I want to begin our discussion of free will and determinism by putting the problem in context. I am treating this problem, of free will and determinism, as though it were part of the philosophy of mind; and my reason for so doing is that our focus is on the will, or our ability to choose, and that ability is a mental process just as perception is a mental process.
For the sake of clarity, let me define a few terms for you, as usual, using an example. Whenever you are confused about some very basic philosophical concept, you will always benefit by considering some clear example. So suppose it is a Sunday morning and you are deciding what to do with the rest of the day. You reflect that youíd really like to see a new movie; but on the other hand it is a pretty nice day out, so maybe it would be better to drive to a lake and have a dinner picnic. So after weighing positives and negatives you decide to go on the picnic, because the movie can be seen anytime soon, but you much more rarely have the time or the nice weather to go on a picnic. And so you pack up the dinner and head out to the lake and have a great time.
Now, letís use this rather mundane example to illustrate what we mean by common terminology. The will is our ability to decide and choose among different options; it is a cognitive process. Another name for the same process is volition. As far as I can tell, "will" and "volition" have the same meaning. So we can say that you use your will or volition in deciding and choosing to go on the picnic. But "using" your will just means going through that mental process, however we might eventually describe it.
The word choice usually means the end result of a decision process. So to use our example, when after thinking about your options, you chose to go on the picnic. That act of choice was an important, perhaps the essential, part of your using your will or volition in that case. If you didnít make a choice, then you werenít using your will. But on the other hand, I donít think that using your will just involves making that final choice. The process you go through leading up to the choice is also properly considered the use of the will or volition. That process is called deciding or deliberating. So before you chose to go to the lake for the picnic, you were deciding, or deliberating. And deciding involves some manner of weighing options, even there are only two options to weigh -- as the song says, "Should I stay or should I go?" It is nonsense to speak of deciding between less than two options.
So much for the purely terminological points about the will. Now a little bit about the word "freedom." This word is highly ambiguous -- we can use it to mean political freedom, or freedom of the will, or simply the lack of something. As in the phrase, "sugar-free." And there are other senses, too, of course. But we are only interested in the sense in which the will is said to be free. Now if, as I said, the will is our ability to decide and choose among different options, then to say that the will is free is to say that our ability to decide and choose among different options is free. To speak of free will is to speak of that ability being free.
So our focus is on only that aspect of the will -- its freedom. And in particular we have two questions to ask: First, what is the freedom of the will? If we had to define "free will," what would we say? And second: Does freedom of the will exist? It does if we can freely decide and choose among different options; so can we?
But before we answer these two questions, I want to address a different question: Why is free will such an important topic? The answer is that it is freedom of the will that makes us morally responsible for what we do. If we had no free will, we could not be morally praised for anything, or blamed for anything. And correspondingly: if we can indeed be morally praised or blamed, then we do have free will. Unless we can decide and choose freely, then we do not really have the right to laud anyone for anything good they do, or condemn for anything evil they do: they werenít responsible for it. It would be like praising the sun for warming up the day, or censuring a tornado for destroying houses. Those things donít have free will, so what would the point be of praising or blaming them?
There is a related principle that philosophers use about a good deal. I mean the principle that "ought" implies "can." In other words, if I ought to do something, that implies that I can do it. After all, if I cannot do it at all, then surely I am not obligated to do it. What sense would there be in saying that I ought to do something that I canít do? If, for example, I see a child drowning in a rushing river and I canít swim, then can I be blamed if I donít jump in and try to save the child? I canít do it. So surely itís not the case that I should do it. Now letís see how free will works in here: if I donít have free will, then whatever I end up doing, I must do; I just donít have any choice in the matter. So then itís never the case that I should do anything other than what I happen to do. Doesnít matter what it is; if Jack murders an old woman, itís not as though he should not have done that. Because he had no choice; if he could held off from killing her, then perhaps he should have. But his will was unfree. So he had to murder her, and we canít say that he ought not to have done so. Ought implies can. A moral obligation requires an ability to do the thing youíre obligated to do; and an ability to do the thing youíre obligated to do requires that you are free to do it. If youíre not free do otherwise, then you lack the ability to do what youíd be obligated to do; which means that you arenít obligated to do it.
