Among these was the Lesser Yellow
Underwing
moth, Triphaena comes.
Richard-Dawkins-Unweaving-the-Rainbow
The second meaning is the theory that the phyla actually branched off from each other during the Cambrian, even during as little as 10 million years within the Cambrian.
This second idea, which I shall call the branch point explosion hypothesis, is controversial.
It is compatible - just - with what I am calling the standard neo-Darwinian model of species divergence.
We've already agreed that, as we trace any pair of modem phyla back in time, we eventually converge upon a common ancestor.
My hunch is that, for different pairs of phyla, we'll hit the common ancestor in different geological eras: say, the common ancestor of vertebrates and molluscs at 8oo million years ago, the common ancestor of vertebrates and echinoderms at 600 million years, and so on.
But I could be wrong, and we can easily accommodate the branch point explosion hypothesis by saying that, for some reason (which is interesting enough to need investigating), most of our backward tracings happen to hit their respective common ancestors during the same relatively short geological period, say, between 540 million and 530 million years ago.
This would have to mean that, at least near the beginning of that 10-million-year period, the ancestors of the modem phyla were nowhere near as different from each other as they are today.
They were, after all, diverging from common ancestors at the time and were originally members of the same species.
The extreme Gouldian view - certainly the view inspired by his rhetoric, though it is hard to tell from his own words whether he literally holds it himself - is radically different from and utterly incompatible with the standard neo-Darwinian model. It also, as I shall show, has implications which, once they are spelled out, anybody can see are absurd. It is very clearly expressed - betrayed might be a better word - in asides in Stuart Kauffman's At Home in the Universe (1995):
One might imagine that the first multicellular creatures would all he very similar, only later diversifying, from the bottom up, into different genera, families, orders, classes, and so on. That, indeed, would be the expectation of the strictest conventional Darwinist. Darwin, profoundly influenced by the emerging view of geologic gradualism, proposed that all evolution occurred by the very gradual accumulation of useful variations. Thus the earliest multicellular creatures themselves ought to have diverged gradually from one another.
So far, this is a fine summary of the orthodox neo-Darwinian view. Now, in a bizarre passage, Kauffman goes on:
But this appears to be false. One of the wonderful and puzzling features of the Cambrian explosion is that the chart was filled in from the top
down. Nature suddenly sprang forth with many wildly different body plans - the phyla - elaborating on these basic designs to form the classes, orders, families, and genera . . . In his book about the Cambrian explosion, Wonderful Life: The Burgess Shale and the Nature of History, Stephen Jay Gould remarks on this top-down quality of the Cambrian with wonder.
As well he might! You only have to think for one moment about what 'top down' filling in would have to mean for the animals on the ground and you immediately see how preposterous it is, 'Body plans' like the mollusc body plan, or the echinoderm body plan, are not ideal essences hanging in the sky, waiting, like designer dresses, to be adopted by real animals. Real animals is all there ever was: living, breathing, walking, eating, excreting, fighting, copulating real animals, who had to survive and who can't have been dramatically different from their real parents and grandparents. For a new body plan - a new phylum - to spring into existence, what actually has to happen on the ground is that a child is born which suddenly, out of the blue, is as different from its parents as a snail is from an earthworm. No zoologist who thinks through the implications, not even the most ardent saltationist, has ever supported any such notion. Ardent saltationists have been content to postulate the sudden bursting into existence of new species, and even that relatively modest idea has been highly controversial. When you spell out the Gouldian rhetoric into real-life practicalities, it stands revealed as the purest of bad poetic science.
Kauffman is even more explicit in a later chapter. In discussing some of his ingenious mathematical models of evolution on 'rugged fitness landscapes', Kauffman notes a pattern that he thinks sounds a lot like the Cambrian explosion. Early on in the branching process, we find a variety of long-jump mutations that differ from the stem and from one another quite dramatically. These species have sufficient morphological differences to be categorized as founders of distinct phyla. These founders also branch, but do so via slightly closer long-jump variants, yielding branches from, each founder of a phylum to dissimilar daughter species, the founders of classes. As the process continues, fitter variants are found in progressively more nearby neighborhoods, so founders of orders, families, and genera emerge in succession.
Kauffman's earlier, more technical book, The Origins of Order (1993), says something similar about life in the Cambrian:
Not only did a very large number of novel body forms originate rapidly, but the Cambrian explosion exhibited another novelty: Species which founded taxa appear to have built up the higher taxa from the top down.
