Can God Guide Evolution?

3 Ways to Reconcile God’s Providence with Evolutionary Randomness

Jul 03, 2025

Evolution is the scientific theory that all living things on earth share a common ancestor, and through the process of descent with modification over several billion years this relatively simple common ancestor gave rise to more complex and diverse species that we see in the fossil record and around us today.1 One mechanism that drives descent with modification is natural selection acting on random mutations.

The idea that the mutations are random creates a tension for theists (like myself) who think that God guided the evolutionary process toward a particular goal, like the creation of human beings.

An excellent new article by the Christian philosopher Michael Bergmann titled, “Could Darwinian Natural Selection Be Divinely Guided?”2 lays out this tension, and offers three potential solutions. Since the article is behind a paywall, I’d like to summarize and reflect on the main ideas, making them accessible to everyone.

The Tension

Bergmann sets up the tension with two theses.

The Randomness Thesis: Humans came into existence by natural selection acting on random genetic mutations.
The Providence Thesis: God intentionally brought humans into existence.

The basic worry is that God can’t guide a random process. That sounds as contradictory as saying that God directed an undirected process.3

Bergmann offers three possible solutions to resolve this tension. He acknowledges that there could be other ways of reconciling the two theses. His goal is simply to show that the two theses are not incompatible.

I’ll briefly summarize each possible solution. The first two solutions provide a model where God acts in very ordinary ways, while the third solution provides a model where God acts in a more special or miraculous way.

Solution #1- The Theistic Hidden Variable View (THV)

There are both deterministic and indeterministic interpretations of quantum mechanics (QM). Albert Einstein favored a deterministic interpretation, according to which the physical universe appears indeterministic only because there are variables that are currently hidden from us. This is the idea behind Einstein’s famous quote that “God does not play dice with the universe.”

Suppose the Hidden Variable interpretation of QM is true, and suppose that God exists. We can call this the Theistic Hidden Variable (THV) view. THV has no problem with the Providence Thesis. God could have set up deterministic laws of nature and initial conditions at the Big Bang that guaranteed the formation of a habitable planet and the evolution of human beings.

What about the Randomness Thesis though? Does THV conflict with it? To answer that, consider the Atheistic Hidden Variable (AHV) view. If a coin is tossed in such a universe, is the result random? In one sense, no, because the past history of the universe guaranteed it. In another sense, yes, because nothing in the physical universe knows or plans for it to land the way that it actually does.

The same can be said of genetic mutations in an AHV universe. Are these random? In one sense, no, because the past history of the universe guarantees each particular mutation. In another sense, yes, and we see this articulated well by the philosopher of biology Elliot Sober. He wrote that a mutation counts as random because, “there is no physical mechanism (either inside organisms or outside of them) that detects which mutations would be beneficial and causes those mutations to occur.”4

Even in an AHV universe, genetic mutations would be random in Sober’s sense. But the same can be said of a THV universe. In both cases, the mutation is guaranteed from the beginning of the universe, and there is no physical mechanism that detects which mutations would be beneficial and causes those mutations to occur. The only difference seems to be that in one universe God planned it, and in the other universe no one did. In both cases though, the gene mutation will be random in Sober’s sense. For that reason, THV has no problem with Randomness Thesis either.

That means that THV shows the compatibility of the Providence Thesis and the Randomness Thesis. This is Bergmann’s first solution.

Solution #2- The Divine Collapse Causation View (DCC)

Bergmann’s second solution appeals to an indeterministic interpretation of QM, specifically what is called “spontaneous collapse”. What is that exactly? Well, here’s a super simplified description.

On this interpretation of QM, there are only probabilities about what an event of a certain kind will cause, and what the location of a particular subatomic particle is. The probability distributions can be represented as a wave function. The wave function periodically "collapses" into a determinate state where causes have definite effects and particles have definite locations. These collapses are spontaneous, not the result of measurement by an observer as the standard Copenhagen interpretation says. Indeterminacy comes in because nothing causes the wave function to collapse to one state rather than another, even though there is a predictable probabilistic pattern to these collapses.

Suppose the spontaneous collapse interpretation of QM is true, and suppose that instead of nothing causing the collapses to occur precisely as they do, God causes the collapses to occur precisely as they do, and this pattern of collapses matches a certain probability distribution over time. Call this the Divine Collapse-Causation (DCC) view. DCC has no problem with the Providence Thesis. Each quantum state in evolutionary history is caused by God, allowing God to guide the evolutionary process to unfold just as it did. If God had caused different collapses, evolutionary history would have unfolded in a different way.

