Phil 101: Compatibilism, Part 1

We said it’s unclear whether Determinism is true about our world. But we want to know whether, if it is true, it implies that you can only choose and do the single things you actually choose and do.

We saw an argument that it does imply this, which many find compelling. But others argue it can be resisted.

We’re going to talk now about views that try to resist the step from “Determinism” to “No free will.” These views argue that in the important senses of words like “can” and so on, Determinism does not imply we can only choose and do what we actually choose and do.

Examples of “can” meaning different things

Recall our discussion of the concept “mother” at the start of term. We said that reflection on this concept may reveal that we use it to mean different things.

“Can” is another concept that seems to mean different things when we use it in different places. Consider the following examples:

  1. Felix can lift weights, but he’s an ex-felon so he can’t vote.
  2. I can’t lift any weights today, I have meetings from dawn to dusk.
  3. Rachel can’t lift weights, because they’ve all been moved to storage.
  4. Arnold can’t lift the table, he’s holding a coffee and a donut and talking on his cell phone right now.
  5. John can already lift 150 pounds, and I think he can lift 200 too — he just needs some extra training.

These examples show the word “can” (and “can’t”) being used in a variety of ways. In (1) we’re talking about what Felix is permitted to do. In (2) through (4) we’re talking about what people have the opportunity to do. In (5) we’re talking about what John now has strength enough to do, and speculating about what strength is achievable for him, with a small-to-moderate amount of extra training.

It’s not an easy matter to identify and articulate what the various senses of “can” used in these examples come to. But I hope it is intuitively clear that the sense in which 1. is saying Felix can’t vote is not the same as the sense in which 2. is saying the speaker can’t lift weights, and so on. Claims of the form “X can do such-and-such” mean a variety of different things. So too with claims like form “It could have happened that Y” and “X can bring it about that Y.” There is not just one way to interpret such claims.

Back to Determinism

Now let’s go back to Determinism.

We were asking whether, if Determinism is true, people can only choose and do the single things they actually choose and do. Now that we’re alerted to the multiplicity of things “can” can mean, we have to ask: can only do that, in what sense of “can”?

Is there some sense of “can” according to which Determinism entails that you can only do the things you in fact do? Perhaps. There may be senses of “can” where that follows. If Determinism is true, then, given the way the world was before you were born, it was causally settled that you would do the things you in fact do. You “can’t” do anything else — in the sense that, given the way the past is, the laws of nature exclude your doing anything else. So in that sense of “can,” the things you in fact do are the only things you can do.

But we’ve seen that the words “can” (and “could” and “able” and so on) mean a variety of different things. What we need to know is not whether there’s a sense of “can” according to which Determinism entails that you can only do the things you in fact do. Perhaps there are other senses of “can” that do not have that consequence. What we need to know is not whether there’s some sense of “can” which lets us say such-and-such, or some sense of “can” which lets us say something else. Our task is harder than that. What we need to do is to somehow identify senses of “can” that we’re independently interested in, and ask whether, with those senses of “can,” Determinism entails that you can only do the things you in fact do. For example, we said at the start of our discussion of free will that the notion of deserving punishment (or blame or resentment or praise or credit or gratitude) seems to involve the idea that you performed an act that you could have refrained from or avoided. Is the sense of “can” involved in that thought compatible with Determinism? This is the question we need to answer.

Compatibilism version 1

We’re going to start by exploring how the Compatibilist thinks notions like “free” and “can” and so on should be understood, so that it turns out that Determinism is still compatible with you acting freely, with it being such that you can do things you’re determined not to do, and so on.

Let’s begin by considering paradigm cases in which you’re not able to act freely, and you don’t seem to be morally responsible for your actions. Then we can try to figure out what all such cases have in common.

Suppose your friend is drowning and needs your help — but you can’t help him because you’re chained to the wall. Or there’s a locked door between you and your friend. Or a high wall which you can’t scale. In all these cases, you are not able to act freely because there is some external impediment which prevents you from doing what you want to do.

Another sort of case is where you want to help your friend, but you can’t reach him in time, because you have a broken leg, or because your legs are paralyzed. These kinds of personal handicaps can also impair your ability to act freely.

A third sort of case is where you are pulled along by ropes, against your will. In the earlier cases, there were obstacles that prevented you from moving in certain ways. Here an external force is being applied to you which forces you to move in a certain way. In this case, too, you are prevented from doing what you want to do.

