Abstract
God is traditionally taken to be a perfect being, and the creator and sustainer of all that is. So, if theism is true, what sort of world should we expect? To answer this question, we need an account of the array of possible worlds from which God is said to choose. It seems that either there is (a) exactly one best possible world; or (b) more than one unsurpassable world; or (c) an infinite hierarchy of increasingly better worlds. Influential arguments for atheism have been advanced on each hierarchy, and these jointly comprise a daunting trilemma for theism. In this paper, I argue that if theism is true, we should expect the actual world to be a multiverse comprised of all and only those universes which are worthy of creation and sustenance. I further argue that this multiverse is the unique best of all possible worlds. Finally, I explain how his unconventional view bears on the trilemma for theism.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Candidate WGMPs pick out a property held to be good-making. Traditional examples include: the presence of free moral agents in the world; the favourable balance of moral actions over immoral ones; the variety of phenomena in the world; and the simplicity of a world’s governing laws.
Candidate WBMPs typically appeal to the presence of unjustified evil or suffering in the world. On the Augustinian view according to which evil is in fact the absence of good (privatio boni), every WBMP would presumably refer to the absence of a WGMP. There may be such WBMPs, and there may also be WBMPs that are the contraries of WGMPs, and there may be other, different, WBMPs. I remain neutral on this; nothing turns on it for my purposes.
It may be that certain good-making properties cease to make worlds better past a certain point, or in certain combinations. The same goes, mutatis mutandis, for WBMPs. So, while the goodness of a world depends on its axiological properties, this dependency may not be simple.
The rival view, which I call contingentism, holds that the actual set of WGMPs and WBMPs might have had different members. One version of this view, called theological voluntarism, holds that God could have willed other properties to be good-making and bad-making than the ones that are so. For a discussion of this view, see Mann (1991, pp. 253–258).
One might deny that all worlds can properly be evaluated with respect to a stable set of WGMPs and WBMPs: this is to hold that there are incommensurable pairs of worlds. (On this, see Mann (1991, pp. 268–273), and Grover (1998).) Alternatively, one might deny that all worlds can be compared. This can be done with or without appeal to incommensurability. Incommensurable worlds, of course, cannot be compared. But even if all worlds are commensurable, there might still be failures of comparability between worlds. For simplicity, I set these issues aside.
For example, see Hoffman and Rosenkrantz (2002, p. 159) and Strickland (2006). Many more treat this as an epistemic possibility.
For example, suppose that w′ differs from w by having one more grain of sand on one beach than w does (and whatever is required for this, and whatever ensues from this). It seems reasonable to suppose that w and w′ are axiologically equivalent.
Kretzmann (1991) offers a clear account of Aquinas’ reasoning in this matter.
This, of course, is a version of the problem of evil. In response, some theists (notably, Adams 1972) have rejected the a priori claim, while others—too many to mention here—have denied the a posteriori claim, or at least suggested that it has not been (or cannot be) justified.
A small qualification: while I here suggest that arguments for the surpassability of the actual world are a posteriori, Ian Wilks has offered a plausible counterexample: a priori introspection of one’s own thoughts, desires, intentions, and the like, might well convince one that the actual world could be better. (In slogan form: male cogito ergo malum est.) In what follows, I ignore this special case. I also set aside other arguments for atheism that have been advanced on EOUW.
I here assume that there is no principled middle ground between claiming that there is exactly one unsurpassable world, and that there are infinitely many.
Nor does God foreknow their outcome. This need not detract from divine omniscience, since it is plausible to hold that there are no truths to be known about (for example) the exact outcomes of random processes.
Molinism is the doctrine according to which there are unalterable contingent truths (known by God) about how libertarian-free creatures would act in various possible circumstances. On Molinism, God can control which libertarian-free creaturely actions occur without causing them: by actualizing the world in which his favoured creaturely actions obtain. This might be thought to favour the view that God determines every feature of a world in actualizing it, and this might be desirable for theists with a robust conception of divine sovereignty. But since theism does not entail Molinism, it remains correct to say that theism does not require the view that God determines every feature of a world in actualizing it. (Besides, Molinists needn’t hold that God determines every feature a world in actualizing it—random processes serve as a counterexample.)
For more on this, see Tom Talbott, “The Best of all Feasible Worlds” (unpublished manuscript). I remain neutral on whether divine world-actualizing activity is best construed to involve many actions, or one ‘spread-out’ action. I also remain neutral on whether divine world-actualizing activity is best understood to be timeless or temporal.
