Sam Harris has a book out called The Moral Landscape: How Science Can Determine Human Values where he discusses among other things the nature of free will. I have not read his book, but I did read his recent posts about free will on The Huffington Post (here, here, and here). His thesis is that from a neurological perspective there is no evidence for the existence of free will in the human brain. He makes a good case, but there are some flaws in his arguments that I'd like to address.
The first and most glaring flaw, which Harris himself may not see as a flaw, is that it is difficult to see how there could be any evidence for free will from a neurological perspective. Analyzing the human brain at that granularity is a bit like analyzing a computer at the level of transistors. Actually, it's quite a lot like analyzing a computer since the brain is basically doing the same thing a computer is doing which is storing and processing information, and sending signals. The main difference is that the brain is far more complex than the average computer and mixes the durability of hardware with the malleability of software. But essentially the same things are going on in both.
The comparison is important because viewed from that level both the brain and the computer are essentially deterministic. The same signal sent to the same part of the brain/computer in the same state should have the same results. I say essentially, because in both cases there is a certain amount of non-determinism in both systems, but despite this they both function under a certain set of rules. In a deterministic system there can not be any room for free will or choice of any kind.
If this was all there was to say about free will, Harris could have made this argument and stopped there.
This leads us to the second problem with his argument, which is that while at the granularity of neurons we have a largely deterministic system, this is not the granularity at which the notion of thoughts have much meaning. This discrepancy was discussed with great eloquence by Douglas R. Hofstadter in his masterful book, Gödel, Escher, Bach: An Eternal Golden Braid. In it he compares the brain to an ant hill in which the ants themselves are simply sending signals back and forth, unaware of the larger purpose of the ant hill (which appears as a sort of consciousness which has befriended an anteater in one of the book's dialogues).
The important distinction in the analogy is between signals (the individual ants), which represent very simple pieces of information, and symbols (the larger activities of the anthill of which the ants are unaware) which are much richer and more complex and exist on a much higher level. It is at the level of symbols at which the notion of though starts to make sense. Thoughts are composed of words and images, both of which are symbols used to represent reality. It is only at this level that we can discuss concepts like will or intention with any clarity.
Now, this does not mean that at the level of symbols and thoughts that the system is any less deterministic. If we build a complex system using simple mechanical components it will still be a machine whether it is carbon or silicon based. On this I think Harris and I both agree with Alan Turing. Computers of a sufficient degree of complexity should be just as capable of thought as human beings are. Rather, I make the distinction in order to point out why Harris' particular argument is flawed.
Finally, and most crucially, I think Harris fails to properly define what free will is. This is understandable because free will, like many philosophical concepts, is very difficult to define. Like the notion of cause and effect, it's very hard to explain what free will means without ending up in an infinite regress of circular definitions. However, this does not mean one should not make a good faith effort at doing so.
I'll stop here for now, and discuss my own opinions on free will in my next post on the subject.
0 comments:
Post a Comment