Free Will - Study Notes
FREE WILL by Thomas Pink
What is the free will problem? Thomas Pink starts his book by contrasting things that are firmly outside our control with other things that we can control. For example we have no control over anything that happened before we were born or the colour of the eyes we were born with. By contrast we do have control over many of our present and future actions. Being in control of what we do and freedom of action is (within limits) a more or less constant feature of our day-to-day experience and of our thinking about ourselves and others.
But are we really in control of our actions? We have to consider the alternative notion that our actions are pre-determined or necessitated in advance, not by ‘fate’ or ‘destiny’ but by prior causes. Suppose that by the time we are born the world already contains causes (such as environment, genetics, upbringing and so on) that determine in advance exactly how we behave and therefore what we are going to do in one situation or another. If these things have been determined all along from the very beginning, how can we possibly be free to act otherwise?
Causal determinism is the claim that everything that happens has already been causally determined to occur, and this includes our own actions. This notion, clearly, is incompatible with the notion of freedom of action which we started with. If causal determinism is true, then freedom of action is an illusion. But our day-to-day beliefs are ‘naturally’ libertarian; that is, we intuitively believe that we are free, and that causal determinism, no matter how cogent it seems in theory, is false. But how can we be sure?
That question is what this book is all about. There are one or two red herrings that might mislead us from the main problem, for example freedom in a ‘procedural’ sense. If you are offered a choice between a glass of gin or a glass of whisky, it is your choice and nobody else’s. To say that you are ‘free’ to choose is literally true, but only in a procedural or a formal sense which neither refutes nor is incompatible with the basic thesis of causal determinism which states that your choice, whatever it is, will be causally predetermined.
Causal determinism received its major endorsement, as a theory, from the physical sciences which made significant progress from the 17th century onwards. Interestingly, in the twentieth century the theory of indeterminacy arose in relation to sub-atomic particles whose motions appeared to be indeterminate. But as Pink says, even if indeterminacy applies at the microscopic level, this may make no difference at the macroscopic level where causal predetermination of our actions remains a serious possibility.
Of course, causal determinism has never actually been proved; it has not been disproved, either. But there is a deeper problem. The converse of determinism is (presumably) indeterminism, the notion that at least some events, including some of our actions, are uncaused and are therefore purely random. If so, this is just as much a threat to the concept of freedom as determinism. How can we be ‘in control’ of our actions if some (or even all) of them are random events? We seem to be between the devil of determinism on the one hand and the deep blue sea of indeterminism on the other.
The question then becomes, what has to be true for an action to be not just a random event, but a genuine action carried out consciously, intentionally and deliberately? An obvious and plausible answer is that the action is carried out for a purpose, which suggests in turn such notions as motivation or desire. This may refute indeterminism, but does it refute determinism? Some philosophers think it does, calling it the compatibilist solution. Compatibilism says that freedom to act is compatible with our actions being caused because our actions are not only initiated but also guided by our desires and motivations.
We shall return to compatibilism later, but in the view of other philosophers the compatibilist answer is not viable for the simple reason that the desires and motivations which guide our actions could be causally predetermined too. Sceptics regard compatibilism as just an attempt to hedge, maintaining that freedom as a concept is inconsistent with determinism, indeterminism and compatibilism and is therefore illusory and impossible, notwithstanding our intuitions to the contrary. So where do we go from here?
In Chapter Two Pink looks at how mediaeval thinking about thought and action might inform our own ideas.
He starts by looking at the notion of purposive action, giving the example of a shark chasing a fish as being purposive. The shark is clearly ‘in control’ of its actions, but Pink asks is it really ‘free’ to act otherwise than it actually does? A clear difference between shark behaviour and human behaviour is our capacity to reason and to deliberate not only about how to act but whether to act. We also have the capacity to understand practical problems and to decide the best solution to a problem, where there can be a reasoned and correct answer. Freedom is closely linked to reason as well as to will and control.
