Читать книгу Nature via Nurture: Genes, experience and what makes us human - Matt Ridley, Matt Ridley - Страница 11

When, as by a miracle, the lovely butterfly bursts from the chrysalis full-winged and perfect…it has, for the most part, nothing to learn, because its little life flows from its organization like melody from a music box.

Douglas Alexander Spalding, 1873¹

Like Charles Darwin, William James was a man of independent means. He inherited a private income from his father Henry, whose father William had amassed $10,000 a year from the Erie Canal. The one-legged Henry used his self-sufficiency to become an intellectual, and spent much of his life shuttling between New York, Geneva, London and Paris with his children in tow. He was articulate, religious and self-assured. His two youngest sons went off to fight in the Civil War, then failed in business and turned to drink or depression. His two eldest sons, William and Henry, were trained almost from birth to be intellectuals. The result was (in Rebecca West’s phrase) that ‘one of them grew up to write fiction as though it were philosophy and the other to write philosophy as though it were fiction’.²

Both brothers were influenced by Darwin. Henry’s novel The Portrait of a Lady was written in thrall to Darwin’s idea of female choice as a force in evolution.³ William’s Principles of Psychology, much of which was first published as a series of articles in the 1880s, contained a manifesto for nativism – the idea that the mind cannot learn unless it has the rudiments of innate knowledge – going against the prevailing fashion for empiricism, the theory that behaviour is shaped by experience. William James believed that human beings were equipped with innate tendencies that were not derived from experience but from the Darwinian process of natural selection. ‘He denies experience!’ wrote James, quoting an imaginary reader. ‘Denies science; believes the mind created by miracle; is a regular old partisan of innate ideas! That is enough! We’ll listen to such antediluvian twaddle no more.’

William James asserted that human beings have more instincts than other animals, not fewer. ‘Man possesses all the impulses that [lower creatures] have, and a great many more besides…It will be observed that no other mammal, not even the monkey, shows so large an array’. He argued that it was false to oppose instinct to reason:

Reason, per se, can inhibit no impulses; the only thing that can neutralize an impulse is an impulse the other way. Reason may, however, make an inference which will excite the imagination so as to set loose the impulse the other way; and thus, though the animal richest in reason might also be the animal richest in instinctive impulses, too, he would never seem the fatal automaton which a merely instinctive animal would be.⁴

This is an extraordinary passage, not least because its impact on early twenty-first-century thought can be said to be almost nil. Very few people, on the side of either nature or nurture, took up such an extreme nativist position in the century to come, and almost everybody assumed for the following hundred years that reason was indeed the opposite of instinct. Yet James was no fringe lunatic. His work has influenced generations of scholars on consciousness, sensation, space, time, memory, will, emotion, thought, knowledge, reality, self, morality and religion – to name just the chapter headings of a modern book about his work. So why does this same book of 628 pages not even have the words ‘instinct’, ‘impulse’ or ‘innate’ in its index?⁵ Why, for more than a century, has it been considered little short of indecent even to use the word ‘instinct’ in the context of human behaviour?

James’s ideas were indeed immensely influential at first. His follower, William McDougall, founded a whole school of instinctivists, who became adept at spotting new human instincts for every circumstance. Too adept: speculation outstripped experiment and before long a counter-reformation was inevitable. In the 1920s the very empiricist ideas attacked by James, embodied in the notion of the blank slate, swept back to power not just in psychology (with John B. Watson and B.F. Skinner), but in anthropology (Franz Boas), psychiatry (Freud) and sociology (Durkheim). Nativism was almost totally eclipsed until 1958, when Noam Chomsky once again pinned its charter to the door of science. In a famous review of a book on language by Skinner, Chomsky argued that it was impossible for a child to learn the rules of language from examples: the child must have innate rules to which the vocabulary of the language was fitted. Even then, the blank slate dominated human sciences for many years. It was not until a century after his book was published that William James’s idea of uniquely human instincts was at last taken seriously again in a new manifesto of nativism, written by John Tooby and Leda Cosmides (see chapter 9).

