Oppy and Modal Theistic Proofs

I argue that Graham Oppy’s attempt to redefend his charge that all modal theistic arguments “must be question-begging” is unsuccessful. Oppy’s attempt to show that theism and modal concretism are compatible is not only tangential for his purposes, it is marred by a misunderstanding of theism, and vulnerable to a counterexample that actually demonstrates incompatibility. Moreover, the notion of begging the question employed by Oppy against the theist is seen to be far too permissive. 1. See Graham Oppy, “Modal Theistic Arguments,” Sophia 32 ( 993): 8, 20. 2. See Graham Oppy, “Reply to Richard Davis,” Philosophia Christi (2009): 423–35. 3. Ibid., 423. 4. See my “God and Modal Concretism,” Philosophia Christi 0 (2008): 57–74. 5. Oppy, “Modal Theistic Arguments,” 20 (emphasis added). 438 PhilosoPhia Christi are just two different accounts of the nature of logical space between which we must choose—the account offered by Lewis, and the account offered by the theist.”6 This is perhaps an oversimplification; still, as Oppy notes, for the purposes of illustrating divergent conceptions of logical space, it seemed to me that there was an advantage in working with Lewis’s modal realism simply because of the vivid picture that it presents of the metaphysics of modality.7 Agreed. Now if T and C are incompatible—that is, if ~(T&C)—it follows that C⇒∼T. Relying upon C in a theistic proof is therefore ill-advised, since it entails atheism. But of course ~(T&C) also entails T⇒~C,8 which raises a different problem. Would it not then be a case of circular reasoning to assume ~C in an MTP? After all, ~C is an entailment of theism itself. The question is whether this is objectionable. And if so, how so? To complicate matters, I noted that there are different accounts of “circularity” on offer; indeed, there is serious and heated debate over which of these best explains why circular arguments should be classified as defective or fallacious.9 Unfortunately, while Oppy’s 993 paper includes the accusation of circularity, it fails to include any account of what he has in mind. I therefore explored various possibilities on his behalf, concluding that the problem could lie neither with logical or epistemic circularity; nor could it be due to Oppy Circularity (OC)—a malady of his own invention. I ended by briefly sketching an MTP immune to all of Oppy’s criticisms. 0 The crucial thing to see is that if theism and concretism are incompatible, that only makes the alleged circularity problem more acute. It is therefore something of a surprise to see Oppy disputing this incompatibility vigorously and at such length. If he is right, my job becomes all the easier (and his, I am afraid, all the harder). But is Oppy right? Well, I do not think so. Oppy’s specific complaints focus on the arguments I gave to show that there is no way for God to exist in a Lewis world. There is no dispute here over this or that premise as one would expect. Rather, what Oppy wants to say is that while theism is inconsistent with Lewisean concretism (C), it is perfectly compatible with concretism more broadly construed (call this “C*”): 6. Oppy, “Modal Theistic Arguments,” 20. . Oppy, “Reply to Richard Davis,” 42 . 8. I use the double-line arrow for entailment (i.e., for logically necessary material implication). . See Douglas Walton, “Epistemic and Dialectical Models of Begging the Question,” Synthese 52 (2006): 237–84. 0. Oppy’s only complaint with this proof is that it is not a modal ontological argument; for it does not include two key premises: ( ) that if a being S is possible, then S is necessary, and (2) that S is possible. This is disputable, but I do not have the space here to adequately deal with this objection. But see my more detailed presentation in “The Conceptualist Argument: A New Defense,” unpublished. See also Quentin Smith, “The Conceptualist Argument for God’s Existence,” Faith and Philosophy ( 994): 38–49. If we combine theism with concretism, then it isn’t true that worlds are maximal physical objects: rather, worlds are mereological sums, of which at most one part is a maximal physical object. Theists who agree with Davis that God might not have created will suppose that there is at least one possible world in which there is no physical part; theists who think that God could not have failed to create will suppose that there are no possible worlds that have no physical part. He then concludes that “Davis’s arguments for inconsistency between theism and concretism fail.” 2 But this is a whopping non sequitur. What follows, at best, is only that my arguments fail to show that T and C* are incompatible. But since that was no part of my project, it is hardly an objection. 