Science in Christian Perspective


Many Worlds Hypotheses: 
A Naturalistic Alternative to Design

Jay Wesley Richards

Princeton Theological Seminary
Princeton, NJ 08542-0803

From: PSCF 49 (December 1997): 218-227

In this paper, I consider the so-called World Ensemble theories or Many Worlds Hypotheses. As I see it, Many Worlds Hypotheses (MWH) arise primarily (but not exclusively) from the realization of the apparent contingency of the visible universe and the way in which it seems fine-tuned" for the existence of complex forms of intelligent life. I am interested in their common function as a metaphysical explanation, and the way MWH function as a sort of "Designer substitute." However, MWH cannot exclude the possibility of theism or design, except as bare assumption, and, much to the chagrin of certain theorists, some might even entail theism.

While Many Worlds Hypotheses (MWH hereafter) occur in various, sometimes contradictory forms, they arise primarily (but not exclusively) from the increasing realization of the apparent contingency of the visible universe and the ways in which it seems "fine-tuned" for the existence of complex forms of intelligent life. The fundamental constants observed in matter, and those thought to obtain in the initial conditions at the origin of the physical universe, look curiously attuned to the development and existence of intelligent life.

I am not interested here so much in the relative merits of the numerous MWH currently proposed, but in their common function as a metaphysical explanation, in which they explain the apparent fine-tuned contingency of the visible universe, and the way MWH function as a sort of "Designer substitute." Where the theist might infer the existence of God from the fine-tuned contingency of the universe or take observations of fine-tuning as confirmation of belief in God, others infer the existence of many, even infinitely many, worlds from these observations. I argue that if the latter is a valid inductive inference, then so is the former. And if MWH function on the same level as theistic belief, and the latter belief is metaphysical, then so is the former. Specifically, many versions of MWH (but perhaps not all) look like deductions from naturalism proposed to accommodate this apparent fine-tunedness. Moreover, MWH cannot exclude theism or design, except as a bare assumption. And in light of a contemporary version of the ontological argument for the existence of God, some might even entail theism (much to the chagrin of MWH theorists). At most, they may function as solace for the naturalist faced with evidence for fine-tuning.

Metaphysical Commitments in Science

In contrast to positivism earlier in this century, a growing body of work in the philosophy and history of science has uncovered the degree to which theories, paradigms, and even metaphysical beliefs and assumptions play a crucial role in scientific research. Without belief in the observability, reality, and general orderliness of nature, science probably would never have developed, and a complete loss in these beliefs would probably presage the demise of science. Moreover, as is now commonly said, scientific data are theory-laden and scientific theories are under determined by the data. To put it contentiously, high-level beliefs infect low-level observations. Many potential theories may be consistent with any finite amount of observational data, and such data will themselves be couched in terms of some specific theory. Even the very relevance of that body of data may be shaped by the broader context in which the observers are working.

While I do not wish to enter this debate here, it does seem clear that MWH theorists take advantage of a blurring of lines between "hard science" on the one hand and "metaphysics" on the other. However, I will not criticize MWH because they are metaphysical. Rather, I wish to show only that they do have a metaphysical function, and, more strongly, seem to stem from metaphysical naturalism. This claim does not entail that there are no other motivations for proposing MWH. My argument is only that metaphysical commitment is a primary motivation. Whether such metaphysical convictions have a place in legitimate science is a question for another time.

Attempts to Avoid the Question

Since at least the time of Hume and Kant, many have insisted that properties such as probability and contingency could not be applied to such a thing as the entire universe. The cosmological and teleological arguments have commonly been thought flawed because they assumed that the universe as a whole needed explaining. While few people would deny that something must have necessary existence in order for anything to exist contingently, the nontheist could always insist that the universe itself was the necessarily existing thing. Its existence, as Aristotle assumed, was infinite in duration. To appeal to a god who would himself be more complex than the universe, in order to account for the existence of the universe, is to violate Ockham's Razor. We should more simply stop the regress of explanation with the universe itself, as the fundamental, necessary brute fact. The teleological argument, on the other hand, was thought destroyed by the awareness that, since we have nothing to compare it with, we cannot rightly say the universe is designed, as we could say, for example, with a watch. Any apparent design in nature, so it has been claimed, is more sufficiently explained by Darwin's natural selection working on random genetic mutation than by recourse to creation by a supernatural being.

Ironically, in the time since many theologians have given up all appeals to such arguments, scientists have been amassing an enormous amount of evidence for the radical contingency and (apparent) fine-tuning of the visible universe. Big Bang cosmology, which implies that the observable universe has a past of finite duration, can be very troubling to those who wish to assign necessary existence to the universe. For while both theist and atheist deny the maxim that everything that exists has a cause for its existence, since both suppose something (either God or the universe) to have necessary and uncaused existence, the intuitive plausibility of another option seems almost irresistible; that is, everything that begins to exist has a cause for its existence.1 A necessarily existing thing by definition does not pop into existence. Of course, one could just deny this new option in the case of the universe, but then one would retain little credibility in appealing to a much more tenuous principle such as Ockham's Razor in denying, say, theism.

