Towards A Good Samaritan World

Friday, September 29, 2006

ON FREE WILL

One reason to reject free will is that one believes in a materialist universe governed by deterministic physical laws, which leaves no room for an entity with the power of choice (in the unreconstructed sense: choice is a form/source of non-predetermined causation such that the future cannot be predicted, even in principle, on the basis of the past). The "compatibilist" position attempts to make free will seem compatible with a deterministic universe by exploiting a certain sloppiness in language: sometimes we will use the phrase "of my own free will" in an especially emphatic sense, to signify not merely an act of choice, but an absence of any unusual constraining factors. For example, if I rob a bank because someone has threatened to kill my family if I don't, I might say "I didn't do it of my own free will, I was compelled," to distinguish it from the usual situation where someone robs a bank in the absence of this dire and very specific threat. This is a different situation, though, than if someone drugs me into a coma, dresses me in a robotic suit that can force pre-programmed movements of my limbs, records and splices my voice to state the necessary words, so that the suit marches me into a bank, reproduces my voice saying "Give me all the money or I'll kill you," grabs the cash, and marches me back out, with me being little more than a spectator to the whole episode. The latter is not a case of choice; the former still is.

But even those without a pre-commitment to materialism might be confused by the way that free will interacts with the fact that people typically have "reasons" for their actions.

If you ask me "Why did you do that?" I will not typically answer "Because I chose to," full stop. I'll say, "I bought that car because I needed to commute to work and it was the cheapest I could find." "I took that job because it was more interesting than my previous one." "I went home from the party early because I was feeling ill."

It's true that sometimes people appear to have no reason for their actions. "I don't know why I did it; there was no reason for it..." But this is certainly not the paradigmatic case of choice, and actually if someone seems to be completely unable to account for their reasons, this unusual circumstance might lead us to doubt whether the person was really acting entirely voluntarily, or whether they were perhaps somehow compelled by psychic forces they did not understand-- "Perhaps," we might ask, "you were recently hypnotized?"

Yet if someone has reasons for their actions, doesn't that mean their choice was predetermined? Given circumstance (a) that I needed to commute to work, and (b) that I wanted to save money, my choice to buy that particular car was predetermined. Given that I was feeling ill, my choice to go home from the party was predetermined, no? No exercise of free will here at all.

A somewhat superficial way of explaining how free will is compatible with having reasons for actions is to say that reasons are not sufficient. Given circumstances (a) and (b), I might still not have bought that car. I might have bought another that was slightly more expensive. I might have decided to get a new job closer to home. I might have decided to use a bike and/or public transit to commute. I might have done nothing, failed to go to work, and been fired.

Of course, if we specified more reasons (c), (d), and (e), my choice might narrow further. One might surmise that this is asymptotic: we might never, in practice, be able to list enough situational elements to make my choice completely determined, but if a truly complete description of the situation were made, it would become clear that I could have done nothing else. (We are speaking of mental reasons here by the way, not physical causes, so this has nothing to do with materialist determinist argument.) But this claim confesses itself to be unproven and unprovable. There is really no reason to think that the range of choices narrows asymptotically to one as the situation is more completely specified.

More profoundly, we have choice with respect to what constitutes "reasons." To a businessman, "it was an opportunity to make money" is the best of reasons for an action. To a monk, this is no reason for an action at all, indeed, if he has taken a vow of poverty, it is a reason to avoid that action. This points towards a level of choice that is deeper and more mysterious than the everyday choices we make: our choice of our own telos, our ultimate purposes. There is something inexpressible, or at any rate incompletely expressible, about these ultimate purposes, and the normal accounting of our actions that we offer when asked "why did you do that?" has a certain arbitrary and incomplete character. We are often aware that we could have answered the question in many different ways; we select the explanation, perhaps, that the specific listener will understand or find appropriate, and we are telling the truth, but not the whole truth. For every one of our actions springs ultimately from these deep purposes that people may know us for years without truly understanding; and those few who have the fullest understanding of these purposes are those whom we call (in a sense a bit stronger than the casual one) friends.

6 Comments:

  • As a matter of fact, the kind of compatibilist view of free will covers the general, day-to-day moral usage, since we generally want to know what people will do, ceteris paribus, to assign moral qualities to that person. Put another way, what we learn about a person from their actions depends on external factors that would tend to coerce that behavior. It means something different to me to find out that someone robbed a bank to save his family from execution as opposed to one who just wanted to feed his family and as opposed to one who just wanted cash.

    That said, most folks - myself included - want free will to do more work than just support conventional moral analysis. We also want a distinction between the moral status of automatons and people, for example. We want people to be *really* responsible for their action in the way that, say, a falling rock that kills a man is not. For that to be the case, real intentionality must be on stage. If a rock could be said to have intentionality, it's only the derived intentionality springing from she who dropped it. Coming closer, a machine programmed to detect warm bodies and shoot them is not a murderer in the ordinary moral sense. It's just a machine with "as-if" intentions programmed into it by a programmer whose intentionality is (if free will exists) original and veridical. In the "robot suit" scenario, Lancelot of course has original intentionality, but since it is not what is bringing about the action, it's in no way "his" action.

    So now the question is if purely material beings can have real, original intentions. Obviously I think they can and do. Not in all cases, of course. I think there are plenty of creatures that appear to make choices and have intentions but really don't (the sphex wasp being a common example of this). Thus "free will" continues to have work to do regarding these more fundamental distinctions. Thus there are at least two varieties of compatibilist free will worth wanting.

