that change the very people classified, creating feedback loops between labels and behavior. This dynamic feature challenges static pictures of natural kinds.

3. Cross-scientific and multi-level kinds.
Another debate addresses whether natural kinds must be microstructural, as suggested by chemical and physical paradigms, or whether kinds operating at higher levels—in biology, psychology, or ecology—can be equally real and explanatory. Discussions of multiple realizability and emergence feed into questions about whether a single microphysical structure is necessary for kind membership, or whether diverse realizations (e.g., of mental states) can belong to the same natural kind.

4. Induction, explanation, and epistemology.
Natural kinds are often linked to the epistemology of science: they are supposed to underwrite reliable inference. Proponents contend that the success of scientific practice—its ability to generate stable laws and predictions—supports some form of natural kind realism. Critics reply that such success can be explained pragmatically, via modeling strategies, idealization, and the flexibility of scientific language, without positing robust metaphysical natural kinds.

Across these debates, the term “natural kind” continues to function as a focal point for questions about how language, thought, and scientific practice relate to the structure of the world, and whether that structure is best understood as pre-given, constructed, or something in between.