No representation of allocentric space has been found in the brain
Critically evaluate this statement.
The question of how animals and humans navigate is a fundamental research problem upon which there has been much experimentation and debate, and so it is necessary to refine the title to a specific point. As Tolman (1948) established that rats can solve spatial problems too complex for a purely stimulus-response system to solve, and that therefore some kind of neural map is necessary for navigation, this essay will basically address the question of whether the brain forms an allocentric (which is independent of the organism) or an egocentric (based on the organism’s own perception of the surroundings) view of the environment. For the purpose of simplicity this essay will concern itself with only the brain of a rat.
This essay will thus discuss the evolution of relevant behaviourist and neurophysiological theories, the most important being O’Keefe (1991); Muller, Kubie, Bostock, Taube and Quirk (1991); and Rolls (1991).
The behaviourist theories proposed that a reward or aversive object/event will motivate a rat to move towards or away from the location along a reward gradient, and this has been shown to be the case with rats in a maze situation (O’Keefe, 1983). Indeed, this situation does not require the rat to have a concept of absolute space; it may depend on associations between cues and responses which are provided by the maze structure itself. However, O’Keefe & Nadel (1978) identified spatial behaviours which they argued would require the existence of an allocentric map: detection of changes within the environment; navigation to the goal from a different starting location; and perhaps most importantly detour behaviour, which required the adaptation of novel behaviour to find the goal after the usual route had been blocked in some way.
Further support was added by Collett et al (1986). They trained gerbils to find seeds located at a central point between two landmarks, and upon moving these landmarks further apart rather than searching at a point equidistant, which would suggest an egocentric view of the environment, they searched at the same distance from each of the landmarks. This would suggest that they had formed a map or reference framework of the environment that was based on cues from this environment. Wilkie & Palfrey (1987) and Zipser (1986) argued that this framework could still be animal-centred, in that it forms an egocentric matrix of the environment and then transforms this as it moves around in the environment, or as the environment itself changes.
Morris (1981) was able to show in that a hippocampal lesion affects a rat’s ability to learn to return to a safe position in a milk maze, it became clear that a neurophysiological approach to the question may be useful.
O’Keefe (1991) suggested that “place cells” in the pyramidal CA1 and CA3 regions, which fire when the rat is in a particular location in the environment, defies the behaviourist argument: the place coding fires at a constant rate (albeit in different locations in CA1 and CA3 regions) independent of the location of the goal, and this firing rate does not change when the goal itself is moved, which suggests that the mapping is not motivational. However, this place coding continues after landmark cues are removed, which suggests that they could be tuned to locations due to the rat’s egocentric concept of the environment, rather than an allocentric map influenced by the environment. Indeed Taube et al (1990) suggested that these place cells were in fact perceptual cells in the post-subiculum tuned to a “local view”: what the rat sees at a particular place where it must make a critical decision about which way to turn to reach a location, because the cell firing was also dependent on the direction the rat was looking in the environment. One limitation of this view is that this directional specificity was much less common in an open unstructured environment.
Muller et al (1991) challenged both the allocentric map and local view models. They found that as well as “place cells” there was apparent in the brain “region cells” which were active in broad and importantly functional regions of the environment. Not only did this suggest that the hippocampal map was distorted in some way, but it also challenges the local view theory in that there would be lots of different “views” from just one region’s firing. This seems to suggest, however, that the hippocampal map does indeed represent a more functional property related to behaviour; for example, a region defined as a wall would coincide with a reason to stop searching in that location for food.
Rolls (1991) examined the functional anatomy of the CA3 cells and suggests an auto-associative matrix memory system. The CA3 system is highly interconnected and there is a relatively high probability (about 3.9%) that a neurone will connect with a neighbouring cell, and the cells respond to a stimulus only from a certain location and then fire others which could then provide an overall map of the environment. This is useful as the a snapshot of a scene will allow retrieval all the relevant locations of objects within the environment, allowing completion in recall. This anatomy also allows Hebbian learning, that is, strongly activated cells will form stronger links with other cells. This means that as the rat learns about its environment different synaptic weights can be attached to different objects or locations.
Rolls (1991) used further anatomical details to suggest how this system could provide an allocentric or egocentric concept of the environment. He hypothesised that as place-coded neurones on the upper layers of the antorhinal cortex are the route of entry of information into the system the CA3 pyramidal cells could make possible calculations of vectors to the location of objects, this is a neurophysiological basis for how animals might represent and navigate in their egocentric environment using the hippocampal region of the brain. Rolls (1991) also suggested that as the subiculum region also receives inputs from brain regions associated with incentives e.g. the amygdala and anterior thalamus, this is a good candidate for the structure responsible for performing goal and aversive movement vectors. This gives credence to behavioural factors involved in map formation, and implies that as this behavioural is goal-oriented, it may also be egocentric.