That is, exemplars of major phyla were present first, followed by progressive filling in at class, order, and lower taxonomic levels. . .
Now, one way of reading this is inoffensive to the point of obviousness. On our 'converging backwards' model it would have to be true that species splittings that are eventually going to become phylum divides would normally precede those that are destined to become divides between orders and lower taxonomic levels. But Kauffman clearly doesn't think he is saying something ordinary and obvious. This is apparent from his statement that 'the Cambrian explosion exhibited another novelty', and from his phrase long-jump mutations'. He thinks he is attributing to the Cambrian something revolutionary. He really does seem genuinely to intend the alternative reading, in which 'long-jump mutations' give rise, on the instant, to brand new phyla.
I hasten to emphasize that these passages of Kauffman's are embedded in a pair of books that are for the most part interesting, creative and uninfluenced by Gould. The same is true of Richard Leakey and Roger Lewin's The Sixth Extinction (1996), another recent book, admirable in most of its chapters, but sadly marred by one, 'The Mainspring of Evolution', which is explicitly and avowedly influenced by Gould. Here are a couple of relevant passages:
It was as if the facility for making evolutionary leaps that produced major functional novelties - the basis of new phyla - had somehow been lost when the Cambrian period came to an end. It was as if the mainspring of evolution had lost some of its power.
Hence, evolution in Cambrian organisms could take bigger leaps, including phylum-level leaps, while later on it would be more constrained, making only modest jumps, up to the class level.
As I have written before, it is as though a gardener looked at an old oak tree and remarked, wonderingly: 'Isn't it strange that no major new boughs have appeared on this tree for many years. These days, all the new growth appears to be at the twig level! ' Just think once again what a 'phylum-level leap' or even a 'modest' {modest? ) class level leap would have to mean. Animals of different phyla, remember, are animals with different fundamental body plans, like molluscs and vertebrates. Or like starfish and insects. A long-jump, phylum level mutation would have to mean that a couple of parents belonging to one phylum mated and gave birth to a child belonging to a different phylum. The difference between parent and offspring would have to be on the same scale as the difference between a snail and a lobster, or a starfish and a codfish. A class level leap would be equivalent to a pair of birds giving birth to a mammal.
Picture the parents gazing wonderingly into the nest at what they have produced, and the full comedy of the notion becomes apparent.
My assurance in ridiculing these ideas is not based simply upon knowledge of the facts of modern animals. Obviously if it were just that, one could retort that things were different in the Cambrian. No, the argument against Kaufman's long jumps, or Leakey and Lewin's phylum level leaps, is a theoretical one, and an extremely strong one. It is this. Even if mutations on this gigantic scale occurred, the products would not have survived. This is fundamentally because, as I have said before, however many ways there may be of being alive, there are almost infinitely more ways of being dead. A small mutation, representing a minor step away from a parent which has proved its ability to survive by virtue of being a parent, has a good chance of surviving for the same reason, and it may even be an improvement. A gigantic, phylum level mutation is a leap into the wild blue yonder. I said that the long-jump mutation we are talking about would be of the same magnitude as a mutation from a mollusc to an insect. But it would never, of course, have been a jump from a mollusc to an insect. An insect is a highly tuned piece of survival machinery. If a mollusc parent gave birth to a new phylum, the leap would have been a random leap, like any other mutation. And the chance that a random leap of that magnitude would produce an insect, or anything with the faintest chance of surviving, is small enough to be discounted totally. The chance of its being viable is impossibly small, no matter how empty the ecosystem, how wide open the niches. A phylum level leap would be a mess.
I do not believe the authors I am quoting really believe what their printed words undoubtedly appear to be saying. I think they were simply intoxicated by Gould's rhetoric and didn't think it through. The whole point of quoting them in this chapter is to illustrate the power to mislead that a skilled poet can unwittingly exert, especially if he has first misled himself And the poetry of the Cambrian as a blissful dawn of innovation is undoubtedly beguiling. Kauffman gets completely carried away by it:
Soon after multicelled forms were invented, a grand burst of evolutionary novelty thrust itself outward. One almost gets the sense of multicellular life gleefully trying out all its possible ramifications, in a kind of wild dance of heedless exploration
At Home in the Universe (1995)
Yes. One does get exactly that sense. But one gets it from Gould's rhetoric, not from the facts of Cambrian fossils nor from sober reasoning about evolutionary principles.
