I remember back in my younger days, taking phsyiological psychology in college, they actually showed circuits in the brain that were theorized to be invovled with the subconcious and evidence supporting the existing theories.
This is stuff that is not taught in psychiatry residency and should be.
+1
There is a lot of relevant neuroscientific and experimental cognitive research which is not taught in all programs of phd clinical psychology either.(well, ofcourse it depends in the PhD and the psychiatric residency i guess. A PhD in a "top-league" university could be more theory/research based

. The same could be for psychiatric residency although a lot of research is highly biased from a molecular-cellular view as far as i know). From a psychological perspective they always teach the traditional same-ol "schools of thought" stuff despite the fact that there have been made huge progressions in various areas e.g. in experimental psychopathology for example.
There is some very interesting research on the "unconsious" (e.g. Lewicki on the complexity of the unconscious, Lotus on repressed memories, Kihlstrom on dissociation and the nature of the unconscious), how "intelligent" really is, if it is like Freud's view (a "repository" of "energy") or more like a highly automized-hyperlearned-network (a lot of evidence about this in modern research) which is based on some striatal and limbic networks (also a lot of evidence about this). Then there is lots of interesting information-processing research about the various alterations in perceptual sensitivity and bias of various mental-health problems, Wegner's experimental research on thought supression and conscious volition, Shallice and Norman's theories about the frontal executive functions and controlled/automatic processes etc.
I think this is the way forward of "mind-science" (well it already is e.g. cognitive/affective science/neuroscience) and there have appeared some very interesting mathematical/probabilistic/Bayesian/non-linear models of brain-function for various processes (perception, action, language-production, emotion). In time, these computational theories would become more "wet" (e.g. involving specific circuits, neurotransmitters and would be more biologically restrained). It would be really interesting (and crazy) if the future of mental health would involve complex maths and psychotherapy (or tMS, or even pharmacotherapy), based on some-kind of complicated non-linear brain-calculus

.
Here is Lewicki and his classic famous experimental research on "non-conscious information processing"
http://www.mwbp.org/research/lewicki/
An interesting paper on the modern science of the unscionscious by the master of the subect-Kihlstrom
http://socrates.berkeley.edu/~kihlstrm/Vienna07_precis.htm
And Wegner's very important research on a lot of crucial subjects (among others)
http://www.wjh.harvard.edu/~wegner/
Ofcourse, if you only do drug-consultations this research would make little difference. But i think its good to know the scientific cutting-edge on mental-health-related underlying processes.