Coupled with Big Bang cosmology has been the abundance of mind-boggling, highly sensitive conditions which appear to obtain in our universe, conditions necessary for the existence of beings such as Homo sapiens. Even granting the most generous assessment of Darwinian explanation, the conditions necessary for either Darwinian evolution, or our continued existence, fall within nearly infinitesimally small parameters. The four fundamental constants observed in physics-the gravitational, electromagnetic, strong, and weak nuclear forces-maintain conditions which, if variant in some instances by even one part in trillions of trillions, would make life of our type physically impossible. Examples of this are myriad and well-documented.2 Here is one example. According to Paul Davies, if the nuclear strong force or electromagnetic force had been different by one part in ten quadrillion, no stars like our own could have formed.3 her in a nearly instant recollapse, or an expansion so strong that no gravitation would have been able to bring hydrogen atoms together for the formation of stars, and thus heavy elements necessary for the life present on our planet.

There is potential trouble in these findings for the materialist who bestows upon the physical universe ultimate reverence: the contingency and even design of nature. During the nineteenth century, the eternity of matter was held by many as a basic pillar on which all of science rested, and the doctrine's crucial role in Marxist materialism was prominent. So we should suspect belief in the eternity and necessity of nature to die hard (necessity and eternity travel together). MWH are able to preserve these materialist commitments while accepting and even insisting on the contingency and radical improbability of the visible universe when considered by itself. Before considering the more interesting option of MWH, however, we should note a few attempts to sidestep the crucial issue of contingency and fine-tuning.

MWH are able to preserve [the eternity of matter] while accepting and even insisting on the contingency and radical improbability of the visible universe when considered by itself.

Some, such as Stephen Hawking, still hope somehow to assert the necessity of our visible universe (we might call it our world to avoid begging any questions about what universe implies). Once a so-called Theory of Everything (TOE) is attained, unifying the four fundamental constants, we will have a core set of equations which will be necessary, and no explanation beyond them will be needed.4 This has such a strong ring of implausibility that it may be a mistake to note its problems. But an ex post facto attribution of necessity to a scientific theory, no matter how powerful or correct, is simply confused. If an internally coherent alternative is conceivable, or if the absence of any TOE is even possible, then its truth or existence cannot be necessary. It would also seem to falter before G–del's incompleteness theorem.5 That a TOE or Grand Unified Theory would contain particular numbers in specific equations should be enough to raise suspicion about attributing necessity to it. Beyond this, one may note that making the universal constants necessary would have the devastating consequence" for science of making experiment unnecessary.6 Armed with a Theory of Everything, we could simply deduce how things will go, making experimentation unnecessary (although a committed Platonist might not shrink from this conclusion).7

Two other strategies often recur where theorists discuss the so-called anthropic principle. Most simply, the anthropic principle notes that human beings will necessarily (in the physical sense) observe a universe consistent with their existence. Barrow and Tipler define it thus:

The basic features of the Universe, including such properties as its shape, size, age and laws of change, must be observed to be of a type that allows the evolution of observers, for if intelligent life did not evolve in an otherwise possible universe, it is obvious that no one would be asking the reason for the observed shape, size, age and so forth of the universe.8

I am not concerned here with the question of whether the anthropic principle is insightful or trivial, or with the way it often makes the broad evolutionary picture true by definition. I mention this "self-selection" aspect of the anthropic principle below, and consider it an important component in MWH. Here I note only its failure as an attempt to avoid the question of the radical contingency of our existence and a universe consistent with it.

For a reason not wholly clear, some partisans of the anthropic principle assume that since only a universe consistent with observers will be observed, we should not be surprised to find ourselves observing such a universe. So Barrow and Tipler state:

We should not be surprised to observe that the Universe is so large. No astronomer could exist in one that was significantly smaller. The Universe needs to be as big as it is in order to evolve just a single carbon-based life-form.

We should emphasize once again that the enormous improbability of the evolution of intelligent life in general and Homo sapiens in particular does not mean we should be amazed we exist at all.9

While these statements underscore the fact that very specific conditions must be met for beings of our sort to exist, they clearly are not explanations. The issue of surprise is not that we observe a universe consistent with our presence, but that that universe is the one which exists.

