Peter Lipton was a Cambidge University philosopher of science and epistemologist. He wrote an important essay on the threat to free will from genetic determinism in The New Brain Sciences
, edited by Dai Rees and Steven Rose.
Lipton captures the causal determinist
analysis of most neuroscientists in his essay, and adds an extensive version of the standard argument against free will
, which he calls "the philosopher's classic free will dilemma."
Lipton sees little hope that any genetic factors can help in the problem of "generic determinism." This is of course quite right. Genetic and environmental factors enter into the "built-in" and adequately determined
factors that are used in the evaluations and deliberations stage of a two-stage free will
Genetic and generic
determinism: a new threat
to free will?
We are discovering more and more about human genotypes and about the connections between genotype and behaviour. Do these advances in genetic information threaten our free will? This chapter offers a philosopher's perspective on the question.
Whether or not genetic discoveries do really threaten free will, many feel threatened, and it is not difficult to see why. If genetic advances enable us to predict with increasing accuracy and reliability what people will do, this seems to undermine the pretensions of individual autonomy and agency. In what sense do I choose for myself what I do, if you can say reliably in advance what that choice will be?
The free will dilemma is a hardy philosophical perennial. After thousands of years of work there is still no generally accepted solution, no clear demonstration that free will really is possible. A philosopher may well wonder how new genetic knowledge could make things any worse, or indeed make things any different.
THE SCEPTICAL DILEMMA AND DIMINISHED RESPONSIBILITY
To see why a philosopher might suspect that genetic information could not possibly make the problem of free will any worse than it already is, we need to consider the classic free will dilemma, an argument with three very plausible premises and a depressing conclusion.
First, everything that happens in the world is either determined or not. Second, if everything is determined, there is no free will. For then every action would be fixed by earlier events, indeed events that took place before the actor was born. Third, if on the other hand not everything is determined, then there is no free will either. For in this case any given action is either determined, which is no good, or undetermined. But if what you do is undetermined then you are not controlling it, so it is not an exercise of free will. Finally, we have the conclusion: there is no free will. The argument has the form: heads or tails, if heads you lose, if tails you lose, therefore you just lose. Either determinism holds or it doesn't, if determinism holds there is no free will, if it does not hold there is not free will, therefore there just is no free will.
The dilemma is remarkably simple, and it packs an immediate punch. Let me nevertheless add a few comments on its structure and elements. The argument is clearly valid, in virtue of its form. To say that an argument is valid is not to say that its conclusion is true, but just that if the premises are all true, then the conclusion must be true as well, or equivalently that it is impossible for all the premises to be true yet the conclusion false. So anyone who wishes to reject the conclusion must also reject at least one of the premises. The argument does not assume any particular facts about our world, which suggests that the problem lies not in our world but in our concept. If the free will dilemma is sound — that is valid and with true premises — it seems to show that the very notion of free will is incoherent, something that could not possibly exist, a round square.
The first premise is indisputable, since it has the tautologous form P or not-P — everything is determined or not everything is determined. (Note that this is not the same as the disputable claim that either everything is determined or nothing is.) Just what determinism entails is a much more difficult question, and there are several different versions of the concept that could be deployed, though the first premise remains a tautology whichever one is chosen. The two most common versions of determinism appeal to causation or to the laws of nature. Thus determinism may be taken to be the view that everything that happens has a cause, or the view that everything that happens follows necessarily from the laws of nature in conjunction with the full state of the universe at any single moment. In fact this yields more than two conceptions of determinism, since the concepts of cause and law have themselves been given diverse philosophical treatment. Thus, some suppose that a cause is a condition sufficient for its effect, while others claim rather that it is necessary, something without which the effect would not have occurred. And while some philosophers have supposed that laws are simple de facto regularities, others have claimed that laws describe what happens by necessity, what could not have been otherwise.
The second premise of the dilemma, which asserts the incompatibility of free will and determinism, lacks the iron-clad security of a tautology, but there are powerful considerations in its favour. Free will seems to entail that the actor 'could have done otherwise', while determinism rules this out. The incompatibility of determinism with 'could have done otherwise' is particularly clear when determinism is defined in terms of the laws of nature (van Inwagen, 1975). If determinism is true, then what I did is entailed by laws of nature along with some particular facts about the state of the world before I was born. To have the power to have done otherwise, I would have to have the power either to change the laws or to change those prenatal facts. Clearly neither is possible.
Those who have tried to show that determinism and free will are nevertheless compatible have typically observed that the claim that my action was determined is compatible with my desires being among its causes and so that I would have acted differently, had my desires been different (Ayer, 1954). But defenders of the second premise reply that this is not enough to show that I could have done otherwise, if my desires are themselves just intermediate links in a long deterministic chain stretching back before my birth. In such a case, that people would have acted differently had their desires been different seems no more to show that they could have done otherwise than would saying that they would have acted differently, had the weather been different. Neither circumstance shows they have the power to change what they would do.