A fourth sort of case is where someone holds a pistol to your head and threatens or coerces you into acting a certain way. We have to be careful here. Not every sort of coercion will count as taking away your ability to act freely. If I threaten to give you a “D” in this course unless you carry out a bunch of assassinations for me, and you go ahead and carry out my orders, you can’t excuse your actions by saying that I forced you to do it. The threat I held over your head was too weak, compared to the badness of your subsequent actions. Suppose, on the other hand, that I hold a much greater threat over your head. Perhaps I threatened to kill you if you did not comply. And suppose all I demand that you do is throw eggs at some Professor in another department. In this case, if you give in to my demands, it’s less clear that you should be held responsible for your actions. When we talk about coercion in these notes, we’ll suppose that we’re always dealing with cases of this latter sort, where the threat always does seem to be relatively strong enough that it might excuse the agent for his or her actions.

In cases of coercion, it also seems to be the case that something (a threat) is preventing you from doing what you want to do. Of course, if a mugger holds a pistol to your head and tells you to hand over your money, there is a sense in which handing over your money is what you want to do. Taking everything into consideration, including what will happen if you don’t hand over your money, it seems that handing over your money is the most desirable option. But there’s also a sense in which you’re being forced to do something you’d rather not do. You don’t want to give the mugger your money; he’s forcing you, by threat, to do something you wouldn’t otherwise choose to do.

A common element to all of these cases is that you’re being forced to do something you don’t want to do, or you’re being prevented from doing something that you do want to do. So we might say, as a tentative first stab, that:

(C1) You do X freely iff:

  1. you do X, and
  2. doing X is what you want and choose to do

“Iff” is philosopher’s shorthand for “if and only if,” or “just in case.”

There is no obvious reason why the kind of freedom we articulate in (C1) should be incompatible with Determinism. The fact that I was causally determined to lecture today does not entail that I don’t want to lecture. (I may very well have been causally determined to want to lecture, too.) The fact that I was causally determined to lecture is compatible with my wanting to, and deciding to, lecture today.

So if this is the kind of freedom we’re interested in, if it’s the kind of freedom which is required for moral responsibility, then freedom does seem to be compatible with Determinism, after all.

This account of freedom is sometimes expressed with the slogan “Freedom is opposed to constraint not to necessity.” What this means is that the opposite of freedom is not causal “necessity” or Determinism. Rather, the opposite of freedom is being constrained to act, that is, not being able to act in the way you want to act, because of a force or threat or chain binding you to the wall. It’s not the mere fact that your act is causally determined that makes it unfree. It has to be caused in certain ways, for it to be unfree. It has to be caused in one of the ways we described, that prevent you from acting in the way you want to act. In his essay “Freedom and Necessity,” the philosopher A.J. Ayer says:

It is not when my action has any cause at all, but only when it has a special sort of cause, that it is reckoned not to be free.

Lemos describes something like this first version of Compatibilism on pp. 22-3. On p. 22 we’re told:

A person is free if and only if he has the power to do what he wants and there are no constraints preventing him from doing as he wants.

One potential difficulty here is that it uses the concept of “power,” and that concept (and related ones like “ability”) are very close to the notions we’re trying to explain. But if we focus on what this account would say about actions that a person actually performs, as (C1) does, then it seems to be saying the same thing about those cases. A bit later, Lemos writes something that looks closer to (C1):

We only hold people responsible when they are doing what they want to do and they are not acting as a result of threat or external physical force. (p. 23)

Evaluation of Compatibilism version 1

So. Is this kind of freedom, the one articulated in (C1), the kind of freedom we’re interested in?

Sometimes people suggest that, even if our actions are determined, still, we can find a certain kind of peace and “liberation” by not struggling against the inevitable. We should resign ourselves to our circumstances. Like dogs on a leash being pulled behind a wagon, we can trot along peacefully or we can resist. Either way we’ll end up at the same destination. At least if we accept our situation and trot along willingly, this thought goes, we can take a kind of responsibility for our inevitable fate. (This image and line of thought date back to the ancient Stoics.)

I don’t know about you, but that doesn’t strike me as being a very satisfying sort of “liberation” or “responsibility for my own fate.” Yet in the case I described, trotting along peacefully behind the wagon is what you want and choose to do. So this dog-behind-the-wagon story is one where you have the kind of freedom articulated in (C1). If we hanker for more than that, we must be hankering for a kind of freedom that (C1) doesn’t yet capture.

The problem with the the dog-behind-the-wagon story is that you’re not following the wagon because you want and choose to follow it. You’d be following it no matter what. You have no choice in the matter. Of course, you may have gone through the process of choosing, and decided to follow the wagon. But you have no choice — in the sense that there’s no other action which you could perform, instead. If you had chosen to stay in one place, instead, you’d be out of luck. Your leash will pull you along after the wagon whether you want to follow it or not. (Conceptually, it also seems to be possible for the opposite to happen: for you to have a choice but you don’t make a choice or go through the process of choosing, for example when you act out of automatic habit.)

