WALTER T. STACE

Walter T. Stace (1886–1967), a professor of philosophy at Princeton University for many years, was the author of several books, including Religion and the Modern Mind (1952), from which this selection is taken. The title is the editors’.

Is Determinism Inconsistent with Free Will?

The second great problem which the rise of scientific naturalism has created for the modern mind concerns the foundations of morality. The old religious foundations have largely crumbled away, and it may well be thought that the edifice built upon them by generations of men is in danger of collapse. A total collapse of moral behavior is, as I pointed out before, very unlikely. For a society in which this occurred could not survive. Nevertheless the danger to moral standards inherent in the virtual disappearance of their old religious foundations is not illusory.

I shall first discuss the problem of free will, for it is certain that if there is no free will there can be no morality. Morality is concerned with what men ought and ought not to do. But if a man has no freedom to choose what he will do, if whatever he does is done under compulsion, then it does not make sense to tell him that he ought not to have done what he did and that he ought to do something different. All moral precepts would in such case be meaningless. Also if he acts always under compulsion, how can he be held morally responsible for his actions? How can he, for example, be punished for what he could not help doing?

691

It is to be observed that those learned professors of philosophy or psychology who deny the existence of free will do so only in their professional moments and in their studies and lecture rooms. For when it comes to doing anything practical, even of the most trivial kind, they invariably behave as if they and others were free. They inquire from you at dinner whether you will choose this dish or that dish. They will ask a child why he told a lie, and will punish him for not having chosen the way of truthfulness. All of which is inconsistent with a disbelief in free will. This should cause us to suspect that the problem is not a real one; and this, I believe, is the case. The dispute is merely verbal, and is due to nothing but a confusion about the meanings of words. It is what is now fashionably called a semantic problem.

How does a verbal dispute arise? Let us consider a case which, although it is absurd in the sense that no one would ever make the mistake which is involved in it, yet illustrates the principle which we shall have to use in the solution of the problem. Suppose that someone believed that the word “man” means a certain sort of five-legged animal; in short that “five-legged animal” is the correct definition of man. He might then look around the world, and rightly observing that there are no five-legged animals in it, he might proceed to deny the existence of men. This preposterous conclusion would have been reached because he was using an incorrect definition of “man.” All you would have to do to show him his mistake would be to give him the correct definition; or at least to show him that his definition was wrong. Both the problem and its solution would, of course, be entirely verbal. The problem of free will, and its solution, I shall maintain, is verbal in exactly the same way. The problem has been created by the fact that learned men, especially philosophers, have assumed an incorrect definition of free will, and then finding that there is nothing in the world which answers to their definition, have denied its existence. As far as logic is concerned, their conclusion is just as absurd as that of the man who denies the existence of men. The only difference is that the mistake in the latter case is obvious and crude, while the mistake which the deniers of free will have made is rather subtle and difficult to detect.

5 Throughout the modern period, until quite recently, it was assumed, both by the philosophers who denied free will and by those who defended it, that determinism is inconsistent with free will. If a man’s actions were wholly determined by chains of causes stretching back into the remote past, so that they could be predicted beforehand by a mind which knew all the causes, it was assumed that they could not in that case be free. This implies that a certain definition of actions done from free will was assumed, namely that they are actions not wholly determined by causes or predictable beforehand. Let us shorten this by saying that free will was defined as meaning indeterminism. This is the incorrect definition which has led to the denial of free will. As soon as we see what the true definition is we shall find that the question whether the world is deterministic, as Newtonian science implied, or in a measure indeterministic, as current physics teaches, is wholly irrelevant to the problem.

