Many contemporary atheists, on the other hand, have a problem where they hold a position their position refutes. As Gutting writes in the New York Times' philosophy blog:
The weakest intellectual aspect of current atheism is its naïve enchantment with pseudo-scientific biological and psychological explanations of why people believe. There are no doubt all sorts of disreputable sources for religious belief, and the same goes for rejections of religion. But it's just silly to say that there's solid scientific evidence that religious belief in general has causes that undermine its claims to truth . . .
I suspect that most atheists think scientific evidence -- evidence that ultimately appeals only to empirically observable facts -- is the only sort of evidence there is.
That may be their assumption, but how do they show that it's correct? It certainly isn't supported by scientific evidence, since that tells us about only what is empirically observable. The question is whether there is anything else.That is to say, this species of atheism is a form of logical positivism.
Logical positivism holds that a statement is only meaningful if it can be verified empirically. By that standard, though, the statement "a statement is only meaningful if it can be verified empirically" is not meaningful, since it cannot itself be verified empirically. The verification standard can't be verified by its standard. There's no test that could be run, no observation one could make, no measuring that one could do that would show whether or not the statement about meaningful statements is a meaningful statement. This is a problem.
Logical positivism can be thought of as a kind of trick to rid the world of metaphysics. It appears to be quite effective, except that it is itself metaphysical.
The story of this philosophical idea of well known. The promise of logical positivists' verificationism, the problems it had, and the alternatives that have been proposed are standard parts of philosophy curricula.
In a lot of atheist discourse about the existence of God, though, there are still these naïve ideas about science and evidence. The arguments against the existence of God often turn out to be grounded in uninterrogated assumptions about what can be a fact. Atheists again and again insist that if God were to exist, God would have to exist in a way that would be meaningful to logical positivists. And, yet, even logical positivism's primary claim was not meaningful to logical positivists.
It's a very weird game, getting rid of all metaphysical claims.
Atheism doesn't have to be like this. There are other sorts that are not dependent on claims about the irrationality of anything that's not strictly empirical. Some atheists are actually bold in making such claims, for example ethical claims about the values of human flourishing. It does seem, though, as Gutting says, that a lot of contemporary atheism is built on this insistence on a logical positivist version of evidence.
A lot of it is just bad philosophy. Though it doesn't have to be.
They can't just keep saying 'there's no empirical evidence' and think they've shown that a theism based on metaphysical reasoning or nonempirical experience is irrational. The core question is whether there is anything beyond the empirical -- some transcendent reality we can call God. I think it can be rational to say there isn't a transcendent reality. But to show that it's irrational to say there is, you'd have to end the impasse in philosophical discussions of theism. That's where atheism falls short . . . people on both sides can be reasonable in holding their positions, but neither side has a basis for saying that their opponents are irrational.That seems right. To say one or the other side of the argument about God's existence is just or simply irrational is going too far. Another conclusion might be that reasonable positions, it turns out, are genially messier and more irrational that one might want to admit.