John Leslie's story of the Firing Squad illustrates the strangeness of this explanation.10 Imagine a man standing before a 50-man firing squad, remaining alive and unscathed after his execution. "Would he reasonably reflect, I shouldn't be surprised that all 50 sharpshooters missed, because if they hadn't, I wouldn't be here to observe it"? Surely not. Similarly, our presence as observers does not dispel amazement at the apparent fine-tuned nature of the visible universe.

More radical arguments occur in works by some such as John Wheeler, who seems to assert that our observation of the universe is itself the cause of the universe's existence, since only that which is observed really exists. Such radical idealism raises a whole host of problems and questions beyond the scope of this essay. Wheeler' proposal is (at least partly) an attempt to account for quantum phenomena; but it has this idealist edge which will probably not endear it to most scientists. So I will not tarry long on it here. However, my avoidance of it should not be taken to imply it is not worth considering.

A final attempt to circumvent the apparent fine-tuning of the universe in such a way as to be consistent with earthly life is an appeal to the Copernican principle. That is, we are charged with anthropocentrism in assuming our form of life and consciousness are the only meaningful forms of life taken. Perhaps life and consciousness can evolve in interstellar dust or in the core of a neutron star. After all, we know so little about even our type of life.

Of course, no one need argue that human life is the only meaningful type of life.11 Ironically, from the same quarters, which insist that human type of life is only conceivable as the result of billions of years of trial-and-error evolution, emerge defenses of the possibility of life existing inside a neutron star or during the milliseconds between the expansion and contraction persisting in other universes. The burden of proof should be on those who propose such implausible life forms; we need only argue that the existence of any plausible life forms will occur in very fine-tuned universes such as the one we observe. M. H. Hart has estimated that even on an ideally habitable planet the chance that living things would develop would probably be lower than 1 in 103,000.12 The motivation to undercut claims such as this may be understandable, but the creation ex nihilo of hypothetical alternate life forms to explain away our apparent uniqueness does not provide a very strong argument against amazement at this uniqueness.

Many Worlds Hypotheses

Having noted these explanations, or evasive maneuvers, concerning the apparent fine-tuned contingent nature of the visible universe, we can now ponder what I think are more interesting and compelling proposals: Many Worlds Hypotheses. Unlike some other proposals, MWH accommodate the troubling observation of a contingent, visible universe and may even bask in the fact. Whereas earlier the contingent objects of the universe were considered part of a necessary, infinite, and eternal universe, developments in Big Bang cosmology and relativity have made attributions of eternity and necessity more difficult to make to the observable universe. On the other hand, MWH postulate that the observable universe, while itself contingent, is a mere part of a vast ensemble of different universes or worlds. We can only observe this world. The Many Worlds theorists and traditional theists both agree that an explanation for a contingent world should be sought outside that particular world, but their strategy is quite different. For the Many Worlds theorists, "the question `Why this universe?' is no longer relevant, because all possible universes exist. The set of universes taken together is not contingent."13

The Many Worlds theorists and traditional theists both agree that an explanation for a contingent world should be sought outside that particular world, but their strategy is quite different.

While theorists have proposed numerous mechanisms for producing multiple worlds, they have several features in common:

(1)Absence of (or very limited) causal contact.

(2)Characteristics between worlds are often very different, for example, different laws.

(3)Worlds are very large, or are becoming very large (not an essential feature).

(4)Worlds or universes apart from our own cannot be known by us in any direct ways.14

So, however we construe them, alternate worlds are generally unobservable by definition. This fact alone is enough to cause some scientists to dismiss all MWH as unscientific speculations. And if science by definition is concerned exclusively with the "directly" observable, this criticism would obviously be correct. But, a great deal of quantum and theoretical physics, as well as cosmology, focuses on subject matter not directly observable. So tagging these theories with the disrepute of being unobservable does not suffice to consign them to the realm of the unscientific." This is not to say that any of the MWH are correct or even plausible, but only to say that the need for explanation with which a unique and contingent universe confronts us is reflected in the various MWH.

The theoretical mechanisms popular at present for hypothesizing a World Ensemble fall into five categories: (1) An oscillating cosmos which goes through an infinite number of cycles, in which there is a Big Bang, an expansion, a contraction, a Big Crunch, and then the process repeats itself; (2) The Everett Interpretation of quantum theory, by which individual, noninteracting worlds split off" from one another as a result of the individual choices" of quantum events; (3) Noninteracting quantum fluctuations in a Superspace, each a separate world or small universe, or perhaps even without a common background Superspace; (4) An infinitely large open" universe in which separate regions are so distant from each other that no contact exists between them; and (5) An Inflationary Cosmos which, although "closed," has an immense collection of different domains, unobservable by any of the others.15 Although all these proposals have their advocates and detractors, they rely on factors which militate against the earlier desire to make our world necessary. Most assume or try to account for (as in the Everett Interpretation of quantum theory) the probabilistic nature of physics at very basic levels. They often use theories of symmetry breaking by which fundamental constants-the very nature of matter as we know it-would vary from world to world. They propose many worlds which are real but very different and unobservable from our own. Our world is merely one among vastly many. The Everett Interpretation of quantum theory reminds us that naturalism and fine-tunedness are not the only motivations for MWH, since the Everett Interpretation was devised at least partly to overcome the collapse of the wave function" in the measuring process.16

[Many World theorists] propose many worlds which are real but very different and unobservable from our own.