So I hope this is enough for you to understand why the issue of whether or not we have free will is very important indeed. If we lack free will we might as well just give up the whole business of praise and blame that goes along with morality. Now perhaps some of you -- not many I suspect, but some of you -- are perfectly happy to get rid of this business of moral appraisal. But are you really quite sure this is the view you want to hold? I mean, should we never praise you for your accomplishments and good deeds, because you just had no free choice in the matter? Should we never blame and reproach and revile criminals when they steal, rape, and murder, because they had no choice but to commit their crimes? If you reflect a little like that Iím sure that the vast majority of you will come to the view that moral evaluation has a very useful function in everyday life. We should not just do away with it. But the only way we can take moral evaluation seriously is if we believe we have free will. So do we have free will?
Now do you see how I just immediately slipped into asking whether we have free will, without asking what it would mean to say that the will is free? Thatís the way that most people proceed. I think this is a great mistake. I think we really do have to know what free will would be, before we could know whether we have it. But it will be instructive for us to ignore what we know we should do first, and instead just forge right on ahead, and see whether we can say we have free will, without knowing what free will is.
So letís begin with the view that says we have no free will. An excellent way to present this view is by stating an argument for it, and this is a perfectly valid argument. This is very similar to the argument presented on p. 100 of the packet. After I state the argument Iíll go over each premise carefully. So here goes:
1. Every event has a set of causes sufficient to explain why it and it not some other event occurred. In brief: Every event is causally determined.
2. Every act of choosing is an event (a mental event).
3. Thus every act of choosing is causally determined.
4. If an act of choosing is causally determined, then it is not free.
5. Therefore, no act of choosing is free. That is, free will does not exist.
Let me define the term determinism as the view that the above argument is correct. In other words, hereís the definition:
Determinism is the doctrine that free will does not exist, because every event, including every act of choosing, is causally determined.
As nasty as the conclusion might appear, this argument seems extremely solid; each premise is, by itself, very hard to deny. So letís go premise-by-premise now.
Premise (1): "Every event has a set of causes sufficient to explain why it and it not some other event occurred. In brief: Every event is causally determined." You may recall "the Principle of Causality" from our discussion of the cosmological argument; that principle we stated as "Every event has a cause." Well, if you liked the principle then, consistency requires you to accept it now. But there is a difference between saying that every event has a cause and that every event is causally determined. This last statement is stronger. So what reason is there to believe that stronger claim, that every event is causally determined?
Itís difficult to give a positive argument for this premise. The best I think I can do is to point to some examples where we assume it, or presuppose it. We definitely presuppose it in science -- or most of science anyway. For example, suppose you want to explain why an eight ball rolled into a pocket. Definitely we require some explanation. So suppose we say it was hit by the cue ball. But that only explains why the eight ball is moving. There are still aspects of the eight ballís movement that are unexplained: for example, its direction. That is explained by the vector and spin of the cue ball when it hits the eight ball. And so on -- we demand explanations for every aspect of the eight ballís movement. Thatís the basic premise of scientific inquiry: that there is a causal explanation to be found for every aspect of a thingís change.
If you believe in free will, you may want to say that there are exceptions to the Principle of Causality. Weíll consider such a move in a moment. But first letís look at the other premises.
Premise (2) says: "Every act of choosing is an event (a mental event)." I donít think anyone would want to object to that. If to act is to do something, and events occur, you have to see is that all doings are occurances. Which is just another way of saying that acts are a kind of event. Anyway, surely we canít reject premise (2).
What about premise (3) then? "Thus every act of choosing is causally determined." But this just follows deductively, and validly I might add, from premises (1) and (2). If (1) and (2) are true, then (3) must be true. If every event is causally determined, and every act of choosing is an event, then it must be true that every act of choosing is causally determined. So the only way to reject (3) is to reject (1) or (2).
