Science in the Soul

by Richard Dawkins

  Sexual selectionists of the Darwin school also have their modern champions. Taking their line through R. A. Fisher in the first half of the twentieth century, modern Darwinian sexual selectionists have developed mathematical models which show that, also paradoxically, sexual selection governed by arbitrary female whim can lead to a runaway process such that the tail – or other sexually selected character – moves dangerously far away from its utilitarian optimum. The key to this family of theories is what modern geneticists call ‘linkage disequilibrium’. When females choose, say, long-tailed males by whim, offspring of both sexes inherit their mother’s whim genes and also their father’s tail genes. It doesn’t matter how arbitrary is the whim, the joint selection on both sexes can lead (at least if you do the mathematical theory in a certain way) to runaway evolution of longer tails, and of female preference for longer tails. So tails can become ludicrously long.

  Cronin’s elegant historical analysis shows that the Darwin/Wallace opposition, in the field of sexual selection, persisted long after the deaths of the original protagonists, right through the twentieth century to today. It is especially pleasing – and might have amused the two men – that both the Darwinian and the Wallacean strands of sexual selection theory, more particularly in their modern forms, have a strong element of paradox. Both are capable of predicting surprising, even zany, sexual advertisements – which, indeed, we see in nature. The peacock’s fan is only the most famous example.

  I said the idea that occurred to Darwin and Wallace independently was one of the greatest, if not the greatest, ever to occur to a human mind. I want to end by giving this thought a universal spin. The opening words of my first book were:

  Intelligent life on a planet comes of age when it first works out the reason for its own existence. If superior creatures from space ever visit earth, the first question they will ask, in order to assess the level of our civilization, is: ‘Have they discovered evolution yet?’ Living organisms had existed on earth, without ever knowing why, for over three thousand million years before the truth finally dawned on one of them. His name was Charles Darwin.

  It would have been fairer, though less dramatic, to have said ‘two of them’ and to have coupled the name of Wallace with Darwin. But let me, in any case, pursue the universal perspective.

  I believe the Darwin/Wallace theory of evolution by natural selection is the explanation not just of life on this planet, but of life in general. If life is ever found elsewhere in the universe, I make the prediction that, however different it may be in detail, there will be one important principle which it shares with our own form of life. It will have evolved, under the guidance of a mechanism broadly equivalent to the Darwin/Wallace mechanism of natural selection.

  I am never quite sure how strongly to put this point.*3 The weak version, of which I am completely confident, is that no workable theory other than natural selection has ever been proposed. The strong form would be that no other workable theory ever could be proposed. Today, I think I’ll stick with the weak form. It still has startling implications.

  Natural selection not only explains everything we know about life. It does so with power, elegance and economy. It is a theory which has evident stature, a stature which really measures up to the magnitude of the problem which it sets out to solve.

  Darwin and Wallace may not have been the first to get an inkling of the idea. But they were the first to understand the full magnitude of the problem, and the corresponding magnitude of the solution which jointly, and independently, occurred to them. This is the measure of their stature as scientists. The mutual generosity with which they settled the question of priority is the measure of their stature as human beings.

  * * *

  *1 In 1858, Charles Darwin was startled to receive a manuscript from what was then the Federated Malay States, written by a little-known naturalist and collector, Alfred Russel Wallace. Wallace’s paper laid out, in every particular, the theory of evolution by natural selection, the theory that Darwin had first thought of twenty years earlier. For reasons that are disputed, Darwin had not published his theory, although he wrote it out very fully in 1844. Wallace’s letter threw Darwin into a tailspin of anxiety. He first thought he should cede priority to Wallace. However, his friends the geologist Charles Lyell and the botanist Joseph Hooker, two elder statesmen of British science, convinced him of a compromise. Wallace’s 1858 paper and two earlier papers by Darwin would be read out at the Linnean Society in London, thereby achieving joint credit. In 2001, the Linnean Society decided to mount a plaque, on the very spot, to commemorate the historic event. I was invited to perform the unveiling, and this is a slightly abridged version of the speech with which I did so. The occasion felt celebratory. It was a pleasure to meet several members of both the Darwin and Wallace families, and in some cases introduce them to each other for the first time.

  *2 I’m referring especially here to Alan Grafen’s clever rendering into mathematical terms of qualitative arguments such as those of Amotz Zahavi. My own attempt to explain these matters is in the second edition of The Selfish Gene, written in a spirit of penance for the unjustified ridicule with which I had treated Zahavi’s ideas in the first edition.

  *3 See the following essay in this collection, ‘Universal Darwinism’.

  Universal Darwinism*1

  IT IS WIDELY believed on statistical grounds that life has arisen many times all around the universe. However varied in detail alien forms of life may be, there will probably be certain principles that are fundamental to all life, everywhere. I suggest that prominent among these will be the principles of Darwinism. Darwin’s theory of evolution by natural selection is more than a local theory to account for the existence and form of life on Earth. It is probably the only theory that can adequately account for the phenomena that we associate with life.

