Eye, Brain, and Vision
Home
Book
Illusions
Biography
Publications
David Hubel's
Home
Book
Illusions
Biography
Publications
HMS
 
     
                                                                                    
   
 



4



     THE PRIMARY VISUAL CORTEX
After Kuffler's first paper on center-surround retinal ganglion cells was published in 1952, the next steps were clear. To account for the properties of the cells, more work was needed at the retinal level. But we also needed to record from the next stages in the visual pathway, to find out how the brain interpreted the information from the eyes. Both projects faced formidable difficulties. In the case of the brain, some years were required to develop the techniques necessary to record from a single cell and observe its activity for many hours. It was even harder to learn how to influence that activity by visual stimulation.


                    TOPOGRAPHIC REPRESENTATION
Even before further research became possible, we were not completely ignorant about the parts of the brain involved in vision: the geography of the preliminary stages was already well mapped out (see the illustration on the next page). We knew that the optic-nerve fibers make synapses with cells in the lateral geniculate body and that the axons of lateral geniculate cells terminate in the primary visual cortex. It was also clear that these connections, from the eyes to the lateral geniculates and from the geniculates to the cortex, are topographically organized. By topographic representation, we mean that the mapping of each structure to the next is systematic: as you move along the retina from one point to another, the corresponding points in the lateral geniculate body or cortex trace a continuous path. For example, the optic nerve fibers from a given small part of the retina all go to a particular small part of the lateral geniculate, and fibers from a given region of the gcniculate all go to a particular region of the primary visual cortex. Such an organization is not surprising if we recall the caricature of the nervous system shown in the figure on page 6, in which cells are grouped in platelike arrays, with the plates stacked so that a cell at any particular stage gets its input from an aggregate of cells in the immediately preceding stage.

 

 

 

 

 

 

 

 

 

 

 

 

 



In the retina, the successive stages are in apposition, like playing cards stacked one on top of the other, so that the fibers can take a very direct route from one stage to the next. In the lateral geniculate body, the cells are obviously separated from the retina, just as, equally obviously, the cortex is in a different place from the geniculate. The style of connectivity nevertheless remains the same, with one region projecting to the next as though the successive plates were still superimposed.
The optic-nerve fibers simply gather into a bundle as they leave the eye, and when they reach the geniculate, they fan out and end in a topographically orderly way. (Oddly, between the retina and geniculate, in the optic nerve, they become almost completely scrambled, but they sort out again as they reach the geniculate.) Fibers leaving the geniculate similarly fan out into a broad band that extends back through the interior of the brain and ends in an equally orderly way in the primary visual cortex. After several synapses, when fibers leave the primary visual cortex and project to several other cortical regions, the topographic order is again preserved. Because convergence occurs at every stage, receptive fields tend to become larger: the farther along the path we go, the more fuzzy this representation-by-mapping of the outside world becomes.
An important, long-recognized piece of evidence that the pathway is topographically organized comes from clinical observation. If you damage a certain part of your primary visual cortex, you develop a local blindness, as though you had destroyed the corresponding part of your retina.
The visual world is thus systematically mapped onto the geniculate and cortex. What was not at all clear in the 1950s was what the mapping might mean. In those days it was not obvious that the brain operates on the information it receives, transforming it in such a way as to make it more useful. People had the feeling that the visual scene had made it to the brain; now the problem for the brain was to make sense of it—or perhaps it was not the brain's problem, but the mind's. The message of the next chapters will be that a structure such as the primary visual cortex does exert profound transformations on the information it receives. We still know very little about what goes on beyond this stage, and in that sense you might argue that we are not much better off.
But knowing that one part of the cortex works in a rational, easily understood way gives grounds for optimism that other areas will too. Some day we may not need the word mind at all.





   
 
 
Contents
Next Page



 
 
 
 
 
The visual cortex in a monkey, stained by the Golgi method, shows a few pyramidal cells—a tiny fraction of the total number in such a section. The entire height of the photograph represents about 1 millimeter.
A tungsten microelectrode, typical of what is used for extracellular recordings, has been superimposed, to the same scale.
The visual pathway, from eyes to primary visual cortex, of a human brain, as seen from below. Information comes to the two purple-colored halves of the retinas (the right halves, because the brain is seen upside down) from the opposite half of the environment (the left visual field) and ends up in the right (purple) half of the brain. This happens because about half the optic-nerve fibers cross at the chiasm, and the rest stay uncrossed. Hence the rules: each hemisphere gets input from both eyes; a given hemisphere gets information from the opposite half of the visual world.

Next Page
Contents