Of course there is a sense in which one can define a word arbitrarily in any way one pleases. But a definition may nevertheless be called correct or incorrect. It is correct if it accords with a common usage of the word defined. It is incorrect if it does not. And if you give an incorrect definition, absurd and untrue results are likely to follow. For instance, there is nothing to prevent you from arbitrarily defining a man as a five-legged animal, but this is incorrect in the sense that it does not accord with the ordinary meaning of the word. Also it has the absurd result of leading to a denial of the existence of men. This shows that common usage is the criterion for deciding whether a definition is correct or not. And this is the principle which I shall apply to free will. I shall show that indeterminism is not what is meant by the phrase “free will” as it is commonly used. And I shall attempt to discover the correct definition by inquiring how the phrase is used in ordinary conversation.

692

Here are a few samples of how the phrase might be used in ordinary conversation. It will be noticed that they include cases in which the question whether a man acted with free will is asked in order to determine whether he was morally and legally responsible for his acts.

JONES: I once went without food for a week.

SMITH: Did you do that of your own free will?

JONES: No. I did it because I was lost in a desert and could find no food.

But suppose that the man who had fasted was Mahatma Gandhi. The conversation might then have gone:

GANDHI: I once fasted for a week.

SMITH: Did you do that of your own free will?

GANDHI: Yes. I did it because I wanted to compel the British Government to give India its independence.

Take another case. Suppose that I had stolen some bread, but that I was as truthful as George Washington. Then, if I were charged with the crime in court, some exchange of the following sort might take place:

JUDGE: Did you steal the bread of your own free will?

STACE: Yes. I stole it because I was hungry.

10 Or in different circumstances the conversation might run:

JUDGE: Did you steal of your own free will?

STACE: No. I stole because my employer threatened to beat me if I did not.

At a recent murder trial in Trenton some of the accused had signed confessions, but afterwards asserted that they had done so under police duress. The following exchange might have occurred:

JUDGE: Did you sign this confession of your own free will?

PRISONER: No. I signed it because the police beat me up.

Now suppose that a philosopher had been a member of the jury. We could imagine this conversation taking place in the jury room.

FOREMAN OF THE JURY: The prisoner says he signed the confession because he was beaten, and not of his own free will.

PHILOSOPHER: This is quite irrelevant to the case. There is no such thing as free will.

FOREMAN: Do you mean to say that it makes no difference whether he signed because his conscience made him want to tell the truth or because he was beaten?

PHILOSOPHER: None at all. Whether he was caused to sign by a beating or by some desire of his own — the desire to tell the truth, for example — in either case his signing was causally determined, and therefore in neither case did he act of his own free will. Since there is no such thing as free will, the question whether he signed of his own free will ought not to be discussed by us.

693

The foreman and the rest of the jury would rightly conclude that the philosopher must be making some mistake. What sort of a mistake could it be? There is only one possible answer. The philosopher must be using the phrase “free will” in some peculiar way of his own which is not the way in which men usually use it when they wish to determine a question of moral responsibility. That is, he must be using an incorrect definition of it as implying action not determined by causes.

Suppose a man left his office at noon, and were questioned about it. Then we might hear this:

JONES: Did you go out of your own free will?

SMITH: Yes. I went out to get my lunch.

15 But we might hear:

JONES: Did you leave your office of your own free will?

SMITH: No. I was forcibly removed by the police.

We have now collected a number of cases of actions which, in the ordinary usage of the English language, would be called cases in which people have acted of their own free will. We should also say in all these cases that they chose to act as they did. We should also say that they could have acted otherwise, if they had chosen. For instance, Mahatma Gandhi was not compelled to fast; he chose to do so. He could have eaten if he had wanted to. When Smith went out to get his lunch, he chose to do so. He could have stayed and done some more work, if he had wanted to. We have also collected a number of cases of the opposite kind. They are cases in which men were not able to exercise their free will. They had no choice. They were compelled to do as they did. The man in the desert did not fast of his own free will. He had no choice in the matter. He was compelled to fast because there was nothing for him to eat. And so with the other cases. It ought to be quite easy, by an inspection of these cases, to tell what we ordinarily mean when we say that a man did or did not exercise free will. We ought therefore to be able to extract from them the proper definition of the term. Let us put the cases in a table:

It is obvious that to find the correct definition of free acts we must discover what characteristic is common to all the acts in the left-hand column, and is, at the same time, absent from all the acts in the right-hand column. This characteristic which all free acts have, and which no unfree acts have, will be the defining characteristic of free will.