Now look at premise (4): "If an act of choosing is causally determined, then it is not free." This is one of the two premises, along with number (1), which has been attacked by people who donít want to accept the conclusion. But before criticizing it, letís see if we canít give some basic account of why itís at least plausible. We could put it like this: if I was determined by outside forces, by influences totally apart from me and my deliberations, to make my choice, then in sense it seems my choice was not really mine at all. It was not free because my whole process of deciding and choosing was just like another cog in a machine that produced my action. I did not play an active role in making the choice, precisely because every aspect of my choice was determined and previously decided by factors that were outside of me and totally beyond my control. Everything that led up to my ultimate choice was controlled by forces by me; so how on earth is it in the slightest possible that my choice was free?
So if we accept premise (4), after having accepted premise (3), the conclusion, that no act of choosing is free, follows deductively. Thereís no avoiding that conclusion if you accept (3) and (4). I want to make a comment now about something our reading says. It says, on p. 100 of the packet (or p. 203 of the Solomon selection) that if our choices are not free, then we do not make a choice. As Solomon says, "If our make-up is such that we could not choose any other, then all of our talk about Ďchoiceí and Ďdecisioní -- and therefore Ďresponsibilityí too -- amounts to just so much nonsense." This is a common sort of thing for philosophers to say, but strictly speaking itís wrong. I mean, does Solomon really mean to be saying that we never go through a process of deciding? Or that we never make any choices? That would be ridiculous! No determinist worth his salt would say anything that absurd. I mean, even if our process of deciding and our choices are totally causally determined, that doesnít imply that there is no deciding or choosing going on. The claim is not that choice doesnít exist; itís silly to deny that it exists. No, the claim is rather that choice is not free. So our decisions and choices exist, but they arenít free -- we are the slaves of forces beyond our control. That, at least, is what the determinist says.
Now if you do not like the deterministís argument, and you believe we do have free will, then you have two options. You can reject either premise (1) or premise (4). Letís begin with premise (1), which is that every event is causally determined. So on the view that weíre going to examine now, some events, at least, are not causally determined; in particular, events of willing are not determined; our decision processes and choices are not, not always, determined. At least sometimes when we choose to do something, the choice is free. The claim neednít be that our choices are always free; only that sometimes they are free. Think of it like this -- what if I flip out and you have to put me in a padded cell, and I very deliberately decide to, I donít know, pound my head on the walls. Then I doubt you would want to say that I have freely decided to pound my head on the walls. Iím in the grip of insanity. So my decisions arenít free, and I canít be blamed if I want to hurt myself or others. After all, that is why we have an insanity defense. So the point is that if you believe that free will exists, you donít have to say that every choice that everyone makes is free. You only have to maintain that some choices that we make are free. Having said that, letís introduce a definition:
Libertarianism is the view that free will exists, because some acts of choosing, are not causally determined.
The word "libertarianism" is ambiguous. It means, in the philosophy of mind, the doctrine that we have free will; but in political philosophy, it indicates strong support of individual liberties and a minimal amount of government in all spheres. These are two totally independent doctrines, and we obviously arenít going to talk about the political doctrine right now.
Libertarianism in the sense weíre talking about it now includes two doctrines, the first that free will exists, and the second that some events (namely, some choices we make) are not causally determined. This second doctrine is called indeterminism. Indeterminism is simply the claim that not every event is causally determined.
In this century, with the advent of quantum mechanics, the branch of physics dealing with subatomic particles, it has become rather common and popular to reject determinism. Now, in order to make out this criticism of determinism Iím going to have to give you another formulation of what determinism says. So hereís the alternative formulation:
Determinism is (alternative definition) the doctrine that, given a complete description of the location and momentum of every particle in the universe at present, together with a complete description of the set of laws that guide those particles, a perfect being could predict with certainty every event that has happened and that will happen in the universe.
