Despite this apparently irrefutable evidence of regional specialization in the brain in connection with speech, movement, seeing, and touch, the American psychologist Karl Lashley produced evidence that Flourens' position remained tenable with respect to some forms of behaviour. By studying maze learning in rats, Lashley showed that the deleterious effects of removing parts of the cerebral cortex depended on the amount of tissue removed rather than on its exact location, a finding enshrined in the principles of mass action and equipotentiality. Lashley's views struck a sympathetic chord with many psychologists of the mid-20th century, who likened the brain to a complex electronic device which can become increasingly unreliable as more of its components are damaged, but which rarely suffers a severe breakdown when a small number of specific localized components are removed.
With hindsight, the often bitter controversy was unnecessary. The view that the brain acts as a whole stems from investigation of complex phenomena, such as learning and remembering complicated tasks involving several of the senses. It is small wonder that a good deal of the brain is involved in such behaviour and, therefore, that damage to any part of it has some effect. The evidence for regional specificity stemmed, by contrast, from investigations of relatively simple actions such as moving a finger, seeing a light in one part of space, or detecting that a particular part of the skin had been touched. The latter are all examples of simple perception or voluntary movements, the former of higher-level cognitive and intellectual behaviour. Even so, the mediation of some complex cognitive abilities can be surprisingly localized. For example, the modern techniques of functional neuroimaging by positron emission tomography (PET) or magnetic resonance imaging (MRI) (see brain imaging), both depending on changes in regional cerebral blood flow, have shown that spatial memory is concentrated in the hippocampus and the recognition of faces in the fusiform gyrus. But let us not forget that this simple view of localization conceals the fact, stressed by Hughlings Jackson in 1879, that localizing the lesion that leads to a 'selective' disturbance of behaviour is not the same as understanding how that bit of the brain works. Jackson's logical argument is just as true today with respect to focal changes in blood flow revealed by functional brain imaging.
The acceptance of the idea of localization of function had one unfortunate effect. It came to be taken for granted that the senses of touch and vision and hearing were mapped on the surface of the brain and that there was a similarly orderly representation of the muscles, as shown in Fig. 1. But why there is a map at all was not recognized as a question of fundamental importance, despite the fact that nature went to enormous trouble to evolve genetic instructions which ensured that the retina of the eye and the surface of the body are represented on the surface of the brain in an orderly map and not higgledy-piggledy. Furthermore, a computer programmed to recognize patterns does not need within its components anything like a geographical map of the original scene. So why does the brain have one?
It became increasingly difficult to sidestep this question with the demonstration from 1970 onwards of multiple maps of the eye in the brain. A map is demonstrated by recording the electrical activity of clusters of nerve cells, determining where a visual stimulus must lie on the retina for it to excite these particular cells, and then moving the recording electrode to another group of cells. Using this procedure in anaesthetized animals, it was shown that the retina is mapped not once but many times in the cortex. The macaque monkey has at least ten mapped representations of the retina, and about twenty others where it is the nature of the stimulus rather than its position that is computed. At least a third of the cerebral cortex in the owl monkey is concerned with the multiple mapped representations of visual space (see primates, evolution of the brain in, Fig. 8).
What is the purpose of such an arrangement, which is not confined to vision, for there are now known to be several topological representations of the surface of the body and the musculature in monkeys? A plausible explanation concerns a well-known physiological phenomenon called lateral inhibition. In the eye itself, adjacent differences in the brightness of colour of the image are given prominence in the nerve signals that leave the eye. This is accomplished by a system of lateral inhibitory connections in the retina which ensure that nerve cells tend to inhibit their immediate neighbours. In an area of uniform illumination or colour, all cells are equally excited by the light and equally inhibited by their neighbours. But where there is a sharp difference in illumination, as at the image of a contour, the highly illuminated cells exert a powerful inhibition on their neighbours in the shade, and the difference in signals sent by the two groups of cells is enhanced. Lateral inhibition cannot create something out of nothing, but it can enhance one feature of the visual image at the expense of another. Lateral inhibition of this kind ensures that edges and contours are prominently coded in the signals from the eye.