The notion that we possess a will arises out of the idea that we have a capacity for rationality, a decision-making capacity. In the middle ages philosophers identified freedom of action with freedom of decision-making, and this is certainly consistent with our ordinary opinions these days. It is up to us how we act only because we can decide for ourselves which actions to carry out as well as how we shall act. We can control which decisions we take, and voluntary actions are the kinds of actions between which we decide one way or the other. In mediaeval terminology, an action of the will results in a voluntary action. It follows that an action can only be free to the extent that the will is free.
A decision, or a series of decisions, leads to the formation of an intention which then causes the performance of the voluntary action(s) decided upon. And with intentions, we also have the concepts of goals and purposes. Our commonsense intuitions on these matters are not so very different from the mediaeval concepts.
The major divergence between mediaeval and modern thinking is that for the mediaeval thinkers, the will and our decision-making capacity were non-physical events and processes. To us, however, the idea of non-physical events and processes is more puzzling; we are more likely to think in terms of physical brain-events such as neurological processes. Our scientific knowledge also leads much more readily to the notion of causal determinism which would have been less likely to have occurred to the mediaeval thinkers.
The mediaevals were more concerned with an entirely different problem which was, basically, theological. If God is the all-powerful ruler of the universe so that everything happens by the will of God and if human actions are no exception to this, how can we possibly be free to act otherwise? We’ve heard this question before, of course, but for a very different reason.
In Chapter Three we return to compatibilism, the notion that freedom to act is compatible with our actions being caused because our actions are not only initiated but also guided by our desires and motivations.
Pink makes the interesting observation that some philosophers advocate compatibilism not because they are trying to preserve the concept of freedom of action as such but because it may not matter after all. Some philosophers seek to make sense of responsibility, especially moral responsibility, even in the absence of ‘freedom’ (though whether this does make sense or not is debatable).
Also, there are two very dissimilar views of ‘freedom’. One is called the rationalist conception of freedom. This means that a free agent is ‘an agent capable of reasoning or deliberating about how to act, and of taking decisions about how to act on the basis of that deliberation’. We encountered this concept a moment ago.
The other view of freedom is the naturalist conception of freedom. To put it in a nutshell, this equates freedom of action with power or control over events. This again is a concept we have encountered before. Whether reasoning takes place or not, it does not differ in kind from the actions of animals who also may have a greater or lesser degree of control over what comes to their attention. In Chapter Three Pink concentrates on the rationalist conception of freedom, leaving the naturalist conception of freedom until Chapter Four.
The rationalist version of compatibilism means that to have real control over our actions involves the capacity to give our actions deliberate guidance and direction. We have to be able to reason about how it is best to act, and to decide how to act on the basis of that reasoning. On this basis to be a free agent is to be a rational agent, and if these are the same, then freedom and rationality should never conflict.
But they often do conflict, and the rationalist conception of freedom can in theory lead to paradox. Pink gives the example where the sensible thing is to take the one medicine which will cure a serious illness but which happens to have a nasty taste. If you are completely reasonable you will do the sensible thing with no chance that you will do otherwise. But if the very fact of your rationality ensures that you take the medicine, how can you ever be free to act otherwise? As Pink says, being fully reasonable seems to involve one’s decisions and actions being determined in advance by ‘reason’. If so, this brings us straight back to determinism where you do not control your actions, some sort of internal rational calculator does, inside your head so to speak. But compatibilism as we understand it means guiding your own actions, not, one would have thought, surrendering that guidance to a rational calculator, either internal or external.
In Chapter Four Pink summarizes the views of the 17th century philosopher Thomas Hobbes (1588-1679), whose views were radically different from those of the mediaeval thinkers.
They held that the human will was a very special decision making capacity, a rational or reason-involving capacity that animals entirely lacked. Hobbes ‘downgraded’ the will: to him, it was nothing more than the desire and determination to satisfy appetites and wants.