Muller et al (1991) go on to conclude that there is a map in the hippocampus, but it is topological rather than metric. They suggest that there is at first a only a behavioural basis for the formation of a neural map, and then a spatial map arises only because rats learn that they do certain things in certain places; thus there are then both behavioural and spatial causes for map formation, and their use is dependent on the circumstance. This suggests that the map is at first egocentric, and then as rats have more time to experience their environment an allocentric map is formed with motivational aspects included.
Muller et al (1991) also suggest that the theory of place cells as an egocentric map is unreliable because of the issue of locomotion. They suggested that the firing of place cells would be in some way connected to the movement of the rat, yet the same place cells fire when the rat is moved around the environment by hand or in a special cylinder carrier. Furthermore, place cells will fire in the specific field, independently of how the rat reached this location. To me this strongly suggests that the place cells are triggered by environmental cues rather than by the movement vectors of the rat, because the rat can no longer be using an egocentric map to calculate movement vectors in order to find its way around, and it must be using an allocentric representation of the environment. However, it is still possible that, as Muller et al (1991) suggest, this allocentric representation was formed after goal-oriented behaviour had formed an egocentric map.
In answer to the main question it seems possible to conclude that there is some strong evidence for an allocentric representation of the environment in rats. Although the neurophysiological examination given by Rolls (1991) strongly suggests that there is a strong motivational aspect to the formation of a spatial map, which suggests that it is egocentric map, there can be a number of different interpretations. Muller et al’s (1991) conclusion that there is a behavioural and allocentric spatial map operating seems to be reasonable. Intuitively it makes sense that an organism’s first representation of its environment would be based on its motivations since they are most important to its survival, but it also seems reasonable that once this map has been formed a general spatial map would be formed, especially as the location of goals in the environment is likely to change, and so the organism must be adaptable to allow it to use its allocentric map to find the new location of the goal. This is backed by O’Keefe & Nadel’s (1978) finding of detour behaviour, and also by O’Keefe (1991) in that place cell firing is not related to distance to goal. It would be interesting to see if a study could show that there is indeed a change between a motivational map and a general spatial map, as a rat becomes more accustomed to its environment.
References
Collett, T.S., Cartwright, B.A. & Smith, B.A. (1986)Landmark learning and visuo-spatial memories in gerbils. Journal of Comparative Physiology, 158, 835-51.
Morris, R.C. (1981) Spatial localization does not depend on the presence of local cues. Learning and motivation, 12, 239-60.
Muller, R.U., Kubie, J.L., Bostock, E.M., Taube, J.S. and Quirk, G.J. (1991) Spatial firing correlates of neurones in the hippocampus of freely moving rats. In J.Paillard (Ed) Brain and Space: OUP.
O’Keefe, J. (1983) Spatial memory within and without the hippocampal system. In Neurobiology of the hippocampus (ed. W.Seifert), pp. 375-403. Academic Press, London.
O’Keefe, J. (1991) The hippocampus cognitive map and navigational strategies. In J.Paillard (Ed) Brain and Space: OUP.
O’Keefe, J. & Nadel, L. (1978) The hippocampus as a cognitive map. Clarendon, Oxford.
Rolls, E.T. (1991) Functions of the primate hippocampus in spatial processing and memories. In J.Paillard (Ed) Brain and Space: OUP.
Taube, J.S., Muller, R.U., Ranck, J.N.Jr. (1990) Head-direction cells recorded from the post-subiculum in freely moving rats. I. Description and quantitative analysis. Journal of Neuroscience, 10, 420-35.
Tolman, E.C. (1948) Cognitive maps in rats and men. Psychology Review, 40, 60-70.
Wilkie, D.M. & Palfrey, R. (1987) A computer simulation model of rats’ place navigation in the Morris water mave. Behavioural Research Methods, Instruments and Computers, 19, 400-3.
Zipser, D. (1986) Biologically plausibly models of place recognition and goal location. Parallel distributed processing. Exploration in the microstructure of cognition. Vol. 2. Psychological and biological models (ed. J.L.McClelland, D.E.Rumelhart, and the PDP Research Group), pp.432-70. MIT Press, Cambridge, Mass.