More of that later. First, a digression on teleology. It was Darwin’s genius to turn the old theological argument from design on its head. Until then, the obvious fact that parts of organisms appear to be engineered for a purpose – the heart for pumping, the stomach for digesting, the hand for grasping – seemed logically to imply a designer, just as a steam engine implied the existence of an engineer. Darwin saw how the entirely backward-looking process of natural selection could none the less produce purposeful design – what Richard Dawkins called the blind watchmaker.⁶ Though in theory it makes teleological nonsense to talk of a stomach having its own purpose, since the stomach has no mind, in practice it makes perfect sense so long as you engage the grammatical equivalent of four-wheel drive, the passive voice: stomachs have been selected to appear as if equipped with purposeful design. Since I have an aversion to the passive voice, I intend to avoid this problem throughout this book by pretending that there is indeed a teleological engineer thinking ahead and planning purposefully. The philosopher Daniel Dennett calls such an artefact a ‘skyhook’,⁷ since it is the rough equivalent of a civil engineer hanging his scaffolding from the sky, but for the sake of simplicity I shall call my skyhook the Genome Organising Device, or GOD for short. This may keep religious readers happy, and allows me to use the active voice. So the question is: how does the GOD build a brain that can express an instinct?

Back to William James. To support his assertion that human beings have more instincts than other animals, James systematically enumerated the human instincts. He began with the actions of babies: sucking, clasping, crying, sitting up, standing, walking and climbing were all, he suggested, expressions of impulse, not imitations or associations. So, as the child grew, were emulation, anger and sympathy. So was a fear of strangers, loud noises, heights, the dark, reptiles. (The ordinary cock-sure evolutionist ought to have no difficulty in explaining these terrors,’ wrote James, neatly anticipating the argument of what is now called evolutionary psychology, ‘as relapses into the consciousness of the cave-men, a consciousness usually overlaid in us by experiences of more recent date.’) He moved on to acquisitiveness, noting the tendency of boys to collect things. He noticed the very different play preferences of boys and girls. Parental love, he suggested, was at least initially stronger in women than in men. He tripped quickly through sociability, shyness, secretiveness, cleanliness, modesty and shame. ‘Jealousy is unquestionably instinctive,’ he remarked.

The strongest of the instincts, he believed, was love. ‘Of all propensities, the sexual impulses bear on their face the most obvious signs of being instinctive, in the sense of blind, automatic and untaught.’⁸ But, he insisted, just because sexual attraction was instinctive did not mean it was irresistible. Other instincts, like shyness, prevent us acting upon every sexual attraction.

So let me take James at his word, provisionally at least, and examine the idea of the love instinct in a little more depth. If he is right, there must be some heritable factor, which gives rise to a physical or chemical change in our brains when we fall in love, that change causing, rather than caused by, the emotion of falling in love. Such as this, from the scientist Tom Insel:

A working hypothesis is that oxytocin released during mating activates those limbic sites rich in oxytocin receptors to confer some lasting and selective reinforcement value on the mate.⁹

Or, to put it more poetically, you fall in love.

What is this oxytocin and why does Insel make such an extravagant claim for it? The story starts in an almost ridiculously unromantic process: urination. Some 400 million years ago, when the ancestors of our species first left the water, they were equipped with a tidy little hormone called vasotocin, a miniature protein made out of a chain of just nine amino acids formed into a ring. Its job was to regulate salt and water balance in the body, and it performed this job by rushing about switching on cells in the kidney or other organs. Fish still use two different versions of vasotocin for this purpose today, and so do frogs. In the descendants of reptiles – and that includes human beings – there are two slightly different copies of the relevant gene lying next to each other, facing different ways (in human beings on chromosome 20). The result today is that all mammals have two such hormones, called vasopressin and oxytocin, that differ at two of the links in the chain.

They still do their old job. Vasopressin tells the kidney to conserve water; oxytocin tells it to excrete salt. But, like vasotocin in modern fish, they also have a role in the regulation of reproductive physiology. Oxytocin stimulates the contraction of muscles in the womb during birth; it also causes milk to be expelled from the ducts in the breast. The GOD is an economiser: having invented a switch for one purpose, he readapts it for other purposes, by expressing the oxytocin receptor in a different organ. But a much greater surprise came in the early 1980s, when scientists suddenly realised that vasopressin and oxytocin had a job to do inside the brain as well as being secreted from the pituitary gland into the bloodstream.

So they tried injecting oxytocin and vasopressin into the brains of rats to see what effect they had. Bizarrely, a male rat injected with intracerebral oxytocin immediately begins yawning and simultaneously gets an erection.¹⁰ So long as the dose is low, the rat also becomes more highly sexed: it ejaculates sooner and more frequently. In female rats, intracerebral oxytocin induces the animal to adopt a mating posture. In human beings, meanwhile, masturbation increases oxytocin levels in both sexes. All in all, oxytocin and vasopressin in the brain seem to be connected to mating behaviour.

