Some Musings About Free Will

When I was on my way to work today, I saw a car being pulled by a pickup truck with an attached tow-rope. To keep the car from careening off course, someone was sitting in the driver's seat and turning the steering wheel to keep the wheels aligned with the truck's. Being me, this got me thinking about free will--specifically, about the relationship between freedom and the ability to have done otherwise in any given situation.

Here's the free will problem briefly. As modern humans, we have two competing intuitions that seem to contradict each other. On one hand, we feel as if our actions are free; I feel like I freely chose to sit down and write this blog post, and that I could just as easily have chosen to do something else (my job, for instance). On the other hand, I also know enough modern physics to know that (at least at the level of cars and cats and computers and people), things don't work that way--I know that the behavior of everything is governed by deterministic physical laws, and see no reason to suppose that I, who am really just a complex physical system, should be any more exempt from those laws than (say) a bouncing ball. This idea--that the current state of any physical system is in principle knowable if you know the state immediately preceding it plus the rules affecting it--is known as causal determinism.


So, at least prima facie, it looks like we've got a contradiction here: it seems to us that our actions are totally free, but as good educated society members in the 21st century, we have a hard time seeing how this can be in light of causal determinism. What's the solution to this puzzle? Philosophers, predictably enough, are divided on the answer. Roughly speaking, there are three camps that philosophers tend to fall into: libertarianism, incompatibilism, and compatibilism.

Libertarianism (not to be confused with the equally crazy political ideology) holds that causal determinism is false, and that we (humans) do indeed have free will. I've yet to see a libertarian account that impresses me as much of anything except wishful thinking (they also tend to make pseudoscientific style appeals to quantum mechanics), so I'm not going to give it much consideration here. It could well turn out that libertarianism is the right position, but to know for certain we're going to need to advance physics, biology, and the rest of the natural sciences a lot more.

Incompatibilism (sometimes called "hard incompatibilism" to distinguish it from libertarianism, which is also technically an incompatibilist position) takes the position that freedom of the will and causal determinism are incompatible and that causal determinism is true. To put that symbolically:

~(D & F)
or
(D --> ~F) & (F --> ~D)

Incompatibilism seems significantly more plausible to me than does libertarianism, but it too involves giving up a strongly held intuition--that our actions are free. This certainly doesn't necessarily mean that incompatibilism is wrong (lots of theories with radically counterintuitive implications turn out to be true; see relativity), but the idea that we are free agents is so strongly held (and seems so obviously true from our subjective frame of reference) that I think it warrants some serious investigation before we abandon it.

Some philosophers, investigating this problem, have endorsed a third view--called compatibilism--which dictates that freedom of the will and causal determinism are not contradictory, as they first appear, but can actually coexist. There are almost as many flavors of compatibilism as there are compatibilists (it's probably safe to call this the dominant view among professional philosophers today), so I'm going to be as brief as I can in characterizing the view as a whole. Compatibilists, in general, reject the traditional analysis of freedom (we'll get there in a moment), and instead embrace what's called the "conditional analysis of 'can.'" Briefly, the conditional analysis (CA) goes something like this.

Suppose we have a man--let's call him Bill--who is trying to decide if he should go to work or stay home and watch TV all day. Under the traditionally compatibilist (i.e. CA) account of freedom, it is accurate to say that Bill can go to work, and it is accurate to say that he can stay home all day. So far so good. What compatibilists mean by 'can,' though, is something very specific, namely: "Some agent x can do y iff it is the case that 'If x wanted to y, x would y' is true." That seems complex at first, but the idea is actually fairly intuitive: for any action (say, going to work), it is fair to say "Bill can go to work" if and only if (there's the 'conditional' part of the 'conditional analysis) it is the case that the proposition "If Bill wanted to go to work, Bill would go to work" is true. This maps out fairly nicely onto our colloquial usage of 'can:' I can get up and dance right now (if I wanted to get up and dance right now, I would), but I can't fly to the moon right now (if I wanted to fly to the moon right now, I still wouldn't). This view is compatible with causal determinism because it is entirely possible that my (and Bill's) desires might be the result of a deterministic system, but that does not matter (according to the compatibilists)--even if my desires are deterministic, as long as I act in accord with them (i.e. I am not being coerced, controlled, etc.), then I am acting freely.

There are, of course, problems with this, and I don't think it's entirely satisfactory. I have to leave work now, though, so more tomorrow on my favorite view: semicompatibilism.