Free Will

[netmouse]

In the comments on yellowmouser's journal here and here I engage the question of whether or not free will is an illusion (I think not). Updated to add: J argues against me here.

What are your thoughts on the question? Is it an interesting question or purely a polemic debate?

Comments

[yaleartificer.livejournal.com]

I think that to say free will is an illusion because of biochemistry is to misunderstand what it is to be free. Free will is the ability to decide things. I choose to type "potato," and voila, it appears. Now, this choice was actually determined by several factors: I wanted to demonstrate that I could make an arbitrary decision. Historically, I have tended to choose words for root vegetables when I choose arbitrary words, and now I do so out of habit. At the same time, I usually choose "rutabaga," and so I decided to do something different this time. All of these factors were encoded in my brain, and my decision process was implemented by the firing of neurons in a physical way. But does it make sense to say that because I had reasons for what I did, that I was not free? No, because to require people to do things for absolutely no reason is too strict a criterion, and is not free will at all. How could we be free if we were never allowed to have reasons for what we did? We must be able to decide things based on reasons if free will is to be meaningful. And those reasons are encoded in a physical substrate: the structure of the neurons that make up my brain. So to say "I did this because the physics dictated it" is the same as saying "I had reasons for doing this," only to look at it at the wrong level of abstraction.

Another interesting point is that there could never be a human or machine that could be certain of another person's actions, because said human or machine's own workings would be a part of the physical system that is affecting the human it is trying to predict. Said human or machine would then have to be able to predict its own behavior, which is classic Halting Problem material. Even if we are more or less deterministic, I'm guessing predicting our actions is not computable.