the terminology runs a little different in psychiatry / psychology than it does in philosophy...

but in psychiatry / psychology it is quite common to see people railing against 'dualism' (where 'dualism means something quite different from what it means in current philosophy of mind). it basically seems to mean 'cartesian dualism' which hasn't been popular since... the 1800's but i guess we will have to forgive the scientists with outdated philosophy as we will have to forgive the philosophers with outdated facts ;-)

it is worth distinguishing between two different (though surely related) senses of 'biological'.

1) genetic
2) neurophysiological

a disorder can be 'biological' in the sense that is is determined by genes (which is just to say that vary the environment fairly much as you will given the genetics certain people just will develop huntingtons if there genes are such that...). a disorder can be 'biological' in the sense that it is determined by neurophysiology - where only an old fashioned dualist (of which there are no theorists left) would maintain that neurophysiology doesn't cause behaviour. the crucial issue here is... what causes the neurophysiology to be the way it is? is it genetic (in the sense of given the genetics the neurophysiology will be fairly similar no matter how you vary the environments) or is it environmental (in the sense of whatever the genetics certain environments will produce neurophysiology such that...)

?

even if we grant that depression is some disorder of the brain (which we will surely grant unless we are good old fashioned 19th century cartesian dualists) it doesn't follow that medication is the most effective way of treating it.

psychotherapy causes neurophysiological changes (and really unless you were an 19th centuary cartesian dualist how did you expect it to lead to behavioural change?)

social intervention causes neurophysiological changes (ditto)

as does psychosurgery and medication.

which is the most effective is an empirical matter.

despite what the drug companies would have us believe...
despite what the APA would have us believe (given the funding source for their conferences and lunches and stationary)

medication simply isn't that effective. future medications may well be... but then the best way to teach someone to cook is likely to be to SHOW THEM HOW for many years to come. why is it that we expect depression (or psychosis for that matter) to be any different???

i'm not sure...

oh yeah... the notion that if it isn't biological (in the sense of being treatable by medication) then it is all your fault. that doesn't follow... but push it enough and consumers will buy in... dammit.