God is not a Good Theory (Sean Carroll)

[Music] [Music] it's very good to be here. um, as you can see from my title, I'm going to be approaching the question of the role that the idea of God has to play in explaining the universe we see from the point of view of a scientist, a physicist, or a cosmologist. now, not the point of view of a professional theologian or philosopher. I do think that the, uh, point of view of a physicist or a scientist is relevant to this question. it is not the only possible point of view one can have. there are aspects of the idea of God that do not reduce to the roles played by a simple scientific theory, but there are also aspects of the idea of God that do play the roles of a scientific theory, and those aspects can be judged by the same criteria that we use to judge theories. in the original schedule, if you have seen that, uh, it it removed the word "good" from my title, so it just said "God is not a theory." but that is the opposite of what I'm trying to say here. so, just so you, uh, know my point of view, I'm saying that God is a theory. the idea of God, uh, to the extent that almost all believers use the idea of God, does have aspects of it that can be judged in precisely analogous ways to the way that we judge, uh, an ordinary scientific theory.

now, I'm very well aware there is a, a thread of thought that says the opposite of that, that says that the idea of God is something separate. it's a different kind of idea than the ideas that we use when we have physics theories. uh, so, you know, God is not reproducible or something like that. so I just want to get around that particular definitional problem by putting forward the idea of "theory" is something very, very broad. I don't want to pick some very, very specific philosophy of science definition of a theory. to me, a theory is just an idea about the universe that may be true or false, and that we should try to figure out how to judge whether or not this particular idea is true or false. um, so, one of the problems with God as a theory is that it is not a very precisely specified theory. that is going to be one of my points, is that, uh, even though it we sometimes try to have God play the role of a physical theory, making predictions and explaining things, pres, uh, preserving some explanatory role, I'm going to argue that it doesn't do a very good job.

now, again, I will, uh, even though I'm not a trained theologian, I certainly know that we have different conceptions of God. in fact, this is what I would argue is one of the check marks against God as a very good theory, is that if you try to say, "well, here's why I don't believe in God," someone else will say, "ah, you just don't understand what God is. God is completely different than that." and I recognize that. I will try to at least explain which versions of God I am talking about. I can't guarantee they are your favorite versions of God. uh, so that we can pick out three. that's not an exclusive list: passive, active, and emerging conceptions of God.

and what I mean by this, under passive, is the kind of roles that are given to God, going back to some of the ancient Greeks, going back at least to Aristotle, a sort of philosophical role for God as something that helps explain, helps sustain the universe, an unmoved mover, a first cause, a necessa, a necessary being, various philosophical notions that are not necessarily empirical in the same way that scientific theories are. this kind of God is one that just is, uh, something that we need in this conception to make sense of the universe, not something that necessarily is is going around poking around in the universe and intervening in it and changing the physical laws.

the active conception of God is a more personal idea. uh, often people will define God as a kind of person, just as often they will say that God is not a person. but there's this conception of God as someone who cares about life here on Earth, cares about human beings, cares both in the sense that, you know, he wants us to succeed, but also sometimes he judges us, and sometimes a little harshly. uh, this, it depends on how far you want to go, but God will, um, pass down moral guidance, will set up an afterlife for you that, you know, you'll be sorted once you die, and so forth. there's various uses that this sort of more active God has, uh, to perform.

and finally, there's, there's what I like to call an emergent idea of God, which is to say that God, not as a fundamental category, but God as a way of talking about aspects of the universe that you could also choose to talk about in different ways. God is the laws of nature, God is the orderliness of the cosmos, God is the universe itself, or God is love, or God is our feeling of awe and transcendence as we approach the universe. and these different conceptions of God have different justifications for them. uh, the passive notion of God, just in the background, can be attempted to be justified on purely logical grounds, just by thinking about how the universe might be. the more active conception of God is going to be essentially an empirical argument that the best way to explain the universe we see is to imagine that God is doing these things in an active sense. whereas the emergent idea of God is a more rhetorical move, and I don't mean that in any disparaging way. it's a way of talking about the universe that we see.

so, I'm actually not going to be talking about this third category. my personal belief, for those of you who are fascinated by it, is that I don't see any need to talk about the universe in that vocabulary, but, you know, if you want to, I won't stop you. what I'm more interested in for the pur, purposes of this talk are the, the, the more, uh, the conceptions of God that introduce God as a separate metaphysical category that we have something to do with. and I'll start with, you know, these more philosophical metaphysical arguments in favor of God. I think that as a scientist, the second category is more relevant, but I do have to get this one out of the way before we can move on to the second one.

