This post is a sequel to Pavel Tichý on Existence. There I explained Tichý's theory as a variation on the Fregean theory and made a start on a critique of it. Here I examine an argument of his for it. He writes,
If existence were a property ascribable to individuals, then the force of such an ascription could only be to the effect that the individual in question is indeed one of the individuals there are. But since any individual is, trivially, one of the individuals there are, all ascriptions of existence would be tautologically true. If existence were properly raised in regard to individuals, then a negative answer to such a question would be self-defeating: it would suggest that no question has in fact been asked and that, accordingly, no answer is called for in the first place. Genuine existence questions would be answerable wholesale and a priori in the affirmative. ("Existence and God," Journal of Philosophy, August 1979, pp. 404-405)
This is an old argument that has been given too many times by too many philosophers, some of them distinguished, for example, by A. J. Ayer in Language, Truth, and Logic, p. 43. The argument can be set forth most perspicuously as follows:
1. If existence were a property of individuals, then affirmative singular existentials would be tautologies and negative singular existentials would be contradictions.
2. It is not the case that affirmative singular existentials are tautologies and negative singular existentials contradictions.
Therefore
3. Existence is not a property of individuals.
Although this oft-repeated argument is valid in point of logical form, it is a very bad argument. I of course concede the minor. The general affirmative existential 'Cats exist' is not a tautology, and neither is the singular affirmative existential 'Felix exists.' And the general negative existential "Three-headed dogs do not exist' is not a contradiction any more than the singular negative existential 'Cerberus does not exist.' The problem resides with the major premise.
What I will argue (and have argued in A Paradigm Theory of Existence, pp. 113-114) is that premise (1) rests on the ancient modal fallacy of confusing the necessity of the consequent (necessitas consequentiis) with the necessity of the consequence (necessitas consequentiae).
(1) derives whatever force it has from a paradoxical feature of reference. Given that there is no reference to the nonexistent, if I say of an existing thing that it exists, it seems I have said nothing, or at least nothing that can fail to be true; and if I say of an existing thing that it does not exist, then it seems I have said something that cannot fail to be false. Thus it seems that attributions of existence to individuals are tautologically and thus necessarily true, and that denials of existence to individuals are self-contradictory and thus necessarily false.
To see why (1) is false we must distinguish between sentences and the propositions they are used to express, and keep in mind the modal fallacy mentioned. One commits the fallacy if one makes the inferential move from 'Necessarily, if p then p' to 'If p, then necessarily p.' For example, it is necessarily true that if I am sitting, then I am sitting. But it does not follow that if I am sitting, then necessarily I am sitting — surely I might not have been sitting.
Sentences of the form a exists (where 'a' is a nonvacuous individual constant) have the peculiarity that they cannot be used to express a falsehood. Sentences of this form can thus be said, in a loose sense, to be necessarily true. But all this means, speaking strictly, is that necessarily, if a sentence S of this form is used, then the proposition expressed by S is true. This, however, does not entail that the proposition expressed by S is necessarily true. To think otherwise is to commit the above modal fallacy.
Although it is true that, necessarily, every use of a sentence of the form a exists expresses a true proposition, it does not follow that every use of a sentence of the form a exists expresses a necessarily true proposition. The necessity of the consequence does not entail the necessity of the consequent.
To put it another way, it is pragmatically impossible for one to use 'Socrates exists' or 'I exist' to express a false proposition; hence it is pragmatically necessary that whenever one uses such a sentence one expresses a true proposition. But the pragmatics of sentence utterance has no bearing on the modal status of propositions. The possibility of my nonexistence is not ruled out by the pragmatic impossibility of my using 'I do not exist' to express a true proposition.
So (1) above is false and Tichý's argument collapses. We have been given no good reason to think that existence is not a property — in a suitably broad sense — of individuals.