Ellen Watson - Even If Free Will Existed, You Wouldn't Want It
Delivered 20 Sep 2006
Last Text Revision: 18 Dec 2006
www.philorum.org

Introduction

The idea that human beings have free will forms the foundation of many people's very conception of human nature. It also forms the grounds of many moral systems, those that rely on the principle that people can be held responsible for their actions because they freely chose to do them.

This conception of human free will is seen to be threatened by a deterministic understanding of the physical world.

Deterministic chain of physical causes.
Deterministic chain of physical causes.

If human beings are part of nature (our physical bodies anyway, if not our minds as well) and nature is determined by the laws of physics (or of other sciences), then doesn't it follow that our actions are not free, and that choice is an illusion?

I'm going to present a way out of this problem. It turns out that when you actually try to define what free will would be, the very concept appears flawed and nonsensical. I will argue that if your definition gets something free enough, it no longer looks anything like will, and vice versa. I will try to convince you to stop wanting free will, and therefore you will no longer be troubled by determinism.

Body of the argument

For those of you who want human beings to have free will, what would that will look like? How would it interact with the causal processes of physical reality, yet still be free?

Well, one way is for acts of will to be completely separate from deterministic causal processes, like in this diagram.

One way an act of will might be free:

Deterministic chain of physical causes with unconnected mental event.
Deterministic chain of physical causes with unconnected mental event.

But that's not what you want, when you want free will. This is a picture of something that has no causal interaction with physical reality at all, but you want the will to affect your actions, and to affect other things in the world. So although if the will was like this, it would be free, it's too free. It doesn't really look like will.

For those of you who want human beings to have free will, I'd say that you want at least a causal relationship from acts of will back into the stream of causal processes. So you want at least something like this diagram.

Uncaused mental event that can cause physical events..
Uncaused mental event that can cause physical events.

But in this picture, the act of will is itself uncaused. It comes out of the blue. It's a good model of a process that is random. An example of a random process is the decay of an atom that has a certain half-life. Let's say that when that atom decayed, your arm went up. That's a process that matches this model, but is not what you want when you want free will. You don't want your acts of will to be random and come from nowhere, you want them to make some sense given what has gone before. So this picture, while also free, doesn't look very much like will.

You want the will to look more like this - connected in both directions to the causal stream.

Caused mental event that can cause physical events.
Caused mental event that can cause physical events.

For example, if you hand bumped something hot, it would cause a decision to move, and then you'd move your hand.

Or maybe that's too simple, because that kind of process could in theory take place entirely in the realm of physical reflexes, without willing.

You probably want an act of will to look like this - say something happens in the physical world like Israel and Hezbollah start bombing each other. You see and hear the news reports about it, and then you consider your own moral and political beliefs and examine the background knowledge you have, and you use your sense of reason to think it through and you decide to attend a peace march, and that decision drives your body there.

Caused mental event that can cause physical events, in a larger chain.
Caused mental event that can cause physical events, in a larger chain.

This looks much more like what we want when we think about the will, and its relation to human action. But look. This chain is entirely connected to the causal flow.

Conclusion

So, I should have convinced you that you don't want the will to be free, in the only ways it can be. You want the will to be causally connected with the world, in both directions - caused by things that happen, and able to cause actions. So the picture of the will that you want is like this:

Mental events just run through the head.
Mental events run through the head.

And what I draw from that is that the causes of human action are no different than the causes of events in the physical universe. The only difference between a willed act and an unwilled act is that in the case of a willed act the causal path happens to go through the head.

Therefore acts of will are just as determined as anything else - and this is what you should want.

I'm happy to draw this distinction, because it matches my intuitions - that there's a difference between the case where my arm goes up because someone has attached electrodes to it - or even because someone has walked over to me and picked it up - and the case where my arm goes up because I decide to raise it. But the difference is not morally interesting at all. And the causal relationships are exactly on a par. So I think it turns out that human actions are not free, they are just as determined as anything else. But if they were free, they wouldn't be what you want, so you shouldn't be bothered by this conclusion.

Alternative approaches

To conclude I'll quickly explore two alternative views that might offer other ways out of the problem - one a different conception of causation in the natural world, and the other a consideration of free will at a different level of explanation.

The first alternative involves a different understanding of causation in the physical world. The argument goes, if causal processes are not deterministic but are chaotic, then effects are fundamentally unpredictable, and so human actions can be as well, even if human beings are purely physical beings.

Events are determined but not predictable.
Events are determined but not predictable.

However, this is not much different from the random example above. I would still maintain that this is not what you want when you want free will. You don't want your actions to follow your decisions unpredictably, you want them to be reliably related. So this alternative approach doesn't help much.

The second alternative is to analyse free will at a different level of explanation. We've been analyzing acts of will as causes of action, but perhaps they are better understood as reasons for action, or motivations for action. Reasons or motivations can result in several different possible future events. For example, you could have a good reason for going to bed early, but still stay up late. Would it help to understand acts of will this way - would it get you the kind of freedom you want?

Reasons, not causes.
Reasons, not causes.

The argument against this approach is that even events described at a higher level of explanation have to be reduced to events in the physical world, and if these are deterministic then at bottom acts of will must be as well. However, if you aren't committed to this theory of reduction, then this alternative approach might have more potential, and deserves further investigation.