It no longer baffles me why philosophers like Elliott Sober and Michael Ruse, both professed nonbelievers, spend their time telling the faithful how they can reconcile God and evolution. To me it seems like a complete waste of time, especially for atheists and agnostics, but Sober explains his reasons at the end of this post. I completely disagree with them, but first let’s revisit the controversy.
Last April, Elliott, a distinguished philosopher at the University of Wisconsin, gave a talk at my university about the possibility of God-guided mutations. His thesis was that science could not rule out the possibility that some mutations had been engineered by God, and although they look random, we can’t rule out that God created a few of them in his desire to get certain forms of life—presumably us. (Note: by “random” mutations, evolutionists mean that mutations arise irrespective of whether they’d make an individual more or less fit. A fruit fly made to live on medium containing high concentrations of toxic alcohol, for instance, doesn’t suddenly experience more mutations for alcohol tolerance.)
Elliott thought that God’s actions might be “hidden variables” in the mutation process. Thus, although most mutations might be random, a few might be made by God to foster adaptation (say, if an ancestral primate obstinately refuses to experience those mutations producing bipedal walking). And we wouldn’t be able to detect God’s actions.
Now Sober doesn’t really believe this happens, but he wants to argue that one can’t rule that out as a logical possibility, nor can we rule it out as an empirical issue—so long as God-driven mutations are sufficiently rare.
Elliott wrote me yesterday calling attention to a new paper on his website about the issue: “Evolutionary theory, causal completeness, and theism—the case of ‘guided’ mutation” (free pdf at the site). Ironically, the paper is intended for a Festschrift for Michael Ruse, his brother in theistic apologetics. It’s short and clear (15 pages of text), and easily accessible to philosopher, biologists, or laypeople interested in science.
The arguments haven’t changed, but I wanted to offer a final critique, and explain why I think Elliott is misguided.
The thesis is “the idea that God intervenes in the evolutionary process by causing this or that mutation to occur in a given time and place.” He emphasizes again that he doesn’t really believe this, but presses on nonetheless, claiming that evolutionary theory “is logically compatible with this type of divine intervention.”
Sober then describes a thought experiment that could show (and in fact real experiments have shown) that mutations look random: if it’s good to be red rather than green, an green organism put in a red environment shows no increase in mutations to red coloration, nor does the reverse situation obtain for a red organism in a green environment. Neverthless, Sober argues that:
These models, old and new, describe the effect of manipulating an organism’s natural environment and how those manipulations affect (or fail to affect) mutation probabilities. None of these models rules out hidden variables. So none rules out supernatural hidden variables. Just as a model can be true without being causally complete, so too can a model be both true and inductively generalizable without being causally complete.
Yes, of course science can’t prove that God didn’t have a hand in some of these mutations, especially if they’re rare. But considering this possibility is a waste of time for five reasons:
1. There is no evidence that God exists—at least a theistic God, which is the type demanded by Sober’s thesis. Ergo, we needn’t consider the rest of his hypothesis. We don’t have to consider that tiny, fire-breathing dragons actually ignite the gasoline in your pistons once every 10,000 ignitions, because we’ve seen no evidence for them.
2. Experiments have repeatedly showed that mutations do appear to be random: we don’t jack up the probably of an adaptive mutation by putting an organism in an environment where such adaptive mutations would be useful. (The immune system, often cited as a counterexample, isn’t: the shufflings of antibodies against the body’s invaders are random, but those shufflings that produce adaptive antibodies are fed back to the genome so that the organism makes more of the useful molecule.) This rules out the possibility that mutations could be massively nonrandom.
Sober would presumably respond that yes, well, they look random, but some rare ones might be caused by God. And of course we can’t logically or empirically rule that out, but there’s no evidence for it. We needn’t consider all logical possibilities in science that have no evidence supporting them, particularly because in this case the biggest piece of evidence—the existence of an interventionist God—is so implausible as to be unworthy of consideration. Theist-apologist are always confusing what is logically possible with what, given the evidence, is probable.
3. If you’re a theist, and thus have some idea of how God works, then you have to ask, “Why would God do it that way, rather than just bringing new species or complex adaptations into existence de novo?” The answer, “God works in mysterious ways,” is not only unsatisfactory but unparsimonious. If you appeal to God’s unknowable ways, then you have to give some evidence for an interventionist God in the first place. The whole business is simply an attempt to rescue God from the palpable fact that his actions have always remained hidden. As Delos McKown said, “The invisible and the non-existent look very much alike.”
4. If you’re going to make an argument that God intervenes rarely to cause an outcome—so rarely that the process looks random—then you might as well argue that God intervenes everywhere in a rare fashion: in the rolling of dice at Las Vegas, at coin-tossings in the Superbowl, and so on. We can’t rule out a rare God-effect there, either, but we don’t see Sober arguing for such things. Why not? Because that idea is scientifically sterile. If a hypothesis is compatible with all conceivable outcomes, it’s not a hypothesis worth entertaining, for there can be no evidence against it. As such, there’s no reason to accept it.
