Copyright © 1999 Massachusetts Institute of Technology
Edward Tolman is credited with the introduction of the term cognitive map, using it in the title of his classic paper (1948). He described experiments in which rats were trained to follow a complex path involving numbers of turns and changes of direction to get to a food box. Subsequently in a test situation the trained path was blocked off and a variety of alternative paths provided. A large majority of the rats chose a path that headed very close to the true direct direction of the food box, and not one that was close to the original direction of the path on which they had been trained. On the basis of such data Tolman argued that the rat had "acquired not merely a strip-map . . . but, rather, a wider comprehensive map to the effect that the food was located in such and such a direction in the room" (p. 204).
This concept of cognitive map has elicited considerable interest over the years since. Tolman's results seem to imply that animals or humans go beyond the information given when they go directly to a goal after having learned an indirect path. That conclusion is strongest when the spatial cues marking the goal location are not visible from the starting position. Rieser, Guth, and Hill (1986) reported a compelling experimental example of such behavior. Blindfolded observers learned the layout of objects in a room. They were trained specifically, and only, to walk without vision from a home base to each of three locations. Naturally they were also able to point from home base to the three locations quickly and accurately without vision. Furthermore, when they walked to any one of the locations, still without vision, they could point as rapidly and almost as accurately to the other two locations as they had from home base. As with Tolman's rats these observers seemed to have acquired knowledge of many of the spatial relations in a complex spatial layout from direct experience with only a small subset of those relations. In that sense they had acquired a cognitive map. More generally, two different criteria are often used to attribute to people maplike organization of spatial knowledge. One, as in the previous example, is when spatial inferences about the direction and distances among locations can be made without direct experience. The other is when it is possible to take mentally a different perspective on an entire spatial layout. This can be done by imagining oneself in a different position with respect to a layout (Hintzman, Odell, and Arndt 1981).
From where do such cognitive maps arise? One answer involves the specific kinds of experience one has with the particular spatial layout in question. For example, Thorndyke and Hayes-Roth (1982) compared observers' organization of spatial knowledge after studying a map of a large building and after experience in actually navigating around the building. They found that map experience led to more accurate estimation of the straight line or Euclidean distances between locations and understandably to better map drawing, whereas actual locomotion around the building led to more accurate route distance estimation and to more accurate judgments of the actual direction of locations from station points within the building. A second answer involves the nature of one's experience with spatial layouts in general. Thus congenitally blind observers who have never had visual experience show less maplike organization of their spatial knowledge than do sighted ones (or at least take longer to develop such organization). Why might this be? One hypothesis (Rieser et al. 1995) is that sighted persons have prolonged experience with optical flow patterns as they move about during their lives. The information from this optical stimulation specifies how the distance and direction of locations with respect to the observer change as they move. Sighted persons use that knowledge to keep track of even out-of-sight locations as they move. Blind persons without the optical flow experience do not do this as well, and, of course, all their locomotion involves moving with respect to out-of-sight locations.
It seems clear that the absence of visual experience puts one at a disadvantage in developing cognitive maps of spatial layout. When is that visual experience important? The literature comparing early and late blinded observers in spatial perception and cognition tasks often reports better performance in the late blind (Warren 1984; Warren, Anooshian, and Bollinger 1973). However, the age boundaries are very fuzzy because the ages used vary from study to study and because the availability of blind participants of specific ages is rather limited. With sighted participants, maplike organization is evident at very early ages ranging from two years (Pick 1993) to six or seven years (Hazen, Lockman, and Pick 1978), depending on the situation.
Another kind of answer to the origin of maplike organization of spatial knowledge comes from the considerable research on underlying brain mechanisms. This research has been particularly concerned with where and how spatial information is represented in the brain and how, neurologically, orientation is maintained with respect to spatial layouts. Lesion studies and studies of single cell recording have been among the most informative. Many human brain damage studies have been concerned with VISUAL NEGLECT, a deficit in which part of a visual stimulus is ignored. Such deficits have often been associated with parietal lobe damage, and the neglect is of the visual space contralateral to the damage. This neglect apparently operates in memory as well as during perception. One example particularly relevant to cognitive mapping has been reported by Bisiach and Luzzatti (1978). Patients were asked to describe a familiar urban scene, when viewed from one end. They described it, but ignored features on the left side. They were then asked to imagine viewing it from the other end. They now reported it including originally missing features that were now on the right side, omitting originally reported features now on the neglected left side.
