Because of the discovery of cells in the visual cortex which can be triggered by stimuli of ever increasing complexity, some people have wondered if things are not leading in the direction of "one cell, one concept" -for example, you have a "grandmother cell" which would fire if, and only if, your grandmother came into view. This somewhat humorous example of a "superhypercomplex cell" is not taken very seriously.
However, it is not obvious what alternative theory seems reasonable. One possibility is that larger neural networks are excited collectively by sufficiently complex visual stimuli. Of course, the triggering of these larger multineuron units would somehow have to come from integration of signals emanating from the many hypercomplex cells. How this might be done, nobody knows.
Just when we seem to be approaching the threshold where "symbol" might emerge from "signal", the trail gets lost- a tantalizing unfinished story. We will return to this story shortly, however, and try to fill in some of it.
Early I mentioned the coarse-grained isomorphism between all human brains which exist on a large anatomical scale, and the very fine-grained, neural-level isomorphism which exists between earthworm brains. It is quite interesting that there is also an isomorphism between the visual processing of cat, monkey, and human, the "grain" of which is somewhere between coarse and fine. Here is how that isomorphism works.
First of all, all three species have "dedicated" areas of the cortex at the back of their brains where visual processing is done: the visual cortex. Secondly, in each of them, the visual cortex breaks up in three subregions, called areas 17, 18, and 19
of the cortex. These areas are still universal, in the sense that they can be located in the brain of any normal individual in any of the three species. Within each area you can go still further, reaching the "columnar" organization of the visual cortex. Perpendicular to the surface of the cortex, moving radially inwards towards the inner brain, visual neurons are arranged in "columns" - that is, almost all connections move along the radial, columnar direction, and not between columns. And each column maps onto a small, specific retinal region.
The number of columns is not the same in each individual, so that one can't find "the same column". Finally, within a column, there are layersin which complex neurons tend to be found. (The hypercomplex neurons tend to be found in areas 18 and 19 predominantly, while the simple and complex ones are found mostly in area 17. )It appears that we run out of isomorphisms at this level of detail. From here on down to the individual neuron level, each individual cat, monkey, or human has a completely unique pattern -somewhat like a fingerprint or a signature.
One minor but perhaps telling difference between visual processing in cats' brains and monkeys' brains has to do with the stage at which information from the two eyes is integrated to yield a single combined higher-level signal. It turns out that it takes place slightly later in the monkey than in the cat, which gives each separate eye's signal a slightly longer time ti get processed by itself.
This is not too surprising, since one would expect that the higher a species lies in the intelligence hierarchy, the more complex will be the problems which its visual system will be called upon to handle; and therefore signals ought to pass through more and more early processing before receiving a final "label". This is quite dramatically confirmed by observations of the visual abilities of a newborn calf, which seems to be born with as much power of visual discrimination as it will ever have. It will shy away from people or dogs, but not from other cattle.
Probably its entire visual system is "hard-wired" before birth, and involves relatively little cortical processing. On the other hand, a human's visual system, so deeply reliant on the cortex, takes several years to reach maturity.
Godel, Escher, Bach: An eternal Golden Braid by Douglas R. Hofstadter.
on page fuck 'em.
However, it is not obvious what alternative theory seems reasonable. One possibility is that larger neural networks are excited collectively by sufficiently complex visual stimuli. Of course, the triggering of these larger multineuron units would somehow have to come from integration of signals emanating from the many hypercomplex cells. How this might be done, nobody knows.
Just when we seem to be approaching the threshold where "symbol" might emerge from "signal", the trail gets lost- a tantalizing unfinished story. We will return to this story shortly, however, and try to fill in some of it.
Early I mentioned the coarse-grained isomorphism between all human brains which exist on a large anatomical scale, and the very fine-grained, neural-level isomorphism which exists between earthworm brains. It is quite interesting that there is also an isomorphism between the visual processing of cat, monkey, and human, the "grain" of which is somewhere between coarse and fine. Here is how that isomorphism works.
First of all, all three species have "dedicated" areas of the cortex at the back of their brains where visual processing is done: the visual cortex. Secondly, in each of them, the visual cortex breaks up in three subregions, called areas 17, 18, and 19
of the cortex. These areas are still universal, in the sense that they can be located in the brain of any normal individual in any of the three species. Within each area you can go still further, reaching the "columnar" organization of the visual cortex. Perpendicular to the surface of the cortex, moving radially inwards towards the inner brain, visual neurons are arranged in "columns" - that is, almost all connections move along the radial, columnar direction, and not between columns. And each column maps onto a small, specific retinal region.
The number of columns is not the same in each individual, so that one can't find "the same column". Finally, within a column, there are layersin which complex neurons tend to be found. (The hypercomplex neurons tend to be found in areas 18 and 19 predominantly, while the simple and complex ones are found mostly in area 17. )It appears that we run out of isomorphisms at this level of detail. From here on down to the individual neuron level, each individual cat, monkey, or human has a completely unique pattern -somewhat like a fingerprint or a signature.
One minor but perhaps telling difference between visual processing in cats' brains and monkeys' brains has to do with the stage at which information from the two eyes is integrated to yield a single combined higher-level signal. It turns out that it takes place slightly later in the monkey than in the cat, which gives each separate eye's signal a slightly longer time ti get processed by itself.
This is not too surprising, since one would expect that the higher a species lies in the intelligence hierarchy, the more complex will be the problems which its visual system will be called upon to handle; and therefore signals ought to pass through more and more early processing before receiving a final "label". This is quite dramatically confirmed by observations of the visual abilities of a newborn calf, which seems to be born with as much power of visual discrimination as it will ever have. It will shy away from people or dogs, but not from other cattle.
Probably its entire visual system is "hard-wired" before birth, and involves relatively little cortical processing. On the other hand, a human's visual system, so deeply reliant on the cortex, takes several years to reach maturity.
Godel, Escher, Bach: An eternal Golden Braid by Douglas R. Hofstadter.
on page fuck 'em.
