Natural kind terms: are they rigid designators in the same sense as proper names, and do theoretical identifications come out necessary?

Saul Kripke, answered from the texts and cited to the page.

The position in brief

Natural kind terms — "water," "gold," "tiger," "heat" — are rigid designators of the kinds they refer to, much as proper names are rigid designators of individuals.¹ The reference is fixed at the actual world by samples (this stuff is water; that stuff is gold) plus the implicit methodological commitment that what matters for being of the kind is its underlying nature, discoverable empirically.

Theoretical identifications between rigidly designating natural-kind terms are thus necessary if true, though typically discovered a posteriori: "water is H2O," "gold is the element with atomic number 79," "heat is molecular motion." A possible world in which something looks like water, behaves like water, and yet is not H2O is a world in which something resembles water but is not water; the same kind is wherever H2O is, regardless of superficial properties.²³⁴

Sources

1. More recent philosophy has followed suit, except that, in the case of both proper names and natural kind terms, it often replaces the notion of defining properties by that of a cluster of properties, only some of which need to be satisfied in each particular case. My own view, on the other hand, regards Mill as more-or-less right about 'singular' names, but wrong about 'general' names.

   Naming and Necessity, p. 136

2. Theoretical identities, according to the conception I advocate, are generally identities involving two rigid designators and therefore are examples of the necessary a posteriori. Now in spite of the arguments I gave before for the distinction between necessary and a priori truth, the notion of a posteriori necessary truth may still be somewhat puzzling.

   Naming and Necessity, p. 148

3. Tigers are a natural kind, and to be a tiger one must be of that natural kind. So it is not the case that any animal looking like this need be a tiger by defi nition. Of course, we may hope or believe that no such perfect ringers exist. Th e less they do, the easier it is to identify species by their appearances.

   Reference and Existence, p. 45

4. I recall reading a couple of references that att ribute to Putnam the idea that only impurities in the gold around us produce the apparent yellow color, and that chemically pure gold would not be yellow. Epistemically, this is certainly a possibility, and an elegant way of bringing out the point that 'gold is yellow' is not a priori certain.

   Reference and Existence, p. 46

Where the argument is made

• Naming and Necessity (1980), Lecture III
• "Identity and Necessity" (1971, Philosophical Troubles Ch 1)

How the argument was received

The central insight — that theoretical identifications between natural-kind terms are necessary a posteriori, and that the natural sciences uncover necessary truths about the world — survives in the post-Kripke and post-Putnam literature as a near-consensus. The technical claim that natural-kind terms are rigid in the same sense as singular terms is more contested. Soames argues at length in Beyond Rigidity Ch 9 that there is no natural concept of rigidity for predicates that does the work N&N attributes to it — that all natural-kind predicates come out rigid, while familiar descriptive predicates like is a bachelor do not, and that theoretical identifications involving rigid predicates must be necessary. Soames offers a more limited vindication; I find the critique persuasive on the formulation while continuing to hold the underlying necessary-a-posteriori claim.