"But surely," you might argue, "science and religion must be compatible. After all, some scientists are religious." One is Francis Collins, head of the National Institutes of Health and an evangelical Christian. But the existence of religious scientists, or religious people who accept science, doesn't prove that the two areas are compatible. It shows only that people can hold two conflicting notions in their heads at the same time. If that meant compatibility, we could make a good case, based on the commonness of marital infidelity, that monogamy and adultery are perfectly compatible. No, the incompatibility between science and faith is more fundamental: Their ways of understanding the universe are irreconcilable.Later in the column there's a paragraph about the high number of agnostic and atheist scientists.
But don't just take my word for the incompatibility of science and faith — it's amply demonstrated by the high rate of atheism among scientists. While only 6% of Americans are atheists or agnostics, the figure for American scientists is 64%, according to Rice professor Elaine Howard Ecklund's book, Science vs. Religion.So what's the problem? In the first paragraph, he's taken the position that you can't prove religion/science compatibility by collecting cases. That's the force of the marriage example. Cases of marital infidelity will not prove that monogamy and adultery are compatible. But note--it goes the other way too. Cases of marital fidelity don't prove that monogamy and adultery are incompatible. If there's really an analogy there, then Francis Collins shows nothing about science/religion compatibility, as Coyne says. But then the atheist scientists show nothing either.
So the problem is the marital infidelity analogy. It forces you to read Coyne as completely dismissing the case of Collins as showing anything at all. Given that analogy, you can't read him as merely saying Collins doesn't prove anything, as he now says (at his blog) was his meaning. He had to mean he showed nothing at all, and couldn't (logically) go on and cite the high rate of non-belief among scientists to prove that religion and science are incompatible.
Diagnosis: dismissing Collins using the marital fidelity example is a standard move. It's in Sam Harris's new book too (p. 160). But you just can't use it if you want to go on to cite data about how often scientists are non-believers. The two moves are...incompatible. It would take careful thought to decide which move is the one to give up.
If we do decide cases tell us something, then clearly we should include both religious and non-religious scientists as evidence, but should we really include the entire population? Coyne seems to think so. He writes--
Further proof: Among countries of the world, there is a strong negative relationship between their religiosity and their acceptance of evolution. Countries like Denmark and Sweden, with low belief in God, have high acceptance of evolution, while religious countries are evolution-intolerant. Out of 34 countries surveyed in a study published in Science magazine, the U.S., among the most religious, is at the bottom in accepting Darwinism: We're No. 33, with only Turkey below us. Finally, in a 2006 Time poll a staggering 64% of Americans declared that if science disproved one of their religious beliefs, they'd reject that science in favor of their faith.If we're going to take scientists, believing and non-believing, as showing something about religion/science compatibility, isn't that because they're especially rational, and especially cognizant of the nature of science? Why take the average person's psychology as telling us whether it's rational to accept both religion and science? I certainly wouldn't take a poll to find out if free will and determinism are compatible, or atheism and morality are compatible. Why resolve the issue of science/religion compatibility that way?
Note: I rewrote this post for greater clarity and eliminated something I'd said about supposing the NIH had a Hindu instead of a Christian head. Whoops! I read Coyne's column and chapter 4 of The Moral Landscape at the same time and got them mixed up. The point about Hinduism is Harris's. It's actually quite compelling. See pg. 162.