Although there is diversity among MWH, I am interested here only in their function as an interpretive and metaphysical explanation for the surprisingly fine-tuned nature of our universe. The appeal of a multiple worlds theory should be fairly obvious. If there are infinitely many universes, the fact that this particular one exists is hardly surprising. This one had to exist somewhere, since it is a member of the set of all possible universes. "What would originally make this universe seem awe-inspiring, since Awe can conceive of so many alternatives to it,"17 becomes less troubling if all those alternatives exist as well. When dealing with infinite sets, counter-intuitively, the near infinitesimally improbable, if at least possible, is not only actual but common, even infinitely common. It is here that the self selection" effect of observation comes in. For only those universes (however improbable) which are consistent with the evolution and existence of observers will be observed. Just like an immense lottery, someone has to win. The winner may be surprised to be the one, but there is clearly nothing surprising that someone won. In the case of universes, observers will exist only in those universes intricately arranged to permit their existence. That which might initially be thought highly improbable is on the strongest MWH, inevitable. Most crudely, an infinite ensemble of all possible worlds entails the actuality of every possibility.

Theorists sometimes qualify this extreme affirmation by insisting that multiple worlds is an ensemble of many or all physically possible worlds, with the laws of physics being the same in all worlds. So Paul Davies says:

The selection of universes on offer is restricted to those that are physically possible. There will be many more universes that are logically possible but contradict the laws of physics...So, although such many-universe theories might provide a selection of alternative states of the world they cannot provide a selection of laws.18

However, such a qualification does not accommodate the words of many multiple worlds theorists. Barrow and Tipler, for example, are unwavering in their interpretation of Everett's Many-Worlds Interpretation of quantum physics: "...the Universe, which is defined as everything that actually exists, is equal to all logically consistent possibilities."19 If Many Worlds theorists accepted Davies' qualification about the inviolability of physical laws, they would be strapped with the continuing trouble of why those laws exist. In order to preserve the notion that these laws occurred randomly, they must be conceived as only one set among many, all arising through some chance mechanism such as symmetry breaking. With Davies' qualification, we still have the presence of contingent physical laws which need explaining. The Many Worlds theorist wants to circumvent that need. Most Many Worlds partisans are willing to give up even the inviolability of physical laws to accommodate the apparent intricate fine-tuning of physical laws (in our world) for life. What these theorists need is a theory which preserves randomness and chance, and yet accounts for the apparent design" of our universe. Infinite worlds are thought to provide this:

If cosmological initial conditions are exhaustively random and infinite then anything that can occur with non-vanishing probability will occur somewhere, in fact, it will occur infinitely often.20

Of course, an infinite number of worlds need not be actual in order to make the existence of ours untroubling. If the likelihood of our universe existing, is, say, 1 in 1010^10, then only 1010^10 worlds need exist to serve this purpose:

For when might our existence be made unpuzzling through a multiplicity of universes and an observational selection affect? Answer: Just as soon as the multiplicity and the variety of those universes were great enough to give a fair chance that at least one universe would contain intelligent observers.21

Interestingly, this point is often missed by many of our theorists, who prefer talk of necessity, logical possibility, and infinite sets. A whole bunch of worlds" just doesn't pack the punch for dispelling troubling questions of design and contingency that a claim about "an infinite variety of worlds" does. "The question of why does this universe rather than that universe is answered by saying that all logically possible universes do exist. What else could there be?"22 While such a reply may satisfy those disposed toward naturalism and away from questions about a Designer, I will argue that it can explain nothing more than is explained by theistic belief, is probably inferior to that belief as an explanation for the apparent fine-tuning of the observable universe, and that such explanations raise additional questions which make them less the ally of the naturalist than might be thought; and, in any event, insofar as MWH act as substitutes for considering the possibility of design and the existence of a Designer, they are metaphysical. This is not to say science should not be involved in metaphysics, but rather that the traditional boundaries between science and metaphysics are becoming increasingly blurred.23 Whether this is a good or bad thing I leave to the reader to decide.

