Today, just like last time, we’re talking about a metaphysical topic that has implications for ethics: free will. We think free will is important for moral responsibility: if someone doesn’t do something of their own free will, they can’t be held morally responsible for it. These ethical implications bleed over into philosophy of religion: many religious believers claim that free will helps deal with the problem of evil, by allowing the blame for the evil in the world to be assigned to human beings or fallen angels. The issue of free will is also a sticking point between different religious groups: for example Calvinists, Christian followers of John Calvin, are notorious for believing that God chooses whether a person will accept salvation or not and therefore whether they will go to heaven or not. The idea that someone could be damned through no free choice of their own horrifies many people, and provides them with a reason for disliking Calvinism. That can actually be extended to a criticism of Christianity as a whole, if you think the Bible teaches predestination. Similar accusations of denying free will have also been leveled against Islam. Beyond these specific worries, the mere suggestion that free will doesn’t exist seems somehow inherently frightening to many people.

So what is free will? I don’t know. You see, a lot of the philosophical debate over free will has been over not whether free will exists, but rather what free will would be, if it existed. There are two basic positions on this issue, and they split over the issue of determinism. If you want to know what determinism is, here you don’t have to worry, because I can define that. Determinism is the view that given the state of the world at any given time, and the laws of nature or whatever principles it is that govern the world, only one future is possible. The deterministic view of the world is the view contained in the Newtonian physics you likely studied in high school: in a common sort of high school physics problem you’re given some objects in some initial condition, and you have to figure out what will happen to them. It’s taken for granted that only one possible outcome is given.

Nowadays, you should be aware, many physicists are convinced that determinism isn’t true, that at the level of quantum mechanics there is some irreducibly random element in the world, such that a given situation could genuinely turn out multiple ways. It’s essential here, though, not to confuse determinism with mathematical chaos. Who here is familiar with fractal geometry, perhaps the Mandelbrot set? That’s mathematical chaos. Chaos, in this sense, isn’t opposed to determinism, but is a special type of deterministic system. In a chaotic system, any given starting point can only turn out one way, but an arbitrarily small change could cause things to turn out another way.

One of the most famous representations of chaos is the “butterfly effect”: the suggestion that a butterfly flapping its wings could cause a hurricane on the other side of the world. I don’t know if this is really possible, in terms of how the weather actually works, but if it were possible it would be consistent with determinism. Any given initial conditions will yield the same result every time. Add a butterfly, you get an entirely different result. However, if you add the exact same butterfly moving in exactly the same way many times, you’d get the same result every time in each butterfly trial, even if the result differed radically from the non-butterfly trials.

If you’re interested in learning more about this in depth, I recommend running some Google searches for chaos, fractal geometry, and the Mandebrot set. Here’s a short summary: the Mandebrot set is an instance of a fractal, an infinitely complex geometric figure. It ties in with chaos in that if you look at some area of the figure closely, and then move the area of focus just a little, you can get something entirely different. I can’t give you the full effect here, it really has to be seen live–or at least with the help of a Java web applet.

Now we get to understand the big divides among philosophers regarding free will: is free will compatible with determinism? Philosophers who say “yes” are compatiblists. Philosophers who say “no” are incompatiblists. Among the incompatiblists, there are those who think we have free will, “libertarians” (same word as the political position, different meaning) and those who don’t, the “hard determinists.” Finally, there are those who think the whole concept of free will doesn’t make any sense, who have been called “hard incompatiblists.” Incompatiblists are going to say “hey, of course if your actions are determined by the prior state of the world, you aren’t free.” Compatiblists, in contrast, argue that for an action to be freely chosen, it must be causally determined in the right way.

Which position should we hold? The logic of the situation is a little complicated. I think the starting point for most people is that we have free will. Starting from that point, we can ask three questions:
1) Do we have good reason to think determinism is true?
2) Do we have good reason to doubt that the compatiblist account of free will makes sense?
3) Do we have good reason to doubt that the libertarian account off free will makes sense?

If we can say “yes” to (1) or (3), but not (2), we’re forced into compatiblism. If we say “yes” to (2), but not (1) or (3), we’re forced into libertarianism. If we say “yes” to (2) plus (1) or (3), only then would we reluctantly give up our initial belief in free will. “Yes” to (1) plus (2) yields hard determinism; “yes” to (2) plus (3) yields hard incompatiblism.

Now let’s look at our questions. Do we have good reason to think determinism is true? The standard line you’re likely to hear from the physics department is that determinism is false. More specifically, you’re likely to be told that in quantum mechanics, a system can exhibit two states at once, but when an observation is made, you get a collapse of the state and one of the two possible states is selected randomly. Famously, physicist Erwin Schrödinger suggested that it should in theory be possible to build a box containing a cat that would be both alive and dead–the idea is that you get some small-scale system in two states at once, and then make whether the cat is given poison dependent on the small-scale system. Only when you look in the box would the cat be alive or dead.

I’ve just taken you into a world of absolutely crazy ideas which we, unfortunately, will not discuss in detail because I don’t understand physics as well as I’d like. But I want to throw you something to chew on: some have argued that the most reasonable interpretation of quantum mechanics doesn’t involve randomness. Rather than say that the cat is randomly selected to be alive or dead, why not say–since you’ve already accepted the crazy notion that the cat is both alive and dead–that the when the scientist opens the box, the cat is seen to be simultaneously alive and dead? Crazy, but no crazier than the initial idea now widely accepted by scientists.

Now, what about compatiblism? When I introduced the concept, I suggested some people will think it just obvious that if your actions are determined by things beyond your control, then you’re not free. This idea has been formalized by Peter van Inwagen as follows: If you have no choice about the fact that P, and you have no choice about the fact that P will result in Q, then you have no choice about Q. Call this “van Inwagen’s principle.” On determinism, you have no choice about the distant past, and no choice about the fact that the distant past can only give rise to a single unique future, therefore you have no choice about the future.

Van Inwagen, here, is relying on the notion that his principle has a good amount of intuitive attractiveness. But it’s not a logical axiom. Not everything works the way van Inwagen claims choice works. If I drop a pencil, the fact that it fell is due to gravity, but its initial position and the fact of gravity aren’t due to gravity. On compatiblism, choice would be something like this. An action can be chosen by you even if the things which contribute towards determining your choice–such as your initial psychological dispositions–aren’t your choice.

What about libertarianism? The basic worry here, I think, is that we don’t call random events acts of free will. If my hand suddenly flails out for literally no reason at all, we wouldn’t call that a free act. It seems like if an event isn’t determined, it must be random, so a non-determined event couldn’t be free. Free will would actually require that determinism be true, at least to a significant degree. Libertarians have responded to this argument by saying that the options aren’t limited to determinism and randomness. This raises the question of what the other option would be. The typical response, especially from van Inwagen, is to claim this as a great mystery, and say that they needn’t have a full-fledged account of free will to know that compatiblism is wrong.

It’s rather hard, then, to get a knock-down answer to any of the questions in my list of three. But still, many people are tempted to give a “yes” to at least one of them. It’s a very puzzling issue. We are tempted to say that when we know the neurological causes of a homicide, the killer is no longer morally culpable in the sense of being able to be sent to prison, so they are committed to a psychiatric institution instead. And there seems to me to be a very strong temptation to think of our own choices as happening outside the chain of causality. It’s a genuinely puzzling issue, and I’m curious to hear what you all have to say about it.