how do you reconcile the use of
evidence for establishing belief
in one domain, and then not require
this in another?
Scientists regularly believe in a theory, or concept, that is not yet proven. Many scientists spend their entire lives proving something. Hence, I wouldn't suggest belief in science has historically relied on evidence or proof.
Now onto the main point, the difference lies with what is provable and what isn't. Many religions believe in a being that is not provable. Or believe in teachings for the betterment of society. These aren't provable or disprovable theories (whether you believe that is their nature or they are intentionally designed that way!) There are religions where one of the central tenets is that science and reason must be in agreement with the religion itself.
Personally I think the two can be seperated, and the problem only comes when religious leaders preach science. Religion, for me, covers a set of unprovable concepts.
I also wouldn't say people are too quick to diffuse the science vs. religion argument, I think pg has a few comments on what causes flame wars! (And a rather excellent essay relating to labels / identity [1]).
I think most of these arguments would go away if people understood some probability theory. "It's just a theory", "It hasn't been/can't be proven/disproven.", etc.
The point is do you think it has a greater than 0.00001% of being true? How can you possibly justify such a high probability estimate? In the space of millions of possible hypothesises, the probability of any one being correct is incredibly small. Especially with Occam's Razor, simple explanations are far more likely to be true. And religion is a very over-complicated explanation.
We use evidence to rule out explanations which can't be true, or more accurately lower their probability (the evidence could just be a coincidence or a fluke, for example.) To raise the probability of a specific religion being true, you would need a lot of evidence. Which simply doesn't exist and in fact we have lots of counter-evidence (things that religion say that turned out to be false for example.)
Scientists do spend their lives looking at things that haven't been proven, but they have to have to think they have a greater than, say, 10% chance of being true to be worth their time.
Now onto the main point, the difference lies with what is provable and what isn't. Many religions believe in a being that is not provable. Or believe in teachings for the betterment of society. These aren't provable or disprovable theories (whether you believe that is their nature or they are intentionally designed that way!) There are religions where one of the central tenets is that science and reason must be in agreement with the religion itself.
Personally I think the two can be seperated, and the problem only comes when religious leaders preach science. Religion, for me, covers a set of unprovable concepts.
I also wouldn't say people are too quick to diffuse the science vs. religion argument, I think pg has a few comments on what causes flame wars! (And a rather excellent essay relating to labels / identity [1]).
[1] - http://www.paulgraham.com/identity.html