Many Worlds Hypotheses and Design

While the motivation for the popularity of MWH cannot answer the question of the truth of any of them, MWH do serve to answer the question of the observable universe's contingency and apparent fine-tuning. Moreover, theorists often assume that if some version of a many worlds theory is correct, no consideration of an actual Designer need arise, since such a question would be superfluous. MWH and the need for a Creator are often presented as mutually exclusive alternatives, and discussions of them are usually framed in such terms. Nevertheless, they are clearly not logical complements. They do not reside in wholly incompatible conceptual territories. The truth of one does not entail the falsity of the other. They are really only alternatives in that they can explain the same phenomena. They may both be compatible with the data; they may be equally adequate empirically.

While MWH and creation by design are not logically contradictory, they are usually alternatives.

On the other hand, MWH and design by a Creator (these are usually the live options) are alternatives. If we had reason to believe one were true, we would not need recourse to the other, since the phenomena calling for explanation would be sufficiently explained. And they are certainly perceived as alternatives by much of the relevant scientific constituency. In fact, fear about a resurgence of questions about design and teleology in nature has inspired much unnecessary suspicion toward findings which indicate the incredible intricacy and improbability of the existence of complex systems such as DNA or life in general. MWH may be a way to concede these observations without re-entering putatively won battles over teleology. John Leslie is explicit about this:

Nowadays, when the God hypothesis is so unpopular, many scientists would initially be very reluctant to accept that Life balanced on such a razor's edge. Hawking's estimate [that an expansion speed decrease at early times by one part in a million million would have meant life's absence] would thus suggest to them that Hawking is wrong. But these scientists could well change their minds when they saw that a varied ensemble of universes and an observational selection effect could do much the same work as God might do.24

So, while MWH and creation by design are not logically contradictory, they are usually alternatives.

Second, MWH do not enjoy greater empirical support or verification than belief in design. Every datum which might reasonably be thought to support the former could equally well support or confirm the latter. The popularity of MWH among scientists may obscure this point. Any calculations on Baye's theorem for the improbability of the intricate equilibrium obtaining in physical constants, necessary for the formation of stars and biological forms, will be support for design at least as much as for many worlds. As explanations for intricacy, apparent fine-tuning, and our presence in this world, MWH and design occupy very similar conceptual territory. (I am not saying here that Christian or theistic belief is merely an explanation or a hypothesis. I am referring here only to the function of such belief as an explanation for contingency, uniqueness, et al. in the physical universe.) The privilege which science and naturalism enjoy in our intellectual context may blind us to the fact that MWH are no more "empirically verified" than is theistic design.

MWH do not enjoy greater empirical support or verification than belief in design.

Of course, they may both be empirically adequate, but that is another matter. Because it is scientists on the whole who are at present the most common advocates of MWH, we may be hoodwinked into bestowing upon these theories the status of "Science," while continuing to dismiss questions of design and creation as theological, metaphysical, or speculative. As John Leslie notes:

Science, remember, does not support the multiple-universes hypothesis rather than the God hypothesis. We do not have, independently of the delicate adjustments which-so science seems to say-were crucial to Life's possibility in this universe of ours, any strong evidence of a World Ensemble.25

Third, and like unto the second, MWH do not enjoy an a priori status of preference over intentional design by a Creator. This presumption of naturalism, we might call it, is presupposed by some advocates of MWH: If any hypothesisCno matter how implausible, unlikely, or unverifiableCcan be proposed to explain the apparent fine-tuning of our universe, then we should prefer it to recourse to divine intention. I can think of no good argument to justify such a presumption, other than a desire to support naturalism contra theism. On this level, we may see most clearly that some advocacy of MWH is metaphysically motivated. This is not, of course, a judgment about the truth or falsity of any such hypothesis. A commitment to materialism or naturalism, like more mundane presuppositions about the orderliness or existence of an external world, can influence research and theory choice in science as in other fields. Which physicist would be more likely to spend more time considering various theoretical models for producing multiple worlds-a Christian or other theist, or a naturalist? I doubt any honest respondent would deny that the latter would be more likely.

Of course, few people would decide so ultimate a question about whether the universe were created as a question of probabilities, or as a mere inference to the best explanation. Moreover, probably few individuals dwell in an epistemic locus from which such a question could be disinterestedly assessed. Nevertheless, I do think there are some good arguments for the theistic belief and against the atheistic belief which motivates, if only implicitly, much multiple world hypothesizing. Of course, I am only concerned with belief in a Creator versus belief in multiple or infinite worlds as an explanation for apparent fine-tuning. If both theism and the existence of many worlds are true, then there is little controversy. But since I judge the bulk of motivation for MWH to be a recognition of apparent fine-tuning plus a commitment to naturalism, I suspect there are few reasons to hold both these ultimate beliefs as true. There seem to be few other reasons-empirical, philosophical, or otherwise-for the postulation of so many countless and unobservable regions. That a proliferation of MWH would follow on the heels of a mass of evidence in science that the observable universe fits into the very narrow parameters necessary for life of a sort even remotely similar to the type of life we know, is a sequence which is surely not coincidental.

