The Spike [or Victoria University College Review 1961]
Shall We Say We are Free?
Shall We Say We are Free?
It has been said that the dispute between libertarians and determinists is not a real one — I intend to argue that this is not so. To do this I shall consider the characteristics of unreal disputes and then the nature of the disputes between the libertarian and the determinist.
When people say a dispute is unreal they are not usually clear just what they mean by this word. They mean, perhaps, that the dispute cannot be settled one way or the other, or that it is about the use of words rather than about the world, or that it is unimportant : but if we tried to accept all these criteria for unreal disputes most arguments would be seen to be both real and unreal. Disputes about how a word is in fact used can be readily settled but are not about things, and, at least sometimes, unimportant; the question of God's existence is not just about words, is important, but may be impossible to settle.
It is rather that an unreal dispute is a badly conducted one, one which cannot be settled because the opponents do not understand why they differ and so try to convince each other by inappropriate arguments — they argue as if they knew what sort of considerations would decide the argument but in fact do not allow any such considerations to decide it. Let us consider an example — suppose two people, A and B, were arguing about whether a certain publication was a book or a magazine: it might be the case that agreement can be reached once B knows a few more facts about the publication. He thought it was a book merely because he did not know that a similar publication appeared at monthly intervals. Such a dispute is a typically real one and can be settled by finding out what are the facts. But an argument may arise even when A and B know all there is to know about this publication and understand thoroughly the use of the words book and magazine. The reason for their difference of opinion is that the publication they are considering has two of the characteristics of books (stiff covers, one author), and two of the characteristics of magazines (appears every month, read by flappers).
Since the difficulty is due to this conflict of criteria it can be settled only by improving the accepted ways of distinguishing books from magazines — A may explain that the criteria for magazines should be more precisely stated. The author of this publication has, he admits, published a novel every three months for some time but he has not yet sold yearly subscriptions. If the dispute is carried on in this way, with A and B both understanding why they differ, it is likely that they will reach agreement — and such a dispute is a real one.
It is, I admit, an argument about what A and B should call this publication, about whether they should say it is a book or not, but it is also about whether it is a book or not. The difference between these two questions is over-emphasized by simpleminded seekers after truth. A dispute about what to call an object seems to them trivial, merely about words, to be decided arbitrarily. This is sometimes the case but more often it is not so. Such people may introduce unreality into the dispute by their misunderstanding of the issue. They may argue like this:page 43
B: Let's say it isn't a book because then it won't get an import licence and such trash shouldn't be imported.'
A: 'But it is a book and it's wrong to lie even for useful purposes.'
A fails to see that this particular publication calls for a new ruling on how to distinguish books from magazines and that it is not possible to decide how this new ruling should go simply by a closer consideration of the object and of how the word book has been used in the past. He thinks that all arguments are either about the facts or about what we shall decide to do — that since this is a factual argument it must be settled by discovering the truth and that considerations of consequence can never help us to discover the truth. B on his side exaggerates by sounding as if such a problem can be decided by considering the future only.
An argument of this conflict of criteria type must take note of all three aspects. They are :
|1||— The way people have till now used the words 'book' and 'magazine'.|
|2||— The characteristics of this publication and of other publications that have appeared in the past and of those that will probably appear in the future.|
|3||— The consequences of redrawing the distinction between books and magazines in a more precise or in a different way.|
Since a classification dispute is about words and objects, 1 and 2 must be considered together. For the question to be asked under 1 and 2 is, 'Is this publication sufficiently like what people are in the habit of calling " books " to justify us in calling it a "book"?' To answer this question we must examine the use of the word 'book' and the characteristics of this publication. We must examine not only this publication, but also the characteristics of others, past, present and to come. For if we have to amend the criteria for distinguishing books from magazines in order to settle this argument it is desirable that our work should not be wasted; it should last, should be useful in resolving future disputes.
There are two kinds of consequences for us to consider, the logical and the practical. If we decide to call our publication a magazine on the grounds that its author publishes a book every three months this may mean that many publications hitherto called books should now be called magazines — or that many other publications will become borderline cases. Both these would be undesirable logical consequences. A practical consequence would be that many desirable publications will now be classified as magazines and so be subject to import restrictions.
So far I hope to have shown that a dispute of this criteria amending type is not merely about words (factor 2), and is not usually trivial (factor 3), and is not to be settled arbitrarily, since it should be settled by weighing the various factors indicated under my headings 1, 2, 3. However, some might still want to deny the reality of the dispute on the grounds that it is never possible to be sure one has found the right solution; there is no final court of reference by which A can be sure B is right — or vice versa. I thoroughly disagree with such a use of the word real'. Most of the interesting disputes have to be settled in ways such as I have outlined. Many factors have to be considered and these all influence the decision, but no facts dictate the decision.page 44
It is, of course, not at all clear in detail what 'real' means, but at least this much is clear — it is a pro word. A real dispute is one worth arguing — interesting disputes are also worth arguing — so it would be odd if all the interesting ones were unreal! Disputes that are definitely settle able by looking up a dictionary or turning the light on are often important but never worth arguing when quick and definite means of settling them are available.
I have examined this type of argument at length because I think the difference of opinion between the determinist and the libertarian is of this nature. It is clear that most people are in the habit of saying that they perform certain actions of their own free will. Usage is definitely on the side of the libertarian, but the determinist claims that this is no longer an appropriate way to talk. The progress of science has showed that all our actions have certain other characteristics which demonstrate that they are not freely chosen. Since we cannot deny that some of our actions also have characteristics in virtue of which we would normally say we acted freely, this famous dispute is one of the conflict-of-criteria type. It is then, if I have succeeded in making my point, a real dispute, and should be considered under the three headings I have just outlined.
