What You Can’t Do with a Multiverse
“Existence, the preposterous miracle of existence! To whom has the world of opening day never come as an unbelievable sight?” – John Wheeler, “At Home in the Universe

Can we make sense of why the universe is this way, rather than some other way? Or will the ultimate laws simply be the end of all our explanations? Is there something about our universe that a deeper explanation could latch onto, something noteworthy or rare or clever or preposterous?

The fine-tuning of the universe for life is certainly a surprising fact, if the audiences I’ve addressed on the topic are any indication. As presented in Chapter 8 of “A Fortunate Universe” (AFU), fine-tuning gives us as a way of comparing naturalism (the physical universe is all that exists) and theism (God exists).

  1. Naturalism gives us no reason to expect physical reality to be this way, rather than some other way, especially with regards to its ultimate laws.
  2. Theism naturally explains (and thus, makes more probable) ultimate laws of nature that permit the existence of moral agents, such as intelligent life forms.
  3. The laws and constants of nature as we know them are fine-tuned – vanishingly few will produce intelligent life.
  4. Thus, the probability of a life-permitting universe is much greater on theism than naturalism.

The problem with this argument, as I pointed out in AFU, is that we don’t have the ultimate laws of nature (premises 1 and 2). Why should we care that the laws as we know them (premise 3) are fine-tuned?

Checking our Expectations

In science, and in everyday life, we often ask: if this idea were true, what would I expect? A geologist thinks: if this canyon was formed by erosion, would I expect such steep cliffs? A parent thinks: if my child had really cleaned their room, would I expect them to panic when I say “I’m coming in to check”?

Naturalists argue against the existence of God in this way. If we are an accidental product of blind nature, then we would expect to exist in a boring, typical, insignificant part of the universe – and here we are, third rock from an average star. We’d expect good and bad things to happen, with no rhyme or reason – and life is profoundly unfair. We’d expect natural forces to run the universe with no exceptions – and miracle claims are rare and dubious. There’s nothing special about our circumstances. So says the naturalist.

These claims are, of course, debatable. They appeal to intuition – not that there’s anything wrong with that, but we would like to do better. We would like to know: if naturalism were true, what kind of universe would be expected to exist? Would it be like this universe? This leads us to The Big Question.

The Big Question: of all the possible ways that a naturalistic physical universe could have been, is our universe what we would expect?

Alas, the Big Question is too big. We can’t handle the set of all possible physical universes. It would involve (at least) every possible mathematical law of nature, even ones we haven’t thought of yet. We seem to have two options.

Option 1 – Abandon the project: we have no idea what kind of universe to expect on naturalism. I think this is a very bad idea for the naturalist. They would have to abandon most arguments for naturalism – they can’t say whether, on naturalism, they expect life to be insignificant, unfair or even orderly enough to do science (i.e. for the universe’s laws to be discoverable).

Option 2 – Find a smaller, answerable question: we try to find a question that reflects the big question, and that we can handle. We need a subset of the set of all possible physical universes.

Looking at the deepest laws of nature that we know, a promising candidate emerges. There are constants in the ultimate laws of nature, such as the strength of the electromagnetic force and the mass of the Higgs boson, and also the freedom to choose the initial conditions of the universe. We can thus pose The Little Question.

The Little Question: of all the possible ways that the constants and initial conditions of the universe could have been, is our universe what we would expect?

The Little Question is reasonable:

  • It has been addressed by physicists for the last 40 years, but not for the purposes of testing naturalism or promoting theism. They were just exploring the consequences of the laws of nature.
  • With some degree of confidence, we can calculate what the universe would be like with different constants and initial conditions.
  • It is systematic. There is a well-defined set of possible constants and initial conditions. We aren’t just checking every possibility that we can think of.
  • It reflects the best physics we have, rather than indefinitely postponing the Big Question until physics is finished.
  • The probabilities we need are provided by Bayesian theory testing (or model selection). More details in this paper of mine.
  • If anything, it is biased in favour of finding universes like ours. Like searching for bears starting at a place where bears were recently sighted, we are looking at other universes starting near our universe.

And the answer to the Little Question is: no. An emphatic, lingering, melodic nnnoooooo with a chuckle in the middle.

This isn’t an argument for the existence of God, yet. To do that, we’d have to answer the question: what kind of universe would we expect God to create, if God existed? That’s a topic for another day. Read AFU for a summary, and Richard Swinburne’s “The Existence of God” for a detailed argument. But, at least, we should have made the naturalist rather uncomfortable.

What about the Multiverse?

Here’s an interesting approach. What if there were a vast ensemble of universes, with varying constants and local initial conditions, of which our universe was just one? Then, the right conditions for life are likely turn up somewhere, and of course we could only observe a universe in which the right conditions prevailed.