The universe, to Hobbes, was a material, deterministic system, wholly physical and mechanistic, and human beings, like animals, are simply material bodies, albeit complex ones. Causal determinism is implicit in his thinking. ‘Freedom’ can be explained entirely in materialist and determinist terms as the absence of obstacles or other hindrances to the satisfaction of what you desire or want to do. ‘Reason’, for Hobbes, is nothing more than our ability to calculate, to make deductions etc. in order to achieve our objectives. Decisions and intentions are no more than forms of desire, and the will is nothing more than the determination to fulfill our desires.
Hobbes defines ‘action’ as what is done voluntarily on the basis of a decision or a want to do it. A decision cannot, therefore, be an action itself. A decision is nothing more than a strong desire which, like a feeling, we experience and then act upon. Human action is basically us being pushed into motion by our wants – wholly mechanistic and deterministic.
His definition of freedom is fully compatible with causal determinism and the only things that remove Hobbesian freedom are, as Pink says, ‘locked cell doors and ropes and chains’ (or, perhaps, the threat of fines or other penalties). Freedom also applies, in Hobbesian terms, to animals; a wild animal free to roam is literally that - free.
Pink remarks that Hobbes is ‘the inventor of the modern free will problem’. In the Middle Ages the free will problem was about how freedom was related to reason and to the concept of an omnipotent and omniscient God. The modern problem is, as we have seen, how is freedom possible if humans are part of a wholly physical world of cause and effect?
Many modern philosophers have been convinced by Hobbes. Not all have, and Pink raises an interesting objection to the Hobbesian equation of freedom with the satisfaction of desires or other objectives. He suggests that desires themselves can be obstacles to freedom. He gives the example of the drug addict who is imprisoned (metaphorically) not by locked cell doors or chains but by his own desires; he lacks the freedom not to take the drug to which he is addicted, and he lacks the freedoms to do the sorts of things the rest of us do, because of his addiction. He has no control over how he acts or what he does; if he successfully obtains the drugs he wants, he is not acting ‘freely’ but only slavishly. Pink could have mentioned that addiction is only one example of how people can be slaves to their own desires, obsessions and so on, and therefore ‘unfree’. This would appear to be a shortcoming in Hobbes’ account of ‘freedom’, though I suspect Hobbes would have shrugged it off. He might simply have said, ‘look, I’m only defining freedom, I’m not passing comment on individuals who frustrate their own freedom for psychological reasons.’
But the main area where Hobbes’ account does not ‘square’ with commonsense is his suggestion that taking a decision is akin to having a feeling which we then act upon. This seems simplistic. The weighing up of different options against different factors and criteria is more akin to a reasoned process, not an impulse or a feeling. Also, a decision, unlike a feeling, an impulse or a compulsion, is something that we are free not to take, and what we decide is within our conscious control.
It is important to consider the difference in meaning between the words ‘free’ and ‘voluntary’ in this particular context. To do something ‘voluntarily’ means that you are doing it deliberately and intentionally, even if you don’t necessarily want to do it (e.g. your tax return). To be ‘free’ to do something means that you have the choice to do it or not to do it, which is different. Freedom and voluntariness often overlap, but they are not the same. Hobbes failed to make the distinction. Freedom applies to decisions; actions are carried out voluntarily (except for accidents, mistakes or through being forced).
In Chapter Five, quite a short one, Pink looks at the notion of responsibility for an action (moral responsibility in particular) without the notion of freedom. We noted a moment ago that this idea is debatable. Some philosophers define responsibility for an action purely in terms of voluntariness, just as Hobbes did, and either ignore freedom as we have just defined it or assimilate it to voluntariness which simply means doing what you intend.
Pink mentions John Calvin, the 16th century Protestant who said we are responsible and blameworthy because we do wrong voluntarily, because we want to, but we are also predestined to sin by necessity. On this basis it sounds as if you could be liable for something you cannot help, although Pink does not discuss this particular angle. Pink also mentions American philosophers such as Harry Frankfurt who simply says that if you deliberately perform an action you’ve decided upon or want to do, then you are responsible for it and the freedom to act otherwise is irrelevant. Cogent, after all.
