Part IV Cosmology and Physics

Does God Love the Multiverse?

DON N. PAGE

A central point of the Judeo-Christian tradition is that God loves everyone. One of the most famous verses in the Bible is in the Gospel of John: “For God so loved the world that he gave his only begotten Son, that whoever believes in him should not perish but have everlasting life” (John 3:16). Furthermore, according to this tradition, God not only loves all people, but he also created them in his own image. As it is written in the Book of Genesis: “So God created man in his own image; in the image of God he created him; male and female he created them” (Gen. 1:27). For us to receive both the love and the image of God, it seems that humanity is unique in some sense – different from the rest of creation.

In light of modern science, however, we have to ask how unique we really are. Some have taken the image of God for humans to imply that God created us individually and separately from other living beings. This is challenged by Darwin’s theory of evolution, which suggests that we were not created separately by an individual act, independent of the creation of the remainder of the earth’s biosphere. Rather, we are related to the rest of life on earth. A new challenge to the beliefs of some Christians has arisen from the idea of a multiverse, that our universe is not unique but instead is one among many.

Parallels between Evolution and Multiverse Ideas

When Darwin proposed evolution, many conservative Christians accepted it as not necessarily contrary to Christianity. One famous example was Benjamin B. Warfield (1851–1921), the conservative Christian theologian and principal of Princeton Seminary from 1887 to 1921. Warfield wrote the chapter on “The Deity of Christ” in The Fundamentals, the source of the term “fundamentalism.” Thus one of the most famous original fundamentalists accepted the possibility of Darwinian evolution, writing:

I am free to say, for myself, that I do not think that there is any general statement in the Bible or any part of the account of creation, either as given in Genesis 1 and 2 or elsewhere alluded to, that need be opposed to evolution.

(Warfield 2000, 130)

However, many Christians later came to oppose evolution, perhaps most famously some other fundamentalists. Although this has a complex history which cannot be told here, one of the central reasons for resistance to evolution by some Christians is their belief that evolution removes one particular design argument for the existence of God, the argument that all of the marvelously many different species of living things on earth had been separately designed and created by God. Nevertheless, evolution did not disprove the existence of God or even of some overall design. Indeed, there are many leading theologians and scientists today who accept both evolution and creation by God.

It seems to me that there may be a parallel development occurring today in cosmology. Before Darwin, some Christians took the marvels of humanity as evidence of separate and individual design. Now, some Christians take the marvels of the fine-tuning of the physical constants of the universe as evidence of theism and often of separate and individual design of these constants by God. Here I wish to argue that this approach to cosmology could be just as mistaken as the approach of those who held to the separate and individual design of human beings. Embracing the multiverse is not yet a popular view among theists, but a minority of theists (e.g., Leslie 1989, 64–65; Leslie 2001, 211–214; Collins 2002; Barr 2003; Swinburne 2011) have broken with tradition and argue that a multiverse could reveal an even more grand design of the universe. If our observable universe is only a part of a much more complex structure, the physical process that would generate such a “multiverse” would have to have suitable basic laws and initial conditions to produce any life at all. The laws and initial conditions would apparently have to be even more special to produce not just life, but life like ours, in view of the order we actually do see around us. Stephen Barr, Gerald Cleaver, Robin Collins, Klaas Kraay, John Leslie, Richard Swinburne, and others claim that since God is infinitely creative, it makes sense to say that he might create a physical reality much larger than the single visible part of the universe or multiverse that we can observe directly (see, e.g., Leslie 1989, 180; Leslie 2001, 18–19, 215–216; Barr 2003, 151–153; Collins 2007, 460–462; Kraay 2010, 361–366). To understand this reasoning, we need to address the apparent fine-tuning of the physical constants in our universe.