A number of areas in the brain have been implicated in spatial information processing by SINGLE-NEURON RECORDING studies in animals. Besides the posterior parietal cortex, the HIPPOCAMPUS has been found to play a particularly important role. Indeed, O'keefe and Nadel (1978) authored a book entitled The Hippocampus as Cognitive Map. Early spatially relevant single cell recording research resulted in the exciting discovery of "place" cells. These cells fire selectively for particular locations in a spatial environment. However, place cells by themselves seem to reflect place recognition but not necessarily information about how to get from one place to another (especially if it is out of sight). An analysis by McNaughton, Knierim, and Wilson (1994) suggests a vector subtraction model that could solve the wayfinding problem. In their prototypic situation, a kind of detour behavior, an animal knows the distance and direction from its home to landmark A. On some occasion it finds itself at an unknown landmark C, from which A is visible but not its home. Their model suggests how hippocampus place cells in conjunction with distance and heading information are sufficient to generate a straight line path to its home. Heading information is potentially available from integration of vestibular stimulation, and there are a variety of visual sources for distance information.
Spatial analogies have frequently been attractive ways of describing nonspatial domains. Examples include kinship relations, bureaucratic organizations, statistical analyses, color perception, etc. An intriguing possibility is that our spatial thinking and the idea of cognitive maps can apply to other domains that are easily described in spatial terms. It is possible to think of cognitive maps of data bases -- and indeed the term cognitive map is often being used more and more metaphorically. It is an empirical question with important practical and theoretical implications to know how well the underlying spatial cognition transfers to such nonspatial domains.
Bisiach, E., and C. Luzzatti. (1978). Unilateral neglect of representational space. Cortex 14:129-133.
Hazen, N. L., J. J. Lockman, and H. L. Pick, Jr. (1978). The development of children's representation of large - scale environments. Child Development 49:623-636.
Hintzman, D. L., C. S. O'Dell, and D. R. Arndt. (1981). Orientation and cognitive maps. Cognitive Psychology 13:149-206.
McNaughton, B. L., J. J. Knierman, and M. A. Wilson. (1994). Vector encoding and the vestibular foundations of spatial cognition: neurophysiological and computational mechanisms. In M. S. Gazzaniga, Ed., The Cognitive Neurosciences. Cambridge, MA: MIT Press.
O'keefe, J., and L. Nadel. (1978). The Hippocampus as Cognitive Map. Oxford: Oxford University Press.
Pick, H. L. Jr. (1993). Organization of spatial knowledge in children. In N. Eilan, R. McCarthy, and B. Brewer, Eds., Spatial Representation. Oxford: Blackwell Publishers, pp. 31-42.
Rieser, J. J., D. A. Guth, and E. W. Hill. (1988). Sensitivity to perspective structure while walking without vision. Perception 15:173-188.
Rieser, J. J., H. L. Pick, Jr., D. H. Ashmead, and A. E. Garing. (1995). Calibration of human locomotion and models of perceptual - motor organization. Journal of Experimental Psychology: Human Perception and Performance 21:480-497.
Samsonovitch, A., and B. L. McNaughton. (1997). Path integration and cognitive mapping in a continuous attractor neural network model. Journal of Neuroscience 17:5900-5920.
Thorndyke, P. W., and B. Hayes-Roth. (1982). Differences in spatial knowledge acquired from maps and navigation. Cognitive Psychology 14:560-589.
Tolman, E. C. (1948). Cognitive maps in rats and man. Psychological Review 55:189-208.
Warren, D. H. (1984). Blindness and Early Childhood Development. New York: American Foundation for the Blind.
Warren, D. H., L. J. Anooshian, and J. G. Bollinger. (1973). Early vs. late blindness: the role of early vision in spatial behavior. American Foundation for the Blind Research Bulletin 26:151-170.
Kitchin, R. M. (1994). Cognitive maps: what are they and why study them? Journal of Environmental Psychology 14:1-19.
Siegel, A. W., and S. H. White. (1975). The development of spatial representation of large-scale environments. In H. W. Reese, Ed., Advances in Child Development and Behavior, vol 10. New York: Academic Press.
Copyright © 1999 Massachusetts Institute of Technology