New theory of hippocampal function: Associated rehearsal of multiplexed coded symbols

Abstract A new theory of the role of the hippocampus in the selective storage of information in long‐term memory is presented. This theory is based on the very recent discovery that neurons in the mammalian cerebral cortex transmit extremely precise copies of patterns of discharge in time when specific sensory inputs are presented, patterns that are interpreted to code for or symbolize specific items of information. The theory incorporates and provides an explanation for both the complex and unique internal structure of the hippocampal formation (the hippocampus and associated dentate gyrus) and the roles of many of the direct and indirect connections that structure makes with other brain structures. It also explains the deficits in learning that result from damage to the hippocampus and/or tracts that provide inputs to (or outputs from) this body as well as the role of the hippocampal formation in mapping the relationship of an individual to objects in its environment. The proposed enplanation is as follows. The hippocampal formation functions as a coordinated structure that, specifically, generates multiple copies of two different kinds of symbols (i.e., specific patterns of trains of nerve discharges in time). These two kinds of patterns are respectively provided by the entorhinal cortex through the perforant‐alvear pathways and by the septal region, through the fornix, one of the two inputs to and the only output from the hippocampal formation. These two separate and different kinds of patterns are used to make multiplexed patterns that are ultimately transmitted to the cingulate gyrus and from there to other cortical memory storage locations. This transmission of amplified representations of different symbols occurs through the fornix. From there they are either transmitted to the mammilary bodies in the hypothalamus and from there to the anterior thalamic nuclei or, alternatively, directly to the anterior thalamic nuclei, bypassing the mammilary bodies. These thalamic nuclei in turn project the information to the cingulate gyrus of the cortex. The effect of the transmission of these mixtures of symbols is to cause the coordinated rehearsal and selective storage of relationships between separate inputs (specifically, patterns of discharge that symbolize different aspects of input) that are of probable significance to the survival of the system. The repeated presentation of such specific combinations of representations (symbols) then causes rehearsal‐consolidation of these symbol associations as more permanent memories. The separate items of information are derived from at least two quite separate sources: (1) inputs to the hippocampal formation, via the alvear (or perforant) pathways; and (2) inputs from the afferent fibers to the hippocampal formation from the septum via the fornix. Such multiplexed patterns are generated as follows: patterns from the septum, by virtue of the time intervals separating impulses, are proposed to select specific sets of hippocampal pyramidal cells as the locations in which the quite different input symbols that are presented via the alvearperforant input pathways are stored . For example, if the former input consists of the pattern in time (of discharges) consisting of / / /, and the second input consists of the pattern///, then, according to the mechanism proposed here, the multiplexed output will consist of the hybrid pattern/// /// ///. This output clearly contains three copies of the two separate patterns in multiplexed form. The existing connections and inputs are analyzed in terms of established neuroanatomy and appear to provide a logical basis for the function of hippocampal formation–to cause the storage of relationships between symbols that stand for: (1) objects; and (2) the positions (and/or other features that define the object's qualities of “meaning”) of these objects with respect to the observer. The structure thus privides a mechanism for selective storage (through selective rehearsal) of important (i.e., affect‐evoking)experiences.