In other words, if you know where everything is, and exactly how it operates, then, if you could do all the necessary calculations, youíd be able to predict everything that happens. Everything would be predictable. And after the fact, everything would have its causal explanation. So you can see the relation between this definition and the earlier one given. Suppose I state the set of causes that is sufficient to explain why a given event has occurred, and not some other event. Well, that would be a complete description of the location and momentum of all the bodies that contributed to the event in question, together with a statement of the relevant causal laws. So I think we can take the two definitions of "determinism" to be pretty much equivalent; if one is true, then the other is true. They stand and fall together.
Now Iím going to give a criticism of the second definition of "determinism," which we can take as equally criticizing the first. One of the basic results of quantum mechanics is the following. There are certain subatomic particles about which we cannot say what their location and momentum is at the same time. If you say that the particle is at one specific location, then you canít say what itís momentum is; and if you specify its momentum, then you canít say what its location is. Well, that means that a "description of the location and momentum of every particle in the universe at present" cannot even be given! Such a description would be impossible, given what we know from quantum mechanics. So there are some predictions that we couldnít make, no matter how powerful our computers were. It just couldnít be done. Therefore, determinism, in the second sense defined above, is false.
Now if we got very deeply into the issues involved here, it would take us a very long time indeed, and weíd have to know a lot of physics. So if you are totally baffled by this argument against determinism, let me refer you to p. 101 of the packet, and give you a really simple summary. Quantum mechanics, a very difficult and technical part of physics, says that there are some very small particles that we canít make predictions about. So we canít say that causal determinism holds for those particles. Thatís the gist of what Iím trying to get you to understand.
The trouble with this argument is that it doesnít show that anything other than those subatomic particles is not determined. I mean, letís suppose itís true that certain subatomic particles are "random," or undetermined. That doesnít mean that anything else is undetermined. All this talk of how things work on the subatomic level doesnít address how things work on the macro level, the level that we all live and operate on. And in particular, itís hard to see what the randomness of some subatomic particles might have to do with the process of volition.
About the most help that quantum mechanics might give to libertarianism is this: there are exceptions to the Principle of Causality, if we are to believe the physicists. So we might say: maybe free will is another exception to this allegedly general rule, that all events are causally determined. If it were an exception, it wouldnít be the only exception. So thatís something at least. It makes the libertarian objection to determinism at least a little more plausible. So much for the libertarianís objection to determinism.
What can be said specifically against libertarianism itself? Letís suppose that, in fact, our choices are not determined at all. Now, whatís the claim there? Apparently, itís that, when we choose, the event of choosing doesnít have a cause. A choice is a causeless event. Or at least, the causes that it does have donít determine the choice. There is still some aspect of the choice that is uncaused. Or perhaps youíd want to say that I am the cause of my choice; but then, to say that my choice is free, what Iíd mean, as a libertarian, is that nothing causes me to choose what I do. So then thereís an element of myself that is uncaused. So the libertarian view, that some acts of choosing are not determined, can be spelled out in various ways. Either a choice has no causes at all; or there are some aspects of the choice that are not caused; or though I entirely cause my choice, nonetheless there is something about myself that is uncaused.
Those are all different ways of understanding what it means to say that a choice is not determined. But something they have in common is that there is a randomness involved in my choice. After all, if your choice is in some sense uncaused, then to that extent what can it be but random? Itís as though God rolls a die and whoops, there you go do something that has no causal explanation. So when you were choosing to go on the picnic instead of to the movies, what made that choice free was this random element. Total unpredictability: thatís what made your choice free.
But doesnít that sound strange? See here, free will is supposed to be what accounts for why we are morally responsible for what we do. But now, if weíre libertarians, weíre saying that itís this random element of a choice that makes the choice free -- and which makes us morally responsible for the choice! But isnít that strange? Why should we be morally responsible for something that is totally random? That doesnít make any sense!
So let me explain the criticism of libertarianism here. Libertarianism says that a choice is free because it is not determined. And this means, as I said, that the choice is, to some extent, random. But if the choice is random, why should we be held responsible for making the choice? Freedom is supposed to be what makes us morally responsible; but why on earth would anyone say that it is causelessness, randomness, which makes us morally responsible?
