Support First Things by turning your adblocker off or by making a  donation. Thanks!

Most often the story is told like this: There is some feature of the world that science is at a loss to explain. Christians rush to claim that this feature can only be explained by God. Science later produces probable non-theistic hypotheses, and the Christians must beat a hasty retreat. In the early nineteenth century, the feature was the complexity of life, the scientific explanation Darwinian evolution.

Atheist pulses quicken at this story, but its lesson is just as received among theologians: We must not put God in the gaps. We must keep our commitments clear of the contracting ignorance of scientific explanations. As Dietrich Bonhoeffer said, we must find God in what we know and not in what we don’t know. But why should we not find God in both?

The question is particularly acute because there are gaps into which the Christian cannot avoid inserting God. Witness the gap between Good Friday and Easter. The God of Scripture acts in history. His transcendence does not preclude—it rather enables—his acting in the world as an unmediated efficient cause, and such action cannot but leave a permanent gap in scientific reconstructions.

There is no natural explanation for the Resurrection: The explanation is that God did it. Here is a gap and a God we all already believe in. The inevitable theological conclusion is that there can be nothing wrong with a God of the gaps, so long as it is not a God made to fit the gaps.

Why, then, disallow God from acting in primordial history? If the Catholic Church sends doctors to Lourdes to examine the veracity of reported healings, why should she not send scientists to the fossil fields? If God remains active and interfering in these last days after the Resurrection, why should we not find him in those first six days, however long exactly they took?

The emergence of life is not by itself an argument for God’s existence. But it is by that fact a possible arena of God’s interfering action. Why not, then, admit him as a possible hypothesis?

Deferring for a moment the scientific appropriateness of a God hypothesis, the theologians’ opposition shifts here from the properly theological, that God somehow does not fit into the gaps, to the apologetic, that it may be ill-advised to build one’s beliefs on such shifting grounds. This is the essence of much Christian suspicion: God has been inserted into the gaps in our knowledge only to be squeezed out again as our knowledge expands; put him in again, have him rendered superfluous again, and belief will suffer.

Acknowledging the worry as apologetic does two things, however. First, it introduces countervailing apologetic concerns. Allowing God to become a kind of vanishing explanation may have cost Christianity, but so too has the methodological atheism that has replaced it. The cultural dominance of science ensures that its methods color our broader view of the world, and its atheism is no exception.

Much has been made of our age’s disenchanted view of things, and at least one form of that—the form that makes it difficult for us to believe in demon possession and miraculous healings and all those other immediate spiritual interventions in our lives—stems directly from our scientism. How much that in turn contributes to a broader atheism, how much it makes difficult also a belief in God, is an open question—but I count even the first difficulty a loss.

Second, if the worry about God in the gaps is an apologetic one, then it is essentially a contextual concern. And here it begins to lose traction as a worry for our times. In a world of the watchmaker God, where appeal to some otherwise unexplainable scientific fact is taken as the singular demonstration of God’s existence, we might rightly worry about the instability of scientific conclusions. But the watchmaker’s reign was of relatively limited compass: For most Christians during all of Christian history, arguments from the gaps only supplemented other reasons for believing in God.

Gap arguments, after all, generally do not even point to the Christian God. The tinkering they take as evidence could equally as well be that of Athena on Olympus as Elohim beyond the world. By their very nature, they stand in need of supplement by other arguments.

And this is borne out today. No one now grants to gap arguments such indispensability, even those who think that some form of them is successful. Theism today is too chastened by the reception of Darwin to be so careless, and there is no reason why it should forget this chastisement, even were it to find such arguments a useful addition to its cause.

Part of what the Enlightenment absolutized was its worries, and part of our absolution from it is putting them back in proper perspective: Gap arguments can be helpful without being taken as either necessary or sufficient for belief. At present, as in the past, such arguments can serve as breaches in the dam of atheism without being meant to clear the river on their own, or then to baptize you in the stream.

There remains the question of whether hypothesizing a God is appropriate for science. Three reasons against it are usually suggested.

The first is that explanations-by-God are surds, that they do not admit of further explanation and so by their nature shut down inquiry. And this is in a certain sense true: Explanations that appeal to acts of free will, especially acts of divine free will, quickly exit the realm of science. To say “Apollo did it” is to cut off further scientific explanation, either because Apollo is governed by a different, supernatural set of laws, or because Apollo’s free acts cannot be explained by laws of efficient causality at all but only by ends and intentions and all the other machinery of final causality.

In that sense, divine-agent explanations do bottom out for the scientist, but that is not quite the same thing as saying that they shut down scientific inquiry. To invoke Apollo is to give an explanation that does not admit of further explanation, but it does not forestall alternate explanations. The sun can be said to rise because Apollo mounts his chariot, and there may be no getting behind his decision to some further cause of the morning; still, we can inquire whether there actually is a chariot, or a decision, or an Apollo.

We can still inquire whether other explanations, implying an Apollo or not, make better sense of the data. The question is still open, in that sense; scientific inquiry is still possible. To posit a God in the gaps simply disallows an infinite chain of explanation.

The second reason given against God hypotheses is that they are too explanatory, that they can, in theory, explain anything. Any gap can be filled by God, so what makes this particular gap particularly indicative of divine intervention? There need to be criteria for when inferences to divine activity are reasonable. This of course is true, and the nature of such criteria is not obvious, depending as it does upon the phenomenon in question.

Still, some such criteria exist. Even the atheist can acknowledge that at some point the hypotheses needed to avoid God might become so improbable as to be unbelievable. If there were no fossil record of evolution, and snakes and birds and human beings had all just appeared suddenly in the past, the case for a divine designer would be a rather strong one.

The atheists who fete evolution do by their feting acknowledge this: It is just because the world is so and not otherwise that they claim belief in God is irrational. But it is not clear that the world is so in all its particulars, as for instance with the origin of life. It is not clear that the same sort of miraculously sudden complexity that would characterize the spontaneous appearance of snakes in the fossil record does not occur at a smaller level in the emergence of life.

This, again, is not an argument that it does. It is an argument that it could, and that we need to investigate the gaps before we can give an answer either way. It is possible , under certain circumstances, that divine activity stands as the best explanation. Allowing that, we can then inquire whether those circumstances are our own.

And this brings in the third reason against admitting God as a hypothesis: We want it too much. We are biased to find intelligence and design where there is none—recall the historical narrative of gods getting squeezed progressively out of the gaps—and that should make us reluctant to do so again. But even granting the worry, it stands only as a warning, not a prohibition.

The bogus Martian canals of the late nineteenth century, later demonstrated to be an optical illusion, are a cautionary tale, but they would not prevent us from inferring the existence of Martians if a rover sent back images of cave paintings. So also with divine design: Caution might reasonably be incorporated into our criteria for inferring intelligence; it does not obviate the need for criteria altogether.

This is enough to undercut the methodological atheism taken for granted among the accepted and aspiring sciences today. If the causal chain for a particular event might include God, our reconstructions of the chain in scientific explanation should not in principle exclude him. For there is at stake here not just apologetics, whether Christian or atheist, but what is always and most fundamentally at stake in scientific questions: the truth.

If a divine explanation is possibly the true explanation, why would we not entertain it? Perhaps we are confident that it will not turn out that way, but then why prejudge the matter? Why exclude a class of possible answers? Why be atheistic as a matter of method? Why not, instead, choose a more catholic science, admitting a broader array of hypotheses, excluding conclusions based not on predetermined criteria but on what best fits the data?

Ross McCullough is a Ph.D. student in theology at Yale University.