3 And this is all beside the point anyway; none of this even slightly advances Oppy’s central claim that a theist’s assumptions about modality “can only be supported” by presupposing theism. (You will notice that in his present contribution, this claim is dropped entirely.) At any rate, how does merely announcing that worlds are mereological sums demonstrate or prove that T and C* are compatible? Well, it does not by itself. So Oppy adds that God’s existing necessarily can be understood in terms of his having a counterpart in each of these mereological sums. To say that God necessarily has his omni-properties means that “in each one of the worlds, the counterpart of God is omnipotent, omniscient, omnigood, eternal, and the creator of the world ex nihilo.” 4 Here we can safely ignore this idea that God necessarily creates, and that each world therefore has a maximal physical object (MPO) as a part. For this implies that God creates by the necessity of his nature, that creation is not the free act of a personal cause or agent, and that matter is logically necessary. Only Spinoza and his followers would dream of saying such things. 5 This is no tenet of classical theism. What about this idea that God has counterparts? What is that supposed to mean? What is a counterpart? The grandfather of counterpart theory, David Lewis, explains: I prefer to say that you are in the actual world and no other, but you have counterparts in several other worlds. Your counterparts resemble you closely in content and context in important respects. . . . But they are not really you. For each of them is in his own world, and only you are here in the actual world. 6 11. Oppy, “Reply to Richard Davis,” 431. 2. Ibid. 13. Oppy’s “Modal Theistic Arguments” knows nothing of worlds as mereological sums, containing at most one maximal physical object as a part. 14. Oppy, “Modal Theistic Arguments,” 18–1 . 5. Thus saith Samuel Clarke. See his A Demonstration of the Being and Attributes of God and Other Writings, ed. Ezio Vailati (Cambridge: Cambridge University Press, 998), 46. 16. David Lewis, “Counterpart Theory and Quantified Modal Logic,” Journal of Philosophy 65 ( 968): 4. riChard davis 439 440 PhilosoPhia Christi Putting this together with Oppy’s suggestion, what we must envision, I suppose, are distinct worlds each containing beings who possess the omni-properties, but who are not identical with each other or the God of the actual world. No doubt many of these worlds will have MPOs, and of course these too will be distinct. But at least one world will contain only a God counterpart. For the sake of simplicity, then, let us imagine there are only three “mereological worlds”: (W ) God + MPO (W2) God2 + MPO2 (W3) God3 Notice that there is a God counterpart in each world; we have also got some worlds with MPO parts and others without. Doesn’t this automatically show that T and C* are compatible? By no means. First, if God ≠ God2 ≠ God3, then what we have is not “necessary being” theism, but rather a bizarre form of polytheism: multiple distinct omni-beings each existing in its own world. Second, Lewis, Oppy, and the “necessary being” theist all hold that the nature and number of possible worlds is a matter of necessity. According to Lewis, for example, “it is not contingent what conditions the entire system of worlds does or does not satisfy.” 7 Oppy concurs: “the nature of logical space cannot be a contingent matter.” 8 But if so, every possible world will exist as a matter of necessity, in which case (if Lewis is right) each of W , W2, and W3 will have a counterpart in the others. But this is not possible. For one thing, W and W2 cannot have counterparts in W3 because W3 is an MPO-less world; as such, it cannot possibly contain (as a mereological part) an MPO resembling those in W and W2. Furthermore, suppose MPO is as big as our cosmos, while MPO2 is quite small—perhaps only the size of a flea. 9 Then it simply will not be possible for MPO2 to contain anything that closely resembles MPO , from which it follows that W will not have a counterpart in W2. Anyway you look at it, Oppy’s attempt to graft God into mereological worlds yields a contingent logical space; and this is not compatible with the modal framework required by “necessary being” theism. So it seems to me that Oppy’s “add on” to Lewis’s concretism is in serious trouble. Not only has Oppy failed to show how T and C* are compatible, we can (I believe) see quite clearly that they are not. 7. David Lewis, On the Plurality of Worlds (Oxford: Basil Blackwell, 986), 25. 18. Oppy, “Modal Theistic Arguments,” 23. 1 . Plantinga points out that on Lewis’s definition of “maximal physical object,” some Lewis worlds will be “no larger than a flea” (Alvin Plantinga, Essays in the Metaphysics of Modality, ed. Matthew Davidson (Oxford: Oxford University Press, 2003), 205).