Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!husc6!bbn!bbn.com!pineapple.bbn.com!barr From: barr@pineapple.bbn.com (Hunter Barr) Newsgroups: comp.ai Subject: Re: More Free Will Summary: Possible flaws in channic's resoning on subjective experience. Message-ID: <730@papaya.bbn.com> Date: 12 May 88 19:08:58 GMT References: <3200017@uiucdcsm> Sender: news@bbn.com Lines: 160 In article <3200017@uiucdcsm> channic@uiucdcsm.cs.uiuc.edu writes: >"How do I know?" is as old a question as western philosophy if not all >philosophy. The reason the question is that old is because subjective >personal experience is undeniable. I know I have free will because that >is my experience. You have seen optical illusions where a straight line is made to appear curved. No-one doubts that your subjective experience tells you that the line is curved. In fact, because all humans are built the same in certain respects (at least those who are able to see and thus can experience optical illusions), all humans will have that same subjective experience when they look at that straight line. The same is true of many other optical illusions, auditory illusions, and, naturally, the illusion that we call free will. No-one says we do not experience it, only that we *mis-understand* what we experience. After I have taken a straight-edge to determine that the line is straight, I may still "feel" that I am looking at a curved line, but I now believe the drawing's caption, which calls it a straight line. No-one has yet defined free-will so that a straight-edge can be found for it. But many humans activities have been shown to be less voluntary than they appear subjectively. For expample: As a child, I thought that I chose what I wanted to eat, and I felt I had completely free will in this matter. I noticed that I always chose the M&Ms over the spinach (no matter what Popeye said), but I still felt that my choice was completely free. Now I know a little about the human digestive track, body-chemistry, and psychology; I can see that these played a big part in my "choices". Such a big part that I now think they completely determined my choices. So it seems useless to talk about free will in this case. Any child will make similar choices, most the exact same ones, because that is how children are built. Only external factors will change the child's choices, such as reward/punishment for choosing one food over another. (I consider learning about the real world to be a form of external reward/punishment. For example, I learned that a steady diet of M&Ms could lead to a scary visit to the Dentist.) Choices we make as adults seem much the same, probably depending on factors we do not yet comprehend, just as we did not comprehend the factors which led us to choose M&Ms as children. As I said, no-one has yet provided a reliable straight-edge for us to use on the free will experience (a definition of free will would help.) But each experience with a choice which we *do* understand (like M&Ms over spinach) provides one little peice of that straight edge. All the little peices in my experience look like they will fit a straight-edge called " determinism." When we learn more about the factors which influence our choices, we will develop more or less confidence in this hypothetical straight-edge. But until we understand most of these factors, we cannot claim to have this ill-defined "free will", whatever our subjective experience. This puts me in agreemnet with someone you quoted who said: >> No one denies that we humans experience free will. But that >> experience says nothing about its nature; at least, nothing ruling >> out determinism and chance. I think this gives me an answer to your question: >From where does this tendency to IGNORE subjective experience when discussing >SUBJECTIVE PHENOMENA originate? My own (and your own) experience of free will >tells me (and you) a great deal about its nature. In fact, this experience is >the most reliable source of information regarding free will. And the >experience is neither deterministic or random. No matter what decision I make, >whether choosing a political candidate, a career, or a flavor of ice cream, >I experience neither that my choice is determined, nor random, NOR some >nebulous combination of the two. I answer that this discussion itself is proof that we do not ignore subjective experience. It does indeed tell us a great deal, and provides essential insight, but it is by no stretch of the imagination our most reliable source of information. Try to adapt your reasoning to the straight line illusion. Following your reasoning, I should not bother to rummage around in my desk for a straight-edge; nor should I doubt in any way my subjective experience that it is curved. Why, you don't want me to think about things at all! You just want me to sit there and "experience" that the line is curved, as if that experience matched reality. What if my ability to better understand reality makes me able to help people? What if my ability to distinguish germ theory from superstition enables me to wipe out smallpox? What if my ability to distinguish straight lines from curved ones enables me to build safer cars? Then maybe my ability to distinguish demermined behaviour from this nebulous "free will" will enable me to wipe out a mental disease. Whatever turns out to be the root our "free will" experience, we will not learn to understand it merely by sitting here and experiencing it subjectively. >Free will explained as an additive blend of determinism and chance >directly attacks the concept of individual responsibility. Can any >machine, based on this theory of free will, possibly benefit society >enough to counteract the detrimental effect of a philosophy which >implies that we aren't accountable for our choices? We say, "A tornado was responsible for destroying my house," and, "John's dog is responisble for this scar on my leg." If we could stop the tornado from touching down elsewhere, we would. Under most circumstances we actually do prevent John's dog from biting anyone else, either by incarcerating it or killing it. The same will always be true of people. Where we can stop them from repeating their offenses, we will. What about one-time offenders? Say someone kills her husband and we don't think she'll do it again. Punish her on the basis of how likely it is that she will commit the same crime again? No. This would produce the damage you fear. For the prevention of crime it is important that punishment be applied predictably and evenly, with as few exceptions as possible. This means we should keep the current idea of responsibility around in Law to do the job it has always done-- support prevention by making sure a crime is punished predictably. This may mean a stricter interpretation of "Ignorance of the Law is no excuse." Yours is an excellent argument against treating insane offenders differently from sane ones. It is too easy to say that anyone who commits a crime has demonstrated mental illness by committing the crime. But if in fact determinism turns out to be true, then we will want to convert all our prisons to mental hospitals. Can we work to prevent crime as the same time as we work to cure the offenders? Hmmm... why is there such a thing as an "insanity defense" in Law? I guess the more we learn about determinism in the mind, the fewer places "free will" has to hide. So psychologists really can claim "She didn't know what she was doing, your Honor." If our ability to diagnose mental illness continues to outstrip our ability to correct it, then we will have the unpleasant necessity of punishing people who really couldn't help it. But this has been happening for thousands of years anyway. I guess the only course I can see is to continue the perceived injustice of punishing the mentally ill, except in cases where we can cure them. Can we cure them now? Not reliably, but... oh, this *is* a hard problem! We may have to continue the age old system of punishment, sacrificing cures for prevention for a very long time after some cures become possible. There are lifetimes of thinking yet to be done on this topic. We may really get started figuring out how to solve crime when we figure out what the free will experience is. > >I feel that I'm on a soap box and I apologize. I do not wish to >single out AI researchers as the cause of the world's problems. >In fact, the great response to this topic indicates many more >people are thinking about these issues than the literature of >the field suggests. I only wish to add the observation that free will, >values, and beliefs are the reason we're all in this game, so while >we're playing we shouldn't pretend these things don't exist or are >any less complicated and miraculous than they really are. > > >Tom Channic >University of Illinois >channic@uiucdcs.cs.uiuc.edu >{decvax|ihnp4}!pur-ee!uiucdcs!channic I have no problem with your getting up on a soapbax-- after all, you left room for me up here next to you! The only way I can get the benefit of your knowledge is to expose my ignorance and spark some true communication. I, myself, apologize for taking up so much space saying what I have to say. Thanks for listening (reading, that is.) ______ HUNTER