While naturalism does not explain every motivation for MWH proposals, MWH are naturalistic answers to these observations. If naturalism is true, some version of MWH is a live option, given the alternatives. As a Christian and theist, I can think of no good reason to think any of the currently popular MWH are true (except perhaps for some principle of plenitude). Most currently proposed, relevant theories-even if true-probably would not produce all logically possible worlds.26 Many seem already to have been disproven, such as dissipative processes making a cyclic universe impossible.27 And the actual existence of an infinite set of entities like worlds may not even be possible.28

Many Worlds Hypotheses and the Ontological Argument

Since some good arguments against MWH and for theism already exist in the literature, I will not recount them here.29 Most of these arguments could be classified as either teleological or cosmological versions. Swinburne's point that Ockham's Razor alone should lead us to prefer theism to multiple worlds30 seems reasonable and correct to me, but of course this opinion does not aspire to a knock-down-drag-out argument for God's existence. In the remainder of this paper, I will call attention to some facets of MWH which come to light when we consider a form of the ontological argument for the existence of God, namely, the modal version formulated by Alvin Plantinga. The intersection of the ontological argument and MWH generate some questions which the infinite worlds enthusiast might wish to consider.

As noted above, the most popular MWH tend to employ the language of necessity, possibility, and infinity. Rather than argue for a large number of different worlds, Many Worlds theorists are fond of insisting on an infinite variety of worlds, of every logically possible world existing at some point. We also noted the motivation for such ontological multiplication: the set of all possible worlds enjoys a necessity which our contingent (visible) universe lacks. Arguments for the mere physically possible, where the same physical constraints and boundary conditions obtain in all universes, cannot answer the question, "Why these laws?"

Not surprisingly, some philosophers such as John Leslie cringe at how scientists offering MWH use the concepts like necessity and possibility.31 The (apparent) explanatory power of infinite worlds will always exceed any finite World Ensemble, and such an infinite set can accommodate any calculations for our world's improbability, no matter how immense. Leslie himself notes this value of explanatory power: "A chief reason for thinking that something stands in special need of explanation is that we actually glimpse some tidy way in which it might be explained."32 An infinite set has a certain pleasing "tidiness" that, say, 703 trillion lacks. Thus arguments for infinite worlds proliferate.33

We should note that the actual existence of all possible worlds plus the observational selection effect is crucial. All worlds are as actual as is ours, but those not conducive to the existence of observers will not be observed.34 The logically possible is the actual. If such were the case, it would certainly make our observation of this world less astounding. Of course, by definition we have no access to these other worlds, and the popularity of an infinite variety should serve as still more evidence for viewing MWH as metaphysically driven. Ironically, the use of all possible worlds by theorists may unintentionally raise some unwelcome questions such as the following: "Just what do `all logical possibilities' include?"

To consider that question adequately, a brief excursus is necessary concerning the ontological argument for the existence of God. The original version of this argument from St. Anselm states that God is that Being than which there can be no greater, the greatest possible being. As such, it should be clear that such a being must exist in reality, and not merely in our understanding alone. For if he existed merely in our understanding, then any existing being would be greater than he (since any existing being is greater than any mere theoretical but non-existent being). If that were the case, then there would be a being greater than the greatest possible Being, which is impossible. Therefore, God must exist in reality.35

Alvin Plantinga's modal version of [the ontological] argument [states]:...That which exists in some possible worlds but not others is logically possible, but its existence is contingent or accidental.

This frustratingly intriguing argument has usually been dismissed since Kant, who charged that Anselm erroneously made existence a predicate or a property.36 However, the truth of this "refutation" is far from obvious, and even if it were, it is not clear that Anselm's argument does treat existence as a predicate.37 Be that as it may, such criticism is irrelevant to Alvin Plantinga's modal version of this argument. Most simply, he restates it in terms of possible worlds. That which exists in some possible worlds but not others is logically possible, but its existence is contingent or accidental. It might not have existed. So the fact expressed by Bill Clinton is the President of the United States is contingent. It could possibly have been otherwise. That which exists in no possible world is impossible. It is necessarily non-existent. So a married bachelor or a square circle exists in no possible world. That which exists in every possible world, such as the truth that 2 + 2 = 4, is said to exist necessarily. There is no possible world in which it could not have existed.

