我最初对自由意志这个问题发生兴趣是在看约翰·埃克尔斯的《脑的进化：自我意识的创生》时，去年，Matt Ridley的 Genome: the autobiography of a species in 23 chapters （《基因组：一个物种的23章自传》）一书再次激起我的兴趣，该书第22章谈论了这个问题，新浪读书上有 该书的节译本 ，其中包括了 第22章的译文 。下面是该章原文：（注：关于“休谟之叉”一词究竟指什么，好像有不同说法，所以我暂时放弃这一术语） CHROMOSOME 2 2  Free Will Hume's fork: Either our actions are determined, in which case we are not responsible for them, or they are the result of random events, in which case we are not responsible for them. Oxford Dictionary of Philosophy As this book is being completed, a few months before the end of a millennium, there comes news of a momentous announcement. At the Sanger Centre, near Cambridge - the laboratory which leads the world in reading the human genome - the complete sequence of chromosome 22 is finished. All 15.5 million 'words' (or so - the exact length depends on the repeat sequences, which vary greatly) in the twenty-second chapter of the human autobiography have been read and written down in English letters: 47 million As, Cs, Gs and Ts. Near the tip of the long arm of chromosome 22 there lies a massive and complicated gene, pregnant with significance, known as HFW. It has fourteen exons, which together spell out a text more than 6,000 letters long. That text is severely edited after tran­scription by the strange process of RNA splicing to produce a highly complicated protein that is expressed only in a small part of the prefrontal cortex of the brain. The function of the protein is, generalising horribly, to endow human beings with free will. Without HFW, we would have no free will. The preceding paragraph is fictional. There is no HFW gene on chromosome 22 nor on any other. After twenty-two chapters of relentless truth, I just felt like deceiving you. I cracked under the strain of being a non-fiction writer and could no longer resist the temptation to make something up. But who am 'I'? The I who, overcome by a silly impulse, decided to write a fictional paragraph? I am a biological creature put together by my genes. They prescribed my shape, gave me five fingers on each hand and thirty-two teeth in my mouth, laid down my capacity for language, and defined about half of my intellectual capacity. When I remember something, it is they that do it for me, switching on the CREB system to store the memory. They built me a brain and delegated responsibility for day-to-day duties to it. They also gave me the distinct impression that I am free to make up my own mind about how to behave. Simple introspection tells me there is nothing that I cannot help myself doing. There is equally nothing that says that I must do one thing and not something else. I am quite capable of jumping in my car and driving to Edinburgh right now and for no other reason than that I want to, or of making up a whole paragraph of fiction. I am a free agent, equipped with free will. Where did this free will come from? It plainly could not have come from my genes, or else it would not be free will. The answer, according to many, is that it came from society, culture and nurture. According to this reasoning, freedom equals the parts of our natures not determined by our genes, a sort of flower that blooms after our genes have done their tyrannical worst. We can rise above our genetic determinism and grasp that mystic flower, freedom. There has been a long tradition among a certain kind of science writer to say that the world of biology is divided into people who believe in genetic determinism and people who believe in freedom. Yet these same writers have rejected genetic determinism only by establishing other forms of biological determinism in its place - the determinism of parental influence or social conditioning. It is odd that so many writers who defend human dignity against the tyranny of our genes seem happy to accept the tyranny of our surroundings. I was once criticised in print for allegedly saying (which I had not) that all behaviour is genetically determined. The writer went on to give an example of how behaviour was not genetic: it was well known that child abusers were generally abused themselves as children and this was the cause of their later behaviour. It did not seem to occur to him that this was just as deterministic and a far more heartless and prejudicial condemnation of people who had suffered enough than anything I had said. He was arguing that the children of child abusers were likely to become child abusers and there was little they could do about it. It did not occur to him that he was applying a double standard: demanding rigorous proof for genetic explanations of behaviour while easily accepting social ones. The crude distinction between genes as implacable programmers of a Calvinist predestination and the environment as the home of liberal free will is a fallacy. One of the most powerful environmental sculptors of character and ability is the sum of conditions in the womb, about which you can do nothing. As I argued in the chapter on chromosome 6, some of the genes for intellectual ability are probably genes for appetite rather than aptitude: they set their pos­sessor on a course of willing learning. The same result can be achieved by an inspiring teacher. Nature, in other words, can be much more malleable than nurture. Aldous Huxley's Brave new world, written at the height of eugenic enthusiasm in the 1920s, presents a terrifying world of uniform, coerced control in which there is no individuality. Each person meekly and willingly accepts his or her place in a caste system - alphas to epsilons - and obediently does the tasks and enjoys the recreations