so, what does it mean, and what is, what is the role played by God as an unmoved mover, as an Aristotelian first cause, uh, God as a necessary being? very often, if you ask theologians, you know, "how would the universe be different if God did not exist?" they will say, "I cannot imagine a universe in which God does not exist, therefore I cannot answer that question." to me, as a scientist, there's a huge problem right from the start with this kind of reasoning, which is that it is the whole strategy is one based on some a priori metaphysics. that is to say, this is, you know, very much armchair philosophizing in, uh, you know, in the best sense of the word, sitting down, thinking about all the possible ways the world can be, and concluding that those ways must somehow involve the idea of God. you don't ever, there's no step in that process in which you actually go out and look at how the universe actually is. you don't need to in this way of thinking about it, because you can just argue logically that God must be part of the universe. it is my firm belief that this kind of reasoning has never taught us anything true and interesting about the actual world. uh, that is not to say that armchair thinking without going out and looking at the world is not useful in any way. it's extremely useful. mathematics, logic, other, uh, branches of philosophy and formal inquiry are not empirical in nature. they don't involve going out and looking at the actual world. they, they reason in an a priori sense, but they also don't reveal interesting truths about the actual world. mathematics reveals consequences of axioms. you say, "I have a certain axiomatic, uh, structure, and I derive theorems on the basis of that." it doesn't tell you which axioms are possibly true. if you want to actually figure out our universe, does it involve some notion of God that is an actual fact about the specific universe in which we live? and I personally don't think that this a priori kind of reasoning is ever going to, uh, get us there.

but let me take it seriously for just a little bit and sort of give you my feelings as a scientist about this. how would we, we refute the claim that God is a necessary being, for example? and I, and I should also admit that I'm happily conflating all sorts of ultimately different kinds of a priori metaphysical arguments: first cause, prime mover, necessary being. these are all subtly different in ways that I'm going to completely ignore. so, sorry about that. uh, they have a similar flavor, and I'm going to treat them similarly. so, if you did believe that God was a necessary being, that you literally could not imagine a universe in which God did not play an important role, in order logically for me to refute that belief, all I need to do is to invent a universe in which God does not exist. you know, if you say, uh, that all swans are white as a logical fact, all I need to do is show you one black swan to show that your logical argument can't be right. so, all we need to do is invent a universe that God does not play a role in, and that would improve your imagination enough so that you could imagine universes where God was not important.

I claim that it is not hard to invent universes, possible universes, hypothetical universes, not necessarily our own, but universes that are self-contained, consistent, coherent, and in which God plays no special role, no role at all. so, here's the universe for, uh, apologies for the math. this universe is three-dimensional space, uh, evolving in time, and in that space, there is one particle that is the whole universe. this particle moves according to Newton's laws of motion in the under the influence of some potential energy. so, here's space, there's time, there's our particle moving. it has an equation of motion, and this goes on forever into the infinite far past or the infinite far future. I claim this is a logically conceivable universe. you can't stop me from imagining that this is the universe. it's clearly not our universe. I'm not trying to claim that there is some complicated transformation that makes our real world look like this, but this could be a universe, and there is no God in this universe. there's just a particle moving.

if you want a slightly more realistic example, uh, here is a universe that I think might actually be right. uh, I replace instead of three-dimensional space, I have a space of states, which is some Hilbert space. that is to say, some quantum mechanical space of possible states, a space of wave functions, if you like. there's also time evolution, so the state is an element of that Hilbert space, and it evolves with time. if you want to know how it evolves, well, it solves this equation, Schrodinger's equation, where this capital H is an operator called the Hamiltonian, which basically tells you how the state evolves with time. so, this universe is one that is mathematically isomorphic to a, a, uh, trajectory in a Hilbert space according to this Schrodinger equation. it goes forever from the infinite far past to the infinite far future. it is self-contained. there's no something outside this universe that is holding it up, sustaining it, keeping it going, causing it to exist, or allowing it to persist. it is simply that mathematical structure, full stop. I actually, like I said, think that this is a plausible mathematical structure for the real world in which we live. clearly, there are some details to be filled in. I need to know what the actual Hamiltonian of the world is, which I don't claim to know, but this is a plausible framework for reality.

and just to give you a final idea of what kinds of universes I'm thinking about here, is a world, a universe, it's one point. this is actually bigger than a point because it's hard to show you a point, so this is a circle, but imagine a mathematical structure that is isomorphic to a single point. I claim that is a possible universe. once again, it is not our universe in any interesting way, but it is a conceivable universe, and once again, God plays no role.

so, what am I trying to get at here? what I'm trying to get at is that the, if you sat down again, a priori, and tried to think about all the possible ways that existence could manifest itself, we know that to at least some extent, there are mathematical ways of talking about existence, talking about the real world. I can easily imagine self-contained and consistent mathematical structures, uh, in which the notion of God or necessary beings or first cause, first causes, play no role whatsoever.

