The Map Inside the Head: Place Cells, Grid Cells, and the Brain's Sense of Where

Close your eyes in a familiar room and you still know where you are. You can point to the door, picture the route to the kitchen, and walk it in the dark without bumping into much. This feels so automatic that it hardly seems to need explaining, but somewhere in your brain there is machinery keeping track of your position, and over the past fifty years neuroscientists have found the actual cells that do it. The discovery is one of the rare cases where an abstract mental faculty, the sense of place, has been traced down to identifiable neurons firing in a particular pattern. It earned a Nobel Prize in 2014, and it changed how we think about both navigation and memory.

The story starts in 1971 with John O’Keefe, working in London with rats. O’Keefe was recording the electrical activity of individual neurons in the hippocampus, a curled structure buried in the temporal lobe, while the animals moved freely around an enclosure. He noticed something that should not have been there if the hippocampus were doing what people then assumed. Certain neurons fired only when the rat was in one particular spot in the box. Move the rat to that corner, and the cell chattered away. Move it elsewhere, and the cell fell silent while a different neuron took over. O’Keefe called them place cells, and the region of space that made a given cell fire became known as its place field. Together, the population of place cells formed something like a map: at any moment, the animal’s location was encoded by which cells were active.

This was a strange and important claim, because it meant the brain was not just reacting to what was in front of it. It was maintaining an internal representation of space, a model of the environment that existed in the firing of neurons. O’Keefe and his colleague Lynn Nadel laid this out in 1978 in a book arguing that the hippocampus is the seat of a cognitive map, a phrase the psychologist Edward Tolman had used decades earlier on purely behavioral grounds. The cells gave the idea a physical home.

But a map needs more than a set of labeled locations. It needs a coordinate system, some way of measuring distance and direction that does not depend on recognizing particular landmarks. A place cell tells you that you are here, but it does not obviously tell you how far here is from there. For a long time it was unclear where that metric could come from. The answer arrived in 2005, from a Norwegian husband-and-wife team, May-Britt Moser and Edvard Moser, who had trained with O’Keefe and then set up their own lab in Trondheim. They went looking just upstream of the hippocampus, in a region called the medial entorhinal cortex, which feeds heavily into it.

What they found there was stranger and more beautiful than place cells. A neuron in the entorhinal cortex did not fire in just one location. It fired in many, and when the researchers plotted all the spots in the enclosure where a single cell was active, the spots formed a regular triangular grid, a pattern of points spaced evenly across the whole floor like the vertices of a tiling of equilateral triangles. They named these grid cells. Each grid cell lays a periodic lattice over the environment, and as the animal crosses the box the cell ticks every time the animal passes through one of the grid’s nodes. Different grid cells have grids of different spacing and orientation, and stacked together they provide exactly what a place system was missing: a built-in ruler, a coordinate mesh that measures distance and direction independently of any particular landmark.

The grid is remarkable partly because it appears to be generated internally. It holds its shape in the dark, when the animal cannot see landmarks, which means the brain is updating it from the animal’s own movement, integrating speed and direction over time to keep a running estimate of position. This kind of dead reckoning, the same trick a sailor uses to track a ship from its heading and speed, is sometimes called path integration, and the grid cells seem to be doing it continuously. Around the same time, researchers identified other pieces of the same system: head-direction cells that act like a compass, firing when the animal faces a particular way; border cells that respond to the edges of the environment; and speed cells that report how fast the animal is moving, feeding the path integrator the rate it needs.

In 2014 the Nobel Prize in Physiology or Medicine went to O’Keefe and the two Mosers for this work, described in the citation as the discovery of cells that form a positioning system in the brain, an inner GPS. The comparison to GPS is apt in spirit but slightly misleading in mechanism. Satellite navigation fixes your position by triangulating signals from outside. The brain’s system, by contrast, builds its estimate largely from the inside, from a combination of self-motion and remembered landmarks, which is why it can keep working when you close your eyes and why it sometimes drifts and leaves you disoriented in an unfamiliar place.

There is a deeper twist that connects all this to memory, and it may be the most consequential part. The hippocampus, home of the place cells, has long been known as essential for forming new memories of events, ever since the famous patient who lost the ability to make new memories after his hippocampus was removed. Why would the same structure handle both space and memory? One influential idea is that they are versions of the same operation. Remembering an event means placing it in a context, a where and a when, and the spatial map may be the scaffolding the brain uses to organize experience more generally. Place cells do not only track current location. When an animal rests or sleeps, sequences of place cells that fired along a recent path replay in fast-forward, as if the brain were reviewing the route, a process thought to help consolidate memory. The map and the memory may be built from the same bricks.

This is not only an academic matter. The entorhinal cortex, where grid cells live, is one of the very first regions damaged in Alzheimer’s disease. That fits an unsettling clinical fact: getting lost, losing the thread of where you are and how to get home, is often among the earliest signs of the disease, appearing before the more familiar failures of memory. Understanding the cells that build the spatial map gives researchers a concrete target, a specific circuit whose breakdown might be detected early or, eventually, protected.

What I find most striking about this line of work is how it collapses a gap that usually seems unbridgeable, the gap between a felt experience and its physical basis. The sense of knowing where you are is about as immediate and personal as experience gets. And yet it turns out to be carried by neurons whose behavior can be recorded, plotted, and described with the geometry of a triangular lattice. A grid cell does not know it is part of your sense of place any more than a single key knows the tune. But assemble enough of them, with the place cells and the head-direction compass and the speed signal, and you get the quiet, reliable awareness that lets you walk through a dark and familiar room and reach for the door exactly where you knew it would be.