Leslie mentions four restrictions on something counting as a universe: (1) absence of causal contact with any other universe; (2) significantly unique in character; (3) large in size; and (4) unknowable from any other universe (1989, pp. 66–67). The first seems to entail the fourth (but not vice-versa, as Leslie notes); the second and third seem unnecessarily-restrictive for my purposes.
Many philosophers have defended various versions of the multiverse. For a historical survey of many-universe hypotheses, see Munitz (1951). Those who take seriously the idea that theism suggests the multiverse include McHarry (1978), Forrest (1981, 1996), Coughlan (1987), Turner (2003), Draper (2004), Hudson (2006), and O’Connor (2008).
It will be worthwhile to distinguish this proposal from Lewis’ modal realism. For Lewis, possible worlds just are isolated, spatiotemporally-interrelated concrete objects. Lewis denies that one possible world can consist of two or more completely disconnected spacetimes—although he concedes that he would prefer not deny to this (1986, p. 71, 74). On the strength of the considerations referred to above, I take seriously the thought that Lewis erred on this point. (For more on this, see Bricker 2001.) I make no further claims concerning the ontological status of possible worlds: I aim to be as neutral as possible between competing accounts.
See Turner (2003, p. 148).
Given the Identity of Indiscernables, at any rate, this is impossible. (See McHarry 1978, p. 133.)
I here assume that bi-location is impossible, but of course this is controversial. Also, I should note that ruling out trans-universe identity does not itself rule out trans-world identity of individuals. But see note 34, below.
In note 3, I suggested that the dependency of the axiological status of a world on its WGMPs and WBMPs need not be simple. Similarly, the dependency of the axiological status of universes on the relevant properties need not be simple. One further point. Some WGMPs can equally be deemed universe-good-making properties. But not all: consider the property comprising many good universes. While this is a plausible world-good-making property, it cannot be a universe-good-making property.
While worlds are necessarily-existing states of affairs, and are hence actualized rather than created, universes are contingently existing spatiotemporal entities, and are thus created rather than actualized.
I remain neutral on this question. In contrast, O’Connor (2008, Sect. 5.1) claims that there are no unsurpassable universes.
Draper 2004, pp. 318–319. Although their cosmogonies are different, the same basic motivation drives Bruno and Kant:
I hold the universe to be infinite as a result of the infinite divine power; for I think it unworthy of divine goodness and power to have produced merely one finite world when it was able to bring into being an infinity of worlds (Bruno, as quoted in Munitz 1951, p. 244).
But what is at last the end of these systematic arrangements? Where shall creation itself cease? It is evident that in order to think of it as in proportion to the power of the Infinite being, it must have no limits at all …the field of the revelation of the Divine attributes is as infinite as these attributes themselves (Kant 1755,138–9).
Turner (2003, p. 147) and Hick (1978, p. 72) attribute this term to Arthur Lovejoy. Lovejoy has in mind “…not only the thesis that the [actual world] is a plenum formarum in which the range of conceivable diversity of kinds of living things is exhaustively exemplified, but also any other deductions from the assumption that no genuine potentiality of being can remain unfulfilled, that the extent and abundance of the creation must be as great as the possibility of existence and commensurate with the productive capacity of a ‘perfect’ and inexhaustible Source, and that the world is better, the more things it contains” (1936, p. 52). Clearly, the principles I suggest are more restricted.
This threshold would be difficult to specify, but could presumably be expressed in the language of WGMPs and WBMPs. Van Inwagen (2006, pp. 95–112) offers an argument that could be deployed to suggest that talk of such a threshold is incoherent. Van Inwagen’s argument is not without its critics, however. See, for example, Jeff Jordan (2003).
O’Connor (2008, Sect. 5.1) expresses doubts about the principle of plenitude, favouring instead a weaker principle which requires God to create at least one universe of every significant type or kind, above some threshold. He does not, however, explain what he means by ‘type’ or ‘kind’, nor does he explain what is wrong with the principle of plenitude.
This principle is modeled on Rowe’s notorious “Principle B” (2004, p. 91). It can be seen as the conjunction of two claims:
-
PP2a
If a being fails to create any universe that is both creatable and worth creating, then ceteris paribus, it is possible for that being to have performed a better world-actualizing action.
-
PP2b
If it is possible for a being to have performed a better world-actualizing action, then, ceteris paribus, it is possible for that being to have been better.
Critics, of course, might attack either claim.
-
PP2a
This sounds more dramatic than it is. God can fail to sustain a universe in a variety of ways. One is, of course, to destroy it altogether. Another is to introduce something into that universe that would otherwise have been lacking, and still another is to remove something from that universe that would otherwise have been present.