Now all of this sounds rather unromantic: urine, masturbation, breastfeeding – hardly the essence of love. Be patient. In the late 1980s, Tom Insel was working on the effect of oxytocin on maternal behaviour in rats. Brain oxytocin seemed to help the mother rat form a bond with its young and Insel identified the parts of the rat brain that were sensitive to the hormone. He switched his attention to the pair bond, wondering if there were parallels between a female’s bond to her young and to her mate. At this point he met Sue Carter, who had begun to study prairie voles in the laboratory. She told him how the prairie vole is a rarity among mice for its faithful marriages. Prairie voles live in couples and both father and mother care for the young for many weeks. Montane voles, on the other hand, are more typical of mammals: the female mates with a passing polygamist, separates quickly from him, bears young alone and abandons them after a few weeks to fend for themselves. Even in the laboratory, this difference is clear: mated prairie voles stare into each other’s eyes and bathe the babies; mated montane voles treat their spouses like strangers.

Insel examined the brains of the two species. He found no difference in the expression of the two hormones themselves, but a big difference in the distribution of molecular receptors for them – the molecules that fire up neurons in response to the hormones. The monogamous prairie voles had far more oxytocin receptors in several parts of the brain than the polygamous montane voles. Moreover, by injecting oxytocin or vasopressin into the brains of prairie voles, Insel and his colleagues could elicit all the characteristic symptoms of monogamy, such as a strong preference for one partner and aggression towards other voles. The same injections had little effect on montane voles, and the injection of chemicals that block the oxytocin receptors prevented the monogamous behaviour. The conclusion was clear: prairie voles are monogamous because they respond more to oxytocin and vasopressin.¹¹

In a virtuoso display of scientific ingenuity, Insel’s team has gone on to dissect this effect in convincing detail. They knock the oxytocin gene out of a mouse before birth. This leads to social amnesia: the mice can remember things, but they have no memory of mice they have already met and will not recognise them. Lacking oxytocin in its brain, a mouse cannot recognise a mouse it has just met ten minutes before – unless that mouse was ‘badged’ with a non-social cue such as a distinctive lemon- or almond-scented smell (Insel compares this to an absent-minded professor at a conference who recognises friends by their name tags, not their faces).¹² Then by injecting the hormone into just one part of the animal’s brain in later life – the medial amygdala – the scientists can restore social memory to the mouse completely.

In another experiment, using a specially adapted virus, they turn up the expression of the vasopressin receptor gene in the ventral pallidum, a part of a vole’s brain important for reward. Pause here to roll that idea around your mind a few times to appreciate just what science can do these days: they use viruses to turn up the volumes of genes in one part of the brain of a rodent. Even ten years ago such an experiment was unimaginable. The result of turning up the gene’s expression is to ‘facilitate partner preference formation’, which is geekspeak for ‘make them fall in love’. They conclude that for a male vole to pair-bond, it must have both vasopressin and vasopressin receptors in its ventral pallidum. Since mating causes a release of oxytocin and vasopressin, the prairie vole will pair-bond with whatever animal it has just mated; the oxytocin helps in memory, the vasopressin in reward. The montane vole, by contrast, will not react in the same way, because it lacks receptors in that area. Female montane voles express these receptors only after giving birth, so they can be nice to their babies, briefly.

So far I have talked of oxytocin and vasopressin as if they were the same thing, and they are so similar that they probably stimulate each other’s receptors somewhat. But it appears that to the extent that they do differ, oxytocin makes female voles choose a partner; vasopressin makes males choose a partner. The male prairie vole becomes aggressive towards all voles except its mate when vasopressin is injected into his brain. Attacking other voles is a (rather male) way of expressing his love.¹³

