Our current model for learning that seems to have the most salience is the neural network approach. We have somewhere between 10 billion to 100 billions neurons in our brains, each of which has ~10,000 distinct connections (synapses) to other neurons. A nerve cell essentially sums up all the incoming impulses and if the total reaches a certain threshold, the cell fires, which then affects all the neurons it connects to, which undergo the same summation process. When a group of neurons all fire together the connections between them tend to strengthen. Over time, a network of related neurons develops. One can imagine neural networks of hundreds to hundreds of millions of cells which form stable structures; this would be the neuro-physiological basis for a mental "object", or symbolic representation; I would add that this would be akin to the template I described my earlier post, Changes: PartI.
As an example, the quality of "dogness" would be represented by a stable network of related neurons. Some might be in the sensory (visual, auditory, olfactory, tactile) areas, others in the language areas, and still others in the emotional circuitry. Seeing a dog would cause this network to fire with all the secondary effects that would follow. Imagining a dog in one's mind would trigger some of the same neurons to fire, though with slightly different connections. Someone who had been traumatized by a dog would have more powerful connections to the part of the network that signals danger; they would be more likely to have an anxious reaction to a strange dog, (even in the absence of any threat from the dog). All sorts of levels of complexity could be encoded (put into memory) in this way. If the original trauma was from the person's very early life, they might not even remember the event which triggered their fear of dogs; all they would know is that dogs terrify them. Since there are lots of dogs around, this could lead to all sorts of problems. The person would have to maintain a high level of vigilance to protect against being surprised by a dog. I would suggest that this situation would represent a neural network that has pseudo-stability. It requires high levels of mental energy to maintain the particular network (anxiety can be exhausting).
If our hypothetical person entered therapy to treat his dog phobia, we might try to desensitize him to the template "dog". We might find memories of frightening dogs in the person's past, and help them see these incidents as being relevant to a child, but not to an adult. We might introduce the person to small, peaceful dogs; some of the connections to the terrifying dog of his template would be weakened. Perhaps we would move up to a Golden Retriever (a type of dog that is nearly impossible for an adult to fear) and more of the connections would fray. The neural network which includes "dog" and "terror" would become less stable. Over time, with enough unraveling of the connections to the terrifying "dog" we would notice an interesting phenomenon. There would come a point where one more intervention, one more recovered memory of a dog, one more recognition of how the terror doesn't belong to "this dog", and suddenly, there would be a moment of insight; the person would realize he has nothing to fear from dogs that are wagging their tails and licking him.
In Psychoanalysis, we call this the "mutative interpretation", usually a series of related comments that lead to structural change. When the patient has an insight like this, they can almost feel the energy being freed up in their minds. It has been described as an ineffable experience, but if we think of what happens in terms of the network suddenly destabilizing and then immediately reorganizing at a lower energy level, it makes a great deal of sense.
I have oversimplified an extraordinarily complex situation, but I think it has some relevance for the question of how people change and how societies change, and I will return to this shortly.
Recent Comments