[ma2rten]

Theories with predictive power are the essence of science and models are optional, not the other way round.

I was thinking about Math or some subfields of Computer Science where you don't need any predictions (I'd think other disciplines have these purely theoretical subfields as well, but I dunno). I am not sure what the formal definition of a model is, but I think you'd agree you need some kind of non-trivial logic argument, otherwise it becomes, well... trivial ("X happens because we assume X happens").

Maybe I should have left that optional-remark or explained it further, it might be the reason I was downvoted in the beginning. First I also had a (semi-)joke in there about gaining citations being the ultimate goal of science, people might have though I was one of those fundamental Christians mocking science. It's very easy to get misunderstood if all people know about you is what they read in a comment of a few lines.