now, I'm, you know, this is not some brilliant new insight. people who do believe in necessary beings and so forth have a response to this. they will say, "sure, you can imagine that, but they're not really legitimate possibilities for what the world can be like." and that's because there's no explanation, fundamentally, there's no reason why the world is like that. the role of God is to ground why this world rather than some other, or indeed, why this world at all, why a world at all. so, the, the response to my examples would be, "sure, you can mathematically write those down, but you haven't explained why that would be the universe." and my response to those arguments is, "so what? who is to say that I'm supposed to explain why there is a universe?" you might prefer that there be an explanation as to why this universe rather than that universe, but you can't prove logically that there must be some explanation. the universe could just be. that is the fundamental, uh, disagreement between people who believe in a metaphysical notion of God and the more empirically minded scientists.

so, let me flesh that out just a little bit. just because we have a universe, is it necessarily true that there is some explanation or reason or cause? I think that the reason why you'd be tempted to think that is because in our actual universe, you associate causes with events. things happen. you go, "I can explain to you why that happened. here is the reason why." now, there's two things to notice about that. one is that there are very often multiple different-sounding reasons that you could give for a single event. why did this football team lose their match? well, because the goalie let in that goal, or well, because they were ill-prepared, or well, because spacetime is four-dimensional. I mean, there's many possible reasons you could give. what counts as a reason depends on the framework in which you are speaking. reasons are not existing once and for all, uniquely, all by themselves. they appear within contexts, and the context in which the whole universe appears is different, whatever it might be, than all of the contexts in which things inside the universe appear. even if it were true, and I don't think that it is true, but even if it were true that everything inside the universe can be associated with a reason or a cause, that doesn't mean that the same thing applies to the universe itself. it may or may not.

and so, the, the second point is that you might want to invent a principle, you might want to invent a way of thinking, the principle of sufficient reason or something like that, that simply demands that things have explanations, that nothing that occurs does so without a reason why. but my response to that would be, "why? why is it necessary that everything have a reason why?" and you could give an argument. you say, "well, when I look at the universe, everything that happens has a reason why." why? but the point is that you're now taking your metaphysical argument and turning it into an empirical one. why is it necessary for things to have reasons or explanations? well, I see that things have reasons or explanations, but that doesn't mean that you can simply draw the conclusion that everything that happens must happen for a reason. uh, the example that I like to use is conservation of energy. because conservation of energy is something, well, you look around, you see a lot of different things happening, a lot of different processes, dynamical happenings in in the world, and you go, "look, every single time, I can associate a number, the energy, which is constant over time. it doesn't change in the process of this dynamical event." so I invent a law of nature, conservation of energy. but the point is, the logic is not, "I see that energy is conserved, therefore energy is conserved." the logic is, "I see that energy is conserved, therefore I make a hypothesis, I imagine that energy is actually conserved, and then I go out and test that hypothesis." it's an empirical one. I look for a logical foundation on the basis of some wider theory, and I actually do tests. I actually, there are experimental tests of energy conservation. it might be wrong. likewise, if you see that everything in the universe happens for a reason, which I, I don't think is actually true, but even if you saw that, you would at best be able to say, "therefore, I hypothesize that perhaps everything happens for a reason, perhaps therefore the universe has a cause or a reason." but that's not simply a metaphysical demand. that is an idea that you should go out and test. so, it goes back to being an empirical claim and should be evaluated accordingly.

just as one more example, uh, there's something called the Kalam cosmological argument that begins by saying that everything that has a beginning has a cause. therefore, it says, the universe has a beginning, and it then says, the universe has a beginning, therefore the universe must be caused. but the response to this is, "maybe not." I don't even need to say, "no, it's not true that everything that has a beginning has a cause." all I need to be able to say is, "maybe there are things that happen without causes." it then becomes an empirical claim that I need to go out and judge, just like I judge every other scientific theory, just like conservation of energy.

so, the conclusion of this kind of reasoning is that we should think about God as an empirically testable hypothesis, not as something that is completely a priori. even the purported metaphysical arguments that God is necessary ultimately come down to contingent empirical claims. so, what we should actually be asking is, does including God in our ontology, does having God play a role in our conception of how the world works, give us the best possible theory on conventional scientific grounds? now, admittedly, we don't understand what those conventional scientific grounds are with perfect clarity. the idea of what is the best possible explanation for something, even within science, is a, a bone of contention. but I'm not going to get into that in any detail. I think we all sort of know it when we see it when it comes to what makes a good scientific theory. we're looking for something that is a simple, coherent theory that explains the largest amount of data. and when it comes to that, it's not hard to imagine aspects of the observed world that you might say, "the best possible explanation for this observed phenomenon is the existence of God." historically, that has certainly been true.

so, motion is one of them. remember, one of Aristotle's conceptions of God is an unmoved mover. and this is the one when I was, uh, an undergraduate at my Catholic university, taking all those religion courses, made the least sense to me. and it took me years later until I finally figured out why it made no sense to me. why you needed to have an explanation for why things are moving. and the answer was that in an Aristotelian metaphysics, you're basing what happens on an Aristotelian physics, in which motion is something that needs to be explained. right? in Aristotelian physics, in order to get something moving, you have to push on it and keep pushing on it. and we see things moving all over the place, so there must be something doing some ultimate pushing in some sense. but then along comes what we now know as conservation of momentum. Avicenna and Newton and Galileo point out that left to their own devices, things do not need to be kept to be pushed constantly to keep moving. things will just keep moving. so the idea that things are moving is not something that needs to be explained. it is just something that is a law of physics. it is an aspect of the world. I think that the, you know, the shift in, uh, theological thinking that sort of implicitly followed the invention of Newtonian mechanics is an interesting and underdeveloped thing to think about.

