Free Life, the Journal of the Libertarian Alliance

In the determinism versus freewill debate it is possible to identify only three consistent positions:

Freewill implies indeterminism but indeterminism does not imply freewill. Or, indeterminism is a necessary but not a sufficient condition for freewill. It is not possible to advocate both determinism and moral responsibility or determinism and freewill. We ought also to distinguish the following two meanings of determinism. We can say that all human behaviour is determined by biological and environmental stimuli. But this need not mean the same thing as the physical world being deterministic. We can have an indeterministic physical world (i.e., one in which there are intrinsically chance-like events) but with human behaviour being wholly governed by these events. In this case, we can still say that human behaviour is biologically and environmentally determined. That is, human behaviour is determined, or caused, by those stimuli but the stimuli themselves may be intrinsically indeterministic. Thus, to say that human behaviour is determined is not necessarily to imply that the world is deterministic.

Now for the determinist case against freewill to succeed it is not enough to show that most human behaviour is determined or that general patterns of behaviour can be predicted. The determinist has to show that all human behaviour is determined.

Sean Gabb thinks that if the world is intrinsically indeterministic this might prove a threat to his case. He is dubious about what the physicists have to say about this. In fact, Mr Gabb need not have worried. Indeterminism alone is insufficient to defeat his case.

Firstly, if human behaviour is governed by random stimuli this is not enough to produce freewill. We do not seem to act randomly. On the whole, we act in a fairly ordered fashion. Though, if we do have freewill, then our behaviour is indeterministic but it is not random.

Secondly, despite Mr Gabb's protestations, indeterminism in the micro-world need not manifest itself as indeterminism in the macro-world. Large numbers can produce regularity. That is why we can use probability theory to make predictions. Thus, if we suppose that the sequence of heads and tails produced by a repeatedly tossed coin is a random process we can still say that, after a large number of tosses, the ratio of heads to tails will be approximately one to one. Chemical reactions are essentially deterministic. So it's quite possible for the brain to be macroscopically deterministic even though there is an underlying indeterminism at the quantum level.

Thirdly, although quantum mechanics is a probabilistic theory, it does also, under certain conditions, assert probabilities of 0 or 1, which imply determinism in those situations.

Of course, a new theory may show that quantum indeterminism is an illusion and that we have complete determinism after all. Yet the fact that our currently most accurate physical theories are indeterministic should give the thoroughgoing determinist pause for thought. Although indeterminism alone cannot refute the determinist case against freewill it does show that, given the same initial conditions and laws, there can be variations in observed behaviour.

However, to say that we have freewill is not to say that our actions are uncaused. Freewill theory asserts that at least some of our actions are not necessitated by antecedent factors. In such cases the ultimate cause of our actions is the will.

Mr Gabb takes offence at the accusation that determinism holds human- beings to be "mere" robots. Yet it is not clear why he should take offence at this. If determinism is true then we are mere robots. And Mr Gabb offers no evidence to dispute this. He just says that we are, in effect, complex robots.

As regards the actual practice of punishing people for crimes, I do not think our procedures need be affected by whether we hold determinism or freewill to be true. (I here take punishment to refer to the practice of incarcerating people or imposing fines on them, etc.) What should be affected is the meaning which we attribute to our actions.

Suppose Smith murders Jones. If we hold Smith's actions to be fully determined we could (consistently) punish him for only the three possible reasons:

If we hold Smith to have freewill we can punish him for the same reasons. The first reason is unproblematic for the freewill case. But the second and third reasons seem, at first, to support the determinist case. Is not deterrence only effective because it determines people's behaviour? Is not punishment as a means of reform only effective because it determines Smith's behaviour?

Man, in acting, takes into account both the laws of nature and the laws of men. In other words, he acts under constraints. The laws of nature and the laws of men are constraints, given that Man aims to satisfy certain ends. Now, in a sense, one can claim that acting under constraints implies determinism. In other words, constraints "determine" that you will not act outside the constraints. But strict determinism requires much more than this. It requires that within a set of constraints it is still not possible to act freely.

Suppose the external environment's constraint set of behavioural possibilities is reduced from the set A, B, C, D to A, B. That is, prior to this reduction it is possible for an individual to choose actions A, B, C, or D. After the reduction he is restricted to choosing from A or B. Then we can, if we like, say that the individual is "determined" not to choose C or D. However, if he is able to choose freely from A or B then we have freewill. Just as we do if, prior to the change in constraints, he is free to choose from A, B, C or D. Of course, determinism claims that the free choice is illusory in either case. But all I am claiming at this stage is that constraints per se do not imply determinism. It should be noted that constraints are typically only constraints, given that Man aims at definite ends. They are not constraints if he aims at other ends. If my end is that I wish to die then the law of gravity is not a constraint to my walking off the Empire State building.

Mr Gabb claims that the alleged fact that determinism implies the absence of moral reponsibility would not refute determinism. He also casts doubt on the alleged fact. He is right on the first count but wrong on the second. I have already shown the second to be wrong in my discussion of the reasons for punishment above. Nevertheless, Mr Gabb seems quite happy to accept that determinism implies the absence of moral responsibility, in any case. Although it is true that this does not refute determinism, in practice it makes the behaviour of most determinists inconsistent. Virtually everybody uses the language of moral condemnation and it is difficult to believe that it is used only to evoke shame in the recipient. Morally condemning someone is senseless if it is held that he cannot do other than what he does.

