ABSTRACT
These days we speak freely of ‘natural kind terms’, indicating that they constitute a special, semantic category of terms. It was not always so. Indeed, prior to Putnam and Kripke’s writings from the 1970s, the label ‘natural kind term’ seems not to have been employed at all. In his well-known paper ‘The Analytic and the Synthetic’ (1962), Putnam does set out to draw some distinctions among the general terms that he takes to be of semantic signifi - cance. In particular he wishes to distinguish so-called one-criterion terms, such as ‘bachelor’, from cluster terms, such as ‘man’ or ‘crow’, where the meaning is given by a cluster of associated properties, none of which are immune from revision. Among the cluster terms, Putnam also suggests, there is a set of terms of special interest to science, the ‘law-cluster terms’, such as ‘energy’. These are set apart by what goes into the cluster, in particular laws and general principles. In his paper ‘Is Semantics Possible?’ (1970), however, natural kind terms appear on the scene. As in the earlier paper, Putnam draws a distinction among the general terms between onecriterion terms and others, but he now drops the talk of law-cluster terms in favour of that of natural kind terms:
A natural kind term . . . is a term that plays a special kind of role. If I describe something as a lemon, or as an acid, I indicate that it is likely to share certain characteristics (yellow peel, or sour taste in dilute water solution, as the case may be); but I also indicate that the presence of those characteristics, if they are present, is likely to be accounted for by some ‘essential nature’ which the thing shares with other members of the kind. What the essential nature is is not a matter of language analysis but of scientifi c construction. (1970: 140)
Putnam’s formulation captures the central elements of the contemporary notion of a natural kind term: the idea that these terms pick out not superfi cial, macro-level properties, but underlying, essential properties the nature of which it is up to science to establish. The notion took on quickly,
in particular after the appearance of Kripke’s Naming and Necessity (1980). By suggesting that natural kind terms, like names, are rigid designators, Kripke seemed to provide the semantic tools for separating out the natural kind terms from other kind terms. The suggested candidates included a rather diverse set of terms: mass terms (‘water’, ‘gold’), count nouns (‘tiger’, ‘whale’) as well as adjectives (‘hot’, ‘loud’). What sets these terms apart, Kripke suggested, is their ‘name-like’ semantic behaviour. In addition, Putnam presented his Twin Earth thought experiment which was also taken to indicate that natural kind terms are set apart: the meanings of these terms are not ‘in the head’, but have to be given an externalist account (1975a). Thus, the idea emerged that natural kind terms are semantically special.1
