Published online by Cambridge University Press: 25 February 2009
The following objection to the ‘ontological’ argument of St Anselm (Proslogion II) has a continuing importance. The argument begs the question by introducing into the first premise the name ‘God’. In order for something to be truly talked about, to have properties truly attributed to it—it has been said—it must exist; a statement containing a vacuous name must either be false, meaningless, or lacking in truth-value, if it is not a misleading formulation to be explained by paraphrase into other terms. In any case the question of the divine existence is begged.
1 A brief statement of the argument follows:
(1) God is a being (or the being) greater than which none can be conceived.
(2) If God did not exist, then there could be conceived a being greater than God, i.e. a being just like God in every respect, only existing—which is absurd.
(3) Therefore, God exists.
2 It is a crucial contributor to the murky doctrine that existence is not a predicate (‘exists’ is not a property). Cf. Gassendi's formulation: ‘Existence a perfection neither in God nor in anything else; it is rather that in the absence of which there is no perfection’ (Plantinga, Alvin (ed.), The Ontological Argument (Garden City, 1965), p. 46).Google Scholar
3 The simplest such paraphrase supplies what is thought to be the implicit conditional ‘if it exists’. Thus ‘perpetual motion machines never stop’ can be read as ‘If perpetual motion machines exist, they never stop’. Such a reading of the first premise of Anselm's argument would make the conclusion read ‘If God exists, He exists’.
4 The difference between these formulations is not of importance in this context, since the ‘greater than’ clause requires that there be at most one greatest conceivable being.
5 Hume, David, A Treatise of Human Nature (ed.) Selby-Bigge, L. A. (Oxford, 1896), p. 67.Google Scholar
6 Nakhnikian, G. and Salmon, W. C., ‘“Exists” as a Predicate’, Philosophical Review, LXVI, 1957, p. 541.Google Scholar
7 I take it that ‘Pegasus does not exist’ does not mean ‘If Pegasus exists, he does not exist’.
8 Cf. e.g., Ryle, Gilbert, ‘Systematically Misleading Expressions’, in Rorty, Richard (ed.), The Linguistic Turn (Chicago, 1965), Ch. 5.Google Scholar
9 For abundant examples of such discourse, cf. Bradley, A. C., Shakespearean Tragedy (London, 1929).Google Scholar
10 Linsky, Leonard, ‘Reference and Referents’, in Caton, Charles E. (ed.) Philosophy and Ordinary Language (Urbana, Illinois, 1963), p. 85.Google Scholar
11 ‘Gertrude was Hamlet's mother’ clearly does not mean ‘If Gertrude existed, she was (would be) Hamlet's mother’. Read as a material conditional, this is vacuous; read as a strong (counterfactual) conditional, it is speculative, except in so far as it is based on the fact that Gertrude (in the play) is Hamlet's mother.
12 My inclination is to think that even in a historical fiction, which seems to be about real persons, everything is simply fictional. Shakespeare's Richard III is as much a fictional character as Gandalf, and the England of Shakespeare's history plays as much a fictional land as Middle Earth. (This does not prevent a story-teller from incorporating all sorts of known facts about a place or a person by referring to a real place or person. Still, all these facts become fictional, as is shown by the power of the writer to deviate from his real model at will.) In the case of a play like The Deputy, which tells a story while at the same time advancing a historical thesis, essentially two different kinds of discourse are going on at once, which may reinforce or (more likely) interfere with one another.
13 Ryle, Gilbert, ‘Imaginary Objects’ Proceedings of the Aristotelian Society, Supp. XII, 1933, esp. pp. 27–28.Google Scholar
14 There is actually a range of cases here, illustrating the problem of establishing identities here. A character who appears in one work as manifesting only—though not necessarily all—the characteristics manifested in another—would provide the strongest case for identity. A not-so-clear case is that of Rosencrantz and Guildenstern of Rosencrantz and Guildenstern are Dead, and of characters in other such elaborations of portions of works, but the case for identity is still quite strong. The case of Sherlock Holmes, even in Sherlock Holmes stories written by others than Sir Arthur Conan Doyle, is also inviting, despite the contradictions to be found in the Sherlock Holmes canon. Faust, who in Goethe is called Heinrich and is saved, and in Marlowe is called Johann and is damned, is an intermediate case: we cannot make the Fausts cohere into a single figure, yet an important point is served by speaking of the same character. Travesties, in which a character has properties wildly different from those which he has in the model, and stories in which only the same name is used, lie at the other end.
15 In the case of a play, the actors take upon themselves the persons of the fictional characters. The names of the characters refer, not (except by a figure of speech) to the actors, but to the persons they bear. In the case of a mime-game, where actors make up the story as they go along, the actors are also the authors, but the situation is otherwise unchanged.
16 For detailed discussion illustrating such vagaries, cf. e.g. Bradley, op. cit., Note I, pp. 423ff. (contradictory time indications in Othello), and Note DD, pp. 484ff. (‘Did Lady Macbeth really faint?’).
17 Frege, G., Foundations of Arithmetic, with a tr. by Austin, J. L., 2nd ed. (Evanston, Illinois, 1968), p. 65.Google Scholar
18 Meinong, Alexius, ‘The Theory of Objects’, tr. Levi, Isaac, Terrell, D. B. and Chisholm, Roderick M., in Chisholm, Roderick M. (ed.) Realism and the Background of Phenomenology (Glencoe, Illinois, 1960), p. 76.Google Scholar
19 For more on Meinong, cf. Chisholm, Roderick M., ‘Editor's Introduction’Google Scholar, §II, op. cit., and Findlay, J. N., Meitwng's Theory of Objects and Values, 2nd ed. (Oxford, 1963).Google Scholar For the relation of his theory to the ontological argument, cf. Kenny, Anthony et al. , ‘Symposium: Descartes' Ontological Argument’, in Margolis, Joseph (ed.) Fact and Existence (Oxford, 1969).Google Scholar
20 Cf. Hodges, Michael, ‘On “Being About”’, Mind, n.s., LXXX, 1971, esp. pp. 5–6, 15–16.Google Scholar
21 This article is drawn from my doctoral dissertation, ‘The Very Idea of an Ontological Argument’. I am indebted to Professor Wallace I. Matson, Mr Hans Sluga, and Mr Ernest Sanchez for stimulating criticisms.