Abstract

Contemporary logicians sometimes discuss questions like ‘What criterion is there for deciding whether a portion of language or a theory is committed to the existence of anything?’, ‘Does mathematics require us to countenance or tolerate abstract entities?’. In the course of these discussions ancient problems about universals reappear in a new dress. Quine, for one, would agree that it is not by our use of general terms that we commit ourselves to the existence of anything, for general terms like ‘red’ and ‘five’ are not names. Nor does the use of names commit one to corresponding entities, for we sometimes use a name non–designatively. “Names,” as Quine says, “are in fact altogether immaterial to the onto–logical issue.” Quine argues that there is only one way in which we can find out what entities a theory commits us to. Given that the theory is expressed in the symbolism of quantification theory, we discover what entities the theory is committed to by examining what entities have to be values of our bound variables if the statements made in the theory are to be true. “To be is purely and simply to be the value of a variable”. I want to argue in this paper that Quine’s criterion of ontological commitment cannot distinguish genuine from bogus ontological commitment and is therefore inadequate.