there is, of course, life. and I think that for many people now and in antiquity, the reason why we, uh, are tempted by supernatural explanations is because living beings look very different than non-living beings. uh, living beings are alive and conscious and they respond in interesting ways. it seems like the simplest explanation that, you know, when something dies, when it goes from being a living object, a living organism, to being a corpse, it seems like something changed. something left it. something is now absent that was there, some sort of spiritual energy, some life force. and that, I think, is is one very obvious motivation in a primitive society for thinking that there's something beyond the merely physical. but nowadays, we think that there's nothing about life or death that cannot, in principle, be explained just by chemistry, just by biology.

there is, of course, design. uh, not just the existence of living creatures, but the wonderful diversity of all the different species we have of creatures here on Earth. there was a, a, a British person called Charles Darwin, who explained that in terms of natural selection. and nowadays, we're still left with some unanswered questions. there's the question of consciousness. what makes a person, uh, have a feeling of what it is to be themselves, to reflect, to think about things? I am hopeful that neuroscience is addressing and will continue to get better at understanding what consciousness really is. I'm not a professional neuroscientist. I'm not going to, uh, talk about that. but given the empirical success of science in giving naturalistic explanations for features of the universe, I see no obstacle to that happening in the case of consciousness.

there's also, finally, the origin of the universe itself. and here, I actually do have some expertise. what does naturalistic scientific modes of explanation have to say about the origin of the universe? it is a traditional role for God to play. if you pick up the Bible, it's right there in the beginning, God is creating the whole universe. so, here's what we know from the modern perspective about the origin of the universe. 13.7 billion years ago, the universe was in a hot, dense, rapidly expanding state, which we call the Big Bang. a lot of the credit for this idea goes to AB George Lemaître, a Catholic priest. after Lemaître helped invent the idea of the Big Bang, the Pope tried to use it, uh, and to try to, um, match on the scientific theory of the Big Bang to a theological view of God creating the universe out of nothing. and Lemaître very wisely advised him against doing that. but nevertheless, the Big Bang has left us with questions about the origin of the universe. the Big Bang is not necessarily the beginning of the universe. we often talk as if it is, but what we're really doing when we, when we talk about the Big Bang is being the beginning is we're using a theory to draw a conclusion when we have no right to believe what that theory is telling us. that theory, of course, is classical general relativity, Einstein's theory of spacetime and gravity. if you extrapolate in classical general relativity back to the Big Bang and keep going backwards, you hit a singularity, a moment when the curvature of spacetime, the expansion rate, the energy density are all infinitely big. so, in classical general relativity, you would say that is a boundary to spacetime. that is a point past which you can no longer go. however, we have absolutely no reason to think that classical general relativity is an appropriate and correct description of what is happening near the Big Bang. we should not think of gen, of the Big Bang as the beginning of the universe. we should think it as the end of our understanding of what is happening. we need to do a better job, at the very least, a theory of quantum gravity, before we can actually draw conclusions about what happened at that moment. right now, it's better to simply say, "we don't know."

can we of course say, can we possibly know? is it possible to imagine theories of the universe in which the beginning of the universe is explained, or do we need to hold that idea separate, giving it an opportunity for God to sneak in there and play a role? so, I don't know what the right theory of the origin of the universe is, but I have no shortage of plausible theories. so, let me just give you some of them. one is that the Big Bang really is the beginning, that the universe was in some sense created from nothing. Stephen Hawking, obviously, very famously has promoted this idea for a long time. but it's not the only one on the market. there is the cyclic universe, a universe which the Big Bang is just a phase that the universe goes through, that it expands and collapses again and bounces an infinite number of times. there is, uh, eternal inflation. we are, Joe already talked about inflationary cosmology, the idea that from a very tiny patch of space, a negative pressure fluid can force the universe to accelerate a huge amount and grow greatly in volume. well, it turns out that that can happen forever, and you can go from one little universe to many, many little sub-universes, and that can again happen an infinite number of times. there's also, uh, my favorite idea, which is that this kind of thing happens, but it happens in both directions of time. so, there is some point in the universe which is sort of a middle, and to the right and to the left of that, the universe grows and gets bigger and creates many offspring universes. any one of these models is, like I said, self-contained. it tells you the whole story of the universe. there is no external influence that is bringing the universe into existence. I'm not advocating any one of these models as correct. I'm merely pointing out that there's absolutely no reason to fall on our swords and say, "we need some help to explain the origin of the universe outside conventional, conventional naturalistic explanations." we don't need God to explain the origin of the universe.