Fine-Tuning in Our Universe

Now it does seem to be true that we could not be here if many of the constants of physics were significantly different, so that in our part of the universe, the constants of physics do in fact seem to be fine-tuned for our kind of life. This is generally agreed upon both by those who attempt to use this fine-tuning to support theism (as in Swinburne 2011), and by many scientists who are usually neutral or opposed to such an attempt (e.g., Carter 1990, 125–133; Carr and Rees 1979, 605–612; Davies 1982; Barrow and Tipler 1986; Rees 2000; Barrow 2002; Bostrom 2002; Susskind 2006; Carr 2007). Of course, no one knows what other forms of life might be possible if the constants of physics were significantly different, but the general consensus seems to be that it would be very difficult to imagine the possibility of any complex life at all existing if certain combinations of the constants of physics were greatly different.

For example, one of the most remarkable fine-tunings is the value of the cosmological constant or energy density of the “dark energy” responsible for the current acceleration of distant galaxies away from each other. Measurements show that the cosmological constant is more than 120 orders of magnitude smaller than unity in certain natural units (called Planck units, obtained by setting to unity the speed of light, Planck’s quantum constant of action, and Newton’s gravitational constant). With the other constants kept fixed, it would be difficult to have a universe with gravitationally formed structures lasting long enough for life if the cosmological constant were even just a few orders of magnitude larger than its observed value. But even if one tuned the other constants to allow the possibility of such structures when the cosmological constant has a value many orders of magnitude larger than its observed value, one still seems to need it to be many orders smaller than unity. So it is hard to see how to avoid at least some significant amount of fine-tuning of this parameter. (Basically, if the cosmological constant were of the order of unity in the natural Planck units, the spacetime of the universe would always have large quantum mechanical fluctuations, and no one knows any plausible way to have persisting complex structures that one could call life in such a case.)

Another constant that is many orders of magnitude away from unity, in this case about 36 orders of magnitude larger than unity, is the ratio of the electrostatic repulsion to the gravitational attraction between two protons (the nuclei at the centers of hydrogen atoms). With other constants kept fixed, it seems that one could not have the types of stars that appear to be necessary for life if this constant differed by much more than even one order of magnitude from its actual value. Again one could try to imagine a universe hospitable to some other form of life when this constant is significantly different by also tuning other constants to an appropriate range, but again it seems that complex life of any form relying mainly on the electromagnetic and gravitational forces would be impossible if this constant were close to unity. (Then it seems that one could not have stars, planets, and living organisms with large numbers of atoms, since the number of atoms in such structures generally scales as a positive power of this constant and would approach some small number near unity if this constant were itself near unity.)

Martin Rees (2000) discusses in much more detail these two constants and four others in our universe that are crucial for its properties. Life as we know it would apparently be impossible if anyone of them were greatly different (with the others held fixed). So although it might not be necessary for all of them to have their observed values, there are some combinations of them that apparently could not be very much different and yet give a universe with life, at least life at all similar to present life on earth. So we need some explanation for why these values are the way they are.

Explanations for Fine-Tuning

There is a general consensus that there appears to be “fine-tuning” of some of the constants of physics in our part of the universe. That is to say, if these constants had been sufficiently different, life as we know it could not have existed. So what is the explanation for this phenomenon? Three general types of explanations are often put forward.

Some suggest that the fine-tuning was done by a separate act of God to allow life. Others say that it is accidental, a fluke. And yet others propose that it arises from a huge multiverse of very many different possible constants of physics. It is also noted in several sources (such as Leslie 1989, 22; Leslie 2001, 211; Bostrom 2002, 11; Barr 2003, 153–154; Carr 2007, 16–17, 27, 411–412, 459–480) that the three explanations are not mutually exclusive, so that virtually any combination of them is logically possible. Also it should be noted that each of these three explanations really stands for a class of explanations, so that one should actually compare specific proposals taken from these classes rather than the classes themselves. For example, theists of different theological convictions might propose different ideas of how God would choose the constants. Those saying that the fine-tuning is a fluke might say that the constants are determined by any number of different mathematical structures that just happened to give biophilic values, or they might propose that there is truly some random process determining the constants in some way not derivable from any simple mathematical structure. And of course there are a huge number of possible multiverse theories. These theories are rapidly growing in favor, though not without a lot of opposition from both theists and nontheists.