There is now incontrovertible evidence from physiology and anatomy that lateral inhibition works in the brain as well as in the eye, and this provides the major reason for the existence of a map of the eye on the cortex of the brain. If the differences in illumination of adjacent parts of the eye are to be accentuated in the cortex, the sensory connections between the nerve cells concerned with the two adjacent parts of the image should be close together. In a map they are as close together as possible, and lateral interactions will be maximally efficient. If there were no map, so that nerve cells concerned with adjacent parts of the image were often far apart in the relevant cortical area, the problem of interconnecting the cells becomes formidable and the average length of a connection would be much greater, about 20 to 30 times greater in visual area 1 of man and monkey. In a map of the sensory surface the lateral interconnections between cells can all be local, and anatomy has shown this to be so.
But why are there many maps for each of the senses rather than just one? The answer is really the same. Inhibitory connections between neighbouring nerve cells of the cortex are now believed to be involved in coding many attributes of the visual image, such as colour, movement, disparity, orientation, size, and spatial periodicity. If all of these were to be attempted within one map, the local interconnections would again have to be longer and the problem of interconnecting the right cells would increase. By having many maps, each small and containing nerve cells concerned only with one or a few of the stimulus attributes just mentioned, the lateral interconnections can be kept as short as possible and the problem of interconnecting the right type of cell is minimized.
This simple idea has much to support it. First, although there are long fibre connections from one part of the brain to another, microscopy has shown that the connections within a particular map are short and predominantly inhibitory. Second, physiology has shown that nerve cells within a particular cortical representation of the eye tend to be concerned with a restricted range of stimulus qualities, such as orientation, distance, size, colour, or movement. Different maps deal with different stimulus qualities. Third, the human corpus callosum contains about 600 million nerve fibres connecting the two sides of the brain. They are grouped from front to back according to destination and function; if they were not, their average length would have to be longer. Fourth, there are many examples of very selective effects on visual perception of localized brain damage. Although they are rare, some patients suffer a highly selective disturbance of the perception of colour or position or depth or motion, as would be expected when the damage is occasionally restricted to one of the visual maps. Functional maps keep connections short and, therefore, keep the brain (and skull) small enough to be born through a narrow birth canal.
Although the different sensory qualities of the visual scene are initially coded in separate visual areas, our visual perception is unitary not fragmented, which means that the timing of the activity of cells in different visual areas must be precisely coordinated. If we look at a moving, spinning, coloured object and the nervous signals in one visual area are out of phase with all the others, some distortion should occur in what is seen. Indeed, fever, toxicosis, and brain damage can all lead to temporary visual perceptual dislocations. For example, in one part of the visual field objects may appear too large or too small, smooth movement may look jerky, contours may be multiplied, position and orientation be greatly misperceived.
Multiple brain maps of sensory and motor systems are now established. They permit the maximum efficiency and economy in the myriad interconnections between nerve cells responsible for analysing sensory signals. Their existence also throws light on what is now seen as an unwarranted controversy about localization of function. The cortical representations of the sensory attributes of stimuli, such as colour, may be confined to a few areas. The cortical events underlying certain complex and cognitive actions are probably so widely dispersed that no brain damage, however great, can either destroy them entirely or leave them wholly unimpaired. See also neuropsychology.
Fig. 1. The diagram on the left shows a slice through one side of the brain, roughly between the ears and through what is called the somatosensory cortex. There is a map of the body surface here. Note that some areas, like face and fingers, have a much larger cortical representation than others, i.e. the map is distorted. The diagram on the right shows the map of the muscles in the motor cortex on a slice of the other side of the brain and at a more anterior portion.
(Published 2004)
— Alan Cowey
- Bibliography
- Cowey, A. (2001). 'Functional localisation in the brain: from ancient to modern'. Psychologist, 14.