however, it's still possible that God is a better explanation than any of the purely physical, uh, naturalistic ones that I was just talking about. so, we should be judging God by the same standards that we judge all the other theories. in my personal opinion, the, the best argument on the basis of empirics in favor of the existence of God comes from the fine-tuning of physical parameters in our universe. uh, it is a, it seems to be true that if you look around in the universe, we characterize what's going on by various numbers: the energy density on average, the expansion rate, physical parameters like the mass of the electron, and so forth. it seems to be the case that if you changed many of these numbers by an appreciable amount, things would be very, very different. I, I said that weakly enough that it's almost certainly true. if you change the mass of the electron by a substantial amount, things would be very, very different. it is plausibly true that for many of these numbers, if you change them by an appreciable amount, life would not possibly exist. that's one fact. it seems likely, or at least possible, that the values that the physical and cosmological parameters we observe around us take on allow for the existence of life, and not many would. furthermore, it seems as if some of these physical parameters are not natural. that is to say, if you just picked them from a bag, mass of the electron, energy density of empty space, expansion rate of the universe, you would get very, very different numbers than what we actually observe. so, you see where the argument is going. you need certain, very specific, the argument says, parameters in the universe for life to exist. these parameters don't seem to be generic. they do not seem to be what you would easily get just by randomly choosing numbers for the universe. and yet, they're right there where you need them in order for life to exist. why were we so fortunate to find ourselves in a universe that seems unnatural, but in a way that is allowing the universe to be hospitable for the existence of life? this is a very natural, sensible place to wonder whether or not there was some designing going on, whether or not God helped the universe along a little bit in picking the appropriate parameters to allow for the existence of life.

the best example in my mind is the cosmological constant, the energy density of empty space, the vacuum energy. this is something that we think is not zero. for, you know, when I was your age, we thought this was zero. when I was in graduate school, uh, but in 1998, we found that the universe is accelerating in a way that is exactly compatible with a small but non-zero energy density in empty space. now, the reason why that's interesting is because you can sit down and ask yourself, if I didn't look at the universe, if I just estimated on the basis of what I know about particle physics and cosmology, what would I guess the value of the vacuum energy should be? my guess is larger than the observed value by a factor of 10 to the 120. that is to say, if it is true that we have a non-zero but small vacuum energy, we did not pick it randomly out of a hat. we think we know what the natural value would be. if we did pick it randomly out of a hat, it is very, very, very different than the value we actually observe. so, either our notion of picking randomly out of out of a hat is very, very wrong, or there is some dynamical explanation for why the value for the vacuum energy is so small.

so, what kind of possibilities might help to explain this apparent fine-tuning of the parameters of nature? well, one is that we just got lucky. um, it is not ever completely clear to me that asking this question is a sensible thing to do. uh, aren't we fortunate that the laws of nature allow us to be here asking how fortunate we are? uh, I'm not sure that that's a sensible question to ask, because if the laws of nature didn't allow us to be here, we would not be asking the question. everyone who asks the question is lucky enough to live in a universe that allows for them to exist. nevertheless, I don't find it quite plausible, or at least, you know, as a working scientist, it sounds like a cop-out to me to say, "well, we just got lucky. the laws of physics allow for us us to be here. isn't that nice? let's move on." I suspect that this apparent fine-tuning of the vacuum energy and other constants of nature are clues. they're telling us something. there's something we don't understand about the universe. we should try to do better. so, I'm going to put aside this possibility.

another possibility is that the parameters are not nearly as finely tuned as we think. when it comes to explaining the existence of life, maybe for very different values of the parameters, there would be other kinds of life. it is extremely presumptuous, I think, to think that we understand perfectly when life could exist. maybe life is quite generic. and then the, the last two more substantive possibilities are, maybe what we're seeing is a selection effect. maybe the parameters of nature are very different from place to place. maybe most places they are utterly inhospitable to the existence of life, but of course, there's an environmental selection. we are only going to find ourselves in those regions of the universe that are hospitable to us existing. therefore, we observe these finely tuned parameters. or someone made it that way. God intervened, made the parameters because God wanted to make a universe where we could exist. what would be the point of a universe without us, after all?

so, let me just give a little bit of lip service to this second possibility, the idea that life is a lot more generic than we think it is. uh, that the parameters don't need to be that finely tuned in order to get something like life. if we were really serious about making the claim that the parameters of the universe are clearly finely tuned in such a way as to allow for the existence of life, this is the kind of argument we would have to make. we would have to say, "consider all possible physical theories, consider all possible cosmological manifestations of those laws of physics, consider all possible ways that life and or consciousness or something like that could exist within those universes, and then calculate the fraction of those universes compared to the total number of possible universes." now, nobody does this. religious people who want to argue that fine-tuning is evidence for God don't do this. physicists and cosmologists who want to argue that the multiverse is necessary to explain things don't do this. nobody does this because it's wildly impractical. our knowledge of all possible laws of physics isn't anywhere close to allowing us to make this argument. and the scientists are just as guilty of this as any of the theologians who want to use this as evidence for the existence of God. there are some hand-wavy arguments that say, you know, "if the mass of the neutron were smaller than the mass of the proton, life could not exist." I personally have almost no faith that we understand what the universe would be like if the parameters of nature were very, very different. so, I'm open-minded about this, but it's clearly a very hard problem. I just want to admit that it's just as much a problem for scientists as for, um, religious believers.

