Christoph Koch writes in Biophysics of Computation (Oxford University Press, 1999, p. 87):
Given an approximate density of 100,000 cells by mm3 in the primate, a synaptic density of 6 x 108 per mm3, a total surface area of about 100,000 mm2 for one hemisphere, and an average thickness of about 2 mm, the average human cortex contains on the order of 20 billion neurons and 240 trillion synapses (2.4 x 1014) [...]
This reasoning is not yet complete as while the number of synapses falls from the argument, the number of neurons doesn't. However, in The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 71 note 5), the same Christoph Koch says:
Given a packing density of 50,000 cells per mm3, a total surface area of 2 x 100,000 mm2, and a thickness of about 2 mm, the average human cortex contains on the order of 20 billion neurons and 200 trillion synapses (2 x 1014).
Also in The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 26), Koch says:
All the visual information that the brain can access is implicitly encoded by the membrane potentials of the more than 200 million photoreceptors in the two eyes.
Since photoreceptors are neurons, those facts combined provide a formal proof (we insist on the terms "formal proof" here) of the elaboration of symbols by the brain. The proof is that if each image is using 200 million neurons, only 100 images (100 x 200 million = 20 billion) can be stored by the entire brain, obviously insufficient. Information needs to be reduced (which we actually know it is, but that's besides the argument here).
This sets the scene for a formal study of symbol construction, starting with a very small number of photoreceptors and going up from there.
Bertrand du Castel
Given an approximate density of 100,000 cells by mm3 in the primate, a synaptic density of 6 x 108 per mm3, a total surface area of about 100,000 mm2 for one hemisphere, and an average thickness of about 2 mm, the average human cortex contains on the order of 20 billion neurons and 240 trillion synapses (2.4 x 1014) [...]
This reasoning is not yet complete as while the number of synapses falls from the argument, the number of neurons doesn't. However, in The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 71 note 5), the same Christoph Koch says:
Given a packing density of 50,000 cells per mm3, a total surface area of 2 x 100,000 mm2, and a thickness of about 2 mm, the average human cortex contains on the order of 20 billion neurons and 200 trillion synapses (2 x 1014).
Also in The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 26), Koch says:
All the visual information that the brain can access is implicitly encoded by the membrane potentials of the more than 200 million photoreceptors in the two eyes.
Since photoreceptors are neurons, those facts combined provide a formal proof (we insist on the terms "formal proof" here) of the elaboration of symbols by the brain. The proof is that if each image is using 200 million neurons, only 100 images (100 x 200 million = 20 billion) can be stored by the entire brain, obviously insufficient. Information needs to be reduced (which we actually know it is, but that's besides the argument here).
This sets the scene for a formal study of symbol construction, starting with a very small number of photoreceptors and going up from there.
Bertrand du Castel
0 comments:
Post a Comment