All this suggests that our concept of freedom should include something about our having alternative choices open to us. We think that, whenever we really do something of our own free will, it has to be the case that we could also have done something else instead. We needed to have (at least some) control over which option we picked.

This is what’s known as a Principle of Alternative Possibilities for notions like acting freely or making a free choice or being morally responsible for your action. It’s sometimes abbreviated as PAP. You had to have picked the option you did because it’s the option you wanted, not because it was the only option there was.

Here is van Inwagen expressing this idea:

To make a negative moral judgment about one of your acts is to evaluate your taking one of the forks in the road of time, to characterize that fork as a worse choice than at least one of the other forks open to you… A negative moral evaluation of what someone has done requires two or more alternative possibilities of action for that person, just as surely as a contest requires two or more contestants. (p. 268)

This Principle of Alternative Possibilities is a powerful idea. (Though as we see in Lemos’ discussion of Frankfurt cases on pp. 28-31, it’s sometimes argued against.)

Let’s consider whether there’s a way for the Compatibilist to incorporate it into their view.

Compatibilism version 2

The Compatibilist thinks the basic insight of the PAP is that had we chosen a different alternative, there should be no obstacle to our pursuing that option, instead.

They think this insight is one they can accommodate. We do that by understanding the phrase “you could have done something else” in a particular way.

When we say “you could have done something else,” the Compatibilist thinks we mean that if you had chosen to do something else, you would have succeeded in doing something else. Nothing but your choice prevents you from doing the other thing. We call this the Compatibilist’s analysis of “could have done otherwise.” (Sometimes it is called the conditional or hypothetical analysis, instead.)

There are several variations on this basic analysis. As I’ve presented it, the Compatibilist says that “Y could have done otherwise” means that if Y had chosen or decided to do otherwise, then Y would have done otherwise. You may instead see Compatibilists saying that “Y could have done otherwise” means that if Y had tried to do otherwise, then Y would have done otherwise.

Here are presentations of this idea from our texts. van Inwagen writes:

According to this [Compatibilist version 2] solution, a future is open to an agent if, given that the agent chose that future, chose that path leading away from (what semed to be) a fork in the road of time, it would come to pass. Thus it is open to me to stop writing this book and do a little dance because, if I so chose, that’s what I’d do. But if Alice is locked in a prison cell, it is not open to her to leave: if she chose to leave, her choice would be ineffective because she would come up against a locked prison door. (p. 272)

Lemos has:

KATE: Many compatibilists believe we could do otherwise [than what we do] even if our acts are causally determined. They provide what’s called a conditional analysis of ‘could have done otherwise.’ So for instance, consider the case of John going to class this morning. He contemplated staying in bed and not going to class, but he got up and went to class… Many compatibilists think such free acts can be determined and the agent still could do otherwise. The compatibilist may say that the phrase ‘the agent could have done otherwise’ just means ‘he would have done otherwise, if he had wanted to or chose to.’ So for instance, even though John’s decision to go to class was causally determined, he still could have done otherwise in the sense that he would have done otherwise if he had wanted to.
PROF. DANIELS: Right, the idea here is that John’s decision to go to class was determined by his stronger desire to improve his GPA. But, even though the decision was determined, he could have done otherwise in the sense that if he’d had a stronger desire to stay in bed then he would have. (pp. 25-6)

In the the dog-behind-the-wagon story, it is not true that you could have done otherwise, in the Compatibilist’s sense. If you had chosen to stand in one place, you would have been dragged along after the wagon anyway. Nor is it true that you could have done otherwise in our other paradigm cases, involving chains or locked doors or paralyzed limbs. Hence, the Compatibilist proposes:

(C2) You do X freely iff:

  1. you do X, and
  2. doing X is what you want and choose to do (let’s keep this condition, for example to help rule out cases of coercion), and
  3. if you had chosen to do something other than X, nothing would have prevented you, so you would have succeeded in doing that other thing

Is freedom in this sense compatible with Determinism? It seems like it is. If we accept this account of what it is to do X freely, then once again, being causally determined to do X seems to be compatible with doing X freely. After all, being causally determined to do X does not prevent me from satisfying condition (b). And it does not seem to prevent me from satisfying condition (c), either. It may be causally determined that I would lecture to you right now. But that’s because (it was causally determined that) I chose to lecture to you. If the world had gone differently, in such a way that I (was causally determined to) choose to stay home today, instead, then I would be at home right now, instead of standing here lecturing to you. So I can satisfy condition (c) of this proposal, even though I am causally determined to act in the way I do.