(1) How the word 'free' is in fact used
The determinist is using ordinary language, so he must mean to deny at least some part of what men ordinarily mean when they claim to act freely, for if he does not he has no excuse for using the words he does. So to understand him we must examine what men mean when they say they act of their own free will. To do this we must consider the actions which may be so described and how they distinguish between them and constrained actions. We say we do of our own free will when there is not a pistol at our backs, when another choice is open to us, when we are sane, not drunk; that we haven't done x of our own free will when someone threatens us, blackmails us, when we have to do x or lose our jobs, when we are neurotically driven to do x, when we act in response to the hypnotist's suggestion, etc.
Some philosophers have claimed that we understand the phrase 'doing x of my own free will 'simply because we can point to examples — e.g. case of a man who asks a girl to marry him without being pressed to do so. They then argue that since we could not understand the phrase unless we could point to at least one free act, the determinist must be wrong. This paradigm case argument is not conclusive. Its weakness is that it presupposes too simple a theory of meaning. We understand the phrase both by pointing to examples of such actions and by listing the characteristics of the actions which we are willing so to describe. Modern science might have shown that though we thought there were actions that had these characteristics we now know they did not. We could still understand the phrase for we would be able to describe what a free action would be like; it would be like the actions that we used to think were free, only it would really possess the characteristics our ancestors falsely attributed to them.
(2) The facts of the case
What are the facts that lead the determinist to think our free acts are not free? Psychological explanations, the hypnotist's powers, and in general the understanding of natural laws that enables scientists to predict events have all in different ways influenced the determinist. Impressed by the fact that some apparently free acts are shown to be due to a hypnotist's suggestion or to a trauma in childhood, he thinks all our acts may be similarly explained. The scientists can accurately predict the movement of the planets so they will one day be able to predict my choices, and if a month ago they could foretell what I shall do today, then I don't decide today what I shall do.
But do any of these new discoveries show that we are more bound than we used to think we were? It is true that since we have learnt about a hypnotist's powers most of us would agree to withdraw our claim that a certain action was free if we learnt that it was done in obedience to a hypnotist's suggestion. I think that even a determinist would agree with us here if we could interest him in what must be to him the less important distinction between not free and very not free actions.
But how do we decide whether or not an act is done in response to the hypnotist's suggestion? For it is not enough that I have been in a hypnotic trance and that the suggestion has been made. If the act that has been suggested is one that is not wildly inconsistent with my usual behaviour and is one for which, if asked, I can give good reasons, then there are no grounds for supposing I did it in response to the hypnotist's suggestion. Even if the act is an unlikely one, I can by giving good reasons for doing it remove the suspicion that I was acting in obedience to the hypnotist. That the hypnotist has power over some people has been established simply because their subjects perform unlikely actions for the doing of which they can give no good reasons.
If then the past is somehow thought of as the great hypnotist, we could only show that we are acting in obedience to its suggestions by the same methods that we show we are acting in obedience to a human hypnotist, and by the test I outlined for this we see that since many of our actions are rational, many of our actions remain free. I think the same point can be made about psychological explanations — my action in buying a house with large rooms is not due to the fright I got at being locked in a tiny cupboard when I was a child unless my action in buying it doesn't fit into a rational pattern. Since we can only identify neurotic acts by contrast with rational acts all our acts can't be neurotic.
The story about feats of prediction seems to me much more complex, and I can here only suggest some approaches to it. First it is not simply the fact that someone can predict what I shall choose to do that shows I am not free. If I am a consistently moral person, or a consistently ambitious person, it may be easy to predict accurately my choices, but I choose freely for all that — or even the more freely for it. I am not as people say at the mercy of my whims'.
The idea that someone's knowing what will happen makes me powerless to alter what will happen, and all its attendant images of time unveiling a future that is already there, needs to be closely examined. The use of the word 'know' feeds this page 46 determinist illusion. No matter how successful a scientist has been in the past I will deny that he knows that x will propose to Mary tomorrow if when tomorrow comes x does not propose. This linguistic fact — ill understood — leads some to think that if the scientist knows that x will propose to Mary tomorrow x must propose tomorrow. This is true only in the sense that if x doesn't propose we will now deny that the scientist knew what we earlier said he knew.
The consequence of man understanding more and more accurately the connections between events at one time and events at a later time is just that man has by this knowledge been able to create his future and not just wait for time to unveil it. Indeed since the existence of causal connections can't be established without experiments we would not be able to predict the future did we not succeed in creating it.
(3) The consequences
I have tried to indicate under (2) what are the aspects of human behaviour that induce the determinist to deny that we ever act freely and why I think he is misguided in so doing. The undesirable logical consequences of the determinist's thesis are obvious from what I have said. It would do away with a useful and workable distinction between two types of human actions. Its practical consequences are also particularly unfortunate. We all know how easy it is to feel tied by our past and know too that the more we think we are tied by our past the more we are so tied. What we say effects what we believe, what we believe effects what we are. True, what I believe about the position of London does not change London much, but what I believe about my ability does make a difference to my ability; more than this, to believe I am powerless makes me, indeed, powerless.
To sum up, the determinist has a case to argue: the new discoveries of our age might have made it appropriate to deny that any human actions are free; but I hope I have shown that they do not in fact do so.