Shorter Ross Douthat: Comparing belief in God to belief in the Celestial Teapot is absurd, because it's like comparing a belief only some people know is absurd to a belief everyone knows is absurd.
(Imagine the justificative possibilities. [omg! did I just make another analogy?!])
[UPDATES below the fold.]
UPDATE: Thanks to Brian Leiter for the linkage.
I'm just in the door and about to be out the door again, so I'll have to be brief and promise to cure any undue glibness later. [UPDATE II: Forgot to make clear that these remarks were intended to be in response to comments below. (And Y, I'm telling you, it was a lion!)]
First, mostly what Jon said. (In this respect too, cf. the second link above.)
Second, the main (though certainly not only) problem with the line of reasoning Ross is using (and I like Ross, despite never agreeing with him, ever) is that it treats a majority belief as Bayesian evidence when it's the competence of that very belief that is in issue. That a majority of people are drawn to a belief can't be a bearer of any verification for that belief when there is independent reason to doubt the reliability of that belief, such as, say, a conspicuous history of error in regards to closely related beliefs (again, see Jon's comment below and the second link above). This business, then, of adverting to the manifest sanity of most religious believers is merely a cheap way to leverage the argument that whatever the majority believes ought to be taken seriously. (So if you say, "My belief ought to be taken seriously because it is believed by the majority," that's twice I can't take your beliefs seriously.)
The main distinguishing feature of the Celestial Teapot or Flying Spaghetti Monster analogues, and what leads those sympathetic to religion to resist them as analogues, is that belief in them clearly could not be desire-driven. Whereas religious beliefs are. (Which is another reason we can't plausibly construe the religious belief of the majority as an independent grounds for taking religious belief seriously.)
Which brings us to one moral of the CTP and FSM analogies: If ninety percent of the population of the planet did come to believe in the CTP & FSM on current evidence, none of the remaining skeptics would take that apparently unmotivated belief change as a reason to lend greater credence to those ridiculous postulates. And this case is immaterially distinct from the case in which such belief was there all along.
Gotta go...
UPDATE III: Thanks too to Will Wilkinson, PZ Myers and Charlie Huenemann for the linkerosity.
FINAL UPDATE: Try this.