All this is astonishing enough, but perhaps the most exciting result to emerge from Insel’s lab concerns the genes for the receptors. Remember that the difference between the prairie vole and the montane vole lies not in the expression of the hormone, but in the pattern of expression of the hormone’s receptors. These receptors are themselves the products of genes. The receptor genes are essentially identical in the two species, but the promoter regions, upstream of the genes, are very different. Now recall the lesson of chapter 1: that the difference between closely related species lies not in the text of genes themselves, but in their promoters. In the prairie vole, there is an extra chunk of DNA text, on average about 460 letters long, in the middle of the promoter. So Insel’s lab made a transgenic mouse with this expanded promoter and it grew up with a brain like a prairie vole, expressing vasopressin receptors in all the same places, though it did not form a pair bond.¹⁴ Steven Phelps then went out and caught 43 wild prairie voles in Indiana and sequenced their promoters: some had longer insertions than others. They varied from 350 to 550 letters in length. Are the long ones more faithful husbands than the short ones? Not yet known.¹⁵

The conclusion to which Insel’s work is leading is devastating in its simplicity. The ability of a rodent to form a long-term attachment to its sexual partner may depend on the length of a piece of DNA text in the promoter switch at the front of a certain receptor gene. That in turn decides precisely which parts of the brain will express the gene. Of course, like all good science, this discovery raises more questions than it settles. Why should feeding oxytocin receptors in that part of the brain make the mouse feel well-disposed to its partner? It is possible that the receptors induce a state a bit like addiction, and in this respect it is noticeable that they seem to link with the D2 dopamine receptors, which are closely involved in various kinds of drug addiction.¹⁶ On the other hand, without oxytocin, mice cannot form social memories, so perhaps they simply keep forgetting what their spouse looks like.

Mice are not men. You know by now that I am about to start extrapolating anthropomorphically from pair-bonding in voles to love in people, and you probably do not like my drift. It sounds reductionist and simplistic. Romantic love, you say, is a cultural phenomenon, overlaid with centuries of tradition and teaching. It was invented at the court of Eleanor of Aquitaine, or some such place, by a bunch of oversexed poets called troubadours; before that there was just sex.

Even though in 1992 William Jankowiak surveyed 168 different ethnographic cultures and found none that did not recognise romantic love, you may be right.¹⁷ I certainly cannot prove to you yet that people fall in love when their oxytocin and vasopressin receptors get tingled in the right places in their brains. Yet. And there are cautionary hints about the dangers of extrapolating from one species to another: sheep seem to need oxytocin to form maternal attachment to their young; mice apparently do not.¹⁸ Human brains are undoubtedly more complicated than mouse brains.

But I can draw your attention to some curious coincidences. A mouse shares much of its genetic code with a human being. Oxytocin and vasopressin are identical in the two species and are produced in the equivalent parts of the brain. Sex causes them to be produced in the brain in both human beings and rodents. Receptors for the two hormones are virtually identical and are expressed in equivalent parts of the brain. Like those of the prairie vole, the human receptor genes (on chromosome 3) have a – smaller – insertion in their promoter regions. Like the prairie voles of Indiana, the lengths of those promoter insertions vary from individual to individual: in the first 150 people examined, Insel found 17 different promoter lengths. And when a person who says she (or he) is in love contemplates a picture of her loved one while sitting in a brain scanner, certain parts of her brain light up that do not light up when she looks at a picture of a mere acquaintance. Those brain parts overlap with the ones stimulated by cocaine.¹⁹ All this could be a complete coincidence, and human love may be entirely different from rodent pair-bonding, but given how conservative the GOD is and how much continuity there is between human beings and other animals, you would be unwise to bet on it.²⁰

Shakespeare was ahead of us, as usual. In A Midsummer Night’s Dream, Oberon tells Puck how Cupid’s arrow fell upon a white flower (the pansy), turning it purple, and that now the juice of this flower

…on sleeping eyelids laid

Will make or man or woman madly dote

Upon the next live creature that it sees.

Puck duly fetches a pansy and Oberon wreaks havoc with the lives of those sleeping in the forest, causing Lysander to fall in love with Helena, whom he has previously scorned, and causing Titania to fall in love with Bottom the weaver wearing the head of an ass.

Who would now wager against me that I could not soon do something like this to a modern Titania? Admittedly, a drop on the eyelids would not suffice. I would have to give her a general anaesthetic while I cannulated her medial amygdala and injected oxytocin into it. I doubt even then that I could make her love a donkey. But I might stand a fair chance of making her feel attracted to the first man she sees upon waking. Would you bet against me? (I hasten to add that ethics committees will – should – prevent anybody taking up my challenge.)