5. Because Sober considers the idea of God-guided mutations one that science can’t reject, our assertion that those mutations don’t occur doesn’t come from science, but from philosophy. Further, the question isn’t within the ambit of science:
If the existence of guided mutations doesn’t show that God exists, then the nonexistrence of guided mutations doesn’t show that God does not exist. Atheists and theists should agree that the biological question is separate from the theological question.
This is the old “the absence of evidence isn’t evidence for absence” argument. And yes, the nonexistence of guided mutations doesn’t prove that God doesn’t exist, but militates against it in a Bayesian way, particularly since there’s no evidence for a guiding force. As others have noted repeatedly, the absence of evidence is evidence for absence if that evidence should be there. There really isn’t a distinction between biological and theological questions here, since the action of the supernatural on mutation rates is a biological question.
I believe firmly that science doesn’t rule out the supernatural a priori (see here, for instance); our concentration on natural rather than supernatural causes for phenomena comes not as an accepted a priori fiat of scientific research, but as a result of centuries of experiments. We have found that invoking the supernatural has never helped us understand anything about the real world, and hence we’ve stopped invoking it because it hasn’t proven useful. As Laplace said, “We don’t need that hypothesis.”
Now some misguided souls define “supernatural” as “that which can’t be investigated by science.” That’s not only tautological but wrong. The supernatural was not only been part of science in its earlier years (natural theology as an explanation of organismal diversity, God’s supposed tweaking of the planetary orbits, and so on), but has also been tested repeatedly (finding out the age of the earth, refutations of precognition, ESP, and the efficacy of intercessory prayer, and so on). None of these studies has shown the slightest evidence for the action of the supernatural. Science can test the supernatural, so long as gods are supposed to affect the universe. (We can’t of course, look for evidence for a deistic, hands-off god.)
In contrast, assuming there are natural and material causes for material phenomena is a strategy that has been immensely productive in science. We don’t invoke the supernatural not because science forbids us to do that, but because it has never helped us understand the universe. Again, “methodological naturalism” is not something science has assumed as a fiat from the outset, but is a research strategy that has been productive. In the same way, plumbers and electricians don’t assume that God causes power outages and plumbing blockages. As Hawking says, “Science wins because it works.” Religion doesn’t work in that way, and never has.
Why, then, does Sober (and his confére Michael Ruse) write papers and book on this topic? It becomes clear at the end of Sober’s paper:
It is important to distinguish the evidential grounds one has for accepting a proposition from the practical reasons one has for asserting it in public. This essay has considered accommodationism under the first heading, but I want to close by saying something about the second. I bother to publish in defense of accommodationism in part because I want to take the heat off of evolutionary theory. The more evolutionary theory gets called an atheistic theory, the greater the risk that it will lose its place in public school biology courses in the United States; if the theory is thought of in this way, one should not be surprised if a judge decides that teaching evolutionary theory violates the constitutional principle of neutrality with respect to religion. Indeed, the risk is more profound, since what happens in public education often has ramifications for what happens in the wider culture. Creationists have long held that evolutionary theory is atheistic; defenders of the theory are not doing the theory a favor when they agree. Atheists who think that evolutionary theory provides the beginning of an argument for disbelieving in God should make it clear that their arguments depend on additional premises that are not vouchsafed by scientific theory or data. Philosophy is not a dirty word.
Evolutionary theory is no more atheistic than are the theories of chemical bonds or plate tectonics. It’s just perceived that way because evolution is the strongest evidence ever adduced against the existence of a theistic god. If you think God created the world and helps bring new species into existence, the observations of evolutionary biology show that you’re mistaken. If that causes people to disbelieve in God, well, so be it. But the theory isn’t more atheistic than any other scientific theory—it’s just seen that way because of its implications for beleivers. (That reminds me of Jessica Rabbit’s statement: “I’m not bad, I’m just drawn that way!”) If evolution is an atheistic theory and shouldn’t be taught in schools on those grounds, then we can’t teach cosmology, physics, or geology, either.
As philosopher Maarten Boudry said in a recent “quote of the week“:
[Robert] Pennock’s concern about the perceived conflict between science and religion is a legitimate one, but muddled philosophical reasoning will do little to avert that conflict. Science educators should not equate evolution with atheism, but neither should they pretend that the conflict between science and religion is wholly imaginary. Most religious believers would find out for themselves in any case.
And no, philosophy is not a dirty word, but Sober’s recent work on God-guided mutations is making it one, at least among scientists. What a waste of a good mind to produce such papers, and how immensely disingenuous to toss believers a life preserver when Sober himself isn’t holding the rope!