that society expects of him or her. The very phrase 'brave new world' has come to mean such a dystopia brought into being by central control and advanced science working hand­in-hand. It therefore comes as something of a surprise to read the book and discover that there is virtually nothing about eugenics in it. Alphas and epsilons are not bred, but are produced by chemical adjustment in artificial wombs followed by Pavlovian conditioning and brainwashing, then sustained in adulthood by opiate-like drugs. In other words, this dystopia owes nothing to nature and everything to nurture. It is an environmental, not a genetic, hell. Everybody's fate is determined, but by their controlled environment, not their genes. It is indeed biological determinism, but not genetic determin­ism. Aldous Huxley's genius was to recognise how hellish a world in which nurture prevailed would actually be. Indeed, it is hard to tell whether the extreme genetic determinists who ruled Germany in the 1930s caused more suffering than the extreme environmental determinists who ruled Russia at the same time. All we can be sure of is that both extremes were horrible. Fortunately we are spectacularly resistant to brainwashing. No matter how hard their parents or their politicians tell them that smoking is bad for them, young people still take it up. Indeed, it is precisely because grown-ups lecture them about it that it seems so appealing. We are genetically endowed with a tendency to be bloody-minded towards authority, especially in our teens, to guard our own innate character against dictators, teachers, abusing step­parents or government advertising campaigns. Besides, we now know that virtually all the evidence purporting to show how parental influences shape our character is deeply flawed. There is indeed a correlation between abusing children and having been abused as a child, but it can be entirely accounted for by inherited personality traits. The children of abusers inherit their persecutor's characteristics. Properly controlled for this effect, studies leave no room for nurture determinism at all. The step­children of abusers, for instance, do not become abusers.1 The same, remarkably, is true of virtually every standard social nostrum you have ever heard. Criminals rear criminals. Divorcees rear divorcers. Problem parents rear problem children. Obese parents rear obese children. Having subscribed to all of these assertions during a long career of writing psychology textbooks, Judith Rich Harris suddenly began questioning them a few years ago. What she discovered appalled her. Because virtually no studies had controlled for heritability, there was no proof of causation at all in any study. Not even lip service was being paid to this omission: correlation was being routinely presented as causation. Yet in each case, from behaviour genetics studies, there was new, strong evi­dence against what Rich Harris calls 'the nurture assumption'. Studies of the divorce rate of twins, for example, reveal that genetics accounts for about half of the variation in divorce rate, non-shared environmental factors for another half and shared home environ­ment for nothing at all.1 In other words, you are no more likely to divorce if reared in a broken home than the average - unless your biological parents divorced. Studies of criminal records of adoptees in Denmark revealed a strong correlation with the criminal record of the biological parent and a very small correlation with the criminal record of the adopting parent — and even that vanished when con­trolled for peer-group effects, whereby the adopting parents were found to live in more, or less, criminal neighbourhoods according to whether they themselves were criminals. Indeed, it is now clear that children probably have more non­genetic effect on parents than vice versa. As I argued in the chapter on chromosomes X and Y, it used to be conventional wisdom that distant fathers and over-protective mothers turn sons gay. It is now considered much more likely to be the reverse: perceiving that a son is not fully interested in masculine concerns, the father retreats; the mother compensates by being overprotective. Likewise, it is true that autistic children often have cold mothers; but this is an effect, not a cause: the mother, exhausted and dispirited by years of unre­warding attempts to break through to an autistic child, eventually gives up trying. Rich Harris has systematically demolished the dogma that has lain, unchallenged, beneath twentieth-century social science: the assump­tion that parents shape the personality and culture of their children. In Sigmund Freud's psychology, John Watson's behaviourism and Margaret Mead's anthropology, nurture-determinism by parents was never tested, only assumed. Yet the evidence, from twin studies, from the children of immigrants and from adoption studies, is now staring us in the face: people get their personalities from their genes and from their peers, not from their parents.1 In the 1970s, after the publication of E .O. Wilson's book Sociobiol­ogy, there was a vigorous counter-attack against the idea of genetic influences on behaviour led by Wilson's Harvard colleagues, Richard Lewontin and Stephen Jay Gould. Their favourite slogan, used as a tide for one of Lewontin's books, was uncompromisingly dogmatic: 'Not in our genes!' It was at the time still just a plausible hypothesis to assert that genetic influences on behaviour were slight or non­existent. After twenty-five years of studies in behavioural genetics, that view is no