Critics might allege that they are implausible in the divine case, or they might insist that they are unmotivated (or defeated) by reflection on cases of human creation and sustenance.
Talbott and van Inwagen have independently suggested that I cannot reasonably claim that TM comprises all universes worth creating and sustaining. Their objection can be put as follows. Consider some person, P, who exists in a universe, U, that is worthy of creation and sustenance. TM includes U, and also includes, inter alia, all other universes which both (i) contain P’s counterparts, and (ii) are worthy of creation and sustenance. So far, so good. But, according to Talbott and van Inwagen, there are still other universes worthy of creation and sustenance which include P herself, and not just P’s counterparts. If any such universe is possible, then it exists in a possible world—but it cannot exist in TM, on my view. Accordingly, TM fails to contain all universes worth creating and sustaining. My response is developed in “Theism and Modal Collapse”, where I argue—on quite independent grounds—that if theism is true, TM is the only possible world that there is. If this surprising view is correct, then a fortiori no individual can inhabit multiple worlds—and the Talbott/van Inwagen objection fails.
Here is an objection. Suppose (for simplicity) that there are exactly 1,000 universes worthy of creation and sustenance in TM, each with a distinct axiological status, and that they are numbered in ascending axiological order: {1, 2, 3…1000}. Compare that to TM+: a world which comprises only the following universes: {101, 102, 103…1000}. The universes in TM+ have a higher average axiological status than the universes in TM, so one might think that TM+ surpasses TM. But while TM+ comprises only universes worth creating, it does not comprise all universes worth creating. So it is natural to think that TM+ could be improved by the addition of universes {1–100}. But, of course, this is just to describe TM, and so TM surpasses TM+. For other arguments in this spirit, see Turner (2003, p. 151) and Draper (2004, p. 319).
One might try to object that the creation of universes which are unworthy of creation might somehow serve to make the entire world better that it otherwise would be (as the addition of some dissonance might serve to make a musical composition more harmonious overall). But it’s difficult to see how such an objection can be sustained, given that universes are causally isolated. Nor, in my view, can aesthetic considerations plausibly be brought to bear in this context.
That theism requires a multiverse is perhaps surprising, since the multiverse is so often invoked in criticisms of the fine-tuning argument for theism. For a good survey of such criticisms and replies, see the introduction to Neil Manson (2003).
There are, of course, other arguments for atheism on this view, but I will focus on this one, since it is undoubtedly the most influential.
Arguments from evil are sometimes divided into the more ambitious—called deductive or logical—which purport to show that God does not exist, and the more modest—called inductive or evidential—which purport to show that, probably, God does not exist. In what follows, I will not insert parenthetical qualifications like this to cover the latter, but what I say should be assumed to apply to them as well.
See van Inwagen (2006, p. 8).
To my knowledge, no philosopher has yet explicitly criticized “multiverse theism” by appeal to evil. But, as I’ve suggested, claims in typical arguments from evil can be offered as reasons for thinking that our universe is not worthy of creation or sustenance, and thereby deployed in arguments against multiverse theism.
A handful of philosophers have defended theism against traditional arguments from evil by appeal to the multiverse. McHarry (1978) claims that a response based on the multiverse can solve the problem of evil. Turner (2004) suggests that it is a partial solution, and O’Connor (2008, Sect. 5.3.1) sees prospects for theodicy here. More elaborate multiverse-based responses can be found in Forrest (1996) and Hudson (2006). Draper (2004) is the only philosopher to criticize the multiverse-based response to the problem of evil, though he does not directly address these authors. Space does not permit consideration of these rather varied positions.
References
Adams, R. (1972). Must God create the best? Philosophical Review, 81, 317–332.
Blumenfeld, D. (1975). Is the best possible world possible? Journal of Philosophy, 84, 163–177.
Blumenfeld, D. (1995). Perfection and happiness in the best possible world. In N. Jolley (Ed.), Cambridge companion to Leibniz. Cambridge: Cambridge University Press.
Bricker, P. (2001). Island universes and the analysis of modality. In G. Preyer & F. Siebelt (Eds.), Reality and Humean supervenience: Essays on the philosophy of David Lewis. Lanham, Maryland: Rowman and Littlefield.
Coughlan, M. J. (1987). Must God create only the best possible world? Sophia, 26, 15–19.
Draper, P. (2004). Cosmic fine-tuning and terrestrial suffering: Parallel problems for naturalism and theism. American Philosophical Quarterly, 41, 311–321.
Forrest, P. (1981). The problem of evil: Two neglected defences. Sophia, 20, 49–54.