If scientists of the calibre of Kauffman, Leakey and Lewin can be seduced by bad poetic science, what chance has the non-specialist?
Daniel Dennett has told me of a conversation with a philosopher colleague who had read Wonderful Lifers arguing that the Cambrian phyla did not have a common ancestor - that they had sprung up as independent origins of life! When Dennett assured him that this was not Gould's intention, his colleague's response was, 'Well then, what is all the fuss about? '
Excellence in writing is a double-edged sword, as the distinguished evolutionary scientist John Maynard Smith has noted, in the New York Review of Books, November 1995:
Gould occupies a rather curious position, particularly on his side of the Atlantic. Because of the excellence of his essays, he has come to be seen by non-biologists as the pre-eminent evolutionary theorist. In contrast, the evolutionary biologists with whom I have discussed his work tend to see him as a man whose ideas are so confused as to be hardly worth bothering with, but as one who should not be publicly criticized because he is at least on our side against the creationists. All this would not matter, were it not that he is giving non-biologists a largely false picture of the state of evolutionary theory.
Maynard Smith was reviewing Dennett's book Darwin's Dangerous Idea (1995), which contains a devastating and, one might hope, terminal critique of Gould's influence on evolutionary thinking.
What really happened in the Cambrian? Simon Conway Morris of Cambridge University is, as Gould fulsomely acknowledges, one of the three leading modern investigators of the Burgess Shale, the Cambrian fossil bed which is the subject of Wonderful Life. Conway Morris has recently published his own fascinating book on the subject. The Crucible of Creation (1998), which is critical of almost every aspect of Gould's view. Like Conway Morris, I don't think there's any good reason to think that the process of evolution was different in the Cambrian from the way it is today. But there is no doubt that a large number of major animal groups are seen in the fossil record for the first time in the Cambrian. The obvious hypothesis has occurred to many people. Perhaps several groups of animals evolved hard, fossilizable skeletons around the same time and perhaps for the same reason. One possibility is an evolutionary arms
race between predators and prey, but there are other ideas like a dramatic change in the chemistry of the atmosphere. Conway Morris finds no support at all for the poetic idea of an exuberant and
extravagant flowering of life in a wild dance of Cambrian diversity and disparity, subsequently pruned to today's more limited repertoire of
animal types. If anything, the reverse seems to be true, as most evolutionists would expect.
Where does that leave the question of the timing of the branch points of the major phyla? Recall that this is a separate question from the undoubted Cambrian explosion of fossil availability. The controversial matter is whether the branch points in the divergence of all the major phyla are concentrated in the Cambrian - the branch point explosion hypothesis. I said standard neo-Darwinism was compatible with this hypothesis. But I still don't think it is at all likely.
One possible way to tackle the question is by looking at molecular clocks. 'Molecular clock' refers to the observation that certain biological molecules change at a rather fixed rate over the millions of years. If you accept this, you can take blood from any two modern animals and calculate how long ago their common ancestor lived. Some recent molecular clock studies have pushed the branch points of various pairs of phyla deep into the Precambrian era. If these studies are right, the whole rhetoric of an evolutionary explosion becomes superfluous. But there is controversy over the interpretation of molecular clock results so far back in deep time, and we should wait for more evidence.
Meanwhile, there is a logical argument which I can assert with more confidence. The only evidence in favour of the branch point explosion hypothesis is negative: there aren't any fossils of many of the phyla before the Cambrian. But those fossil animals that have no fossil ancestors must have had ancestors of some kind. They can't have sprung from nothing. Therefore there must have been ancestors that didn't fossilize, absence of fossils does not mean absence of animals. The only question that remains is whether the missing ancestors going back to the branch points, who must have existed, were all compressed into the Cambrian, or whether they were strung out through the previous hundreds of millions of years. Since the only reason to suppose that they were compressed into the Cambrian is the absence of fossils, and since we have just proved logically the irrelevance of that absence, I conclude that there is no good reason at all to favour the branch point explosion hypothesis. Doubtless it has great poetic appeal.