Another way of resisting the second premise is to question the connection between free will and 'could have done otherwise'. The desire being a cause of the action — which determinism allows — is clearly insufficient for free will. The addict is a model of someone whose free will has been compromised, though the addict desires the drug and that desire affects behaviour. But it has been suggested that what rules out free will in such cases is not that everything is determined, or that the agent could not have done otherwise, but rather that the addict does not have desires that are related to each other in the right way. For example, it has been claimed that the addict lacks free will because his desire for the drug is determined by the drug itself, rather than by higher-order commitment to wanting the drug (Frankfurt, 1971). Even if the addict is strangely happy to crave the drug, the craving is caused by the drug, not by the desire to crave. Ultimately, we all have desires we do not choose, but on this view what enables those of us who are not addicts to enjoy free will is that many of our desires are maintained because they are themselves desired. The defender of the second premise will not be satisfied by this, however, and will insist that the harmony of our mental economy is not enough to make room for the possibility of free will, if that entire economy and the actions it generates were determined by things that occurred before we were born.
The third premise of the dilemma is that free will is not compatible with indeterminism either. Supposing that some of my actions or their causes are themselves uncaused or ungoverned by deterministic law may allow that my actions could have been otherwise, but it does lot seem to allow that I could have done otherwise. Indeterminism does not allow the agent to control her actions in the way free will requires. I do not exercise free will if my arm spontaneously rises, nor is the situation any more promising if we construe an indeterministic process as one that is irreducibly probabilistic, rather than one that is entirely random. We do not create room for free will by leaving desires undetermined or by loosening the link between desire and action.
The simplest explanation for the conspicuous absence of philosophical progress on the problem of free will is that the sceptical dilemma is sound: free will really is impossible. If that is so, then the answer to our questions about genetic information is simple, if pathological. Nothing can threaten what could not exist anyway. If free will is impossible full stop, then it is something genetic knowledge can neither reduce nor destroy.
But we may be unable to accept the sceptical dilemma, even if we cannot see exactly what is wrong with it. As Isaac Bashevis Singer is reported to have said, 'Of course I believe in free will. I can't help it.' Our disposition to treat others as free agents seems impervious to argument. The dilemma may show that our full-blooded conception of free will is incoherent, and that we must pare it down if we are to believe in something that might exist. The big question is whether this process would leave us with something still strong enough to support the use we make of the concept, and the connections we make between judgements of freedom and judgements of responsibility and dignity.
Here as elsewhere in philosophy, I think that we ought to be opportunists, willing to vary our standards to suit our purposes. Free will is not the only area where powerful reasons are given for incredible conclusions. In the theory of knowledge, for example, all the best arguments seem to show that we have no justification for what we are quite sure we do know, that the sun will rise tomorrow or indeed that anything exists outwith our minds. Taking those arguments seriously helps us to illuminate our cognitive practices, but it is also important to vary the setting on the 'sceptic dial'. Supposing the worst — that we can know almost nothing — is a way of revealing some strata of our belief practices, but for other philosophical purposes we must take some layers of our knowledge for granted.
Similarly, while for some philosophical purposes we may wish to assume that free will is indeed impossible, for others we should suppose that people do sometimes act freely. To assess the impact of genetic information on free will, it is important to consider the radical perspective of the free will dilemma, which challenges the notion of free will under any circumstances. This will save us from claiming that genetic information is a particular threat to free will because it would deprive us of something that, as we can see from the sceptical dilemma, we never had anyway. But if we are accurately to assess the impact of biomedical developments, it is also important to consider the more conventional perspective, which allows that there is a distinction to be drawn among the things we actually do, between those actions that are free and those that are not.
The conventional distinction between free and unfree behaviour treats free will as a default condition which may be compromised in various ways. Addictive behaviour is one sort of case. Certain people lack normal inhibitory mechanisms and so are unable to control their desires. Some people are unable properly to recognise or characterise the nature of some of their own actions. Here one thinks of cases of serious psychological impairment, but it is worth noting that there is also a version of this phenomenon that afflicts us all. Our actions invariably have effects we are in no position to identify: we do things unintentionally, and these are not done by our own free will. It is also worth emphasising how common are the cases both of loss of inhibitory mechanism and of ability properly to identify one's actions, as the problems of excessive drinking illustrate. On the assumption that heavy drinking is not itself always addictive behaviour, we have here also the important complication of cases where one freely chooses to make oneself unfree. And our free will may be compromised in other ways besides. Should the acquisition of genetic information be added to the list?