I’ve previously written against NOMA, Stephen Jay Gould’s assertion that “science” and “religion” are completely compatible because they represent two incommensurable domains of inquiry. But there’s at least as much of a problem with the other extreme, the view of New Atheists like Richard Dawkins that the two are completely incompatible because “science” refutes “religion.” (Few seriously assert incompatibility in the other direction, to reject science. Creationists, for example, typically proclaim their acceptance of science except where it conflicts with the Bible – thus the popularity of intelligent design, sold as a scientific theory.) Both of these views, to my mind, are almost painful in their oversimplification of the matter. There is incompatibility between certain parts of each domain. Many beliefs called “religious” are perfectly compatible with the evidence from controlled hypothesis testing; many aren’t. In the “scientific” domain, the only views I can think of that are incompatible with all “religious” belief are those which involve scientism: the belief that the only valid forms of knowing are based on the practice of science. (It’s worth stating repeatedly that this belief cannot possibly itself be based on the practice of science, and is therefore self-refuting.)

New Atheists often don’t want to admit this point. When they accept common-sense views at odds with their exultation of science as the only true way of knowing, they do it by equivocating on their definition of “science.” One finds the point in a recent exchange on P.Z. Myers’s blog. Responding to Larry Moran, Myers attacks what he calls:

the bizarre claim that “No scientist that is also a decent human being subjects all her/his beliefs to scientific scrutiny.” I think otherwise. There is a naive notion implicit in that statement that scientific scrutiny is somehow different from critical, rational examination. I’d argue the other way: no decent human being should live an unexamined life.

“Critical, rational examination,” eh? If that’s all science is, then every theologian is a scientist par excellence. I don’t think that’s a claim the New Atheists want to be making. Rather, the “science” they are defending is a) completely empirical, and b) based on the controlled experimental testing of hypotheses. So John Pieret responds to Myers by saying:

Really? What tests did you do on yourself to see if you love your wife and children? Hormone testing, eegs, what? Thinking about things is not “science” per se. Science is empiric investigation. Nor is the question whether “love” can be scientifically investigated, the question is whether individual scientists do it before they decide who they love.

Myers’s response:

John, yes, we carried out a long period of empirical investigation. It’s called “dating”. Both my wife and I studied the problem carefully, and if I’d been a jerk or she’d tormented me cruelly, we’d probably have reached the rational decision that we shouldn’t marry.

I really don’t understand how people can fail to recognize that we do carry out critical examinations of others and ourself. Love doesn’t just pop into existence in the absence of knowledge or experience.

And as I predicted, you do have a naive view of what “scientific” means. It does not mean hormones and eegs. You don’t have to put on a lab coat to do it. It’s simple, rational, evidence-based thinking.

An excellent point by Chris Schoen skewers Myers’s attempted defence:

We’re all aware that the practice of science, while it perhaps has some blurry edges, generally relies not just on empirical observation, but also on the testing of hypotheses, and also to the related practices of replicating the results of such tests, and publishing such results for the scrutiny of other scientists. Eliding any number of these steps is a sure way to have your findings (or “findings”) mocked. And it is on these shoals that most “pseudo-sciences” founder. There is plenty of what a lawyer would call circumstantial evidence for things like ESP and homeopathy. What there is not, in support of these phenomena, is hypothesis testing, controlled experiment, and peer review.
…
No doubt the probability of denial was bound to increase in proportion to how personal the counterfactual is (your wife.) But it is remarkable how much a scrupulous scientist has left out of his definition. White lab coats aside, without hypothesis testing and publication and replication of results, Myer’s courtship is about as scientific in its method as UFOlogy. Probably less, given the number of publications devoted to the latter. Which is not to say, of course, that PZ’s love is not real, or that his knowledge of it is flawed.

Pieret and Schoen do a solid job of demonstrating that Myers’s love for his wife is not based on “science” – not, at least, on the kinds of criteria that scientists use to distinguish science from pseudoscience. In the further comments to Myers’s post, he and his defenders try to argue that Myers’s love was still better than “religion” because it was based on empirical evidence.

But this hardly satisfies. When one is dealing with individual issues in particular lives, the evidence can lead to conclusions that would be unscientific in any sense of science accepted by New Atheists. A grad-school colleague of mine, who was proclaimed a reincarnated lama in Tibet, told me that he as a child had been able to recite things he had no way of knowing without his being a lama. Based on the evidence of his life alone, rebirth was the best explanation. He had based this view on the empirical evidence of his life. I don’t imagine it would hold up under hypothesis testing in controlled conditions; but it was based on as much empirical evidence as Myers’s love for his wife.

Beyond this point, I don’t think it can be said too many times that empiricism is self-refuting. Can statements only be true if they can be empirically tested, even in the sense that Myers tested his love for his wife? Well, the statement “statements can only be true if they can be empirically tested” cannot be empirically tested. Therefore, if it is true, it is false. The appeal to empirical evidence won’t get you out of the hard work of assessing the logic of individual claims made by both “science” and “religion.”