9
THE SELFISH COOPERATOR
Wonder. . . and not any expectation of advantage from its discoveries, is the first principle which prompts mankind to the study of Philosophy, of that science which pretends to lay open the concealed connections that unite the various appearances of nature.
ADAM SMITH, The History of Astronomy' (1795)
The medieval bestiaries continued an earlier tradition of hijacking nature as a source of moral tales. In its modern form, in the development of evolutionary ideas, the same tradition underlies one of the most egregious forms of bad poetic science. I refer to the illusion that there is a simple opposition between nasty and nice, social and antisocial, selfish and altruistic, tough and gentle; that these pairs of binary opposites all correspond to the other pairs, and that the history of evolutionary controversy about society is described by a pendulum swinging back and forth along a continuum between these opposites. I am not denying that there are interesting issues to be discussed hereabouts. What I am criticizing is the 'poetic' idea that there is a single continuum and that worthwhile arguments are to be had between vantage points along its length. To invoke the rainmakers yet again, there is no more connection between a selfish gene and a selfish human than there is between a rock and a rain cloud.
To explain the poetic continuum I am criticizing, I might as well borrow a line from a real poet, Tennyson's 'Nature, red in tooth and claw', from In Memoriam (1850), widely assumed to be inspired by On the Origin of Species but actually published nine years earlier. At one end of the poetic continuum are supposed to stand Thomas Hobbes, Adam Smith, Charles Darwin, T. H. Huxley and all those, such as the distinguished American evolutionist George C. Williams and today's advocates of 'the selfish gene', who emphasize that nature really is red in tooth and claw. At the other end of the continuum are Prince Peter Kropotkin the Russian anarchist and author of Mutual Aid (1902), the gullible but immensely influential American anthropologist Margaret Mead, and today a spate of authors reacting indignantly to the idea that nature is genetically selfish, of
whom Frans de Waal, author of Good Natured (1996), is representative.
De Waal, a chimpanzee expert who understandably loves his animals, is distressed at what he mistakenly sees as a neo-Darwinian tendency to emphasize the 'nastiness of our apish past'. Some of those who share his romantic fancy have recently become fond of the pygmy chimpanzee or bonobo as a yet more benign role model. Where common chimpanzees often resort to violence, and even cannibalism, bonobos say it with sex. They seem to copulate in all possible combinations at every conceivable opportunity. Where we might shake hands, they copulate. Make love not war is their watchword. Margaret Mead would have warmed to them. But the very idea of taking animals to be role models, as in the bestiaries, is a piece of bad poetic science. Animals are not there to be role models, they are there to survive and reproduce.
Moralistic devotees of the bonobo are apt to compound this error with an outright evolutionary falsehood. Probably because of their powerful 'feelgood factor', bonobos are often claimed as more closely related to us than common chimpanzees are. But this cannot be, so long as we accept, as everybody does, that bonobos and common chimpanzees are more closely related to each other than either is to humans. You need no more than that simple and uncontroversial premise to conclude that bonobos and common chimpanzees are exactly equally closely related to us. They are connected to us via the common ancestor that they share and we don't. Certainly we may resemble one of the two species more than the other in some respects (and very probably resemble the other in other respects), but such comparative judgements absolutely cannot be reflections of differential evolutionary closeness.
* I should explain that Margaret Mead is 'gullible but influential' because a large section of American academic culture enthusiastically adopted her rose-tinted environmentalist theory of human nature which, it later transpired, she had built on a somewhat insecure foundation: systematic misinformation fed her as a joke by two mischievous Samoan girls, during her brief period of fieldwork in their island. She didn't stay in Samoa long enough to learn the language well, unlike her professional nemesis, the Australian anthropologist Derek Freeman, who uncovered the whole story years later in the course of a more detailed study of Samoan life.
De Waal's book is full of anecdotal demonstrations (which should surprise nobody) that animals are sometimes kind to each other, cooperate for mutual good, care for each other's welfare, console each other in distress, share food and do other heartwarmingly good things. The position I have always adopted is that much of animal nature is indeed altruistic, cooperative and even attended by benevolent subjective emotions, but that this follows from, rather than contradicts, selfishness at the genetic level. Animals are sometimes nice and sometimes nasty, since either can suit the self-interest of genes at different times. That is precisely the reason for speaking of 'the selfish gene' rather than, say, 'the selfish chimpanzee'. The opposition that de Waal and others have erected, between biologists who believe human and animal nature is fundamentally selfish, and those who believe it is fundamentally 'good- natured', is a false opposition - bad poetry.