I think that the problem here is in the libertarian notion of what free will is. Free will doesnít mean random or causeless will. Well then what is free will? Remember that we started out by saying that we should have begun by examining what free will is. I think I have made my point now. Perhaps if we had done started out by asking what free will is, then we wouldnít have gotten into this confusion. So letís ask what free will is.
At the same time we ask this question, we are going to be examining premise number (4), which is: "If an act of choosing is causally determined, then it is not free." Maybe a choice can be determined and free at the same time. Donít just assume that thatís a contradiction, before you have figured out what "free will" means. Maybe when we learn what free will is, we will see that premise (4) is wrong. Maybe once we have fixed on the right notion of free will, we wonít think that free will is incompatible with the willís being determined.
C. Compatibilism and the definition of "free will."
In fact, we may as well now define a third position about free will and determinism, which combines elements of determinism and libertarianism:
Compatibilism is the view that free will exists, and that acts of choosing are causally determined.
This view is called "compatibilism" because it implies that the existence of free will is compatible, or consistent, with causal determinism. So compatibilists reject premise (4) of the deterministís argument; just because an act of choosing is causally determined, that doesnít mean itís not free. It could still be free, even though itís determined. We simply have to have the correct conception of "free will."
So to support compatibilism, we will have to do as I said, and figure out once and for all what "free" does mean. Letís begin with a common formulation due to G. E. Moore. Moore, the English philosopher writing near the beginning of this century, said that to say one chooses freely is to say that one could have done otherwise. So if you freely choose to go on the picnic, you could have gone to the movies instead, or done something else entirely. If you freely chose to sign up for this course, that means you could have signed up for a different course, or for no course at all.
Now I donít think this conception of freedom is very helpful, because itís not clear enough. We could interpret "could have done otherwise" too many different ways. We can always ask: "could have" in what sense? In one way, even if Iím determined, I could have chosen otherwise; the very fact that I had other options itself means those other options were there before me, and I could have chosen them for all anybody knew. In another way, if Iím determined, I couldnít have chosen otherwise, because a choice that is totally determined would be a choice that I must make. So maybe itís true that, whenever we choose freely, we could have chosen otherwise; but this canít work as a definition, because itís just not clear.
So letís look at another definition: to choose freely is to choose without compulsion. This makes some sense; no one compelled you to go on a picnic; no one compelled you to take this course. So you chose freely. A definition like this is give on p. 102 of the packet. But now consider this problem: if I go insane, and you want to say I no longer have free will, itís not like thereís any person who is compelling me to choose to things. At best you might say itís my insanity which compels me to make my choices. And in the same way, we want to say that little children canít be held responsible for everything they do; their wills are not fully free, not yet. But again, itís not any person who is compelling them to do what they do. Itís just that a toddler canít go through a process of decision and choice that weíd want to call "free." If we have free will, we grow into it; so freedom has to be something that we can develop.
Still, thereís something right about the proposal that a free choice is one that isnít compelled; in other words, the process of volition canít be constrained in certain ways, if itís going to be free. A free will is one that isnít be constrained in various ways: by force from other people; by insanity; by drugs; or even, in the case of a little child, by immaturity. But you canít just say that a free choice is one that is not compelled or constrained, period. Because that leave open the question: What sorts of things can compel or constrain the will? Only some of those things make the will unfree. So hereís the big question: What sorts of things, by constraining our ability to use our will, make the will unfree?
So here is my proposal -- Iíll just tell you what it is:
Oneís choice is free iff it is the result of a decision process (whether a snap decision or a long deliberation procedure) which is not constrained by sufficient immaturity or impairment (as by insanity).
