Pages

Monday, October 02, 2006

Against Epiphenomenalism.

The critical feature of epiphenomenalism, as we have seen, is that our mind states have no reciprocal effect on our bodies. That is, our brain states give rise to mind states-- especially qualia-- that are irreducible to simply physical descriptions. But our reactions to the outside world are fully determined by processes within the brain-- physical processes that obey physical, causal laws. The qualia of the mind, then, have no impact on our behavior.

In some sense, I can almost buy this viewpoint, as unusual as it might be. I can comprehend how the sensation of pain is irrelevant for my bodies flinching away from a dangerous situation. If I put my hand on a hot stove, I can understand how the heat will physically affect my nerve endings, which will send a particular signal to the sensory processor in my brain, which will be appropriately connected to the motion-control center of my brain, which will be programmed (through the slow arduous refinement of past evolution) to react by flinching my hand away from the heat source. At no point in this process does the "Ouch, I'm in pain!" feeling do any work.

However...

There are some experiences for which I believe this kind of story simply cannot be told. The best example, in my opinion, is our experience of music. I cannot fathom how sound waves impinging on my eardrum, in turn sending an electrical signal to a part of the brain, can produce a feeling of pleasure-- even if that "pleasure" feeling is simply a reaction in another part of the brain. I cannot understand how some music would produce pleasure, while other music-- which under the epiphenomenalist's view, is ultimately just a different pattern of electrical signals sent from the eardrum-- would produce revulsion. It just doesn't make sense. I don't know if that is a knockdown argument against epiphenomenalism, but it's enough for me to drop it completely.

In the area of music-- and really, in our experience of any kind of art-- the qualia of our experiences have to be doing the bulk of the work. I just can't see it any other way.