122-dialogue on differing asian SOME COMMENTS ON A MODAL ONTOLOGICAL ARGUMENT Brian Garrett Australian National University, Australia An ontological argument for God’s existence is any argument which attempts to prove the existence of God by reason alone. Its only resources are concepts (especially the concept of God) and logical inferences involving those concepts. In contrast, other kinds of argument for God’s existence, such as the Cosmological or Telelogical arguments, rest on contingently true premises concerning, e.g., the existence of the physical universe or the complexity of organisms within it. These arguments do not purport to prove the existence of God by reason alone. One version of the ontological argument is the modal ontological argument. It is called ‘modal’ because it makes essential use of modal notions such as possibility and necessity. These notions are anyway integral to understanding God: it is traditionally held that God is a being who exists necessarily, and not merely contingently. (In possible worlds talk, a necessary being is one who exists in every possible world; a contingent being is one who exists in some but not all possible worlds.) Here is the simplest version of a modal ontological argument, with one premise and one conclusion: (1) It is possible that a necessary being exist1; so (2) A necessary being exists. The argument is valid (at least in standard modal logics such as S5): if there is a possible world containing a necessary being (i.e., a being who exists in every possible world), such a being exists in this world, and thus actually exists. Hence, if the argument is to be faulted, the fault must lie with premise (1). Note that this is an ontological argument: a necessary being is claimed to exist because, and only because, it is possible that a necessary 22 Prajñâ Vihâra, Volume 5, Number 2, July-December, 2004, 22-26 © 2000 by Assumption University Press being exist. Premise (1) is not an empirical, contingent premise: if true, it is conceptually or necessarily true. I There have been three main types of objection to this argument: (a) The idea of a ‘necessary being’ is incoherent; (b) We have no reason to accept (1) in preference to its opposite; (c) The argument is ripe for parody, so cannot be sound. None of these objections strike me as compelling. In which case, unless there are other objections, we should accept our little argument. II (a) According to the first line of objection, ‘necessary’ can be predicated only of sentences or propositions, not of objects. We cannot meaningfully talk of necessary beings, or of contingent ones. Russell advocated this view.2 However, developments in logic and semantics since Russell have allowed us to make good sense of de re modal claims (that is, claims of the form ‘x is necessarily F’ or ‘x is possibly F’).3 And we seem to have little difficulty understanding the claim that numbers necessarily exist. If these remarks fail to convince, it’s worth pointing out that many philosophers are willing to acknowledge the coherence of talk of necessary objects, yet refuse to accept any ontological argument for God’s existence. Such philosophers can be considered the target audience of this note. III (b) Surely, it might be said, whatever modal intuitions support (1) equally support: (1-) It is possible that a necessary being not exist; from which we could conclude: (2-) A necessary being does not exist. Since (2) contradicts (2-), and since (1-) is no less plausible than (1), we should have no confidence in our original argument. Brian Garrett 23 Comment It is perhaps not obvious how we are to understand (1-). If it is to imply (2-), it must be read as: (1-*) It is possible that no necessary beings exist4. Why prefer (1) to (1-*)? Premise (1) is supposed to be a conceptual truth, true simply in virtue of the generosity and abundance of the possible. What kind of general principle would generate (1)? A natural suggestion might be: (A) For any coherent or non-contradictory concept F, it is possible that there are Fs (i.e., there is a possible world containing Fs). From (A), assuming that the concept of a necessary being is coherent, (1) follows. I do not say that (A) is inescapable, but it is plausible. What analogous general principle would deliver (1-*)? Presumably: (B) For any coherent or non-contradictory concept F, it is possible that there are no Fs (i.e., there is a possible world containing no Fs). Yet (B) is not a plausible principle: the concept the even prime number is coherent, yet, assuming that numbers exist necessarily, there is no possible world which fails to contain that number (which is the number 2). Hence there is an asymmetry which favours our premise: it’s not true that that we have just as much reason to believe (1-*) as to believe (1).5 24 Prajñâ Vihâra IV (c) Ever since Gaunilo objected to Anselm, ontological arguments have been subjected to parody ripostes. These are arguments of the same form as ontological arguments, yet with obviously absurd conclusions. We are invited to draw the conclusion that the ontological arguments they parody cannot be cogent either. Thus: (3) It is possible that a necessary island exist; so (4) A necessary island exists. (4) is absurd: our world plainly does not contain an island which exists necessarily. All earthly islands are contingent beings. Since this argument exactly mimics our ontological argument, that argument must be hopeless. Comments (i) We can observe that this objection is hardly complete: at best it tells us that our argument is wrong; it does not tell us where it goes wrong. But there is a more fundamental point. (ii) It is not just that earthly islands are contingent beings. Islands are not the kind of being that could be necessary. They have a beginning and an end; they are part of the causal swim; they can be destroyed and created. In short, they depend for their existence on contingencies. Hence we can, quite justifiably, deny premise (3). The concept of a necessary island is not coherent, and hence principle (A) cannot be used to support (3)6 V I conclude that our ontological argument has not been refuted: therefore, we may reasonably believe that a necessary being exists simply because it is possible for such a being to exist. Brian Garrett 25 ENDNOTES 1 In possible worlds talk: There is a possible world containing a necessary being. 2 For example, in a debate with F. Copleston. (See ‘Must God Exist?’ in Philosophy in the Open (ed.) G. Vesey (Open University Press, 1978), pp. 114 – 121.) Russell objected to talk of necessary beings on the grounds that all necessity is analytic and that existence is not a predicate. However, the first claim is false (many now accept non-analytic necessities such as ‘water is H 2 O’ and ‘Tully is Cicero’), and the second irrelevant (our argument makes no assumption about the logical form of ‘x exists necessarily’.) 3 See, e.g., Saul Kripke, Naming and Necessity (Blackwell, Oxford, 1980). 4 That is: there is a possible world containing no necessary beings. 5 This point seems to have escaped some commentators. See, e.g., J. L. Mackie, The Miracle of Theism (OUP, 1981) pp. 55 – 64. 6 This shows that it is not always an easy matter to determine whether a concept is coherent. It takes varying degrees of reflection to realise that (e.g.,) married bachelor, necessary island, and largest prime number, are incoherent concepts. 26 Prajñâ Vihâra