Does the supposed fine-tuning of the fundamental constants prove that the universe was created?

The short answer

No (at least not yet), because we do not know (1) how many other kinds of universes can support intelligent life, (2) how many physical constants are fundamental, (3) how probable the values of any truly fundamental constants are without a creator, (4) how probable the values of any truly fundamental constants are with a creator.

The longer answer

I. How many kinds of universes can support intelligent life?

Physicist Victor Stenger (1995) has argued that many more possible universes admit the potential for intelligent life (not necessarily carbon-based) than is generally thought. Physicist Sean M. Carroll cautions that since we cannot deduce even the basic properties of atomic nuclei in our own universe from the laws of subatomic physics alone, it is somewhat premature to make claims about the probability of life in universes with different laws (Carroll 2003).1

II. How many physical constants truly are fundamental?

The number of supposedly fundamental constants has diminished over time as physics has become more unified. Sometimes fundamental constants are added, but the trend has been toward reduction. Right now, there are thought to be 26, but fundamental physics is not finished—there still is no consensus on a theory of quantum gravity, for instance—so there could be any number of fundamental constants, including one or zero, once all of the dust has settled.

III. How probable are the values of the fundamental constants if there is no creator?

Let us assume hypothetically that all 26 of the constants currently considered fundamental truly are fundamental. Let us also assume hypothetically that we know that life could not exist unless the values of each of these constants fell into a very small range. How would we then determine the probability of these constants having the values they do? Philosopher Robin LePoidevin asks the right questions:

What determines the probability of [a] lamp’s coming on is a conjunction of the various states of affairs obtaining and the laws of physics. Altering any of these will alter the probability. But if the probability of events is determined in part by the laws of physics, what can it mean to talk of the probability of the laws of physics themselves? If we judge that it was extremely improbable that the charge on the proton should have been 1.602 x 10-19 coulomb, against what background are we making this judgment? What do we suppose is determining the probability of this value? (LePoidevin 1996:49-50).

Most of us react instinctively to this kind of concern by invoking the principle of indifference, which just means that we assume that equally-sized ranges of values are equally probable. This probably is the main reason why the fine-tuning argument initially seems impressive: we see the presumably small ranges of life-permitting values and reflexively think “small = improbable.” However, the principle of indifference actually does not work in this context, since any attempt to view a continuum of values indifferently founders upon a mathematical problem pointed out by the mathematician Johannes von Kries (Vuletic 2000).

There are only two ways to establish a background for the probabilities of the constants of nature: either by directly observing all existing universes and taking stock of their constants, or else by appealing to a physical theory that explains how fundamental constants come to have the values they do. The first possibility is out of the question for now (we don’t even know whether other universes exist, much less have a way to observe them) and there is not yet any consensus about the second (this would require at minimum that a consensus is reached on a theory of quantum gravity). However, some of the main contenders for the second are theories that predict the existence of a multiverse of universes, the values of whose constants range across every possible value—in such scenarios, a universe like ours is guaranteed.2

IV. How probable are the values of the fundamental constants if there is a creator?

Suppose one of your friends—a long-time believer in pixies—wins the lottery and then claims that his win proves that pixies exist. He reasons this way: “The probability of winning the lottery if there are no pixies is tiny. But I won. So, there almost certainly are pixies.” Where does your friend go wrong? One problem is that your friend doesn’t seem to realize that the probability of his winning the lottery is, for all we can tell, equally low if there are pixies. For his win to have a chance at providing evidence that there are pixies, he must first show that the probability of his winning given the existence of pixies is higher than the probability of his winning given that there are no pixies—it is not enough simply to show that the latter probability is low.

Now, suppose your friend grants the above but says, “Well, look, we can expect pixies to want to use their magic to help out people who believe in them, so the probability of my winning the lottery if there are pixies is higher than the probability of my winning the lottery if there are not.” Where does your friend go wrong now? One problem is that he doesn’t seem to realize that a pixie could just as easily want to use its magic to prevent people from winning the lottery. When you take all possible pixies into account, it is no longer at all clear what the relevant probabilites are.

Suppose your friend then says, “I’m only talking about pixies who like me—about the kind that would want me to win the lottery. Surely, the probability of my winning the lottery is higher if they exist than if they don’t.” He is right about that, but of course we are still not convinced because the more he restricts the kinds of pixies he is talking about, the more he reduces the initial probability that the pixies he is interested in exist.

Likewise, for the fundamental constants to provide evidence for a creator (much less prove that there is one), it would not be enough to show that the values of the constants are improbable if there is no creator: one of the other things one would have to show is that their values are more probable if there is a creator. But how probable is it that a creator would want to create life at all? And beyond that, how probable is it that a creator would want to create life in the form of physical beings that live in a universe governed by laws like ours, when a god would be able to create any logically coherent universe? A god would have absolutely no problem creating and sustaining life in a universe whose physical constants were not life-supporting. Heaven, for instance presumably is filled with intelligent beings but is not governed by physical laws at all, much less laws the values of whose constants are fine-tuned to support life.

When one considers the full variety of motivations a creator could have and the full powers it could have its disposal, it no longer is clear at all what any of the relevant probabilities are. If we make few assumptions about what a creator would want, then there is no reason to think that the probability that a creator would want to make a universe like ours is high. If we start making assumptions about the kind of universe a creator would want to make, then the former probability rises, but the initial probability that such a creator exists decreases. By the time we restrict our attention all the way to a creator who wants to create a universe like ours containing beings like us, it is no longer clear that the argument has any force.

V. Summary by Carl Sagan

The late astronomer Carl Sagan eloquently brings together most of the points above when he explains that

deducing that the laws of Nature and the values of the physical constants were established (don’t ask how or by Whom) so that humans would eventually come to be…sounds like playing my first hand of bridge, winning, knowing that there are 54 billion billion billion possible other hand that I was equally likely to have been dealt…and then foolishly concluding that a god of bridge exists and favors me, a god who arranged the cards and the shuffle with my victory foreordained from The Beginning. We do not know how many other hands there are in the cosmic deck, how many other kinds of universes, laws of Nature, and physical constants that could also lead to life and intelligence and perhaps even delusions of self-importance…Clearly we have not a glimmering of how to determine which laws of Nature are “possible” and which are not. Nor do we have more than the most rudimentary notion of what correlations of natural laws are “permitted.” (Sagan 1994:34-35)


1 But see footnote 2.

2 I would like to point out that the von Kries problem cuts both ways: it also means that if the range of life-permitting values for some constant is large, that does not in itself entail that the probability of the constant falling in that range is high. I believe this shows objections of the sort in Part I to be misguided, but I include them for completeness since presumably not everyone will agree with me about the force of the von Kries problem.


Carroll SM. 2003. Why (almost all) cosmologists are atheists. Spotted 09 Jan 2015.

LePoidevin R. 1996. Arguing for Atheism: An Introduction to the Philosophy of Religion. London: Routledge.

Sagan C. 1994. Pale Blue Dot. New York: Random House.

Smolin L. 1997. The Life of the Cosmos. Oxford: Oxford University Press.

Stenger V. 1995. The Unconscious Quantum. Buffalo, NY: Prometheus.

Vuletic MI. 2000. Book Review: Nature’s Destiny. Philo 3(2): 89-103.