Last week-end I drove to Derby, about 15 miles from where I live, in order to meet my wife at the railway station. I got to Derby 30 minutes before her train, having allowed plenty of time for rush-hour traffic delays. Even so, I very nearly missed her arrival.The problem was not the traffic, which hardly slowed me down at all, but an error in what neuroscientists would call my cognitive map, the information I use to find my way about. I took a wrong turning on my way into the town and wasted my half-hour safety margin careering through an endless maze of one-way streets. Eventually I managed to find a passer-by who could direct me to the station.
Research on the brain systems that allow us to work out where we are and how to get where we want to go has taken some important steps forward over recent months. For years it had seemed that the rat, subject of countless laboratory experiments on thebrain, might be different from man. The hippocampus, the brain structure which seems to be involved mainly in producing the rat's cognitive map appeared to be more concerned with memory for events in humans. However, a clever experiment on London taxi-drivers, legendary for their comprehensive cognitive map of London's streets, has reconciled rat and man by showing that the human hippocampus is involved in navigation. This reassures us that the detailed picture of hippocampal function that is being built up by testing rats and building robots, may also be relevant to man.
Journeys can be completed successfully without either cognitive maps or helpful passers by. Migrating songbirds fly thousands of miles over terrain they have never seen before, guided by the stars and by the earth's magnetic field. However, to find our way through a cluttered environment a compass is not enough, we need a map. Rats behave as if they use cognitive maps, and neurones in the hippocampus behave as if they encode parts of the map.
John O'Keefe, of University College London, was the first to propose that the hippocampus provides the rat with a cognitive map. In the early seventies {with Jonathan Dostrovsky now at the University of Toronto} he discovered "place" neurones in the hippocampus, which appear to tell the rat where it is. Each place neurone becomes active when the rat visits a particular location, the "place field" of the neurone. The activity of a hundred or so of the half-million neurones in the hippocampus can tell the rat where it is with a very high degree of precision.
Recent experiments show that place neurones pick up signals from landmarks in the environment. When the frame of reference provided by distant landmarks surrounding the rat is rotated, the place fields rotate too. {Last year, with Neil Burgess at University College O'Keefe showed that} some place cells monitor the rat's distance from the walls of its box. When the walls are moved further apart, the place-fields can stretch in the same direction, or split into two, one associated with each wall.
One of the best tests of a rat's cognitive map is the milk maze, designed in the early eighties by Richard Morris of the University of Edinburgh. The rat is placed in a large pool of opaque milky water, with a submerged platform. The rat soon learns to use visible landmarks surrounding the pool to find its way to the platform, but if the hippocampus is damaged, or if its function is impaired by drugs, the rat can no longer do the task. This suggests that the rat uses the hippocampus, with its inbuilt cognitive map, to orient itself with respect to the landmarks in its environment.
Investigations of the human hippocampus have relied heavily on patients whose hippocampi have been affected by accidental damage or degenerative brain disease including Alzheimers' disease. Unlike the rats, these patients tend to have problems that go far beyond difficulties in finding their way about: often they have severe difficulty remembering events and facts.
However, last year Eleanor Maguire, also of University College found that the human hippocampus is activated when volunteers try to learn the layout of a town by watching a film taken by a camera moving therough the streets. This year she showed that when London taxi drivers talk themselves through an imaginary taxi jiourney, their right hippocampus is activated.
These results suggest that the human hippocampus is involved both in forming and in using cognitive maps. O'Keefe suggests that the wider role of the the hippocampus in human memory may have evolved with our ability to encode time as well as space. Our memory for events may originally have been a sort of scribbling pad in the margin of our cognitive atlas.
My own cognitive map now contains a new landmark to help me find Derby station. It is a street map on the parcel shelf of the car.
The author is professor of psychology at the University of Nottingham
Published in The Financial Times, November 15th 1997