Now imagine a being in whom maximal greatness is exemplified, who possesses whatever properties are great-making, such as power, knowledge, goodness, and so on. Such a being most traditional theists would identify as God (though this obviously would not exhaust God's properties or existence, especially for the Christian). Now clearly a being who possesses maximal greatness, if such greatness were to be maximal, would have to exist in every possible world, that is, he would exist necessarily. Note this does not make existence a property, since a non-existent being has no properties. Rather, existence is a condition for having properties: "So existence and necessary existence are not themselves perfections, but necessary conditions of perfection."38 Note also, that unlike most MWH, possible worlds here are not actual worlds, but are merely possible "states of affairsCpossible with respect to the actual world."39

MWH begin to look like little more than sophisticated ways to assert naturalism-at-any-cost

Since maximal greatness by definition will entail existence in every possible world, any being which possesses maximal greatness will be necessarily existent. Now consider the proposition: (1) Maximal greatness is possibly exemplified, i.e., there is a world W in which maximal greatness is exemplified. But, since maximal greatness entails necessary existence, if it is even possibly exemplified (in one or more worlds), then it is exemplified in every possible world; and this world being one among the set of all possible worlds, maximal greatness is instantiated in this world. Since we would designate God as the being possessing maximal greatness, if he exists as such in any possible world, he exists in every possible world, including this one. In short, (2) Maximal greatness is instantiated in some possible world entails the truth of (3) Maximal greatness is instantiated in this world.40

This argument is clearly valid, but is it sound? Since I believe God exists, I think it is. But how about the atheist? Is he compelled to believe in God on the basis of this argument? Probably not, because the atheist is free to deny the truth of (1) Maximal greatness is possibly exemplified. Of course, this is tantamount to arguing that the existence of God is impossible, that he is necessarily non-existent. Even the most rock-ribbed atheist will probably not want to make such a strong claim.

Having briefly considered this modal ontological argument, we can see its relevance for discussions about MWH. It leads us to ask: Just what is logically possible? In considering (1), there appear no obvious logical flaws in it. It is not self-contradictory. It does not violate any logical principles, excluded middle or otherwise. It surely appears to be a logically possible state of affairs. It is not even dependent on the far stronger claim of some Many Worlds theorists that every possible world is actual. Ironically, if (1) is true, then the strong MWH which make the possible actual-reifying the possible-may even strengthen this argument for God's existence, since no one could claim that it merely "defines God into existence"41 (a complaint I think is incorrect anyway for this version of the argument). This would obviously frustrate the aspirations of most Many Worlds theorists; but such can be the penalty for toying with notions such as possibility, necessity, and infinite sets. MWH, that strained appeal for naturalism-at-all-costs, may actually be complicit in hiding an argument for God's existence!

However, all this talk about all possible worlds being actual does not settle the most burning issue raised by the alternatives of MWH versus design by a Creator: "What is logically possible?" We still have the problem of compossibility. Some states of affairs which, taken on their own, may appear to be internally coherent and logically possible are incompatible with other such states of affairs when they are considered together. For instance, (4) A maximally perfect being exists on whom everything depends appears logically possible. But even its possible truth is incompatible with (5) A universe exists which is dependent on nothing outside it, even though (5) also appears to be logically possible when considered in isolation. However, the conjunction of (4) and (5) together is not a possible state of affairs.

So we are still left with the troubling question of whether (4) or (5) is the actually logically possible state of affairs. How one judges that will undoubtedly depend on how one judges the truth of some version of either theism or naturalism (if these are the two main options in this context). But such a judgment, while it may include evidence of apparent fine-tuning, cannot be deduced necessarily from such evidence, since both theism and MWH seem compatible with it. This is not to say that this evidence in conjunction with other "evidence"-whether historical, philosophical, empirical, experiential, revelatory, or otherwise-may not suggest the truth of some version of theism over naturalism (or vice versa), but only to say that the judgment concerning the possible truth of either (4) or (5) cannot be deduced (so that every reasonable person would have to agree) from the evidence of apparent fine-tuning. In this light, MWH begin to look like little more than sophisticated ways to assert naturalism-at-any-cost, since their defenders clearly do not include a designed universe in the supposedly infinite set of worlds. Why such discrimation? While MWH clearly attempt to work out the implications of naturalistic doctrine, their partisans fail if they aspire to assert an argument against another belief such as theism.


Thus, I conclude that while MWH may be a source of some solace for the atheist and naturalist, they present no compelling arguments for their truth over against theism in general or Christian belief in particular. Moreover, if MWH aspire to make theistic belief irrelevant or superfluous, they fail. At most, they are alternatives to such belief, when one is presented with the overwhelming evidence for apparent fine-tuning in the observable universe.

At the very least, we can conclude that the many popular MWH should be classified as much as metaphysics as physics, as their ability to serve as a substitute for theistic belief indicates. The Christian and theist who see such hypotheses simply as scientific theories which will be discarded at the first sight of falsifying evidence will fail to recognize their metaphysical character.

