Relational versus Nonrelational Memory

Many theories of memory have been proposed over the past few decades (e.g., episodic versus semantic; declarative versus procedural).  Eichenbaum, Otto, and Cohen (1994, 1992) make a distinction between relational and nonrelational memory.  Their theory states that the hippocampal formation is important for relational memory (i.e., spatial and episodic memory: memory for information that is complex and requires interpretation and organization), while the parahippocampal region is important for nonrelational memory (i.e., semantic memory and stimulus-stimulus associations: memory for specific facts and stimuli).  The hippocampal formation includes the hippocampus proper, fornix, subiculum, and dentate gyrus, whereas the parahippocampal region includes the perirhinal and entorhinal cortices.  We have conducted studies looking at dissociations between the hippocampal formation and the parahippocampal region with regard to relational and nonrelational memory. 

Specifically, we have used an interesting task, called the Bunsey Digging Task to examine this distinction.  Basically, the task requires rats to dig in small cups of scented sand to retrieve buried Froot Loops cereal (FL).  This task can be made either relational (Bunsey and Eichenbaum, 1996) or nonrelational.  In the nonrelational study we conducted (Hanson, Bunsey, and Riccio, 2002), rats learned to discriminate between two cups with different scents (e.g., cocoa versus cinnamon) and made the correct choice by digging in the cup that always contained FL (e.g., cocoa) and not in the cup that never contains FL (e.g., cinnamon).  Rats with hippocampal formation lesions were able to correctly solve this task, presumably, because the hippocampal formation is not needed for nonrelational information.  However, rats with parahippocampal lesions were severely impaired on this task.  Interestingly, it appears that the parahippocampal region may be important for processing or storing this type of information, with little help from the hippocampal formation.   

Other questions of interest: 

• Do complete hippocampal formation or parahippocampal region lesions produce different effects (i.e., more disruption) than incomplete lesions?
• What dissociations exist between hippocampal formation-lesioned rats and rats with lesions to the parahippocampal region on a spatial task, like the Sand Maze?
• Are dissociations between the hippocampal formation and the parahippocampal region the same when different components of these systems are lesioned (e.g., fornix vs. perirhinal, subiculum vs. entorhinal, etc.)?

Back to Top