  My concern is not with the details of other planets. I shall not speculate about alien biochemistries based on silicon chains, or alien neurophysiologies based on silicon chips. The universal perspective is my way of dramatizing the importance of Darwinism for our own biology here on Earth, and my examples will be mostly taken from Earthly biology. I do, however, also think that ‘exobiologists’ speculating about extraterrestrial life should make more use of evolutionary reasoning. Their writings have been rich in speculation about how extraterrestrial life might work, but poor in discussion about how it might evolve. This essay should, therefore, be seen firstly as an argument for the general importance of Darwin’s theory of natural selection; secondly as a preliminary contribution to a new discipline of ‘evolutionary exobiology’.

  What Ernst Mayr called the ‘growth of biological thought’ is largely the story of Darwinism’s triumph over alternative explanations of existence. The chief weapon of this triumph is usually portrayed as evidence. The thing that is said to be wrong with Lamarck’s theory is that its assumptions are factually wrong. In Mayr’s words: ‘Accepting his premises, Lamarck’s theory was as legitimate a theory of adaptation as that of Darwin. Unfortunately, these premises turned out to be invalid.’ But I think we can say something stronger: even accepting his premises, Lamarck’s theory is not as legitimate a theory of adaptation as that of Darwin because, unlike Darwin’s, it is in principle incapable of doing the job we ask of it – explaining the evolution of organized, adaptive complexity. I believe this is so for all theories that have ever been suggested for the mechanism of evolution except Darwinian natural selection, in which case Darwinism rests on a securer pedestal than that provided by facts alone.

  Now, I have made reference to theories of evolution ‘doing the job we ask of them’. Everything turns on the question of what that job is. The answer may be different for different people. Some biologists, for instance, get excited about ‘the species problem’, while I have never mustered much enthusiasm for it as a ‘mystery of mysteries’. For some, the main thing that any theory of evolution has to explain is the diversity of life – cladogenesis. Others may require of the
ir theory an explanation of the observed changes in the molecular constitution of the genome.

  I agree with John Maynard Smith that ‘the main task of any theory of evolution is to explain adaptive complexity, i.e. to explain the same set of facts which the eighteenth-century theologian William Paley used as evidence of a Creator’. I suppose people like me might be labelled neo-Paleyists, or perhaps ‘transformed Paleyists’. We concur with Paley that adaptive complexity demands a very special kind of explanation: either a Designer as Paley taught, or something such as natural selection that does the job of a designer.*2 Indeed, adaptive complexity is probably the best diagnostic of the presence of life itself.

  Adaptive complexity as a diagnostic character of life

  If you find something, anywhere in the universe, whose structure is complex and gives the strong appearance of having been designed for a purpose, then that something either is alive, or was once alive, or is an artefact created by something alive. It is fair to include fossils and artefacts since their discovery on any planet would certainly be taken as evidence for life there.

  Complexity is a statistical concept. A complex thing is a statistically improbable thing, something with a very low a priori likelihood of coming into being. The number of possible ways of arranging the 1027 atoms of a human body is obviously inconceivably large. Of these possible ways, only very few would be recognized as a human body. But this is not, by itself, the point. Any existing configuration of atoms is, a posteriori, unique; as ‘improbable’, with hindsight, as any other. The point is that, of all possible ways of arranging those 1027 atoms, only a tiny minority would constitute anything remotely resembling a machine that worked to keep itself in being, and to reproduce its kind. Living things are not just statistically improbable in the trivial sense of hindsight; their statistical improbability is limited by the a priori constraints of design. They are adaptively complex.

  The term ‘adaptationist’ has been coined as a pejorative name for one who assumes, in Richard Lewontin’s words, ‘without further proof that all aspects of the morphology, physiology and behavior of organisms are adaptive optimal solutions to problems’. I have responded to this characterization elsewhere; here, I shall be an adaptationist in the much weaker sense that I shall only be concerned with those aspects of the morphology, physiology and behaviour of organisms that are undisputedly adaptive solutions to problems. In the same way, a zoologist may specialize on vertebrates without denying the existence of invertebrates. I shall be preoccupied with undisputed adaptations because I have defined them as my working diagnostic characteristic of all life, anywhere in the universe, in the same way as the vertebrate zoologist might be preoccupied with backbones because backbones are the diagnostic character of all vertebrates. From time to time I shall need an example of an undisputed adaptation, and the time-honoured eye will serve the purpose as well as ever – as it did, indeed, for Darwin himself, and for Paley: ‘As far as the examination of the instrument goes, there is precisely the same proof that the eye was made for vision, as there is that the telescope was made for assisting it. They are made upon the same principles; both being adjusted to the laws by which the transmission and refraction of rays of light are regulated.’