Here’s the important question. Can we use the multiverse to attack the Big Question? Can we use physical models of a multiverse to pose a “Little Question 2.0”, which better approximates the Big Question? Can we we systematically explore an unbiased set of possible universes/multiverses?

No, for several reasons.

  • There is no standard multiverse model whose parameters we can vary. We have a menagerie of models.
  • The ultimate laws of nature as we know them (the standard model of particle physics, general relativity, the standard model of cosmology) were proposed to explain the data of our universe. Multiverse models were proposed with at least one eye on solving fine-tuning problems. We have studied bespoke, proof-of-concept, cherry-picked examples of multiverse models. These are not enough to tell us what we would expect of a typical multiverse. Multiverse models that rely on cosmic inflation illustrate my point. Such models must begin by proposing the existence and properties of the inflaton field, and must add by hand all the physics of varying constants.
  • We have trouble handling probabilities within a given multiverse model – there is a sprawling, unkempt literature on “the measure problem”, for example. So there are dim prospects for anyone who wants to assign probabilities across broad classes of models.

In short, the multiverse – even if it exists – is irrelevant to addressing the Big Question. It might turn out that, in a systematic set of all multiverses, our universe is still not what the naturalist would expect. But we don’t know. No one knows the identity of the field that causes inflation, what its properties are, what is the possible range of those properties, what is the likely initial state of the field, what the mechanism is that varies the other constants (such as the mass of the electron and the strength of electromagnetism) across the multiverse, or what the distribution of that variation is – and that’s just for models of the multiverse that depend on inflation. Those are supposed to be the best models.

If a convincing, natural, elegant, glorious multiverse model appears tomorrow, then we’ll pose a “Little Question 2.0properly and try to answer it. Until then, the multiverse isn’t appealing to evidence we have; it’s appealing to a theory we don’t have.



Get updates on new posts, upcoming live discussions, and more.

Leave a Reply

2 Comment threads
6 Thread replies
Most reacted comment
Hottest comment thread
2 Comment authors
Luke BarnesDarren Recent comment authors
newest oldest most voted
Notify of

“Until then, the multiverse isn’t appealing to evidence we have; it’s appealing to a theory we don’t have.”

Sure. But we do have math that demonstrates that some form of multiverse is a possibility. The problem with trying to make this sort of argument, is that you are jumping ahead of yourself. You haven’t demonstrated that god is even possible in the first place.

And if any of the multiverse hypothesis turn out to be even remotely accurate, the probability we would see our exact universe is 100%.

Luke Barnes
Luke Barnes

There is no internal contradiction in the idea that a perfect being exists, or that there exists a maximally powerful, all knowing, perfectly good being. Possibilities are cheap. I would also contend that attempted proofs of the logical inconsistency of the idea of God, fail. But if you’ve got a favourite, have at it. The probability that we would see our universe, given a multiverse theory, is absolutely not 100%. It’s a scientific theory. We calculate likelihoods. The theory predicts a wide range of actual universes with a variety of conditions. What fraction of them will observe what we observe?… Read more »


“There is no internal contradiction in the idea that a perfect being exists, or that there exists a maximally powerful, all knowing, perfectly good being. Possibilities are cheap.” You are talking epistemic possible. Like you said, epistemic possibly is a useless standard. I’m talking possible in the real world. Just because it is internally consistent, doesn’t mean it is possible in the real world that it actually exists. And that is what you need to demonstrate before you can compare a god hypothesis to a naturalistic hypothesis. Well, to compare in any meaningful way anyway. We can compare made up… Read more »

Luke Barnes
Luke Barnes

“Just because it is internally consistent, doesn’t mean it is possible in the real world that it actually exists.”

What further principle do you propose, that disqualifies some internally consistent worlds from being “actually possible”? In other words, why are some internally consistent worlds possible and some impossible?

“You don’t think it is 100% likely to get our universe given an infinite number of tries?” That’s not the question. The question is what a particular observer is likely to see.


<> There are two problems with epistemic possibility. The first is that when something it epistemically possible, it is also epistemically impossible at the same time. And that is because of the second problem. Epistemic possibility is basically a form of the argument from ignorance. ~I can’t think of any reason it is impossible, therefore it is possible~. And that goes both ways. ~I can’t think of any way it is possible, therefore it is impossible.~ I’m not concerned about demonstrating impossibility, since that isn’t the claim. You are claiming that god is possible, so that is the only thing… Read more »


Mental note to self, do not respond with ‘s to quote the previous sections you are responding to. They hide what is inside the ‘s

The first in my response is to your first paragraph response, the second is to your second paragraph response.


Second not, the ‘s mean <>’s ….. assuming I got it right this time……

The Fine Tuning of Cosmological Constants – Australian Apologist

[…] of this, but for now the paper will have to do. Helpfully, the author of the paper also wrote this blog post in […]