Wednesday, January 20, 2010

Brain and the Computer by John von Neumann

The Brain - Computer Debate

The brain - computer comparison was investigated in a book by John von Neumann back in the 1950s. In it, the father of all computers looked at the neuro anatomy of brains. He concluded that "whatever computational regime governs the brain it must be fundamentally different from the one governing von Neumann machines". After a detailed and skillful exploration of the brain at the end of the book he concludes "that the brain has a different physical organization and uses a computational strategy that is different from the von Neumann architecture".

The brain is different in that "a solution to the problem of the brains slowness of its neuronal activities and given the low accuracy of its typical representation" is compensated by the "logical depth" only imaginable as "massively parallel analog machine". Because it is "the acquired global configuration of those many millions, nay trillions of synaptic connections that embodies whatever knowledge and skills the brain may have aquried" that makes it what it is and subsequently what it does. Along the sidelines of computer science evolved empirical neuroscience.

Advancement in electron & confocal microscopy, patch clamping, electro - and magnet encephalography, CAF scans, PET scans, MRI and fMRI scans are the technical instruments responsible for its rise as cognitive neuroscience and the insights they give into the workings of the brain. The American Association for Advancement of Science Journal issued an article, which was coauthored by Eric Richard Kandel Nobel Laureate, and subtitled “Breaking down scientific barriers to the study of brain and mind”. The article covers very interesting ground but ends up with the conclusion that the neuroscience of higher cognitive processes is only begining. One of the leading Neuroscientists Randy Gallistel pointed out that “we clearly do not understand how the nervoussystem computes or even the foundation of its ability to compute even for the small set of arithmetic and logical operations that are fundamental to any computation”. Thus we seem to be today as far away from a more meaningful understanding than Hans Lukas Teuber, one of the founders of contemporary cognitive Neuroscience was 50 years ago, when he wrote that "it may seem strange to begin with the claim that there is no adquate definition of perception and to end with the admission that we lack a neurophysiological theory. Althou this was the most that could be said".

What governs von Neumann machines are computer languages. At the heart of these languages lies formal logic. Ever since formal logic and logical syllogism encapsulate connectedness in reasoning, philosophers and mathematicians have entertained the idea to reduce matter to pure form. About a millenia later just as Aristotle created the abstraction of logic by focusing attention on the form of Syllogisms rather than their meaning of their propositions, Boole a young english mathematician, abstracted algebra by looking closely at its rules of operation wherby he showed that they formed a consistend system themselves that didn't have to apply to numbers at all. He showed how numbers could be used to reduce logical propositions to the form of equations that could be solved according to ordinary algebraic rules. Boolen then demonstrated the generality of his system by showing how it could be used to derive ANY true conclusion logically contained in any given set of propositions. Thus, the goal of building a formal system that could generate and prove all laws of science or mathematics was shown to be unatainable. Aristotle's investigation into the old problem posed by the destinction between form and matter ended with the attempts of formal mathematics that proved that its unsolvable. Form would never totaly replace meaning. With the Syllogism at the center, formal languages and computer technologies formed a symbiant relation. What makes such a computer usable is the language governing its operation. But unlike a natural language will govern the operations of von Neuman Machines, so are all formal languages deviced inappropriate in governing the human mind/brain.

No comments:

Post a Comment