There's a interesting discussion going on over at Backreaction over a craziness factor for new theories. The proposal is that one should check how crazy (different to the standard model) is a new theory, and if it is a bit crazy that's good, but if it is very crazy it is bad. As I've said in a few comments over at Backreaction, I think this is misguided, since in my opinion the sole criteria for judging a new theory should be whether it "predicts nature better and is simpler" than the old models. These things can be determined objectively, whereas the agreement of the theory with human expectations is subjective, and should not be used to judge it.
There's a song, by Seal, that contains the line: "but we're never going to survive, unless we get a little crazy", and that's right in my view. The deeper workings of the cosmos probably would seem mad, if suddenly revealed, relative to our quaint conceptions, so our minds are going to have to learn to 'think crazy' relative to our current modes of thought, to understand the universe as we move out into new regimes, as we already do a little to understand relativity and quantum mechanics. Our world view is better than that of the ancient Greeks, but they would see our worldview as bizarre. Plato would have difficulty imagining that people are standing upside down in the antipodes. Who knows what parochial views we are a victim to?
Some will always try to fit the universe into our present notions of sanity, or stay close to them, by adding patches where they can, but this will eventually be inadequate. I'm not saying that we should deliberately try to be crazy, but I am saying that subjective measures like apparent craziness should not be considered when judging theories. Crazy or not, if they predict nature, and are simpler, then let them stand.There is a pleasure sure, in being mad, which none but madmen know.
- John Dryden.