694

Is being uncaused, or not being determined by causes, the characteristic of which we are in search? It cannot be, because although it is true that all the acts in the right-hand column have causes, such as the beating by the police or the absence of food in the desert, so also do the acts in the left-hand column. Mr. Gandhi’s fasting was caused by his desire to free India, the man leaving his office by his hunger, and so on. Moreover there is no reason to doubt that these causes of the free acts were in turn caused by prior conditions, and that these were again the results of causes, and so on back indefinitely into the past. Any physiologist can tell us the causes of hunger. What caused Mr. Gandhi’s tremendously powerful desire to free India is no doubt more difficult to discover. But it must have had causes. Some of them may have lain in peculiarities of his glands or brain, others in his past experiences, others in his heredity, others in his education. Defenders of free will have usually tended to deny such facts. But to do so is plainly a case of special pleading, which is unsupported by any scrap of evidence. The only reasonable view is that all human actions, both those which are freely done and those which are not, are either wholly determined by causes, or at least as much determined as other events in nature. It may be true, as the physicists tell us, that nature is not as deterministic as was once thought. But whatever degree of determinism prevails in the world, human actions appear to be as much determined as anything else. And if this is so, it cannot be the case that what distinguishes actions freely chosen from those which are not free is that the latter are determined by causes while the former are not. Therefore, being uncaused or being undetermined by causes must be an incorrect definition of free will.

What, then, is the difference between acts which are freely done and those which are not? What is the characteristic which is present to all the acts in the left-hand column and absent from all those in the right-hand column? Is it not obvious that, although both sets of actions have causes, the causes of those in the left-hand column are of a different kind from the causes of those in the right-hand column? The free acts are all caused by desires, or motives, or by some sort of internal psychological states of the agent’s mind. The unfree acts, on the other hand, are all caused by physical forces or physical conditions, outside the agent. Police arrest means physical force exerted from the outside; the absence of food in the desert is a physical condition of the outside world. We may therefore frame the following rough definitions. Acts freely done are those whose immediate causes are psychological states in the agent. Acts not freely done are those whose immediate causes are states of affairs external to the agent.

20 It is plain that if we define free will in this way, then free will certainly exists, and the philosopher’s denial of its existence is seen to be what it is — nonsense. For it is obvious that all those actions of men which we should ordinarily attribute to the exercise of their free will, or of which we should say that they freely chose to do them, are in fact actions which have been caused by their own desires, wishes, thoughts, emotions, impulses, or other psychological states.

In applying our definition we shall find that it usually works well, but that there are some puzzling cases which it does not seem exactly to fit. These puzzles can always be solved by paying careful attention to the ways in which words are used, and remembering that they are not always used consistently. I have space for only one example. Suppose that a thug threatens to shoot you unless you give him your wallet, and suppose that you do so. Do you, in giving him your wallet, do so of your own free will or not? If we apply our definition, we find that you acted freely, since the immediate cause of the action was not an actual outside force but the fear of death, which is a psychological cause. Most people, however, would say that you did not act of your own free will but under compulsion. Does this show that our definition is wrong? I do not think so. Aristotle, who gave a solution of the problem of free will substantially the same as ours (though he did not use the term “free will”), admitted that there are what he called “mixed” or borderline cases in which it is difficult to know whether we ought to call the acts free or compelled. In the case under discussion, though no actual force was used, the gun at your forehead so nearly approximated to actual force that we tend to say the case was one of compulsion. It is a borderline case.