longer tenable. Genes do influence behaviour. Yet even after these discoveries, environment is still massively important - probably in total more important than genes in nearly all behaviours. But a remarkably small part in environmental influ­ence is played by parental influence. This is not to deny that parents matter, or that children could do without them. Indeed, as Rich Harris observes, it is absurd to argue otherwise. Parents shape the home environment and a happy home environment is a good thing in its own right. You do not have to believe that happiness determines personality to agree that it is a good thing to have. But children do not seem to let the home environment influence their personality outside the home, nor to let it influence their personality in later life as an adult. Rich Harris makes the vital observation that we all keep the public and private zones of our lives separate and we do not necessarily take the lessons or the personality from one to the other. We easily 'code-switch' between them. Thus we acquire the language (in the case of immigrants) or accent of our peers, not our parents, for use in the rest of our lives. Culture is transmitted autonomously from each children's peer group to the next and not from parent to child - which is why, for example, the move towards greater adult sexual equality has had zero effect on willing sexual segregation in the playground. As every parent knows, children prefer to imitate peers than parents. Psychology, like sociology and anthropology, has been dominated by those with a strong antipathy to genetic explanations; it can no longer sustain such ignorance.2 My point is not to rehearse the nature-nurture debate, which I explored in the chapter on chromosome 6, but to draw attention to the fact that even if the nurture assumption had proved true, it would not have reduced determinism one iota. As it is, by stressing the powerful influence that conformity to a peer group can have on personality, Rich Harris lays bare just how much more alarming social determinism is than genetic. It is brainwashing. Far from leaving room for free will, it rather diminishes it. A child who expresses her own (partly genetic) personality in defiance of her parents' or her siblings' pressures is at least obeying endogenous causality, not somebody else's. So there is no escape from determinism by appealing to socialisa­tion. Either effects have causes or they do not. If I am timid because of something that happened to me when I was young, that event is no less deterministic than a gene for timidity. The greater mistake is not to equate determinism with genes, but to mistake determinism for inevitability. Said the three authors of Not in our genes, Steven Rose, Leon Kamin and Richard Lewontin, 'To the biological determinists the old credo "You can't change human nature" is the alpha and omega of the human condition.' But this equation - determinism equals fatalism — is so well understood to be a fallacy that it is hard to find the straw men that the three critics indict.3 The reason the equation of determinism with fatalism is a fallacy is as follows. Suppose you are ill, but you reason that there is no point in calling the doctor because either you will recover, or you won't: in either case, a doctor is superfluous. But this overlooks the possibility that your recovery or lack thereof could be caused by your calling the doctor, or failure to do so. It follows that determin­ism implies nothing about what you can or cannot do. Determinism looks backwards to the causes of the present state, not forward to the consequences. Yet the myth persists that genetic determinism is a more implac­able kind of fate than social determinism. As James Watson has put it, 'We talk about gene therapy as if it can change someone's fate, but you can also change someone's fate if you pay off their credit card.' The whole point of genetic knowledge is to remedy genetic defects with (mostly non-genetic) interventions. Far from the dis­coveries of genetic mutations leading to fatalism, I have already cited many examples where they have led to redoubled efforts to ameliorate their effects. As I pointed out in the chapter on chromo­some 6, when dyslexia was belatedly recognised as a real, and possibly genetic, condition, the response of parents, teachers and govern­ments was not fatalistic. Nobody said that because it was a genetic condition dyslexia was therefore incurable and from now on children diagnosed with dyslexia would be allowed to remain illiterate. Quite the reverse happened: remedial education for dyslexics was developed, with impressive results. Likewise, as I argued in the chapter on chromosome 11, even psychotherapists have found gen­etic explanations of shyness helpful in curing it. By reassuring shy people that their shyness is innate and 'real', it somehow helps them overcome it. Nor does it make sense to argue that biological determinism threatens the case for political freedom. As Sam Brittan has argued, 'the opposite of freedom is coercion, not determinism.'4 We cherish political freedom because it allows us freedom of personal self­determination, not the other way around. Though we pay lip service to our love of free will, when the chips are down we cling to determinism to save us. In February 1994 an American named Stephen Mobley was convicted of the murder of a pizza-shop man­ager, John Collins, and sentenced to death. Appealing to have the sentence reduced to life imprisonment, his lawyers offered a genetic defence. Mobley came, they said, from a long pedigree of crooks and criminals. He probably killed Collins because his genes made him do it. 