To understand this definition we need only see how it applies in various cases of free and unfree choices. When you chose to go to the lake, the reason that was a free choice, Iím saying, is that it was the result of your own decision process, and that process was sufficiently mature and unimpaired. On the other hand, when they lock me up in the asylum, and I have lost my free will, itís just because my decision process, my faculty of volition, is sufficiently impaired to make you all say that I lack free will. Or suppose that someone slips a dangerous drug into your morning coffee, and you decide that it would be a dandy idea to go out on High Street and play in traffic. Well in that case, you didnít freely choose to play in traffic, just because your ability to deliberate about what you are doing is impaired by the drug.
One problem with this definition is that this word "sufficient" is vague. How much impairment, and of what sorts, would be enough to make my choice unfree? If Iíve had a couple of beers, is that enough? Probably not. What if I am forced to drink a six-pack in a half an hour -- would that be enough? Probably. But there are going to be borderline cases, where we donít know whether we should say I have free will or not, according to this definition. Still I think thatís OK, because the concept of free will is itself one that has borderline cases. In some cases, we canít say for sure whether or not someone choses freely or not. Think of someone who pleads insanity in a criminal trial. We say that if he knew the nature of his actions then he was insane; but in some cases itís just not perfectly clear whether a person knows the nature of his actions. We canít say whether his decision processes were sufficiently impaired by his insanity.
But my claim is that this definition does at least get this right, that the freedom of the will depends on whether our ability to deliberate is sufficiently mature and unimpaired. And if the word "sufficiently" is vague, which it is, then I will also claim that our concept of free will is vague in just the same way.
This definition of free will has an enormous advantage over the others that I gave. The advantage is that I can give a good argument for it. So let me give you this argument. Remember now that we are morally responsible for what we do because we have free will; unless we do in fact have free will, we canít be praised or condemned for what we do. So itís free will, if we have it, that makes us morally responsible beings.
Well in that case, we had better get a definition of "free will" that accounts for the fact that weíre responsible. Just keep this in mind: if weíre free, then weíre responsible. So whatever freedom is, it has to be something that makes us responsible for what we do. The thing that makes us free has to be something that makes us praiseable and blameable, as it were.
In that light, look again at our definition of "free choice" given above: a free choice is the result of a decision process which is not constrained by sufficient immaturity or impairment. So let me ask you: on this account of freedom, is freedom something that makes us responsible for what we do? Is having a mature, unimpaired decision process what makes us responsible for what we do? I think so. If I am a rational adult, I can be praised and blamed for what I do; and the reason that I can is precisely that I am a rational adult. So, I think we can infer that free will is a will unimpaired by sufficient immaturity and impairment such as drug-induced states, or brainwashing, or insanity.
And if this is what free will is, then we can safely reject premise (4) of the deterministís argument. If an act of choosing is causally determined, then it could still be free; if I am a rational adult, with faculties unimpaired, then Iím free, regardless of whether there is a complete causal explanation for my choices.
Now just for the fun it look back at the definition of "free will" that the determinist and libertarian seem to agree on. They seem to agree that the will is free when it is uncaused or random. But honestly, is it the fact that itís uncaused or random that explains why rational adults are responsible for what they do? Of course not. Itís the fact that they have a fully-functioning faculty of deliberation, which they could use, if they wanted to. We are not free and responsible beings because our will is uncaused; we are free and responsible beings because our will is operational and fully-functioning. So free will is compatible with determinism. I will be so bold as to say that this is my own position, but that isnít to say, of course, that you may not quite reasonably hold some contrary position.
There is a position opposed to compatibilism, namely incompatibilism. I will conclude this chapter with a very brief discussion of this view. Here is a definition:
Incompatibilism is the view that the existence of free will is incompatible with the claim that all events are causally determined.
Anyone who accepts premise (4) of the deterministís argument is an incompatibilist. So determinists are incompatibilists, and they deny the existence of free will. And libertarians are incompatibilists as well; but they deny that all events are causally determined. What these incompatibilists have in common is the assumption that if a choice is free, then at least some aspect of it is uncaused. And we just got done debunking that assumption. I doubt there are any good arguments for it. So I hope youíll forgive me if I end up on a somewhat dogmatic note, and say that incompatibilism seems to me pretty obviously false.