I argued above that determinists can punish someone for only three possible reasons and that freewill advocates can punish someone for the same three reasons. However, there are certain things we can say about the nature of the implemented punishment. Determinists have no reason for implementing any more severe punishment than is necessary to achieve the objectives stated above, and compassion ought to lead them to implement less than perfect punishments, if perfect punishments would be more severe ones.

I claim that it is a fact that determinism implies lack of moral responsibility. If so, then it seems cruel to torture or execute a criminal, if what the criminal does is beyond his control. Determinists certainly have no logically valid reason for doing this if it can be shown that more severe punishments do not more perfectly satisfy the three reasons for punishment. They have a logically valid reason, if more severe punishments are more effective, but it does seem rather cruel, given the criminal's lack of responsibility for his actions. The determinist would be saying, in effect: "You are not responsible for your actions. It's not your fault, but I am going to torture and/or execute you, in any case." That would be just revenge. But revenge seems out of place in the determinist's picture of the world.

On the other hand, the freewill advocate can logically implement more severe punishments, whether these are effective or not, since his view is that the criminal deserves it. Revenge is fully consistent with the postulation of freewill.

When determinists argue for determinism they often claim that we can explain the general behaviour of people, often stretched out over many years. So we can explain why Smith is a criminal based on his poor background and inadequate upbringing. If the anti-determinist then points out that Jones, from the same poor background and with the same inadequate upbringing, is not a criminal, the determinist claims he can account for this also. He points to extra factors which apply in Jones's case but not in Smith's. It is not hard to do this since, obviously, no two individuals are subjected to exactly the same environmental stimuli.

(1) It is not enough for the determinist to be able to explain general patterns of behaviour, such as Smith's criminality. He needs to be able to show how every single action of Smith's is determined, from his decision to be a criminal to his decision to embark on an episode of burglary at this instant rather than that instant. A determinist may be able to explain why I like maths but no determinist can explain why I choose to do some maths today but not tomorrow, or vice-versa.

(2) Often, when we claim to have explained someone's actions we have done no such thing.

In regard to (2) suppose it is known that I object to taking a job in the defence industry. Suppose, further, that I am offered such a job and I have been unemployed for several years, so that my finances are running low. If I accept the job the determinist will claim to be able to explain this. He will say that, although I object to working in the defence industry, I choose to make an exception because I need the money. My need for money overrides my principles and can be seen as "determining" my actions.

On the other hand, if I reject the job the determinist will claim to be able to explain that also. He will say that I reject the job because my principles override my need for money. My principles can be seen as "determining" my actions.

But these are ex-post rationalisations of observed behaviour. What needs to be explained is why, in the one case, my principles override my need for money and, in the other, they do not. The best explanation that determinism can give for this is that the "concatenation of events" determines which has the ascendancy. Maybe, but this is no explanation. Anything whatever can be explained by asserting that it is so because of the concatenation of events.

If, on the other hand, we assert that my behaviour is willed we need not look for any further explanations since freewill theory asserts that the will is the ultimate cause (in cases in which we are exercising freewill).

If two alcoholics have tried to give up alcohol and one succeeds while the other does not it is perfectly reasonable to attribute the success of the former to his greater willpower, though it is not necessarily the case that this be the reason. But the sort of explanation I described above for determinism will not do.

Mr Gabb claims that the fact that we feel free is no good reason for upholding freewill. If all I can offer as an argument for freewill is simply the statement - "I feel free" - then Mr Gabb is surely right. However, I can assert more than this. And it may not amount to a proof but it certainly amounts to a challenge - one which I expect not to be able to be answered.

Now, I am quite prepared to admit that much of our behaviour is determined. Ironically, the argument for freewill from introspection works precisely because we do experience behaviour which is determined. Such behaviour provides a contrast, by which we can infer the existence of freewill. I here reproduce the concluding arguments of my essay The Reality of Freewill.1

There are certainly many conscious experiences which involve no element of freedom on the part of the subject. For example, suppose I accidentally burn my hand with hot water. Then my immediate reaction is to withdraw it. I am conscious of the pain and I am conscious of my response. Yet, under these circumstances, I clearly have no control over my behaviour. Thus there certainly can be conscious behaviour which is not free. But suppose that I know in advance that I am going to be burnt. I could then choose to leave my hand under the hot water (as an exercise in pain resistance, for example) or I could choose not to. But if all behaviour is supposed to be completely specified by biological and environmental factors it is difficult to see how my response can be anything other than the conscious and automatic one. My decision to leave my hand put is clearly not caused by the hot water stimulus and nor is it a matter of chance. In short, it is freely willed.

Consider, now, the activity of reading, say, a philosophical treatise. One can have the conscious experience of reading through several pages of text and not making an effort to understand what is read. (Or one can have the unconscious experience of not having read anything at all, i.e., that of daydreaming.) When, however, an effort is required it is clear that this necessitates voluntary action. If you are trying to understand vector calculus for the first time it is not "biology" or "environment" which enables you to do so. It is you. Similar arguments can be applied to the whole range of human experience. In general, new activities require voluntary effort to master them. It is only subsequently that these activities become automatised.

Materialists are unable to give a coherent account of human achievement or creativity by attributing them solely to electro-chemical reactions in the brain. They can point to no factors external to the human will which would enable them to provide a criterion of demarcation between action and reaction. Until they are able to do so the foregoing arguments must be regarded as decisive.

Mr Gabb claims that determinism contains all the available data. Well, it doesn't.

1. Kevin McFarlane, The Reality of Freewill, Philosophical Notes No. 10, Libertarian Alliance, London, 1989.