Perhaps most suspiciously, MWH create a condition whereby no amount of evidence, no matter how intricate, could ever serve as evidence of a Creator, since in a set of all possible worlds, every possible state of affairs will be actualized somewhere, no matter how improbable, intricate or complex. Even a world like our own, but which abruptly emerged in, say, six days enjoys membership in the club of possible worlds; so even that world would be no evidence for a Creator. The explanatory mischief of MWH begins to appear a little strong. The naturalist committed to a MWH must maintain that (1) the existence of God (a maximally perfect being) is logically impossible, and that (2) no amount of evidence for apparent design could count as evidence for actual design. Under these restrictions, an omnipotent Creator would be incapable-short of vetoing human cognitive freedom-of leaving traces of his existence in the physical world which could be recognized as such. This seems a little stubborn.42



1See William Lane Craig and Quentin Smith, Theism, Atheism, and Big Bang Cosmology (Oxford: Clarendon Press, 1993), 3B76, for more on this point.

2For an extensive description of many of these, see John D. Barrow and Frank J. Tipler, The Anthropic Cosmoloical Principle (New York: Oxford Univeristy Press, 1966), 219B556.

3In William Lane Craig, "The Teleological Argument and the Anthropic Principle," The Logic of Rational Theism, ed. Craig and MacLeod (Lewiston: Edwin Mellen Press, 1990), 129.

4Stephen W. Hawking, A Brief History of Time (New York: Bantam Books, 1988).

5Roger Trigg, Rationality and Science (Oxford: Blackwell, 1993), 184B9.

6Stanley Jaki, God and the Cosmologists (Washington, DC: Regnery Gateway, 1989), 96.

7I owe this point to an anonymous referee of this journal.

8Barrow and Tipler, 1B2.

9Ibid., 3, 565.

10John Leslie, Universes (London: Routledge, 1989), 13B14.

11For more discussion of extraterrestrial and alternate life forms, see Barrow and Tipler, 576B677, and Leslie, 11ff.

12Leslie, 131B2.

13Paul Davies, The Mind of God (New York: Simon and Schuster, Touchstone, 1992), 190.

14Leslie, 69B70.

15Ibid., 6B8.

16I owe this point to a referee for this journal.

17C. Pantin, quoted in Barrow and Tipler, 83.

18Davies, 219.

19Barrow and Tipler, 105.

20Ibid., 7.

21Leslie, 140.

22Barrow and Tipler, 496.

23Craig, 146.

24Leslie, 192.

25Ibid., 102.

26See Quentin Smith, "World Ensemble Explanations," Pacific Philosophical Quarterly 67 (January 1986): 73B86.

27Again I owe this detail to a referee for this journal.

28For an excellent argument on the impossibility of the existence of an actual infinite set, see William Lane Craig, The Kalam Cosmological Argument (Barnes & Noble, 1979).

29See, for example, W.<|>L. Craig in Craig, McLeod, 127B53; Craig and Smith, 3B76, 92B107; 141B60, 218B31, 279B300; Robert Prevost, Probability and Theistic Explanation (Oxford: Clarendon Press, 1990); and Richard Swinburne, "Argument from the Fine-Tuning of the Universe," in Physical Cosmology and Philosophy, ed. John Leslie (New York: Macmillan Pub. Co., 1990), 154B73.

30Swinburne, 166B72.

31Leslie, 129.

32Ibid., 129.

33Examples are common and blatant: "[I]t is argued that all logically possible universes exist in an ensemble of disjoint universes." In D.<|>W. Sciama, "The Anthropic Principle and the Non-Uniqueness of the Universe," in The Anthropic Principle, ed. F. Bertola and U. Curi (Cambridge: University Press, 1993), 107.

34Leslie, 14, 94, 123B4, and Sciama, 107.

35See the reprint of Anselm's argument in Louis P. Pojman, ed., "The Ontological Argument," Philosophy of Religion: An Anthology (Belmont: Wadsworth Pub. Co., 1987), 51B3.

36See Kant's charge in "A Critique of the Ontological Argument," Ibid., 53B7.

37Alvin Plantinga, The Nature of Necessity (Oxford: Clarendon Press, 1974), 196B7; CCC, God and Other Minds (Ithaca: Cornell University Press, 1967; reprint 1990), ch. 2.

38Plantinga, The Nature of Necessity, 214.

39Ibid., 54.

40Ibid., 213B7.

41Davies, 188.

42Such intransigence might even be irrational, since we clearly infer design in mundane circumstances. But MWH logic could be used to rule out all inferences that will end with the intentions of an intelligent agent, even in the human sphere. For the structural properties of inferring design, see William A. Dembski, The Design Inference: Eliminating Chance through Small Probabilities. Doctoral Dissertation, University of Illinois at Chicago, 1996.