Science in the Soul

by Richard Dawkins

  A similar point arises about Marxism. You can espouse an academic theory of history which predicts the dictatorship of the proletariat. And you can follow a political creed which values the dictatorship of the proletariat as a good thing that you should work to encourage. Many Marxists as a matter of fact do both, and a disconcerting number, arguably including Marx himself, cannot tell the difference. But logically, the political belief in what is desirable does not follow from the academic theory of history. You could consistently be an academic Marxist believing that the forces of history drive inexorably towards a workers’ revolution, while at the same time voting High Tory and working as hard as possible to postpone the inevitable. Or you could be a passionate Marxist politically, who works all the harder for the revolution precisely because you doubt the Marxist theory of history and feel that the longed-for revolution needs all the help it can get.

  Similarly, evolution may or may not have the quality of progressiveness that Julian Huxley, as an academic biologist, supposed. But whether he was right or not about the biology, it clearly is not necessary that we should imitate this kind of progressiveness in drawing up our systems of values.

  The issue is even starker if we move from evolution itself, with its alleged progressive momentum, to Darwin’s mechanism of evolution, the survival of the fittest. T. H. Huxley, in his 1893 Romanes Lecture, ‘Evolution and Ethics’, was under no illusions, and he was right. If you must use Darwinism as a morality play, it is an awful warning. Nature really is red in tooth and claw. The weakest really do go to the wall, and natural selection really does favour selfish genes. The racing elegance of cheetahs and gazelles is bought at huge cost in blood and the suffering of countless antecedents on both sides. Ancient antelopes were butchered and carnivores starved, in the shaping of their streamlined modern counterparts. The product of natural selection, life in all its forms, is beautiful and rich. But the process is vicious, brutal and short-sighted.

  As an academic fact we are Darwinian creatures, our forms and our brains sculpted by natural selection, that indifferent, cruelly blind watchmaker. But this doesn’t mean we have to like it. On the contrary, a Darwinian society is not the sort of society in which any friend of mine would wish to live. ‘Darwinian’ is not a bad definition of precisely the sort of politics I would run a hundred miles not to be governed by, a sort of over-the-top Thatcherism gone native.

  I should be allowed a personal word here because I am tired of being identified with a vicious politics of ruthless competitiveness, accused of advancing selfishness as a way of life. Soon after Mrs Thatcher’s election victory of 1979, Professor Steven Rose wrote, in New Scientist, as follows:

  I am not implying that Saatchi and Saatchi engaged a team of sociobiologists to write the Thatcher scripts, nor even that certain Oxford and Sussex dons are beginning to rejoice at this practical expression of the simple truths of selfish genery they have been struggling to convey to us. The coincidence of fashionable theory with political events is messier than that. I do believe though, that when the history of the move to the right of the late 1970s comes to be written, from law and order to monetarism and to the (more contradictory) attack on statism, then the switch in scientific fashion, if only from group to kin selection models in evolutionary theory, will come to be seen as part of the tide which has rolled the Thatcherites and their concept of a fixed, 19th century competitive and xenophobic human nature into power.

  The ‘Sussex don’ referred to was John Maynard Smith, and he gave the apt reply in a letter to the next issue of New Scientist: What should we have done, fiddled the equations?

  Rose was a leader of the Marxist-inspired attack of the time on sociobiology. It is entirely typical that, just as these Marxists were incapable of separating their academic theory of history from their normative political beliefs, they assumed that we were incapable of separating our biology from our politics. They simply could not grasp that one might hold academic beliefs about the way evolution happens in nature, while simultaneously repudiating the desirability of translating those academic beliefs into politics. This led them to the untenable conclusion that, since genetic Darwinism when applied to humans had undesirable political connotations, it must not be allowed to be scientifically correct.*17

  They and many others make the same kind of mistake with respect to positive eugenics. The premiss is that to breed humans selectively for abilities such as running speed, musical talent or mathematical dexterity would be politically and morally indefensible. Therefore it isn’t (must not be) possible – ruled out by science. Well, anybody can see that that’s a non sequitur, and I’m sorry to have to tell you that positive eugenics is not ruled out by science. There is no reason to doubt that humans would respond to selective breeding just as readily as cows, dogs, cereal plants and chickens. I hope it isn’t necessary for me to say that this doesn’t mean I am in favour of it.