But Pink maintains that this over-simplifies, saying that commonsense cannot give up the notion of freedom and replace it with voluntariness because we are not only responsible for our actions but also for our prior decisions. We can sometimes take the wrong decision, or a silly one. He agrees with Hobbes that decisions are not taken ‘voluntarily’, but for a very different reason from Hobbes - they are taken freely, in the knowledge that an alternative decision is available. He disagrees with Hobbes that decisions, like feelings, ‘come over us’ in a way we cannot help. This would exclude freedom and responsibility altogether.
In Chapter Six Pink takes a closer look at self-determination.
This must, he says, take the form of freedom of choice, which cannot be the same as voluntariness for the reasons given a moment ago. We have also seen that control over how we act depends on our actions not being causally determined in advance by prior events outside of our control. This is the problem of causal determinism which we looked at earlier. But if an action is causally undetermined, its occurrence must depend on simple chance. This is the randomness problem – for chance does not amount to genuine freedom at all. However, Pink maintains there is a third possibility that there can be some events, that is, actions which are neither causally pre-determined nor merely chance events because they occur under our control. This is to be the argument of the remainder of Pink’s book.
As well as the randomness problem, the other problem is the exercise problem, as Pink calls it. The question is, how is libertarian freedom something we can exercise? The answer is that an action is something done for a purpose, as we have already seen, and it is this that makes the action intelligible as a deliberate doing and that allows us to explain why the action was performed. Actions are not blind reflexes, as he puts it.
Hobbes took the view that the purposes for which we act come from our prior desires which cause us voluntarily and deliberately to act the way we do. This account is at first glance plausible and coherent. But as Pink points out, Hobbes’ theory is by its very nature an account of action as an effect of our desires or wants – but these are prior occurrences outside of our control. So we are straight back to causal determinism, which we saw earlier as a direct threat to the whole idea of libertarian freedom to act.
Now we come to a red herring, a proposed solution to this problem which doesn’t really work. It says that we can distinguish between prior causes that determine that their effects must occur, but perhaps there are other causes that may influence whether their effects occur without strictly determining that those effects will occur. A good example Pink quotes is smoking as a cause of cancer. It often causes cancer, but not always. It is probabilistic.
But as Pink says, the distinction between determining and probabilistic causes does not really help the causal determinism issue. Even if libertarian freedom is affected by causal influences without necessarily determining what we do, that is enough, in Pink’s view, to negate the libertarian concept of full freedom of action.
So what is the way out? One way is to give up the whole idea of libertarian freedom and follow Hobbes. On the other hand, why not abandon the Hobbesian conception of action which seems to be the source of all the problems? Pink suggests that through decision and control we find genuine action, in a form entirely independent of prior causation. He devotes the remainder of the book to developing this theme.
Chapter Seven is entitled ‘Self-determination and the Will’. Summarizing, the ‘exercise problem’ as he calls it arises with Hobbes’ theory of action in which action is characterized as an effect of our desires. But we cannot help having thoughts and feelings as and when they occur to us. They are prior occurrences beyond our control. It follows that ‘freedom is removed’. Pink therefore develops an alternative to Hobbes’ theory of action.
The challenge he faces is how to explain how action can occur ‘in uncaused form’, as he puts it, without falling into the theoretical trap of indeterminism, that is, of actions being purely random events. He maintains that sometimes we can decide to do things without any particular desire or motivation having ‘pushed us so to decide’; in fact decisions are possible without any prior desire or motivation at all. Also, no evidence has been produced that a decision is a passive event that ‘comes over us’ like a feeling. In fact we often experience it very differently.
Pink goes on to argue that the one thing that makes an action a genuine action as distinct from just an event is that it has a purpose. But where does purpose come from? In some cases it does come from desires or wants, obviously, but not in all cases. Many actions have no obvious connection with wants or desires at all. Pink argues for a ‘practical reason based’ theory of action. This means we are capable of deliberating or reasoning about which action to perform, and this does not in itself presuppose the necessity for actions to be based on desires. The distinctive feature of action is ‘goal-directness’, as he puts it, plus your appraisal of how likely or unlikely it is that your proposed action will succeed. Your decision can therefore be sensible or silly, practical or impractical, and so on.