Some multiverse theories seem to me to be too general to be plausible, such as the idea of David Lewis (1990) that all logical possibilities actually exist, or the original idea of Max Tegmark (1998, 1–51) that all mathematical structures have physical reality. These seem to leave it unexplained why what we see has the order that it does, whereas a random possibility from all logical possibilities or from all mathematical structures would surely be far more chaotic (Leslie 1989, 97–98; Leslie 2001, 23–30; Barr 2003, 156; Vilenkin 2006, 203). However, there might be other multiverse theories that better explain the order we observe, perhaps arising naturally out of elegant but specific laws of nature.

One natural way to get a multiverse is to have a universe so large that highly varied conditions occur somewhere. Another is from Everett many-worlds (DeWitt and Graham 1973), that all the quantum possibilities are actually realized. However, those possibilities do not necessarily give varying constants of physics.

One scenario that seems more hopeful is to get multiverses from inflation (Linde 1990, 292–317; Guth 1997, 245–252; Vilenkin 2006, 203–205), which is a very rapid exponential expansion of the early universe that may make the universe enormously larger than what we can observe of it. If the inflationary scenario can include phase transitions, and if the constants of physics can differ across phase transitions, inflation tends to produce all such possibilities.

Recently it has been realized that string/M theory apparently leads to a huge multiverse of 10⁵⁰⁰ or so different vacua or sets of constants. This would be enough for the constants we see to randomly occur somewhere (maybe once per 10²⁰⁰ vacua or so). Then perhaps 10³⁰⁰ or so vacua would reasonably allow for what we see.

If only one universe in 10^M could fit our observations, but if 10^N different universes exist in the multiverse, then it might not be surprising that what we observe exists if N > M. So if N is around 500 as suggested by string/M theory and M is around 200, so then indeed N > M. However, the actual numbers are scarcely known. We really don’t yet know whether N > M in string/M theory, but it does seem plausible. If this is the case, then what we see could be explained without its having to be individually selected.

One might still ask whether the multiverse explanation always works, assuming that it has enough universes (e.g., N > M). Is it sufficient to explain what we see by a multiverse theory in which there are enough different conditions that ours necessarily occurs somewhere? I would say not. Rather, it seems to me there is the further requirement that the conditions we observe should not be too rare out of all the conditions that are observed over the entire multiverse. A theory making our observations extremely rare should not be considered a good theory.

Good theories should be both intrinsically plausible and fit observations. Intrinsic plausibility is quantified by what is called the a priori probability of the theory, the probability that one might assign to it from purely theoretical background knowledge, without considering any observations. The fit to observations is quantified by the conditional probability of the observation given the theory, what is called the likelihood. Then the probability of the theory after taking into consideration the observation, what is called the a posteriori probability of the theory, is given by Bayes’s theorem as being proportional to the product of the a priori probability and the likelihood.

I take the a priori probabilities of theories (intrinsic plausibilities before considering the observations) to be subjective but to be generally assigned higher values for simpler theories, by the principle known as Occam’s razor. However, the recognition of simplicity depends on one’s background knowledge, which itself depends on the laws of physics.

The likelihood of a theory is itself neither the a priori nor the a posteriori probability of the theory, but rather the conditional probability, not of the theory, but of the observation given the theory. A theory that uniquely gives one’s observation would have unit likelihood but might have very low a priori probability.

For example, consider an extreme solipsistic theory that only one’s actual momentary observation exists, not anyone else’s or even any of one’s own in either the past or the future, and perhaps not even that an external world exists at all. This theory would predict that observation with certainty if it were correct. (If the theory were true, certainly the observation would be that single one predicted by the theory.) Therefore, for that observation the likelihood is unity. However, such an extreme solipsistic theory, giving all the details of one’s observation or conscious perception without an external world giving other observations, would surely be highly complex and so would be viewed as extremely implausible, much more implausible than an alternate theory in which the observation resulted from the existence of an external world that also gives other observations. Therefore, this extreme solipsistic theory would be assigned very low a priori probability.