  There are those who will accept the feasibility of physical eugenics but dig their trench before mental eugenics. Maybe you could breed a race of Olympic swimming champions, they concede, but you will never breed for higher intelligence, either because there’s no agreed method of measuring intelligence, or because intelligence is not a single quantity varying in one dimension, or because intelligence doesn’t vary genetically, or some combination of these three points.

  If you seek refuge in any of these lines of thought, it is once again my unpleasant duty to disillusion you. It doesn’t matter if we can’t agree how to measure intelligence, we can breed for any of the disputed measures, or a combination of them. It might be hard agreeing a definition for docility in dogs, but this doesn’t stop us breeding for it. It doesn’t matter if intelligence is not a single variable, the same is probably true of milking prowess in cows and racing ability in horses. You can still breed for them, even while disputing how they should be measured, or whether they each constitute a single dimension of variation.

  As for the suggestion that intelligence, measured in any way or in any combination of ways, does not vary genetically, it more or less cannot be true, for the following reason whose logic requires only the premiss that we are more intelligent – by whatever definition you choose – than chimpanzees and all other apes. If we are more intelligent than the ape which lived six million years ago and was our common ancestor with chimpanzees, there has been an evolutionary trend in our ancestry towards increased intelligence. There has certainly been an evolutionary trend towards increased brain size: it is one of the more dramatic evolutionary trends in the vertebrate fossil record. Evolutionary trends cannot happen unless there is genetic variation in the characteristics concerned – in this case brain size and presumably intelligence. So, there was genetic variation in intelligence in our ancestors. It is just possible that there isn’t any longer, but such an exceptional circumstance would be bizarre. Even if the evidence from twin studies*18 did not support it – which it does – we could safely draw the conclusion, from evolutionary logic alone, that we have genetic variance in intelligence, intelligence being defined in terms of whatever separates us from our ape ancestors. Using the same definition, we could, if we wanted to, use artificial selective breeding to continue the same evolutionary trend.

  I would need little persuading that such a eugenic policy would be politically and morally wrong,*19 but we must be absolutely clear that such a value judgement is the right reason to refrain from it. Let us not allow our value judgements to push us over into the false scientific belief that human eugenics isn’t possible. Nature, fortunately or unfortunately, is indifferent to anything so parochial as human values.

  Later, Rose joined forces with Leon Kamin, one of America’s leading opponents of IQ-measuring, and with the distinguished Marxist geneticist Richard Lewontin, to write a book in which they repeated these and many other errors.*20 They also acknowledged that the sociobiologists wanted to be less fascist than our science, in their (mistaken) view, ought to make us, but they (equally mistakenly) tried to catch us in a contr
adiction with the mechanistic interpretation of mind which we – and presumably they – follow.

  Such a position is, or ought to be, completely in accord with the principles of sociobiology offered by Wilson*21 and Dawkins. However, to adopt it would involve them in the dilemma of first arguing the innateness of much human behavior that, being liberal men, they clearly find unattractive (spite, indoctrination, etc.)…To avoid this problem, Wilson and Dawkins invoke a free will that enables us to go against the dictates of our genes if we so wish.

  This, they complain, is a return to unabashed Cartesian dualism. You cannot, say Rose and his colleagues, believe that we are survival machines programmed by our genes, and at the same time urge rebellion against them.

