Is Naturalism Self-Defeating?

In a review of Richard Dawkin’s The God Delusion for Christianity Today, Notre Dame philosopher Alvin Plantinga argues that naturalism is “self-defeating and cannot rationally be believed.” If, as naturalists claim, there’s no god guiding the evolutionary process, then there’s no reason to think our cognitive faculties are reliable in giving us true beliefs about the world. Since we can’t trust our cognitive faculties, any conclusion we reach about the world is untrustworthy, including the claim that evolution is unguided. Therefore naturalism about evolution (and everything else) is self-defeating and must be given up. For us to trust our own beliefs (and we must, mustn’t we?) God must exist, and must have guided evolution. 

This argument, what’s become known as the Evolutionary Argument Against Naturalism is wonderfully neat, a rationalist’s delight and reminiscent of other armchair proofs of God, such as St. Anslem’s ontological argument. Why then aren’t naturalists immediately persuaded?  Are we being dense or stubborn, unwilling to admit a colossal mistake?  Possibly – no one likes being proven wrong. Plantinga says

The naturalist can be reasonably sure that the neurophysiology underlying belief formation is adaptive, but nothing follows about the truth of the beliefs depending on that neurophysiology. In fact he'd have to hold that it is unlikely, given unguided evolution, that our cognitive faculties are reliable. It's as likely, given unguided evolution, that we live in a sort of dream world as that we actually know something about ourselves and our world. (original emphasis)

So on Plantinga’s view we have some additional non-neurophysiological capacity conferred by guided evolution that makes our cognitive processes reliable, something that makes us truth-trackers in a way that merely physical creatures can’t be. What is this, one wonders, and how does it operate?  He says that

From a theistic point of view, we'd expect that our cognitive faculties would be (for the most part, and given certain qualifications and caveats) reliable. God has created us in his image, and an important part of our image bearing is our resembling him in being able to form true beliefs and achieve knowledge.

Not knowing the mind of God (no one does, after all), Plantinga can’t offer much detail about God’s (and our) supernatural truth-tracking ability, but we know it exists. After all, we do track truth (otherwise we can’t trust our own reasoning, right?), so therefore we must have this ability, and only God could give it to us.

Naturalists, of course, would like this filled in a bit. We love having things explained to us, and absent a convincing, transparent account of what our supernatural cognitive capacity is and how it operates, we tend to doubt its existence, even if this leaves our truth claims perilously unsupported, according to Plantinga. Somehow, we aren’t particularly bothered by the rationalist argument that merely physical creatures can’t track truth, since we want, justifiably, we think, a story about how supernatural creatures do better. This story, connected to the rest of science, would do much in our view to establish the truth - the fact - of non-physical cognition. Facts, for us, are secured by the role they play in transparent explanations. The empiricist in us trumps the rationalist. This same bias in favor of explanatory clarity prompts naturalists to look for a less exalted and less mysterious, that is, naturalized, notion of truth-tracking that can give us a modicum of confidence, if not absolute certainty, that we’re making sense of the world.

That notion, of course, is exactly the idea of adaptive belief that Plantinga dismisses. We’ve been selected by evolutionary pressures – a pitiless cognitive arms race –  to become better and better modelers of the world outside our heads. How do we judge the truth of the model? Pragmatically, by seeing the consequences of acting on it. Do our beliefs get confirmed by further experience, or not? If they do, this is what validates them as true, and it’s all we need as a criterion of truth: the effectiveness of beliefs in predicting what we’ll encounter. If there’s a further criterion of truth then we want to know how it works and what further warrant it adds to the practical test of making our way successfully in the world. Otherwise we’ll go on trusting our model’s conclusions – our reasoning and beliefs – and justifiably so by our lights.

One such belief is that evolution is an unguided process, a conclusion that’s been continuously validated by the best, most economical and predictive scientific theory of how species develop: natural selection. Each fact in the theory is motivated by playing a role in a transparent explanation, and each fact has thus far been confirmed by further investigation. Any fact that doesn’t is dropped from the theory. The truth of the theory is a matter of how it predicts and conforms to further intersubjective evidence gathered by the scientific community, while hewing to canons of explanatory transparency and parsimony. A guiding intelligence isn’t (thus far) needed to account for the origin of species. There isn’t any independent evidence for such a being, there’s no specification of its characteristics, and its mode of action is obscure, to put it politely. So God plays no useful role in scientific accounts of evolution. Of course you can always add God in as an initiator – a remote controller, in some sense. But this adds nothing transparent or predictive to our understanding of speciation, so doesn’t appear in the theory.

More generally, assertions of God’s existence have no theoretical (or practical) value in making testable predictions about the world. That many people nevertheless believe in God suggests that the belief is driven by non-cognitive factors, such as fear of death, the desire to reunite with loved ones in the hereafter, needs for affiliation and community, etc. (regarding which, see the New York Times Magazine cover story on Why Do We Believe?). In fact, a scientific understanding of belief in God might show it to be exactly the sort of irrational disconnection from reality that Plantinga thinks naturalism entails!

In any case, we can see that Plantinga is mistaken when he says in his review that “naturalism implies that evolution is unguided.” It isn’t naturalism, but science that holds evolution to be unguided, as far as it can tell, according to its methods. Naturalism, as a worldview, is simply to take the claims of science as ontologically dispositive in specifying what ultimately exists. This means Plantinga is also mistaken to say that “ in conflict with a premier doctrine of contemporary science [evolution].” If science, conforming to its canons of explanatory adequacy, were to demonstrate a role for God in guiding evolution, then naturalists would happily accept that conclusion. But of course in that case God would have been naturalized, integrated into a scientific account of the world.

Plantinga raises the possibility that, if our cognitive capacities are merely physical, we’re living in a dream. This possibility, like that of global, brain-in-a-vat skepticism, is something naturalists are happy to live with if the only cure is to posit God. That cure is worse than the disease, since it requires we give up our commitment to explanatory transparency as the touchstone of factual truth. Indeed, we naturalists are constitutionally unable to abandon our demand for clear evidence-based explanations, even if that leaves us (possibly) vulnerable to rationalist critiques. And we think we have a reasonable question that the critique should answer if it’s to command our assent: how, precisely, does a supernaturalist, non-physicalist scheme for securing the reliability of our beliefs improve on “mere” adaptation?

None of this, of course, will cut the least bit of ice for Plantinga and other supernaturalist rationalists (such as John F. Haught), since their commitment isn’t to explanatory transparency but to discovering unimpeachable foundations for reason (and ethics, another story). Such foundations, they argue, cannot be supplied by a world whose ultimate constituents are inherently mindless. Only in God can we trust.;

So be it. Some people want certainties of the sort that God can deliver, some don’t. Some want explanations of the sort that science delivers, some don’t. So long as naturalists and supernaturalists don’t demonize one another, we can all get along, enjoying some good arguments as we go.

TWC, March 2007