okay, so let's go on to the multiverse versus God. let's see if we can actually adduce evidence for or against the existence of God from the fact that the parameters are nominally finely tuned. and here, just to make things look sciency, we use Bayes' theorem. Bayes' theorem is the way that we take our expectations for what should be true and update them on the basis of evidence that we obtain, that we attain about the universe. so, what we mean is that we have some observations, some data D, and we have a set of possible theories that might be true, T sub I. and then we define different ways of attaching probabilities to these theories being true. so, P of X given Y is the probability that X is true once we have found out that Y is true. and then, very crucially, we have the prior probability for a given theory. we need to say that not all theories are created equal. we give a larger prior probability to the theories that seem more powerful, more simple, more elegant. uh, theories that require a lot of special pleading and fine-tuning are given smaller prior probabilities. that's just the way that science works. and then Bayes' theorem tells us that once we observe our data, the probability we assign to our theory being the right explanation for that data comes from taking the prior probability that theory was true before we got the data and then multiplying it by the probability that the data would be true in that theory. so, if the theory is true, what's the chance that the data is going to happen? that is the probability the theory is true, given that we've observed that data. and this is just a normalization factor on the bottom. so, basically, all this is telling you is that the thing you need to keep in mind is that when we're judging our theories, we need to calculate two numbers. we need to calculate the, the likelihood that theory was right even before we looked at the universe. then we have to update that likelihood by saying, "now that we've looked at the universe, how does that probability change?"

so, let's apply this notion to, uh, both God and the multiverse. starting with the multiverse. the idea that in many different regions of space, laws of physics are very different. we find ourselves only in the, uh, hospitable regions of space. so, most people agree that the probability that life could exist if the multiverse is true is substantial. it is close to one. if there really is a multiverse that is very, very big, that has all sorts of things going on, then somewhere there, life is going to be able to exist. and any people that are talking about this universe are going to exist in those little sub-places where the life exists. that's more or less uncontroversial among people who are both pro and anti-multiverse. the controversy is on what prior probability we should give to this theory. this is our judgment on how good a theory the multiverse is in the first place. uh, should we penalize the idea of the multiverse because it is not simple? it is ontologically extravagant. you're imagining all of this stuff that you will never observe even in principle, just to explain a few parameters in our little patch of observable universe. so, this was put, uh, forcefully by Richard Swinburne, a theologian, who says, "to postulate a trillion trillion other universes, a trillion trillion is amusingly tiny compared to the actual number of universes that we generally do postulate, but to postulate a trillion trillion other universes rather than one God in order to explain the orderliness of our universe seems the height of irrationality." so, height of irrationality is a fanciful way of saying that the prior probability should be very, very small. Swinburne is saying, "look, this, you're just postulating many, many, many unseen entities to explain one or two features of your observable world. that's not good science," is what he is saying.

of course, the, the problem with this is that one does not postulate trillions of trillions of universes. that is not the way it works. the multiverse, and this is the crucial point, the multiverse is not a theory. the multiverse is not something that people sat around and said, "hey, you know what would be neat is there were like a billion universes or a trillion or 10 to the 500 different universes? wouldn't that be interesting?" that is not how the idea came about. rather, the multiverse is a consequence of other theories. so, it's not that you postulate a bazillion different universes. it's that you postulate a small set of possible laws of physics, and these laws of physics predict the existence of many, many different universes. in the early 1980s, Vilenkin and Linde and others invented what we call eternal inflation. that if you believe inflation is possible, inflation is a very well-grounded physical scenario. we don't know whether it's true or not, but a very plausible scenario for explaining the observed homogeneity and isotropy of the universe. what Vilenkin and Linde showed is that if you believe that, then in a wide swath of possible inflationary models, inflation never ends. it ends in some region of the universe, but elsewhere it keeps on going, and it keeps creating new regions of universe. it's not something that you need to put in. it very generically comes out of your attempt to explain the observed homogeneity of our universe. then, in the late 1990s, string theorists realized that as you go from 10 or 11 dimensions of space, space-time, that string theory posits, down to the four that we know and love, you can do that. but if you can do it in one way, you can do it in 10 to the 500 different ways. and all of these different ways to compactify the hidden dimensions of space give us different low-energy laws of physics. this is called the string theory landscape. all sorts of different possible parameters that you can imagine getting, a huge, huge number of them. so, string theory says that the laws of physics could be very different from place to place. and inflation says it's easy to physically create regions which actually have those laws. now, both of these schemes are optional. we don't know that they are absolutely right. they're not established. but the point is that the number of things that you are postulating is not trillions of trillions. it's more like two or three. you don't, uh, you don't argue against statistical mechanics of the atmosphere in this room on the basis of, "you need a lot of molecules in this room to explain this." you don't count theories. you don't count against theories because of the number of elements in a certain set that the theory, uh, invokes. you don't say that the integers are ontologically extravagant because there's an infinite number of them. it's a simple pattern that generates all of the integers. there's only a small number of kinds of molecules of air in this room. there's only a small number of physical principles that underlie the multiverse. the way to judge the likelihood of the multiverse scenarios, to judge the plausibility of inflation and string theory, or some different ways of getting the same kind of thing, but once you have these dynamical laws, you get the many, many, many different regions of the universe. you don't separately postulate them. they don't count as a strike against the theory.