It is now widely understood that altruism at the level of the individual organism can be a means by which the underlying genes maximize their self-interest. However, I don't want to dwell on what I have expounded in earlier books such as The Selfish Gene. What I would now re-emphasize from that book - it has been overlooked by critics who appear to have read it by title only - is the important sense in which genes, though in
one way purely selfish, at the same time enter into cooperative cartels with each other. This is poetic science, if you like, but I hope to show that it is good poetic science which aids understanding rather than impedes it. I shall do the same with other examples in the remaining chapters. The key insight of Darwinism can be expressed in genetic terms. The genes that exist in many copies in the population are the ones that are good at making copies, which also means good at surviving. Surviving where? Surviving in individual bodies in ancestral environments. That means surviving in the environment typical of the species: in a desert for camels, up trees for monkeys, in the deep sea for giant squids, and so on. The reason individual bodies are so good at surviving in their environments is mainly that they have been built by genes that have survived in the same environment for many generations, in the form of copies.
But never mind deserts and ice floes, seas and forests; they are only part of the story. A far more prominent aspect of the ancestral environment in which genes have survived is the other genes with which they have had to share a succession of individual bodies. The genes that survive in camels will, to be sure, include some that are particularly good at surviving in deserts, and they may even be shared with desert rats and desert foxes. But more importantly, successful genes will be those that are good at surviving in an environment consisting of the other genes that are typically found in the species. So, the genes of a species become selected to be good at cooperating with each other. Genetic cooperation, which is good scientific poetry whereas universal cooperation is not, will be the subject of this chapter.
The following fact is often misunderstood. It is not the genes of any given individual that cooperate particularly well together. They have never been together before in that combination, for every genome in a sexually reproducing species is unique (with the usual exception of identical twins). It is the genes of a species at large that cooperate, because they have met before, often, and in the intimately shared environment of the cell, though always in different combinations. What they cooperate at is the business of making individuals of the same general type as the present one. There is no particular reason to expect the genes of any particular individual to be especially good at cooperating with one another when compared with any other genes of the same species. It is largely a matter of accident which particular companions the lottery of sexual reproduction has drawn for them from the gene pool of the species. Individuals with unfavourable combinations of genes tend to die. Individuals with favourable combinations tend to pass those genes on to the future. But it is not the favourable combinations themselves that are passed on in the long term. Sexual reshuffling sees to that. Instead, what are passed on are the genes that tend to be good at forming favourable
combinations with the other genes that the species gene pool has to offer. Over the generations, whatever else the surviving genes may be good at, they will be good at working together with other genes of the species.
For all we know, particular camel genes might be good at cooperating with particular cheetah genes. But they are never called upon to do so. Presumably mammal genes are better at cooperating with other mammal genes than with bird genes. But the speculation must remain hypothetical, because one of the characteristics of life on our planet is that, genetic engineering aside, genes are mixed only within species. We can test watered-down versions of such speculations by looking at hybrids. Hybrids between different species, when they exist at all, often survive less well, or are less fertile, than pure-bred individuals. At least part of the reason for this is incompatibilities between their genes. Species A genes that work well against a genetic background or 'climate' of other Species A genes do not work when transplanted into Species B, and vice versa. Similar effects are sometimes seen when varieties or races within one species hybridize.
I first understood this while listening to lectures by the late E. B. Ford, legendary Oxford aesthete and eccentric founder of the now neglected school of Ecological Geneticists. Most of Ford's research was on wild populations of butterflies and moths.