I am assuming that, unlike most mammals, human beings are basically monogamous, like prairie voles, and not promiscuous, like montane voles. I base this assumption on the testicle-size argument enunciated in chapter 1; on the ample evidence from ethnography that, though most human societies allow polygamy, most human societies are still dominated by monogamous relationships; and on the fact that human beings usually practise some paternal care – a characteristic feature of the few mammal species that live as social monogamists.²¹ Furthermore, as we have liberated human life from economic and cultural straitjackets, such as arranged marriage, we have found monogamy growing more dominant, not less. In 1998 the most powerful man in the world, far from treating himself to a gigantic harem, got into trouble for having an affair with one intern. The evidence for long-term, exclusive (but sometimes cheated-on) pair bonds as the commonest pattern in human relationships is all around you.

Chimpanzees are different. Long-term pair bonds are unknown, and I predict that they have fewer oxytocin receptors in the relevant parts of their brains than human beings, probably as a result of having shorter gene promoters. The oxytocin story lends at least tentative support to William James’s notion that love is an instinct, evolved by natural selection, and is part of our mammal heritage, just like four limbs and ten fingers. Blindly, automatically and untaught, we bond with whoever is standing nearest when the oxytocin receptors in the medial amygdala get tingled. One sure way to tingle them is to have sex, although presumably chaste attraction can also do the trick. Is this why breaking up is hard to do?

Having oxytocin receptors does not make it inevitable that somebody will fall in love during his life, nor predictable when it will happen, or with whom. As Niko Tinbergen, the great Dutch ethologist, demonstrated in his studies of instincts, the expression of a fixed, innate instinct must often be triggered by an external stimulus. One of Tinbergen’s favourite species was the stickleback, a tiny fish. Male sticklebacks go red on the belly in the breeding season, when they defend small territories in which they build nests, which attract females. Tinbergen made little models of fish and caused them to ‘invade’ the territory of a male fish. A model of a female elicited the courtship dance of the male, even if the model was astonishingly crude; so long as it had a ‘pregnant’ belly, it excited the male. But if the model had a red belly, it would trigger an attack. It could be just an oval blob with a crudely drawn eye but no fins or tail: still it was attacked just as vigorously as if it were a real male rival – so long as it was red. One of the legends of Leiden, where Tinbergen first worked, is that he noticed his sticklebacks would threaten the red post-office vans that drove past the window.

Tinbergen went on to demonstrate the power of these ‘innate releasing mechanisms’ to provoke an instinct in other species, notably the herring gull. Herring gulls have yellow beaks with a bright red spot near the tip. The chicks peck at this spot when begging for food. By presenting newborn chicks with a series of models, Tinbergen demonstrated that the spot was a powerful releaser for the begging action, and the redder it was the better. The colour of the beak or the head of the bird mattered not at all. So long as it had a contrasting spot near the tip of the bill, preferably in red, it would elicit pecking. In modern jargon, scientists would say that the chick’s instinct, and the adult’s beak spot had ‘co-evolved’. An instinct is designed to be triggered by an external object or event. Nature plus nurture.²²

The significance of Tinbergen’s experiments was to reveal just how complex instincts could be, and yet how simply triggered. The digger wasp he studied would dig a burrow, go and catch a caterpillar, paralyse it with a sting, bring it back to the burrow and deposit it with an egg on top, so that the baby wasp could feed on the caterpillar while growing. All of this complex behaviour, including the ability to navigate back to the burrow, was achieved with almost no learning, let alone parental teaching. A digger wasp never meets its parents. A cuckoo migrates to Africa and back, sings its song and mates with one of its own species without as a chick ever seeing either a parent or a sibling.

The notion that animal behaviour is in the genes once troubled biologists as much as it now troubles social scientists. Max Delbruck, pioneering molecular biologist, refused to believe that his colleague at Caltech Seymour Benzer had found a behavioural mutant fly. Behaviour, he insisted was too complex to reduce to single genes. Yet the idea of behaviour genes has long been accepted by the amateur breeders of domestic animals. The Chinese started breeding mice of different colours in the seventeenth century or earlier and they produced a mouse called the waltzing mouse, famous for its dance-like gait caused by an inherited defect in the inner ear. Mouse breeding then caught on in Japan in the nineteenth century and thence spread to Europe and America. Some time before the year 1900 a retired schoolteacher in Granby, Massachusetts, by the name of Abbie Lathrop, took up the ‘mouse fancying’ hobby. Soon she was breeding different strains of mice herself in a small barn adjoining her property and selling them to pet shops. She was especially fond of what were by then known as Japanese waltzing mice, and she developed several new strains. She also noticed that some strains got cancer more often than others; picked up by Yale University, this hint became the basis of early studies of cancer.

