A main point of my article, as I see it, is that we can solve Putnam's problem, as articulated in the first paragraph of section three, without recourse to the definition of “natural-kind term” as “rigid designator of a natural kind”. I had three main objections to this definition:

(a) It makes the classification of a term as a natural-kind term dependent on one's metaphysics, i.e., on the status given to natural kinds. However, Putnam's argument seems to be independent of such metaphysical considerations, and the sort of natural kinds it establishes (if any) should be “read off its face”, not set down in advance (section 2).

(b) It permits the derivation of “exotic necessary truths” such as “If water is H20 then necessarily water is H20” (sections 9 and 10).

(c) Putnam's main point appears to be about the independence of a term's extension from a linguistic community's beliefs. Why should this point affect the theory of designation? Kripke's argument about names establishes the non-descriptiveness of names while leaving undisturbed the classical conception of names designating individuals. Why should we not take his and Putnam's parallel arguments as establishing the non-descriptiveness (non-connotation) of natural-kind terms while leaving undisturbed the classical conception of general terms designating classes (section 10)?