  What’s the problem? Without going into the difficult philosophy of determinism and free will,*22 it is easy to observe that, as a matter of fact, we do go against the dictates of our genes. We rebel every time we use contraception when we’d be economically capable of rearing a child. We rebel when we give lectures, write books or compose sonatas instead of single-mindedly devoting our time and energy to disseminating our genes.

  This is easy stuff; there is no philosophical difficulty at all. Natural selection of selfish genes gave us big brains which were originally useful for survival in a purely utilitarian sense. Once those big brains, with their linguistic and other capacities, were in place, there is no contradiction at all in saying that they took off in wholly new ‘emergent’ directions, including directions opposed to the interests of selfish genes.

  There is nothing self-contradictory about emergent properties. Electronic computers, conceived as calculating machines, emerge as word processors, chess players, encyclopedias, telephone switchboards, even, I regret to say, electronic horoscopes. No fundamental contradictions are there to ring philosophical alarm bells. Nor in the statement that our brains have overtaken, even overreached, their Darwinian provenance. Just as we defy our selfish genes when we wantonly detach the enjoyment of sex from its Darwinian function, so we can sit down together and with language devise politics, ethics and values which are vigorously anti-Darwinian in their thrust. I shall return to this in my conclusion.

  One of Hitler’s perverted sciences was a garbled Darwinism and, of course, eugenics. But, uncomfortable though it is to admit it, Hitler’s views were not unusual in the first part of this century. I quote from a chapter on ‘The New Republic’, an allegedly Darwinian utopia, written in 1902:

  And how will the New Republic treat the inferior races? How will it deal with the black? how will it deal with the yellow man?…those swarms of black, and brown, and dirty-white, and yellow people, who do not come into the new needs of efficiency? Well, the world is a world, and not a charitable institution, and I take it they will have to go…And the ethical system of these men of the New Republic, the ethical system which will dominate the world state, will be shaped primarily to favour the procreation of what is fine and efficient and beautiful in humanity – beautiful and strong bodies, clear and powerful minds.

  The author of this is not Adolf Hitler but H. G. Wells,*23 who thought of himself as a socialist. It is stuff like this (and there’s lots more from the Social Darwinists) that has given Darwinism a bad name in the social sciences. And how! But, again, we must not attempt to use the facts of nature to derive our politics or our morality one way or the other. David Hume is to be preferred to either of the two Huxleys: moral directives cannot be derived from descriptive premisses, or, as it is more colloquially put, ‘You can’t get an “ought” from an “is” ’. Where then, on the evolutionary view, do our ‘oughts’ come from? Where do we get our values, moral and aesthetic, ethical and political? It is time to move on from the values of science to the science of values.

  The science of values

  Have we inherited our values from remote ancestors? The onus is on those who would deny it. The tree of life, Darwin’s tree, is a vast, bushy thicket of thirty million twigs.*24 We are one tiny twig, buried somewhere in the surface layers. Our twig sprouts from a small bough alongside our ape cousins, not far from the larger bough of our monkey cousins, within view of our more distant cousins, cousin kangaroo, cousin octopus, cousin staphylococcus. Nobody doubts that all the rest of the thirty million twigs inherit their attributes from their ancestors, and by any standards we humans owe to our ancestors much of what we are and what we look like. We have inherited from our forebears – with greater or less modification – our bones and eyes, our ears and thighs, even, it is hard to doubt, our lusts and our fears. A priori there seems no obvious reason why the same should not apply to our higher mental faculties, our arts and our morals, our sense of natural justice, our values. Can we exclude these manifestations of high humanity from what Darwin called the indelible stamp of our lowly origins? Or was Darwin right when he remarked more informally to himself in one of his notebooks, ‘He who understands baboon would do more towards metaphysics than Locke’? I shall make no attempt to review the literature, but the question of the Darwinian evolution of values and morals has been frequently and extensively discussed.