At the other extreme, consider the simple theory that predicts all possible observations equally (arguably a consequence of something like the modal realism of David Lewis (1990)). Since this theory is so simple, it might be assigned a high a priori probability, but then because of the enormous number of observations it predicts with equal probability (presumably infinitely many), it would give very low likelihood (presumably zero).

The Growth of Our Knowledge of the Universe

Our whole growth of knowledge about the universe has been an expansion of its scope. As one grows as an infant, one rapidly grows beyond the view that one’s present observation is all that is real, as one develops memories about the past and anticipations of the future. One then goes beyond solipsism and gains an understanding that other persons or observers exist as well. In the early stage of human development, there was the focus on one’s family, which was then gradually extended to one’s tribe, one’s nation, one’s race, and, one might hope, to all humans. But then when one further considers what other conscious observations may be going on, one might well believe that consciousness extends to other creatures, such as other animals.

Of course, one’s direct observation never extends beyond one’s own immediate conscious perception, so one can never prove that there are past or future perceptions as well (and I know philosophers who do not believe the future exists). Similarly, one can never directly experience even the present conscious perceptions of another, which engenders the problem of other minds in philosophy. Nevertheless, most of us believe that we have fairly good indirect evidence for the existence of other conscious experiences, at least for other humans on earth with whom we can communicate, though it is logically possible that neither they nor any external world actually exists. (My belief is that it is much simpler to explain the details of my present observation or conscious perception or experience by assuming that an external world and other conscious experiences also exist, than by assuming that just my own momentary conscious perception exists.)

We may now extend the reasoning to suppose that if the universe is large enough, it will also include conscious extraterrestrials, even though we do not have even indirect evidence for them that is so nearly direct as our (inevitably still indirect) evidence for other conscious beings on earth. We can further theorize that if the universe is so large that there never will be any contact between its distant parts and our part, there still might be other conscious beings not in causal contact with us, so that we never could communicate with them to get, even in principle, the indirect evidence of the same qualitative nature that we have for other humans here on earth with us.

A next step might be to postulate conscious beings and experiences in other universes totally disconnected from ours, so that even if one could imagine traveling faster than the speed of light, there would simply be no way to get there from here; the two parts would be in totally disconnected spacetimes. A similar situation would occur for putative conscious experiences in other branches of an Everett “many-worlds” wavefunction or quantum state. From accepting the existence of such disconnected observers, it hardly seems an excessive additional step to imagine observers in universes or parts of the multiverse with different constants of physics. One might even imagine observers in entirely different universes, not related to ours in the way an entire multiverse might be related by having one single overarching set of natural laws.

So in this sense, the idea of a multiverse seems to be rather a natural extension of our usual ideas of accepting a reality beyond one’s immediate conscious perception, which is all the experience for which one has direct access. All the rest of one’s knowledge is purely theoretical, though one’s brain (assuming it exists) is apparently constructed to bring this knowledge into one’s awareness without one’s needing to be consciously aware of the details of why one seems to be aware of the existence of other conscious beings.

Despite the naturalness of the progression of ideas that leads to multiverse theories, there are various objections to the multiverse theory. However, none of the objections seem to me to be convincing, as there are highly plausible rebuttals to the objections.

Objections to Multiverse Ideas

A scientific objection to a multiverse theory might be that the multiverse (beyond our observed part, which is within one single universe) is not observable or testable. But if one had precise theories for single universes and for multiverses that gave the distributions of different conditions, one could make statistical tests of our observations (likely or unlikely in each distribution). Unfortunately, no such realistic theory exists yet for either a single universe or a multiverse, so I would agree that at present we simply do not have any good theories for either to test.