What about the Randomness Thesis though? Does DCC conflict with it? Once again, it’s helpful to consider an atheistic version of the spontaneous collapse interpretation. Could genetic mutations be random according to it? Yes. The universe would unfold in an indeterministic way with collapses happening according to predictable patterns. Now consider a DCC universe that is physically identical to that one. In both, the pattern of spontaneous collapse influencing gene mutation will be the same, and the particular collapses and mutations will be the same, making the mutations in both of these universes random in Sober’s sense. For that reason, DCC has no problem with the Randomness Thesis either.

That means that DCC shows the compatibility of the Providence Thesis and the Randomness Thesis. This is Bergmann’s second solution.

Solution #3- The Scientifically Undetectable Miracles View (SUM)

Bergmann’s third solution proposes that on some occasions, God tinkered with the gene mutation process in a way that is consistent with the probabilistic patterns scientists have learned to expect, making God’s involvement undetectable by science. Call this the scientifically undetectable miracles (SUM) view. Why would God do it this way? Perhaps He doesn’t want to interfere with creaturely attempts to learn through science.

Whatever the reason, this view has no problems with the Providence Thesis. God could have tinkered in a way that kept evolution on track to intentionally bring about the existence of humans. What about the Randomness Thesis? Does SUM conflict with it? No, the mutations would still be random in Sober’s sense mentioned above.

That means that SUM shows the compatibility of the Providence Thesis and the Randomness Thesis. That is Bergmann’s third solution.

The table below summarizes the main features of each solution Bergmann offered.

With those three solutions on the table, here are some of my own reflections.

My Final Thoughts

I think Bergmann has successfully shown that the Providence Thesis and the Randomness Thesis are compatible. All three of his solutions are at least possible, and that’s enough.

Though the bits about quantum mechanics in the first two solutions sound a bit technical, the main point is to show that God can providentially guide the evolutionary process in ordinary ways. None of the solutions Bergmann offers struggle to secure the Providence Thesis.

The main challenge for all three solutions was to secure the Randomness Thesis as well. Ultimately, the key to do that was to clarify what one means by the word “random” when speaking of “random mutations.” Once Sober’s definition is on the table, it becomes clear why the mutations would be random, whether or not you have a deterministic or indeterministic interpretation of QM, and whether or not God is acting in ordinary or miraculous ways.

It’s worth mentioning that I don’t see how one could offer scientific evidence for a sense of “random” that goes beyond Sober’s sense, claiming that there is no non-physical mechanism or person like God bringing about these genetic mutations. Science doesn’t have the tools for that sort of thing. In solution #2 and #3, God’s actions are undetectable by science since they fit within the probabilistic pattern you’d find in an atheistic universe with our evolutionary history. And in solution #1, mutations would be random much the same way a coin toss would be random.

Of the three solutions that Bergmann offers, I’m not inclined to accept solution #1, because I am not a determinist. I’m not a determinist because I believe that humans have free will, and I believe that free will is incompatible with determinism. This leaves me with solutions #2 and #3. Given those options, I more inclined to go with solution #3 since it does not depend on any particular interpretation of QM. Some may criticize me for favoring this solution though since it makes God’s miraculous work in guiding evolution scientifically undetectable, and unfalsifiable. There are parallels here to Antony Flew’s parable of the invisible gardener5, and I certainly don’t want to sound like the obstinate explorer in the parable.

Here are three quick points in response to this unfalsifiability worry.

First, the goal of solution #3 was simply to demonstrate the compatibility of the Providence Thesis and the Randomness Thesis. It succeeds in doing that, whether the solution is unfalsifiable or not.
Second, it’s not clear to me that a view is problematic simply for being scientifically unfalsifiable. The falsifiability requirement would undermine most of philosophy! What matters more is how well the view fits with the total evidence we have, not just scientific data.
Third, perhaps there are good reasons why God might choose to guide evolution in an undetectable way. I already hinted at one such reason above, though more can and should be said here. At least some responses to divine hiddenness in general will extend to this particular issue.

Does all of this solve every tension that exists between Christianity and evolution? Of course not! One may wonder

Why God chose to use an evolutionary process that results in so much pain and suffering.
Whether evolution debunks religious belief.
How evolution could possibly fit with the creation narrative in the opening chapters of Genesis.

These are questions I hope to take up in future posts.

To wrap up this post, let’s return to the main question I set out to answer. Can God guide evolution? Yes.6

The Analytic Christian

Discussion about this post