But it was Lathrop’s link to Harvard that uncovered the link between genes and behaviour. William Castle of Harvard bought some of her mice and started a mouse laboratory. Under Castle’s student Clarence Little, the main mouse laboratory moved to Bar Harbor, Maine, where it still is – a giant factory of inbred mouse strains used in research. Very early on, the scientists began to realise that different strains of mice behaved in different ways, too. Benson Ginsburg, for instance, found out the hard way. He noticed that when he picked up a mouse of the ‘guinea-pig’ strain (named for its coat colour), he often got bitten. He was soon able to breed a new strain that had the coat colour but not the aggressive streak: proof enough that aggression was somewhere in the genes. His colleague Paul Scott also developed aggressive strains of mice, but bizarrely, Ginsburg’s most aggressive strain was Scott’s most pacific. The explanation was that Scott and Ginsburg had handled the mice differently as babies. For some strains, handling did not matter. But for one strain in particular, C57-Black-6, early handling increased the aggressiveness of the mouse. Here was the first hint that a gene must interact with an environment if it is to have its effect. Or, as Ginsburg put it, the road from the ‘encoded genotype’ the mouse inherits to the ‘effective genotype’ it expresses passes through the process of social development.²³

Ginsburg and Scott both later went on to work with dogs, Scott proving by crossing experiments between cocker spaniels and African basenjis that play-fighting in puppies is controlled by two genes that regulate the threshold for aggression.²⁴ But it did not need science to prove the inheritance of behaviour in dogs: that was old news to dog-breeders. The point of dogs is that they come in different behavioural types: retrievers, pointers, setters, shepherds, terriers, poodles, bulldogs, wolfhounds – their very names denote the fact that they have instincts bred into them. And those instincts are innate. A retriever cannot be trained to guard livestock and a guard dog cannot be trained to herd sheep. It’s been tried. In the process of domestication, dogs have kept incomplete or exaggerated elements of wolf behaviour development. A wolf will stalk, chase, pounce, grab, kill, dissect, and carry food, and a wolf pup will practise each of these activities in turn as it grows up. Dogs are wolf pups frozen in the practising stage. Collies and pointers are stuck in the stalking stage; retrievers are stuck with carrying and pit bulls with biting: each is a frozen mixture of different wolf-pup themes. Is it in their genes? You bet: ‘Breed-specific behaviours are irrefutable,’ says dog chronicler Stephen Budiansky firmly.²⁵

Or go ask the cattle-breeders. I have in front of me a catalogue of dairy bulls designed to entice me into ordering some semen by mail. In enormous detail it describes the quality and shape of the bull’s udder and teats, its milk-producing ability, its milking speed and even its temperament. But surely, you point out, bulls don’t have udders? On every page there is a picture of a cow, not a bull. What the catalogue is referring to is not the bull himself but his daughters. ‘Zidane, the Italian No 1,’ it boasts, ‘improves frame traits and fixes on tremendous rumps with ideal slope. He is particularly impressive in his feet and leg composites with excellent set and terrific depth of heel. He leaves faultless udders, which are snugly attached with deep clefts.’ The characters are all female, but the attribution is to the sire. Perhaps I would prefer to buy a straw of semen from Terminator, whose daughters have ‘great teat placement’, or Igniter, a bull that is a ‘milking speed specialist’ whose daughters ‘display great dairy character’. I might wish to avoid Moet Flirt Freeman, because although his daughters have ‘tremendous width across the chest’ and give more milk than their mothers did, the small print admits they are also slightly ‘below average’ in temperament – which probably means that they tend to kick out when being milked. They are also slow milkers.²⁶

The point is that cattle-breeders have no qualms about attributing behaviour to genes, just as they attribute anatomy to genes. Minute differences in the behaviour of cows they confidently ascribe to the semen that arrived through the mail. Human beings are not cows. Admitting instinct in cows does not prove that human beings are also ruled by instinct, of course. But it demolishes the assumption that because behaviour is complex or subtle, it cannot be instinctive. Such a comforting illusion is still rife within the social sciences; yet no zoologist who has studied animal behaviour could believe that complex behaviour cannot be innate.