This is the forty-third post in my Theories of Knowledge and Reality series. Follow the link for more on the series and for links to other entries as they appear. In the last post, I presented the reasons why someone might be an incompatibilist about freedom and determinism and why someone might be attracted therefore to a libertarian account of free will. In this post, I'll present some problems with libertarian views of freedom.
Libertarians consider determinism the end of the road when it comes to human freedom. If determinism is true, we simply aren't free. So they deny determinism. But does that mean we are free? It doesn't automatically follow from the denial of determinism that anyone actually is free. Just as the compatibilist will have to explain what compatibilist freedom is, why it should count as genuine freedom, and how we can have it if determinism is true, so the libertarian will need an account of freedom that shows how we can have it if indeterminism is true and why it should count as genuine freedom. Libertarians have generally thought that this is much easier than what compatibilists have to overcome given the main incompatibilist argument in the last post. But the problems presented against libertarian views of freedom are not surface-level criticisms. There are deep worries about how freedom of the sort the libertarian thinks we have could arise in the kind of world libertarians think we live in.
The problem arises in its first form in debates among the ancient Greek philosophers. Epicurus was the first philosopher we know of to endorse what we now call a libertarian view in a clear, explicit way. (Some think Aristotle might have been a libertarian, but the evidence there is at least somewhat ambiguous. He didn't formulate a view in this sort of way, at least.) One problem Epicurus faced from his contemporaries was that he was also a materialist and an atomist, believing that everything is made up of indivisible atoms, bouncing around and hitting into each other. Everything we see ultimately is derived from different combinations of atoms. How does free will fit into this? According to his follower Lucretius, Epicurus had an explanation of how freedom appears within such a seemingly-deterministic system. Atoms occasionally swerve. This swerving is not caused by anything prior to my action, so it explains how my action is not determined. So determinism is false, and my choices can be free if they are caused by these swerving atoms.
Lucretius took this explanation to be good enough, but the ancient Skeptics and Stoics found it insufficient. If the atom swerves randomly, meaning that nothing caused it to swerve, then I am just as much out of the loop as I was with the determinist’s story. If nothing caused it to swerve, then its swerving is not under my control, so I still have no choice about what I do as a result. This account of freedom just won’t do. It seems as if denying determinism isn’t enough to show that we’re free. We might not be free even if determinism is false, because our actions could be caused by random events that still aren't under our control.
So libertarians want an account of how I can be free in an indeterministic universe. It can’t be that I am caused by some uncaused, random event. That’s as bad as determinism. I need to be the cause of what I am doing. Since the time of the Epicureans, libertarians have generally responded to this problem with a view that is called agent causation. As far as I've been able to tell, the first person to suggest this was Alexander of Aphrodisias, a follower of Aristotle from Asia Minor (now Turkey) in the late second to third century. Alexander saw the problem in Epicurus' account of freedom based on swerving atoms and concluded that freedom must not be caused by something outside of us at all. Whether it is caused by predetermined events or random events, it is still caused by something outside our control. To be truly free, our choices must be caused within us.
But Alexander couldn't say that our choices are caused within us in the way that the Stoics, who were compatibilists and determinists, thought. They saw our choices as resulting from prior events within us, which eventually do trace back to events outside our control. An incompatibilist couldn't accept such a thing. So Alexander suggested that the cause of our actions is not some event within us, which could be the sort of thing that in turn had been caused by some earlier event. He had to avoid that possibility completely. His solution was to distinguish between two kinds of causes.
Most philosophers think of causes as events. An event, i.e. something that occurs, causes another event. That’s what causation is all about. Examples of events would be football games, elections, the Great Depression, World War II, brain events, hand-wavings, class lectures, concerts, the first ten years of my life, my mouse’s clicking on the file to open it, etc. The firing of neurons in my brain could cause my hand to rise. This is called event causation. But what if I cause some event to take place that leads to my action? This is called agent causation. It’s not that some event causes me to do this. I, an agent (using the classic sense of the term as someone who acts), cause some decision to happen, and that leads me to do the action. Then I’m free, since I caused it. If this view is correct, something can cause events without itself being an event. I cause the action, but nothing within me causes it.
So my action isn't caused by predetermined events or by random events, both of which would be outside my control on an incompatibilist view. Instead, my action is caused by me. What could be a better thing to say if we want to account for my being the one responsible for my action? So far so good. But view does seem rather mysterious. In a physical world, what is the ‘I’ that causes this? Dualism (which I’ll cover in the next unit of these notes) would explain what I am – an immaterial soul, causing something in the physical world. Given materialism (that nothing but the physical world exists), what am I that is causing this event?
For that reason, many materialists think they must deny this view to be materialists. But the issue is actually completely independent of whether we are merely physical beings or whether we have immaterial minds or souls. After all, even if I am just an immaterial soul, there can be events within my mind that cause me to do things. A dualist could be a determinist (and as we'll see in the next section, G.W. Leibniz formulated a very interesting version of dualism that requires determinism or something very much like it). Furthermore, dualism allows for events within your mind to cause you to do things. The issue of whether you are just a collection of events causing further events or whether you are some agent who causes things in addition to what the events within you cause is still an issue for dualists. Dualism adds nothing to the issue that should make a libertarian happier to be a dualist than a materialist. The issues are just completely separate.
As it turns out, Peter van Inwagen is both a libertarian and a materialist (at least about the human mind – he doesn’t think God is a physical being, but he thinks we are). So he doesn’t want to say we have non-physical minds. But he isn't happy with agent causation as an account of human freedom, and his reasons have nothing to do with materialism. John Locke pushed something like this sort of argument, but it was much more carefully formulated in the work of Jonathan Edwards and David Hume.
If I am the cause of my decision, then there’s a particular event I can pick out – my causing my decision. That seems to be a genuine event. Is that event caused or uncaused? If it’s uncaused, then I am not in control, so it better be caused. But it better not be caused by something outside my control. So it must have been caused by some previous event that was also under my control. Then we have another event, and we can ask about that one as well. That one must also have been caused by some other event under my control. If you keep going, you either get an infinite regress, with this infinite past set of events all within my control, or else at some point something outside my control enters in. The first seems absurd – that any one choice I make requires an infinite past series of events all under my control. The second is bad for libertarianism – something outside my control caused what I do. Only a compatibilist could say that.
I do see one possible way out for someone endorsing agent causation. Do we have to admit that the event of my decision must have been caused by some previous event? That’s not what the agency theorist should say. The agency theorist should say that the event of my decision was caused by me. Then what about the event of its being caused by me? That’s just not a legitimate event. This may be the best way an agency theorist would handle this objection. You can put words to describe that as an event, but it's not an event, and so there's no meaningful question to be asked about whether it is caused or what caused it. It's difficult to see why this might be independently plausible, because it sure sounds like an event, but such an ad hoc move might at least allow the libertarian view to be internally consistent.
Peter van Inwagen is one libertarian who is unconvinced. In the end, he leaves freedom as a mystery, but he says compatibilists and those who deny free will also have mysteries to deal with that are worse. His mystery is better than denying the incompatibilist argument from the last post (as compatibilists have to do) or denying that we are free (as hard determinists do), so he’ll live with this lesser mystery. The question is ultimatelty whether the mystery he's chosen to adopt is worse than the mystey of one of the other views. Compatibilists disagree with him on this, and I'll turn to that in the next post.