[krageon]

I'd contend that the most useful ground state for most of us, being people with no knowledge in the field whatsoever, is to assume that the DSM is broadly correct. From there you can, given some evidence, conclude that parts might not be.