Theoretical and Research Issues: Models and Mechanisms of Dissociation
Dissociation may seem like a historical aberration, a throw-back to earlier and more primitive models of the mind. Yet these disorders are surprisingly congruent with information process-ing–based theories of mental function. For example, connection-ist and parallel distributed processing models (Rumelhart and McClelland, 1986) take a bottom-up rather than a top-down ap-proach to cognitive organization. Traditional models emphasize a supraordinate structure in which broad categories of informa-tion structure the processing of specific examples of those cat-egories, that is, the category “sweet” must exist to make sense of “sugar”, “candy” and “jelly”. In the parallel distributed process-ing models, subunits or neural nets process information through patterns of cooccurrence of input stimuli that lead to activation patterns in these neural nets, which produce pattern recognition. The output of one neuronal system becomes the input to another, thereby gradually building up integrated and complex patterns of activation and inhibition. A bottom-up processing model sys-tem has the advantage of accounting for the processing of vast amounts of information and the ability to recognize patterns with approximate information. Nevertheless, such models make the classification and integration of information problematical.
Information seems to be processed on the basis of the cooc-currence of patterns of activation rather than its appearance in a predefined category. Therefore, in parallel distributed processing system models, failures in integration of mental contents are the-oretically likely to occur. Inappropriate but apparent similarities may appear when activation patterns are similar and, conversely, no two pieces of information are necessarily connected. There have been models created to explain psychotic, dissociative andmood disorders, based on abnormal or defective neuronal associ-ation network patterns. These neural models assumed that when there are problems with the processing of input information (a model for traumatic input), the brain is more likely to have dif-ficulty achieving a coherent and balanced output. This could then lead to the development of dissociation of information and data manifested in the subject’s inability to process smoothly all of the incoming information.
There are two broad categories of memory known as explicit and implicit, declarative and procedural, or episodic and semantic. These two basic memory systems serve different functions. Ex-plicit or episodic memory involves recall of personal experience identified with the self, for example, “I went dancing last night”. The second type is known as implicit or procedural memory. This involves the execution of routine operations, such as driving a car, or typing on a keyboard. Most of these rather automatic operations could be carried out with little conscious awareness, but yet with a high degree of proficiency. These two types of memory seem to reside in different cerebral anatomical localiza-tions. Episodic memory seems to be primarily associated with limbic system function, primarily involving the hippocampal formation and mamillary bodies. On the other hand, procedural memory appears to be a function of basal ganglia and cortical functioning.
The fact that there are separate memory systems may ac-count for certain types of dissociative phenomena. For example, the automaticity observed in certain types of dissociative disor-ders reflect the separation of self-identification associated with explicit memory from routine activity in implicit or procedural memory. It is thus not at all foreign to our mental processing to act in an automatic way devoid of explicit self-identification. Fu-ture research on the neurobiology of memory may well provide insights into the functional disintegration of memory, perception, identity and consciousness seen in dissociative disorders.