INTRODUCTION
First, a conceptual model is proposed and discussed. Second, the research history of the field is drawn upon to illustrate the evolution of protocols and explain elements of the emerging model. Third, an overview of our clinical results is given that depicts the use of the proposed "generalized approach" for a number of mental disorders. These results were obtained with a relatively limited set of clinical protocols that evolved out of our extrapolation of new methods from the original research. From these results emerges a need to explain how such broad efficacy can be achieved. It is postulated that the EEG feedback technique not only promotes particular functional states of the brain, but more generally exercises neural mechanisms by which the fundamental functions of arousal, attention and affect are managed by the central nervous system (CNS). Rhythmicity in the EEG is seen as a key variable in the coordination of cortical activity, and clinical improvement is traceable to improved neuroregulation in those basic functions by appeal to their underlying rhythmic mechanisms. Current models of brain function are used to explain both the frequency and the spatial specificity of the EEG biofeedback process.

It will be demonstrated in the following that EEG biofeedback cuts across the bestiary of clinical diagnostic categories that has been devised over the last thirty years, demonstrating an ability to remediate a multiplicity of diagnoses with a limited set of protocols. It acts directly on underlying physiological mechanisms, and presupposes a considerable functional plasticity of the brain, a concept that has only recently become a significant subject of inquiry within the research community. Such plasticity appears as "noise" in the search for the genetic basis of behavior, and as such has asserted itself primarily in its negative implications for such work. It will be argued that EEG biofeedback affects brain function at the network level, and a preoccupation with processes at the molecular, membrane, or even cellular level is not particularly illuminating for brain function at the higher levels. The implicit assumption of the bottom-up approach of the neurosciences seems to be that a viable conceptual model of the network cannot be constructed until we know the detailed workings of all the parts. Yet EEG biofeedback has proven to be a valuable clinical tool (as the data cited will show), and has stimulated the creation of a conceptual model based on such a top-down, systems approach. .

EEG biofeedback can be best understood, and its relevant mechanisms discerned, by viewing the brain through the action of its web of inhibitory and excitatory feedback networks. Such networks require explicit mechanisms to manage them, integrate them, and assure their functional integrity. These networks must meet global stability criteria irrespective of what neurochemical implementation nature has, by evolutionary happenstance, devised. No doubt the technique impacts very specific neuromodulatory mechanisms, which remain undefined at this time. The clinical work can nevertheless proceed fruitfully on an empirical basis. Thus, EEG biofeedback is deemed to address itself to the core issue of control, with specificity at the network level, and yet with considerable generality in terms of clinical implications.