Forrest, P. (1996). A speculative understanding of evil. In God without the supernatural (Chap. 8, pp. 213–256). Ithaca: Cornell University Press.
Grover, S. (1988). Why only the best is good enough. Analysis, 48, 224.
Grover, S. (1998). Incommensurability and the best of all possible worlds. The Monist, 81, 648–667.
Grover, S. (2003). This world, ‘Adams Worlds’ and the best of all possible worlds. Religious Studies, 39, 145–163.
Grover, S. (2004). Rival creator arguments and the best of all possible worlds. Sophia, 43(2004), 101–114.
Hick, J. (1978). Evil and the God of love (revised ed.). New York: Harper and Row.
Hudson, H. (2006). Hyperspace and theism. In The metaphysics of hyperspace (pp. 163–181). Oxford: Oxford University Press.
Jordan, J. (2003). Evil and Van Inwagen. Faith and Philosophy, 20, 236–239.
Kant, I. (1755). Universal natural history and theory of the heavens (W. Hastie, Trans.). Ann Arbour: University of Michigan Press, 1969.
Kraay, K. (unpublished manuscript). Theism and modal collapse. Available at www.ryerson.ca/~kraay/.
Kretzmann, N. (1991). A particular problem of creation. In S. MacDonald (Ed.), Being and goodness (pp. 229–249). Ithaca: Cornell University Press.
Leslie, J. (1989). Universes. New York: Routledge.
Lewis, D. (1986). On the plurality of worlds. Cambridge: Blackwell.
Lovejoy, A. O. (1936). The great chain of being. New York: Harper and Row.
Mann, W. (1991). The best of all possible worlds. In S. J. MacDonald (Ed.), Being and goodness: The concept of God in metaphysics and philosophical theology (pp. 251–277). Ithaca: Cornell University Press.
Manson, N. (2003). God and design: The teleological argument and modern science. New York: Routledge.
McHarry, J. D. (1978). A theodicy. Analysis, 38, 132–134.
Munitz, M. K. (1951). One universe or many? Journal of the History of Ideas, 12, 231–255.
O’Connor, T. (2008). The scope of contingency. In The necessary shape of contingency: Theism and ultimate explanation (Chap. 5). Melbourne, Australia: Wiley-Blackwell.
Plantinga, A. (1974a). The nature of necessity. Oxford: Clarendon Press.
Plantinga, A. (1974b). God, freedom, and evil. Grand Rapids: Eerdmans.
Reichenbach, B. (1982). Evil and a good God. New York: Fordham University Press.
Rowe, W. (1993). The problem of divine perfection and freedom. In E. Stump (Ed.), Reasoned faith (pp. 223–233). Ithaca: Cornell University Press.
Rowe, W. (1994a). The problem of no best world. Faith and Philosophy, 11, 269–271.
Rowe, W. (1994b). Can God be free? Faith and Philosophy, 19(2002), 405–424.
Rowe, W. (2004). Can God Be free? Oxford: Oxford University Press.
Schlesinger, G. (1977). Religion and scientific method. Dordrecht: Reidel.
Sobel, H. (2004). Logic and theism: Arguments for and against beliefs in God. Cambridge: Cambridge University Press.
Strickland, L. (2006). God’s problem of multiple choice. Religious Studies, 42, 141–157.
Swinburne, R. (1979). The existence of God. Oxford: Clarendon Press.
Talbot, T. (unpublished manuscript) The best of all feasible worlds.
Tegmark, M. (2003). Parallel universes. Scientific American, 288, 41–53.
Turner, D. (2003). The many-universes solution to the problem of evil. In R. Gale & A. Pruss (Eds), The existence of God (pp. 1–17). Ashgate: International Research Library of Philosophy.
van Inwagen, P. (2006). The problem of evil. Oxford: Oxford University Press.
Wielenberg, E. (2004). A morally unsurpassable God must create the best. Religious Studies, 40, 43–62.
Acknowledgements
I thank Paul Bali, N. Ballantyne, Jim Dianda, Peter Forrest, Luke Gelinas, David Hunter, Ed Luk, Jonathan Strand, Tom Talbott, and an anonymous Philosophical Studies referee for helpful comments on earlier drafts. Thanks also to Michael Almeida, Paul Draper, Hud Hudson, and Timothy O’Connor for sharing their prepublication work on this topic with me. I am grateful for the generous research support I received from Ryerson University, and from the Social Sciences and Humanities Research Council of Canada, in Fall 2006.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kraay, K.J. Theism, possible worlds, and the multiverse. Philos Stud 147, 355–368 (2010). https://doi.org/10.1007/s11098-008-9289-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-008-9289-y