'He' was not responsible; he was a genetically determined automaton. It to be thought that he had none. So does every criminal who uses the defence of insanity or diminished responsibility. So does every jealous spouse who uses the defence of temporary insanity or justifiable rage after murdering an unfaithful partner. So does the unfaithful partner when justifying the infidelity. So does every tycoon who uses the excuse of Alzheimer's disease when accused of fraud against his shareholders. So indeed does a child in the playground who says that his friend made him do it. So does each one of us when we willingly go along with a subtle suggestion from the therapist that we should blame our parents for our present unhappiness. So does a politician who blames social conditions for the crime rate in an area. So does an economist when he asserts that consumers are utility maximisers. So does a biographer when he tries to explain how his subject's character was forged by formative experiences. So does everybody who consults a horoscope. In every case there is a willing, happy and grateful embracing of determinism. Far from loving free will, we seem to be a species that positively leaps to surrender it whenever we can.5 Full responsibility for one's actions is a necessary fiction without which the law would flounder, but it is a fiction all the same. To the extent that you act in character you are responsible for your actions; yet acting in character is merely expressing the many deter­minisms that caused your character. David Hume found himself impaled on this dilemma, subsequently named Hume's fork. Either our actions are determined, in which case we are not responsible for them, or they are random, in which case we are not responsible for them. In either case, common sense is outraged and society impossible to organise. Christianity has wrestled with these issues for two millennia and theologians of other stripes for much longer. God, almost by defin­ition, seems to deny free will or He would not be omnipotent. Yet Christianity in particular has striven to preserve a concept of free will because, without it, human beings cannot be held accountable for their actions. Without accountability, sin is a mockery and Hell a damnable injustice from a just God. The modern Christian consensus is that God has implanted free will in us, so that we have a choice of living virtuously or in sin. Several prominent evolutionary biologists have recently argued that religious belief is an expression of a universal human instinct — that there is in some sense a group of genes for believing in God or gods. (One neuroscientist even claims to have found a dedicated neural module in the temporal lobes of the brain that is bigger or more active in religious believers; hyper-religiosity is a feature of some types of temporal-lobe epilepsy.) A religious instinct may be no more than a by-product of an instinctive superstition to assume that all events, even thunderstorms, have wilful causes. Such a super­stition could have been useful in the Stone Age. When a boulder rolls down the hill and nearly crushes you, it is less dangerous to subscribe to the conspiracy theory that it was pushed by somebody than to assume it was an accident. Our very language is larded with intentionality. I wrote earlier that my genes built me and delegated responsibility to my brain. My genes did nothing of the sort. It all just happened. E. O. Wilson even argues, in his book Consilience,6 that morality is the codified expression of our instincts, and that what is right is indeed - despite the naturalistic fallacy — derived from what comes naturally. This leads to the paradoxical conclusion that belief in a god, being natural, is therefore correct. Yet Wilson himself was reared a devout Baptist and is now an agnostic, so he has rebelled against a deterministic instinct. Likewise, Steven Pinker, by remaining childless while subscribing to the theory of the selfish gene, has told his selfish genes to 'go jump in a lake'. So even determinists can escape determinism. We have a paradox. Unless our behaviour is random, then it is determined. If it is determined, then it is not free. And yet we feel, and demonstrably are, free. Charles Darwin described free will as a delusion caused by our inability to analyse our own motives. Modern Darwinists such as Robert Trivers have even argued that deceiving ourselves about such matters is itself an evolved adaptation. Pinker has called free will 'an idealisation of human beings that makes the ethics game playable'. The writer Rita Carter calls it an illusion hard-wired into the mind. The philosopher Tony Ingram calls free will something that we assume other people have — we seem to have an inbuilt bias to ascribe free will to everybody and everything about us, from recalcitrant outboard motors to recalcitrant children equipped with our genes. I would like to think that we can get a little closer to resolving the paradox than that. Recall that, when discussing chromosome 10, I described how the stress response consists of genes at the whim of the social environment, not vice versa. If genes can affect behaviour and behaviour can affect genes, then the causality is circu­lar. And in a system of circular feedbacks, hugely unpredictable results can follow from simple deterministic processes. This kind of notion goes under the name of chaos theory. Much as I hate to admit it, the physicists have got there