so, uh, so I would conclude, so I don't know what the answer is here. what we're trying to get here is, what is P of Multiverse? what is the a priori probability you should give on the possibility of a multiverse existing? I wouldn't say it's close to one necessarily, but I wouldn't say it's 1 over 10 to the 500 either. you don't decrease the probability every time you add a new universe. you decrease the probability, the conceptual, uh, complication of your theory, every time you add a new idea. and by that measure, the multiverse might not be of order, uh, uh, almost almost probability one, but it's not vanishingly unlikely either. it's something is very sensible to consider.

so, now let's turn to God. what is the prior probability for God, and the probability that God is true given the data? uh, so I'm going to talk about the prior probability later. let's talk about the probability that God is the best explanation of the data that we see, given that God exists. do we expect to see this kind of universe? so, I have no trouble believing that if God exists, then life should exist. I am willing to grant, uh, the God theorists that the kinds of gods that we're talking about here are the kinds that would make it possible for us to exist. I don't think that that is a, a difficulty. it's the other data, the data other than the fact that life exists, that gets God into trouble as a hypothesis. and this is a very, very difficult thing to do, right? what we're trying to do here is ask ourselves the question, "if we didn't know anything about the actual universe, but we had a theory that says that there's this thing called God that created the universe, that cares about us human beings, that has an large or infinite amount of power, and so forth, what would we expect the universe to be like?" it's very hard to do this because God is a little vague, frankly. God does not make unambiguous predictions. and we know what the universe looks like. so, we tend to say, "well, God would do it just like that, just like we see the universe looking." but that is not at all fair. we need to answer as honestly as we possibly can, what would we expect the universe to look like if God were responsible for it?

so, my favorite example of this is the low entropy of the early universe. this was the, uh, subject of the last Oxford minicourse on philosophy and cosmology. the arrow of time. in the modern conception of the arrow of time, we understand that the microscopic laws of physics are perfectly reversible, the same running forwards and backwards in time. but the reason why macroscopic physics has such a pronounced directionality is that there's that boundary condition in the early universe. it has a much, much lower entropy than it would if you again picked the universe randomly out of a hat. a lower entropy by a factor of about 10 to the minus 10 to the 120. so, this is not a typo. so, the cosmological constant is fine-tuned, one part in 10 to the 120. this is one part in 10 to the 10 to the 120. so, this is enormously more fine-tuning involved to explain the low entropy of the early universe. and crucially, the cosmological constant, the vacuum energy, you can make the argument that if it were larger by a couple of orders of magnitude, that life would not exist. I'm not sure I buy that argument. but you could make that argument for the entropy. however, that argument utterly fails. there's no sense whatsoever in which

The entropy of the early Universe needed to be that low just for life to exist. It could be enormously higher. It could be almost of order one compared to this, and we would still be fine. So, if you really were trying to be consistent, I would argue that the probability that the entropy of the early Universe could be that low, given that God created the universe in order for us to be here, is of order 10 to the 120. So that is big evidence against God.

If we didn't know what the universe were really like, and we just said God created the universe in order for us to be here, we would predict an early Universe with enormously higher entropy. Another way of saying this is, here is the Hubble Ultra Deep Field. This is what you get when you point your camera at the sky and just take a picture. If your camera is attached to the Hubble Space Telescope, you see all of these little blobs, each one of which is a galaxy comparable in size to our Milky Way galaxy. So, every one of these galaxies has billions of stars, of order 100 billion stars. There are 100 billion such galaxies in the universe. None of them are necessary for us to exist here on Earth.

If you were just going to make the universe and you were going to solve fine-tuning problems enough for life to exist, there's no reason to make any of these galaxies. All of the matter, all the degrees of freedom in those galaxies should be in thermal equilibrium. If you're going to claim that your explanation for fine-tuning is that God made it that way in order for life to exist, you strongly predict that this should be blank, or at least just a chaotic mess. It should not have all these other galaxies in there.

