Subscribe to Computing Intelligence

Friday, April 25, 2008

Scientist Appreciation: Donald Hebb

Donald Hebb was a fairly interesting fellow who I am quite a fan of (though I only superficially know of his work). I am a fan for some silly reasons, such as the fact that he was Canadian like me (and Canada does seem to have a deficit of famous scientists, since all our clever people seem to go into comedy). However, I am also a fan for the style and content of his scientific work.

Now, maybe it's because I'm tired and my brain is a little woozy from that exam I wrote this morning (which, incidentally, could definitely have gone better. Guessing on a non-multiple choice exam never bodes well), but I'm going to have to ask any readers to bear with my lack of structure and proper flow. I promised to have this written today, and I've already slept away the afternoon, so dishevelled, slightly ornery, hungry, and a little confused or not, I'm going to just have to forge ahead.

All right, that complete detour is over with, so back to Donald Hebb. One of the very nice things about him is his last name seems to be one of those wonderful last names uniquely suited to claiming ownership of ideas. Hebbian sounds like a proper word, in ways that my last name never could. More importantly than having a last name that is easy to turn into an adjective, Hebb also did some very important work that was worth claiming ownership of. His most famous work is now known as Hebbian Learning, and was basically the idea that a network of interconnected neurons could alter their synapses in such a manner to become strongly associated. Once such associations were created, the activation of just a portion of the neurons would lead to them all activating, and this could be one method of learning and memory. While, like most ideas in psychology, this is not the entire story, there are two major aspects of it that I am a huge fan of. The first is that it is mathematically reasonable, something that Hebb set about demonstrating (Hebbian learning was vitally important in spawning modern work on neural netowrks). The second, and equally important, is that it is rooted in physiological evidence, something Hebb also worked on demonstrating. Those are two aspects that I find unfortunately lacking in many works of psychology, and the fact that Hebb recognized the need for both makes him one of my favourite psychologists. It also helps that, at least according to Wikipedia, he and I agree on Freud lacking rigour.

Unfortunately, I will never get to meet Hebb to give him my appreciation, since he died a few months after I was born.