Another objection is that a multiverse is not a clear consequence of any existing theory. Although it is beginning to appear to be a consequence of string/M theory, that is not yet certain, which is why there can be theorists like David Gross who are still holding out hope that string/M theory might turn out to be a single-universe theory after all, possibly enabling theorists (if they could perform the relevant calculations) to fulfill their wildest dreams of being able to calculate the constants of physics uniquely from some simple principles. One first needs to make string/M theory into a precise theory and calculate its consequences, whether single universe or multiverse. And if that theory gives predictions that do not give a good statistical fit to observations, one needs to find a better theory that does.

A philosophical objection to a multiverse theory is that it is extravagant to assume unfathomable numbers of unobservable universes. This is a variant upon the psychological gut reaction that surely a multiverse would be more complex than a single universe, and hence should be assigned a lower a priori probability. But this is not necessarily so. The whole can be simpler than its parts, as the set of all integers is quite simple, certainly simpler than nearly all the (arbitrarily large) individual integers that form its parts.

As a further rebuttal of the accusation of extravagance, a theist can say that since God can do anything that is logically possible and that fits with God’s nature and purposes, then there is apparently no difficulty for God to create as many universes as he pleases. God might prefer elegance in the principles by which he creates a vast multiverse over paucity of universes, that is, economy of principles rather than economy of materials.

Another philosophical objection to multiverses is that they can be used to explain anything, and thereby explain nothing. I would strongly agree with this criticism of multiverse theories that are too vague or diffuse, which do not sufficiently restrict the measure on the set of observations to favor ordered ones such as what we observe. There is a genuine need for a multiverse theory not to spread out the probability measure for observations so thinly that it makes our observation too improbable. So this objection would be a valid objection to vast classes of possible multiverse theories, but I do not see that it is an objection in principle against a good multiverse theory. Certainly not just any multiverse theory is acceptable, and even if simple single-universe theories do not work for explaining our observations, it will no doubt be quite a challenge to find a good multiverse theory that does succeed.

Most of the objections I have raised and attempted to answer so far would apply both to theistic and nontheistic scientists. However, if one is a theist, one might imagine that there are additional objections to multiverse theories, just as some theists had additional objections to Darwin’s theory of evolution beyond the scientific objections that were also raised when that theory had much less support.

For example, a theist might feel that a multiverse theory would undercut the fine-tuning argument for the existence of God. I shall not deny that it would undercut the argument at the level of the constants of physics (though I think there would still be such a design argument from the general apparently elegant structure of the full laws of nature once they are known). However, the loss of one argument does not mean that its conclusion is necessarily false.

I personally think it might be a theological mistake to look for fine-tuning as a sign of the existence of God. I am reminded of the exchange between Jesus and the religious authorities recorded in the Gospel of Matthew (12:38–41):

Then some of the scribes and Pharisees answered, saying, “Teacher, we want to see a sign from you.” But He answered and said to them, “An evil and adulterous generation seeks after a sign, and no sign will be given to it except the sign of the prophet Jonah. For as Jonah was three days and three nights in the belly of the great fish, so will the Son of Man be three days and three nights in the heart of the earth. The men of Nineveh will rise up in the judgment with this generation and condemn it, because they repented at the preaching of Jonah; and indeed a greater than Jonah is here.”

In other words, I regard the death and Resurrection of Jesus as the sign given to us that he is indeed the Son of God and Savior he claimed to be, rather than needing signs from fine-tuning.

Another theistic objection might be that with a multiverse explanation of the constants of physics, there is nothing left for God to design. But God could well have designed the entire multiverse, choosing elegant laws of nature by which to create the entire thing. In any case, whatever the design is, whether a logically rigid requirement, a simple free choice God made, or a complex free choice God made, theists would ascribe to God the task of creating the entire universe or multiverse according to this design.

A third more specifically Christian objection might be that if the multiverse (or even just our single part of the universe) is large enough for other civilizations to have sinned and needed Christ to come redeem them by something similar to his death on the cross here on earth for our sins, then his death may not sound as unique as the Bible says it is in Romans 6:10: “For the death that he died, he died to sin once for all; but the life that he lives, he lives to God.” The Bible, however, was written for us humans here on earth, so it seems unreasonable to require it to describe what God may or may not do with other creatures he may have created elsewhere. We could just interpret the Bible to mean that Christ’s death here on earth is unique for our human civilization.