And I know that, um, you know, many religious believers would look at a picture like this and go, you know, oh my goodness, the glory and stupendousness of God's creation. I look at a picture like this and say, how in the world can you think that the reason for this is to let us be here? We are very tiny compared to a very big universe. So, there's, there's of course responses to this kind of argument. People have thought of it. They say, well, maybe you can't make us, maybe you can't form our solar system without forming all those other galaxies and so forth. This is just false. Everything we know about physics tells us that none of those other galaxies is necessary to explain what we have in our neighborhood here.

So, you say, well, maybe the reason why you get all those other galaxies of this really lower entropy than you ever needed is because God is sort of, um, uh, procedurally thrifty. God does not just make, I mean, so remember it's God we're talking about. God would have no trouble making just a Milky Way if that's what God chose to do. But you might say, well, God wants to play by his own rules. So, God plays by the laws of physics. So, God makes use of some physical mechanism for universe creation which, as a byproduct of making conditions that allow for us to exist, also make a 100 billion other galaxies. So, that is plausible.

But you see what's happening. You see that by trying to explain this feature that we would not have predicted, we are pushed to create an essentially physical, naturalistic, scientific explanation. We are removing all the usefulness from God. We're saying that the reason why the universe looks like that is because there's some physical mechanism that makes it look like that, and that's what God used. But that's a story I can tell without invoking God at all. There's a physical mechanism that made the universe look like this, and then I can just stop. So, the advantage that you might have had in invoking God in the first place has gone away.

This kind of argument that a priori you would not have expected the entropy of the early Universe to be so low if God were the explanation is the same kind of argument that has been used to argue against God for millennia. Now, the probability that if God really existed, the world wouldn't look like that is what we're talking about. This is the problem of evil. If God existed, we wouldn't expect there to be evil. This is the problem of random suffering. Even if you argue that evil is necessary for free will and human choice and so forth, there's no reason to, you know, have natural disasters strike down innocent children or anything like that. God could prevent that from happening. It's not what you would expect a priori.

My favorite is the problem of instructions. If I didn't know what the universe were like, but I thought that the universe were designed and created by a caring, omnipotent God who wanted us to do well, the first prediction I would make would be that God would explain himself to us very clearly, as a textbook author who has read his amazon.com reviews. I would expect God's textbook to be perfectly clear. It might not be easy to follow the instructions, but I would know what the instructions were. It would have been very easy for God, in some holy scripture, to give us some clues to say, you know, matter is made of atoms, the universe is billions of years old, people of different races and sexes and genders and sexual orientation should be treated equally, governments should derive their power from the consent of the government. A whole bunch of things that God could have told us and chose not to. That's a prediction that I, I would have made otherwise.

Finally, I like to say it. This is one way I like to say it. Imagine you're a theologian in a world where there wasn't evil or random suffering, and where God had given perfectly clear instructions. There were holy scriptures that said, you know, be nice to each other, blah, blah, blah, blah. If you were that theologian, would you take the absence of injustice in the world as evidence against the existence of God? I think you would not. I think you'd go, well, this is what I would expect God to do. He made a just society for us. If you wouldn't count the absence of injustice as evidence against the existence of God, then you should count the presence of injustice as evidence against the existence of God.

There's a couple of ways to salvage the idea that God would have made the data like this. One is to just erase God's fingerprints everywhere, to say, you know what, God really likes to do, act like God doesn't exist. God likes to obey the laws of physics. Again, that's possible, but then you're removing the usefulness of God. And the other strategy is vagueness, to deny that we have any reason to expect anything of God, to deny that we can make predictions for what the universe could be like.

So, I have a quote here from Terry Eagleton. He's first trying to say that God is the condition of possibility, the answer to why there's something rather than nothing, a very abstract, ethereal sounding concept. But then he likes to say, well, God can have regrets. God is an artist who does things out of love, and therefore you can't predict what God is going to do. The problem is, you can't have it both ways. You can't both get credit for explaining the finely tuned value of, let's say, the cosmological constant, and yet say that God makes no predictions for other things that we observe about the universe.

The multiverse theorists, at least, try. Here is a plot from a paper where people who think about the multiverse are predicting the relic density of dark matter axions in the universe on the basis of the multiverse hypothesis. I don't especially believe this is a reliable prediction in any sense, but at least they are trying. I've never seen a picture like this in a theology journal. No one ever tries to use the God hypothesis. They can try to claim credit for the low vacuum energy that allows us to exist. No one ever tries to calculate the predicted density of axions under the God theory.

So, that is my conclusion, which is that as ontologically extravagant as the multiverse seems to be, God, God is much more problematic. It's a whole separate category. It's not a kind of physics, a kind of natural process. It's a whole other kind of thing. It is ill-defined, unnecessary as far as I can tell, and seemingly you need to sort of take or leave different predictions that you might make depending on how they fit. It's on the table as a logical possibility, but I think as good scientists, we've learned over the last few hundred years that we can do better in explaining the universe. Thank you.

[Music]

[Applause]

[Music]