  If a similar instrument were found upon another planet, some special explanation would be called for. Either there is a God, or, if we are going to explain the universe in terms of blind physical forces, those blind physical forces are going to have to be deployed in a very peculiar way. The same is not true of non-living objects, as William Paley himself conceded.

  A transparent pebble, polished by the sea, might act as a lens, focusing a real image. The fact that it is an efficient optical device is not particularly interesting because, unlike an eye or a telescope, it is too simple. We do not feel the need to invoke anything remotely resembling the concept of design. The eye and the telescope have many parts, all coadapted and working together to achieve the same functional end. The polished pebble has far fewer coadapted features: the coincidence of transparency, high refractive index and mechanical forces that polish the surface in a curved shape. The odds against such a threefold coincidence are not particularly great. No special explanation is called for.

  Compare how a statistician decides what P value*3 to accept as evidence for an effect in an experiment. It is a matter of judgement and dispute, almost of taste, exactly when a coincidence becomes too great to stomach. But, no matter whether you are a cautious statistician or a daring statistician, there are some complex adaptations whose ‘P value’, whose coincidence rating, is so impressive that nobody would hesitate to diagnose life (or an artefact designed by a living thing). My definition of living complexity is, in effect, ‘that complexity which is too great to have come about through coincidence’. For the purposes of this essay, the problem that any theory of evolution has to solve is how living adaptive complexity comes about.

  In his 1982 book The Growth of Biological Thought, Ernst Mayr helpfully lists what he sees as the six clearly distinct theories of evolution that have ever been proposed in the history of biology. I shall use this list to provide me with my main headings in this essay. For each of the six, instead of asking what the evidence is, for or against, I shall ask whether the theory is in principle capable of doing the job of explaining the existence of adaptive complexity. I shall take the six theories in order, and will conclude that only Theory 6, Darwinian selection, matches up to the task.

  Theory 1. Built-in capacity for, or drive towards, increasing perfection

  To the modern mind this is not really a theory at all, and I shall not bother to discuss it. It is obvously mystical, and does not explain anything that it does not assume to start with.

  Theory 2. Use and disuse plus inheritance of acquired characters

  This is the Lamarckian theory. It is convenient to discuss it in two parts.

  Use and disuse

  It is an observed fact that on this planet living bodies sometimes become better adapted as a result of use. Muscles that are exercised tend to grow bigger. Necks that reach eagerly towards the treetops may lengthen in all their parts. Conceivably, if on some planet such acquired improvements could be incorporated into the hereditary information, adaptive evolution could result. This is the theory often associated with Lamarck, although there was more to what Lamarck said. Francis Crick, in 1982, said: ‘As far as I know, no one has given general theoretical reasons why such a mechanism must be less efficient than natural selection.’ In this section and the next I shall give two general theoretical objections to Lamarckism of exactly the kind which, I like to think, Crick was looking for. First, the shortcomings of the principle of use and disuse.

  The problem is the crudity and imprecision of the adaptation that the principle of use and disuse is capable of providing. Consider the evolutionary improvements that must have occurred during the evolution of an organ such as an eye, and ask which of them could conceivably have come about through use and disuse. Does ‘use’ increase the transparency of a lens? No, photons do not wash it clean as they pour through it. The lens and other optical parts must have reduced, over evolutionary time, their spherical and chromatic aberration; could this come about through increased use? Surely not. Exercise might have strengthened the muscles of the iris, but it could not have built up the fine feedback control system which controls those muscles. The mere bombardment of a retina with coloured light cannot call colour-sensitive cones into existence, nor connect up their outputs so as to provide colour vision.

  Darwinian types of theory, of course, have no trouble in explaining all these improvements. Any improvement in visual accuracy could significantly affect survival. Any tiny reduction in spherical aberration may save a fast-flying bird from fatally misjudging the position of an obstacle. Any minute improvement in an eye’s resolution of acute coloured detail may crucially improve its detection of camouflaged prey. The genetic basis of any improvement, however slight, will come to predominate in the gene pool. The relatio
nship between selection and adaptation is a direct and close-coupled one. The Lamarckian theory, on the other hand, relies on a much cruder coupling: the rule that the more an animal uses a certain bit of itself, the bigger that bit ought to be. The rule occasionally might have some validity but not generally, and, as a sculptor of adaptation it is a blunt hatchet in comparison to the fine chisels of natural selection. This point is universal. It does not depend on detailed facts about life on this particular planet. The same goes for my misgivings about the inheritance of acquired characters.

  Inheritance of acquired characters

  The first problem here is that acquired characters are not always improvements. There is no reason why they should be, and indeed the majority of them are injuries. This is not just a fact about life on Earth. It has a universal rationale. If you have a complex and reasonably well-adapted system, the number of things you can do to it that will make it perform less well is vastly greater than the number of things you can do to it that will improve it. Lamarckian evolution will move in adaptive directions only if some mechanism – presumably selection – exists for distinguishing those acquired characters that are improvements from those that are not. Only the improvements should be imprinted into the germ line.