Notice how decisions and actions are very different from desires or wants which in many cases have no direct or indirect connection with anything you’re going to do at all. A decision is itself an action with a goal, or a motivation with an object. We therefore have a model of intentional action which characterizes it not as the effect of a desire which we experience, but as a practical exercise of reason. To put it another way, we are not just reacting or responding to events which occur around us, we are making events for reasons. As he puts it, ‘I can be the free creator of my own aims and purposes’.
The important point he makes is that freedom begins with an understanding of possible goals and of controlling which of these goals you choose. This can be understood perfectly clearly without reference to prior causes. And all of this is consistent with our original, libertarian concept of freedom of action, and therefore the ‘exercise problem’ is solved.
Pink notes that many modern philosophers write as if it were ‘just obvious’ that action is caused by desires. He might also have mentioned the influence of behavioural psychology which seems to reduce human behaviour to just a set of conditioned reflexes, or the influence of Freudian psycho-pathology in which, like Punch and Judy, we are jerked hither and thither, like puppets, by our complexes, obsessions and what have you.
Moving on to Chapter Eight, Pink poses a question: ‘is freedom a causal power?’ Action is no longer to be understood as an effect of passive desires; actions can occur as uncaused decisions without being any the less genuine and deliberate. But the randomness problem still remains. Pink still needs to show how we can distinguish between causally undetermined freedom and the operation of mere chance. According to the sceptics there is no middle way between causal predetermination and events occurring through pure chance. Pink maintains there is a third possibility, that an event can occur through the exercise of our freedom because we are exercising control over it.
However, this possibility is ruled out if freedom is only another causal power – which would, by implication, bring us back to determinism. Of course, freedom is a ‘power’ of some kind, but as Pink says, it is a power we have that determines how we will act. It is a special power which only ‘rational beings such as we humans can possess’ and ‘which can only be exercised through action’. But this requires explanation, of course. ‘Causal power’ is found throughout nature. Even mindless objects possess ‘causal power’ – a gale can bring down trees. Is human freedom only a further instance of ‘causal power’ per se?
Pink suggests otherwise, because causal power just on its own does not mean the same as control. A wind blowing against a tree does not ‘control’ whether or not the tree collapses. Control includes intention, and it is a power that can be exercised in more than one way, to determine whether something will happen or not. So it is with freedom and action. Causal power just in itself and the power of freedom are different and it is a mistake to confuse them, or to ‘reduce’ or ‘explain’ the one in terms of the other as Hobbes did.
Compatibilists have claimed that freedom, the power we have over our actions, is basically a causal power of our desires. After all, any obstacle to the satisfaction of our desires is by its very nature an obstacle to the exercise of freedom. So it is tempting to define freedom in terms of desires being satisfied without obstruction. But this is simplistic. Goal directed action need not be caused by desires, and what might prevent the satisfaction of our desires could be none other than our own deliberate decisions to refrain. And as Pink asks, what is ‘freedom threatening’ about this? On the contrary, it is an exercise of freedom. Freedom is more accurately seen as a control over which of our desires are to be pursued or passed over. Pink might have added that an individual hell-bent on the satisfaction of his desires is literally out of control; he is not a free agent at all.
So – what is freedom a causal power of? Pink suggests there is only one possible answer – it is a power exercised directly by our own selves. It is a two-way power, a power to do or to refrain, so it is exercisable in more than one way (very different from a gale blowing against a tree). It is a power possessed and exercised not by events influencing us but by us influencing events (other things being equal). It is a libertarian theory of action called agent-causation. It is therefore a special case of ‘causal power’, not just any old form of causal power like a gale which can hardly be said to control itself. It solves the randomness problem because we are the causes of our actions; they are not simply ‘functions’ of events beyond our control. If Pink is correct, then he has refuted both causal determinism and indeterminism and has vindicated libertarianism and commonsense.
RHS, 2012.