It is typical of bipolar patients that, untreated, they tend to break into recurrent cycles of (hypo)mania and depression. The cycle periods are sometimes very regular, although the cycling period can vary from a few days to over a year. Where such regular cycling manifests itself, adding an anti-depressant tends to shorten the cycle period. Normal people are also prone, in appropriate circumstances, to depression and (hypo)mania. Typically here the depression or (hypo)mania does not reach such extreme levels as in the bipolar patient, and typically, after some time (which can sometimes be years) the person returns to normal psychological functioning. To anyone with any experience of negative feedback control mechanisms, this bipolar cycling looks very like a homeostatic mechanism with inadequate damping. A homeostatic mechanism is one which tries to reach a given state (in this case normal psychological functioning) by means of negative feedback (in this case by producing more of the brain chemicals which are too low, less of those which are too high). An important component of a properly functioning homeostatic mechanism is the damping factor, which can be likened roughly to friction in the mechanism. Too much and the response is sluggish, and tends never quite to reach the desired state. Too little and the response over-reaches itself, and the system oscillates about the desired point, eventually settling down after a period of ever-diminishing swings. Even less damping factor and the mechanism does not settle down, but once pushed off-centre it goes into ever more extreme swings until something breaks or some natural limit is reached. I said the damping factor was roughly like friction. In fact, to perform properly, it should be viscous friction, which is rate dependent. But a good engineer's bodge when faced with an out-of-control homeostatic mechanism which tends to break into oscillations is simply to add friction. This will stop the cycling, and reduce the tendency to break into cycles. Add enough, and the system will become stable and stop having a tendency to cycle. Since in bipolar disorder we are dealing with mood swings, then anything which dampens affective responses should help to stabilise the system. This is what most of the old mood controllers do

...consider that the body releases CRF when it feels danger, which indirectly leads to cortisol.  This prepares the body to fight or flee.   So every time I'm stressed out, I'm gonna go manic/depressive ??!! (sorry, couldn't resist!)

Perhaps this is because control theory is usually taught to electronic engineers, often taught to ecologists, and sometimes even to economists, but never to psychiatrists. Look up the work by Arnold J. Mandell, MD  on that very subject!   He got a McArthur Foundation Award for that work. Mandell AJ, Knapp S, Ehlers C et al:  The stability of constrained randomness: Lithium prophylaxis at several neurobiological levels.  In Posx RM, Ballenger JC (eds):  Neurobiology of Mood Disorders, p 744.  Baltimore, Williams & Wilkins, 1984 He's done stuff since then, of course. And for the Lay Persons,  he is written up in a book called The 3 Pound Universe  by Hooper & Teresi  (check the spelling!) -Bolt