Conclusion

In conclusion, multiverses are serious ideas of present science, though certainly not yet proven. They can potentially explain fine-tuned constants of physics but are not an automatic panacea for solving all problems; only certain multiverse theories, of which we have none yet in complete form, would be successful in explaining our observations. Though multiverses should not be accepted uncritically as scientific explanations, I would argue that theists have no more reason to oppose them than they had to oppose Darwinian evolution when it was first proposed.

God might indeed love the multiverse.

References

Barr, Stephen. 2003. Modem Physics and Ancient Faith. Notre Dame, IN: University of Notre Dame Press.

Barrow, John D. 2002. The Constants of Nature. New York: Pantheon Books.

Barrow, John D. and Tipler, Frank J. 1986. The Anthropic Cosmological Principle. Oxford: Clarendon Press.

Bostrom, Nick. 2002. Anthropic Bias: Observation Selection Effects in Science and Philosophy. New York: Routledge.

Carr, Bernard, ed. 2007. Universe or Multiverse? Cambridge: Cambridge University Press.

Carr, Bernard J. and Rees, Martin J. 1979. The Anthropic Principle and the Structure of the Physical World. Nature, 278, pp. 605–612.

Carter, Brandon. 1990. Large Number Coincidences and the Anthropic Principle in Cosmology. In J. Leslie, ed. Physical Cosmology and Philosophy. New York: Macmillan, pp. 125–133.

Collins, Robin. 2002. Design and the Many-Worlds Hypothesis. In William Lane Craig, ed. Philosophy of Religion: A Reader and Guide. New Brunswick, NJ: Rutgers University Press, pp. 130–148.

Collins, Robin. 2007. The Multiverse Hypothesis: A Theistic Perspective. In Bernard Carr, ed. Universe or Multiverse? Cambridge: Cambridge University Press, pp. 459–480.

Davies, Paul. 1982. The Accidental Universe. Cambridge: Cambridge University Press.

DeWitt, Bryce and Graham, R. Neill, eds. 1973. The Many-Worlds Interpretation of Quantum Mechanics. Princeton, NJ: Princeton University Press.

Guth, Alan. 1997. The Inflationary Universe: The Quest for a New Theory of Cosmic Origins. Reading, MA: Addison-Wesley.

Kraay, Klaas J. 2010. Theism, Possible Worlds, and the Multiverse. Philosophical Studies, 147, pp. 355–368.

Leslie, John. 1989. Universes. London: Routledge.

Leslie, John. 2001. Infinite Minds: A Philosophical Cosmology. Oxford: Oxford University Press.

Lewis, David. 1990. On the Plurality of Worlds. Cambridge: Blackwell.

Linde, Andrei. 1990. Particle Physics and Inflationary Cosmology. Chur, Switzerland: Harwood.

Rees, Martin. 2000. Just Six Numbers: The Deep Forces that Shape the Universe. New York: Basic Books.

Susskind, Leonard. 2006. The Cosmic Landscape: String Theory and the Illusion of Intelligent Design. New York: Little, Brown & Company.

Swinburne, Richard. 2011. Bayes, God, and the Multiverse. In J. Chandler and V. Harrison, eds. Probability in the Philosophy of Religion. Oxford: Oxford University Press.

Tegmark, Max. 1998. Is “The Theory of Everything” Merely the Ultimate Ensemble Theory? Annals of Physics, 270, pp. 1–51.

Vilenkin, Alex. 2006. Many Worlds in One: The Search for Other Universes. New York: Hill & Wang.

Warfield, Benjamin Breckinridge. 2000. Evolution, Science and Scripture: Selected Writings. Mark A. Noll and David N. Livingstone, eds. Grand Rapids, MI: Baker Books.