695

Here is what may seem like another kind of puzzle. According to our view an action may be free though it could have been predicted beforehand with certainty. But suppose you told a lie, and it was certain beforehand that you would tell it. How could one then say, “You could have told the truth”? The answer is that it is perfectly true that you could have told the truth if you had wanted to. In fact you would have done so, for in that case the causes producing your action, namely your desires, would have been different, and would therefore have produced different effects. It is a delusion that predictability and free will are incompatible. This agrees with common sense. For if, knowing your character, I predict that you will act honorably, no one would say when you do act honorably, that this shows you did not do so of your own free will.

Since free will is a condition of moral responsibility, we must be sure that our theory of free will gives a sufficient basis for it. To be held morally responsible for one’s actions means that one may be justly punished or rewarded, blamed or praised, for them. But it is not just to punish a man for what he cannot help doing. How can it be just to punish him for an action which it was certain beforehand that he would do? We have not attempted to decide whether, as a matter of fact, all events, including human actions, are completely determined. For that question is irrelevant to the problem of free will. But if we assume for the purposes of argument that complete determinism is true, but that we are nevertheless free, it may then be asked whether such a deterministic free will is compatible with moral responsibility. For it may seem unjust to punish a man for an action which it could have been predicted with certainty beforehand that he would do.

But that determinism is incompatible with moral responsibility is as much a delusion as that it is incompatible with free will. You do not excuse a man for doing a wrong act because, knowing his character, you felt certain beforehand that he would do it. Nor do you deprive a man of a reward or prize because, knowing his goodness or his capabilities, you felt certain beforehand that he would win it.

25 Volumes have been written on the justification of punishment. But so far as it affects the question of free will, the essential principles involved are quite simple. The punishment of a man for doing a wrong act is justified, either on the ground that it will correct his own character, or that it will deter other people from doing similar acts. The instrument of punishment has been in the past, and no doubt still is, often unwisely used; so that it may often have done more harm than good. But that is not relevant to our present problem. Punishment, if and when it is justified, is justified only on one or both of the grounds just mentioned. The question then is how, if we assume determinism, punishment can correct character or deter people from evil actions.

696

Suppose that your child develops a habit of telling lies. You give him a mild beating. Why? Because you believe that his personality is such that the usual motives for telling the truth do not cause him to do so. You therefore supply the missing cause, or motive, in the shape of pain and the fear of future pain if he repeats his untruthful behavior. And you hope that a few treatments of this kind will condition him to the habit of truth-telling, so that he will come to tell the truth without the infliction of pain. You assume that his actions are determined by causes, but that the usual causes of truth-telling do not in him produce their usual effects. You therefore supply him with an artificially injected motive, pain and fear, which you think will in the future cause him to speak truthfully.

The principle is exactly the same where you hope, by punishing one man, to deter others from wrong actions. You believe that the fear of punishment will cause those who might otherwise do evil to do well.

We act on the same principle with nonhuman, and even with inanimate, things, if they do not behave in the way we think they ought to behave. The rose bushes in the garden produce only small and poor blooms, whereas we want large and rich ones. We supply a cause which will produce large blooms, namely fertilizer. Our automobile does not go properly. We supply a cause which will make it go better, namely oil in the works. The punishment for the man, the fertilizer for the plant, and the oil for the car are all justified by the same principle and in the same way. The only difference is that different kinds of things require different kinds of causes to make them do what they should. Pain may be the appropriate remedy to apply, in certain cases, to human beings, and oil to the machine. It is, of course, of no use to inject motor oil into the boy or to beat the machine.

Thus we see that moral responsibility is not only consistent with determinism, but requires it. The assumption on which punishment is based is that human behavior is causally determined. If pain could not be a cause of truth-telling there would be no justification at all for punishing lies. If human actions and volitions were uncaused, it would be useless either to punish or reward, or indeed to do anything else to correct people’s bad behavior. For nothing that you could do would in any way influence them. Thus moral responsibility would entirely disappear. If there were no determinism of human beings at all, their actions would be completely unpredictable and capricious, and therefore irresponsible. And this is in itself a strong argument against the common view of philosophers that free will means being undetermined by causes.