Friday, September 26, 2008

Classical conditioning in terms of Hebb's postulate.

Ever since Ivan Pavlov's research at the turn of the last century demonstrated conditioned reflexes, we have had strong evidence that certain cues can trigger responses via conditioned association. But can these behavioral understandings be reduced to supervene upon neurological notions?

As Ivan Pavlov researched the digestion of dogs from a physiologists point of view, he came to understand psychological notions we take as canonical. He provided dogs with meat powder to initiate salivation. Soon, he realized the lab technician in the dog's presence alone would cause the dogs to salivate. He prompted the dogs with a number of subsequent cues: bells, whistles, metronomes, as well as environmental visual stimuli along with the meat powder, only to recreate the salivation response with the cue sans meat powder. Thus we have derived notions of classical conditioning in which an unconditioned, natural stimulus in simultaneous occurrence with another stimulus can condition the latter stimulus to evoke the same natural response as the former, natural stimulus.

In the late 40s, Donald Hebb provided his theory on neural networks and the biological process of living nervous systems. His theory consisted in positing notions of synaptic plasticity, which describes the change in strength between neural connections. Often, Hebb's postulate is summarized as "cells that fire together, wire together." Or, a presynaptic cell A, which consistently fires with postsynaptic cell B, will consistently increase in strength. This notion has been instrumental in understanding neural networks as memory hypotheses and learning called Hebbian learning. When seen on a general scale, this process seems not unlike the model of classical conditioning described above.

No comments: