Wednesday, December 17, 2008

I've been reading a book about computation and information/quantum information theory. I've been thinking about the brain as a computer.



Use learning a language as an example. A computer program could provide the necessary codes for translating words from an unknown language input into words from a known language output. [If input aqua is 0011111 then output 1100000 for water.] This is what a human brain does as well when a person is first learning a language, but this kind of crude translation is not considered true understanding or fluency. A human brain eventually learns to directly substitute the words from the new language for words from the old one, doing away with the translational code entirely. As far as I know, a computer program could not do this.

On the other hand, say there is a computer program which reads more like this: if input equals that liquid stuff coming out of the tap, output equals aqua [oo11111] or water [1100000] depending on second input of either Latin or English. The second input is thought of as the language used in a conversation. This is plausible, but doesn't yet explain the transition between the two "programs", which the human brain is clearly capable of.

No comments: