Chapter 11 – Psychobiological Models of Hippocampal Function in Learning and Memory1

Publisher Summary Many models and theories have been proposed over the past few decades that attempt to characterize the role of the hippocampal region in learning and memory. The anterograde amnesia that follows human hippocampal region damage is most saliently characterized by an inability to acquire new episodic information, the kind of information about individual events and experiences that is accessible to conscious control. The aspect of autoassociative memory systems seems ideal for implementing a spatial processor, in which the broad memory of a place should be evoked by any of several views of the area, even if some of the usual cues are missing. Perhaps most pervasive is the idea that the fast, temporary storage in an autoassociator is an important component of an episodic or declarative memory system, in which arbitrary patterns are stored. This chapter emphasizes that in model simulations, self-regulated suppression of recurrent collaterals does suffice to allow switching between storage and recall states in an autoassociative network. It reviews several computational models of hippocampal function in learning and memory.