first. Pierre-Simon de LaPlace, the great French mathematician of the eighteenth cen­tury, once mused that if, as a good Newtonian, he could know the positions and the motions of every atom in the universe, he could predict the future. Or rather, he suspected that he could not know the future, but he wondered why not. It is fashionable to say that the answer lies at the subatomic level, where we now know that there are quantum-mechanical events that are only statistically pre­dictable and the world is not made of Newtonian billiard balls. But that is not much help because Newtonian physics is actually a pretty good description of events at the scale at which we live and nobody seriously believes that we rely, for our free will, on the probabilistic scaffolding of Heisenberg's uncertainty principle. To put the reason bluntly: in deciding to write this chapter this afternoon, my brain did not play dice. To act randomly is not the same thing as to act freely — in fact, quite the reverse.8 Chaos theory provides a better answer to LaPlace. Unlike quantum physics, it does not rest on chance. Chaotic systems, as defined by mathematicians, are determined, not random. But the theory holds that even if you know all the determining factors in a system, you may not be able to predict the course it will take, because of the way different causes can interact with each other. Even simply deter­mined systems can behave chaotically. They do so partly because of reflexivity, whereby one action affects the starting conditions of the next action, so small effects become larger causes. The trajectory of the stock market index, the future of the weather and the 'fractal geometry' of a coastline are all chaotic systems: in each case, the broad outline or course of events is predictable, but the precise details are not. We know it will be colder in winter than summer, but we cannot tell whether it will snow next Christmas Day. Human behaviour shares these characteristics. Stress can alter the expression of genes, which can affect the response to stress and so on. Human behaviour is therefore unpredictable in the short term, but broadly predictable in the long term. Thus at any instant in the day, I can choose not to consume a meal. I am free not to eat. But over the course of the day it is almost a certainty that I will eat. The timing of my meal may depend on many things — my hunger (partly dictated by my genes), the weather (chaotically determined by myriad external factors), or somebody else's decision to ask me out to lunch (he being a deterministic being over whom I have no control). This interaction of genetic and external influences makes my behaviour unpredictable, but not undetermined. In the gap between those words lies freedom. We can never escape from determinism, but we can make a distinction between good determinisms and bad ones - free ones and unfree ones. Suppose that I am sitting in the laboratory of Shin Shimojo at the California Institute of Technology and he is at this very moment prodding with an electrode a part of my brain some­where close to the anterior cingulate sulcus. Since the control of 'voluntary' movement is in this general area, he might be responsible for me making a movement that would, to me, have all the appear­ance of volition. Asked why I had moved my arm, I would almost certainly reply with conviction that it was a voluntary decision. Professor Shimojo would know better (I hasten to add that this is still a thought experiment suggested to me by Shimojo, not a real one). It was not the fact that my movement was determined that contradicted my illusion of freedom; it was the fact that it was determined from outside by somebody else. The philosopher A. J. Ayer put it this way:9 If I suffered from a compulsive neurosis, so that I got up and walked across the room, whether I wanted to or not, or if I did so because somebody else compelled me, then I should not be acting freely. But if I do it now, I shall be acting freely, just because these conditions do not obtain; and the fact that my action may nevertheless have a cause is, from this point of view, irrelevant. A psychologist of twins, Lyndon Eaves, has made a similar point:10 Freedom is the ability to stand up and transcend the limitations of the environment. That capacity is something that natural selection has placed in us, because it's adaptive ... If you're going to be pushed around, would you rather be pushed around by your environment, which is not you, or by your genes, which in some sense is who you are. Freedom lies in expressing your own determinism, not somebody else's. It is not the determinism that makes a difference, but the ownership. If freedom is what we prefer, then it is preferable to be determined by forces that originate in ourselves and not in others. Part of our revulsion at cloning originates in the fear that what is uniquely ours could be shared by another. The single-minded obses­sion of the genes to do the determining in their own body is our strongest bulwark against loss of freedom to external causes. Do you begin to see why I facetiously flirted with the idea of a gene for free will? A gene for free will would not be such a paradox because it would locate the source of our behaviour inside us, where others cannot get at it. Of course, there is no single gene, but instead there is something infinitely more uplifting and magnificent: a whole human nature, flexibly preordained in our chromosomes and idio­syncratic to each of us. Everybody has a unique and different, endogenous nature. A self.