    Now, many people want free will to also have a radical causation variety, but I think this is poorly motivated, like the desire for there to be a giant space unicorn made of glass hiding behind Saturn. Novel, I suppose, and it hints of other possible truths that are more interesting, but sort of worthless in its own right.

    By Blogger Nato, at 11:18 AM  

  • re: "the kind of compatibilist view of free will covers the general, day-to-day moral usage, since we generally want to know what people will do, ceteris paribus, to assign moral qualities to that person."

    Um, no. Day-to-day moral usage is not a matter merely of prediction. It is at least as much a matter of whether to assign praise and blame. If stranger A and stranger B both bump into me on the subway and don't apologize, their actions were the same, and I have no interest in predicting their future actions, since I never expect to see either of them again. Nonetheless, if I know that stranger A did not realize he had bumped into me, whereas stranger B bumped into me deliberately out of spite, I will forgive A but feel indignant with B. It is precisely the assignation of praise and blame, and, prior to that responsibility, which constitutes moral analysis.

    If I am trying, in a more neutral and rational way, to infer a person's characteristics in order to predict their actions for my own advantage, that is a sort of strategic analysis, not moral analysis at all. There are some similarities between strategic and moral analysis, some overlaps, but they are nonetheless quite distinct conceptually. But I could engage in strategic thinking without my moral feelings being engaged at all ("Brown is clever but a liar, so I'll ask Jones, the honest fool, I know he'll tell me the truth without realizing it's against his interest to do so") or engage in moral analysis without any strategic analysis.

    This seems to be secondary however since Nato concedes that another meaning of free will is wanted. He writes:

    "So now the question is if purely material beings can have real, original intentions. Obviously I think they can and do."

    As far as I can tell, Nato makes no argument in favor of this point. Against the idea of "free will of the radical causation variety," he also makes no argument except to call it "poorly motivated" and "worthless" and comparing it to a giant glass unicorn in space.

    Maybe if Nato made some attempt to argue how free will is compatible is a deterministic universe I would have some hope of conceiving how someone could think free will being non-predetermined is "poorly motivated." For the moment, it still seems to me that free will in a deterministic world is a self-evident contradiction in terms, and that free will is pretty much the most well-motivated claim that exists. It is certainly more well-motivated than any belief whatsoever about the physical world, seeing as those beliefs are all based on a tenuous process of induction from sensory experience.

    By Blogger Lancelot, at 12:13 PM  

  • One comment about self-telos and purpose. If you were raised in Saudi Arabia, you would most likely be Muslim (among other things). Now, it's not a certainty you would never arrive at the same telos you have now, but I'm tempted to say that certain differences in your character would be inevitable. Basically, you are a product of your circumstance, both genetic and geographic. Where does free will fit into that? You can't choose your parents, you can't choose where you're born, you can't choose your gender or other physical characteristics, you can't choose what language and society you grow up in (though your parents can to a more or less narrow extent). The thoughts that you have and the language you think them in are entirely dependent on what you've been exposed to in life. So in a sense, you're right. How can we (materialists) assign blame or praise to people who are in circumstances beyond their control? And honestly, when a crime is committed or I've been slighted or whatever, I try to think about the confluence of circumstances that brought the offender to that point, and that process makes me have compassion for them and forgive them (in general). I think forgiveness is a direct result of the realization that choices and free will are limited/determined by circumstance. Otherwise, how could you truly forgive someone for an Evil act? If no circumstance compelled someone to do Evil works, yet they did them anyway, wouldn't that be ultimately unforgiveable? Not only that, but that person would be inherently Evil, correct? That's just one more thing that I can't believe in, the idea that someone could be inherently Evil. I'd be interested to hear how you reconcile that in your own beliefs.

    By Blogger Thomas Reasoner, at 8:06 PM  

  • re: "One comment about self-telos and purpose. If you were raised in Saudi Arabia, you would most likely be Muslim (among other things). Now, it's not a certainty you would never arrive at the same telos you have now, but I'm tempted to say that certain differences in your character would be inevitable. Basically, you are a product of your circumstance, both genetic and geographic. Where does free will fit into that?"

    In my own personal case, I was raised as a Mormon, and later left the Mormon church and became Russian Orthodox. Of course birth has a major effect on what you come into contact with, your influences... it provides you with a choice set, as it were. But that choice set is quite vast really. People end up in all sorts of unexpected places in life that you couldn't have predicted based on their birth.

    By Blogger Lancelot, at 9:09 PM  

  • Some asseverations:
    1)Empirical descriptions are always predictive.

    2)Knee-jerk assignment of moral content is first order, "natural" strategic analysis.

    3)Radical causation by itself is worthless because it doesn't give us a way to assign stable moral content nor does it enable a non-question begging account of veridical intentionality.

    4)To give an effective compatibilist argument for free will in the varieties worth wanting, I only have to give accounts of determinism allowing stable moral assignment and intentionality worthy of the title.

    By Blogger Nato, at 9:17 AM  

  • For 4) Dennett wrote a whole book: Freedom Evolved. A book length, evolutionary compatibilist account of free will should provide for plenty of apoplexy for anyone who wants to insist on traditional joins along which to carve the universe.

    By Blogger Nato, at 1:46 PM  

Post a Comment

<< Home