We remember things from our own first-person perspective, the way we experienced them. The hippocampus is the region where these egocentric, episodic memories are made. Yet the hippocampus also encodes spatial information in a decidedly non-egocentric manner, using place cells and grid cells to form a generalized map of the external environment. How does one region of the brain manage both of these?
James Knierim's lab at Hopkins has reconciled this seeming discrepancy in a manner Talmudic in its neatness: there's no contradiction! Egocentric spatial information is encoded by one population of neurons in the hippocampus, and allocentric spatial information is encoded by a different population of cells in the hippocampus. Voila, problem solved.
Neuroscientists like Knierim use the word "allocentric" to denote a way of coding spatial information that defines the location of an object relative to other objects, as opposed to relative to the self. (Psychologists use it to describe someone whose interest, attention, and actions are focused on others rather than themselves.) Either way, its opposite is egocentric: spatial information (and/or one's interest) with the self as the primary reference.
Knierim's lab recorded the activities of two different neuronal populations in rats running through mazes. Both cell populations were previously known to provide sensory information to the hippocampus, where episodic memories are made. One provides an objective spatial framework, an internal representation of the external world in which the experience to be remembered could take place. The other was thought to provide the content—the "what" of the memory—to place inside that mental matrix.
This content is obviously egocentric. Knierim's new work showed that, contrary to predictions, these content encoding neurons also fired in a location-dependent way, just in a very subjective manner. The don't fire when the rat passes over certain spots in the environment, like grid cells do; they fire when the rat looks at certain objects, placing those objects in space but in an egocentric framework.
It is impossible to get out of our own heads. Different people remember the same events differently. We can't help it; we see those events from different places, quite literal different points of view. And, at least in rats, those egocentric viewpoints get encoded into long term memories.