  Here’s the fundamental logic of Darwinism. Everybody has ancestors but not everybody has descendants. We have all inherited the genes for being an ancestor, at the expense of the genes for failing to be an ancestor. Ancestry is the ultimate Darwinian value. In a purely Darwinian world, all other values are subsidiary. Synonymously, gene survival is the ultimate Darwinian value. As a first expectation, all animals and plants can be expected to work ceaselessly for the long-term survival of the genes that ride inside them.

  The world is divided into those for whom the simple logic of this is as clear as daylight, and those who, no matter how many times it is explained to them, just don’t get it. Alfred Wallace wrote about the problem*25 in a letter to his co-discoverer of natural selection: ‘My dear Darwin – I have been so repeatedly struck by the utter inability of numbers of intelligent persons to see clearly, or at all, the self-acting and necessary effects of natural selection…’

  Those who don’t get it either assume that there must be some kind of personal agent in the background to do the choosing; or they wonder why individuals should value survival of their own genes, rather than, for instance, the survival of their species, or the survival of the ecosystem of which they are a part. After all, say this second group of people, if the species and the ecosystem don’t survive, nor will the individual, so it is in their interests to value the species and the ecosystem. Who decides, they wonder, that gene survival is the ultimate value?

  Nobody decides. It follows automatically from the fact that genes reside in the bodies that they build and are the only things (in the form of coded copies) that can persist from one generation of bodies to the next. This is the modern version of the point Wallace was making with his apt phrase ‘self-acting’. Individuals are not miraculously or cognitively inspired with values and goals that will guide them in the paths of gene survival. Only the past can have an influence, not the future. Animals behave as if striving for the future values of the selfish gene simply and solely because they bear, and are influenced by, genes that survived through ancestral generations in the past. Those ancestors that, in their own time, behaved as if they valued whatever was conducive to the future survival of their genes, have bequeathed those very genes to their descendants. So their descendants behave as if they, in their turn, value the future survival of their genes.

  It is an entirely unpremeditated, self-acting process which works so long as future conditions are tolerably similar to past. If they are not, it doesn’t, and the result is often extinction. Those who understand this understand Darwinism. The word Darwinism, by the way, was coined by the ever-generous Wallace. I shall continue my Darwinian analysis of values using bones as my example, because they are unlikely to ruffle human hackles and therefore distract.

  Bones are not perfect; they sometimes break. A wild animal that breaks its leg is unlikely to survive in the harsh, competitive world of n
ature. It will be especially vulnerable to predators, or unable to catch prey. So why doesn’t natural selection thicken bones so that they never break? We humans, by artificial selection, could breed a race of, say, dogs, whose leg bones were so stout that they never broke. Why doesn’t nature do the same? Because of costs, and this implies a system of values.

  Engineers and architects are never asked to build unbreakable structures, impregnable walls. Instead, they are given a monetary budget and asked to do the best they can, according to certain criteria, within that constraint. Or they may be told: the bridge must bear a weight of ten tons, and must withstand gales three times more forceful than the worst ever recorded in this gorge. Now design the most economical bridge you can that meets these specifications. Safety factors in engineering imply monetary valuation of human life. Designers of civilian airliners are more risk-averse than designers of military aircraft. All aircraft and ground control facilities could be safer if more money was spent. More redundancy could be built into control systems, the number of flying hours demanded of a pilot before he is allowed to carry live passengers could be increased. Baggage inspection could be more stringent and time-consuming.

  The reason we don’t take these steps to make life safer is largely one of cost. We are prepared to pay a lot of money, time and trouble for human safety, but not infinite amounts. Like it or not, we are forced to put monetary value on human life. In most people’s scale of values, human life rates higher than non-human animal life, but animal life does not have zero value. Notoriously, the evidence of newspaper coverage suggests that people value life belonging to their own race higher than human life generally. In wartime, both absolute and relative valuations of human life change dramatically. People who think it is somehow wicked to talk about this monetary valuation of human life – people who emotionally declare that a single human life has infinite value – are living in cloud-cuckoo-land.