Among these was the Lesser Yellow Underwing moth, Triphaena comes. This moth is normally yellowish brown, but there is a variant called curtisii which is blackish. Curtisii is not found in England at all; however, in Scotland and the Isles curtisii coexists with the normal comes. The curtisii dark colour pattern is nearly completely dominant to the normal comes pattern. 'Dominant to' is a technical term, which is why I can't just say 'dominates'. It means that hybrids between the two look like curtisii even though they bear the genes of both. Ford caught specimens from Barra in the Outer Hebrides, west of Scotland, and from one of the Orkney islands, north of Scotland, as well as from the Scottish mainland itself. Each of the two island forms looks exactly like its opposite number at the other island site, and the dark curtisii gene is dominant on the two islands, as well as on the mainland. Other evidence shows that the curtisii gene is the very same gene in all localities. In view of this you'd expect that, when you cross- bred specimens from different islands, the normal dominance pattern would hold up. But it doesn't, and this is the point of the story. Ford caught individuals from Barra and mated them with individuals from Orkney. And the dominance of curtisii completely disappeared. A full range of intermediates turned up in the hybrid families, just as if there was no dominance. What seems to be going on is this. The curtisii gene does not, in itself, encode the formula for the coloured pigment by which we distinguish the moths, nor is dominance ever a property of a gene on its own. Instead, like any other gene, the curtisii gene should be thought
of as having its effects only in the context of a suite of other genes, some of which it 'switches on'. This suite of other genes is part of what I mean by 'genetic background' or 'genetic climate'. In theory, any gene could therefore exert radically different effects on different islands, in the presence of different suites of other genes. In the case of Ford's Yellow Underwings, things are a little more complicated, and very illuminating. The curtisii gene is a 'switch gene' which has what looks like the same effect on both Barra and Orkney, but it achieves it by switching on different suites of genes on the different islands. We notice this only when the two populations are cross-bred. The curtisii switch gene finds itself in a genetic climate which is neither one thing nor the other. It is a mixture of Barra genes and Orkney genes, and the colour pattern that either suite could produce, on its own, breaks down.
What is interesting about this is that either the Barra mixture or the Orkney mixture can put together the colour pattern. There is more than one way of achieving the same result. Both of them involve cooperating suites of genes, but they are two different suites and the members of each suite don't cooperate well with the other. I take this to be a model for what often goes on among working genes within any gene pool. In The Selfish Gene, I used a rowing analogy. A crew of eight oarsmen needs to be well coordinated. Eight men who have trained together can expect to work well together. But if you mix four men from one crew with four from another equally good crew, they don't gel: their rowing fails apart. This is analogous to the mixing of two suites of genes which worked well when each was with its previous companions, but whose coordination breaks down when each is pushed into the foreign genetic climate provided by the other.
Now at this point many biologists get carried away and say that natural selection must work at the level of the whole crew as a unit, the whole suite of genes, or the whole individual organism. They are right that the individual organism is a very important unit in the hierarchy of life. And it really does display unitary qualities. (This is less true of plants than of animals, who tend to have a fixed set of parts, all neatly parcelled inside a skin with a discrete, unitary shape. Individual plants are often harder to delimit as they straggle and vegetatively propagate themselves through meadows and undergrowth. ) But, however unitary and discrete an individual wolf or buffalo, say, may be, the package is temporary and it is unique. Successful buffaloes don't duplicate themselves around the world in the form of multiple copies, they duplicate their genes. The true unit of natural selection has to be a unit of which you can say it has a frequency. It has a frequency which goes up when its type is successful, down when it fails. This is exactly what you can say of genes in gene pools. But you can't say it of individual buffaloes. Successful buffaloes don't become more frequent. Each buffalo is unique. It has a frequency of
one. You can define a buffalo as successful if its genes increase in frequency in future populations.
Field Marshal Montgomery, never the humblest of men, was once heard to remark, 'Now God said (and I agree with Him) . . . ' I feel a bit like that when I read of God's covenant with Abraham. He didn't promise Abraham eternal life as an individual (though Abraham was only 99 at the time, a spring chicken by Genesis standards). But he did promise him something else.
And I will make my covenant between me and thee, and will multiply thee exceedingly . . . and thou shalt be a father of many nations . . . And I will make thee exceeding fruitful, and I will make nations of thee, and kings shall come out of thee. (Genesis 17
Abraham was left in no doubt that the future lay with his seed, not his individuality. God knew his Darwinism.
To resume, the point I am making is that genes, for all that they are the separate units naturally selected in the Darwinian process, are highly cooperative. Selection favours or disfavours single genes for their capacity to survive in their environment, but the most important part of that environment is the genetic climate furnished by other genes. The consequence is that cooperating suites of genes come together in gene pools. Individual bodies are as unitary and coherent as they are, not because natural selection chooses them as units, but because they are built by genes that have been selected to cooperate with other members of the gene pool. They cooperate specifically in the enterprise of building individual bodies. But it is an anarchistic, 'each gene for itself kind of cooperation.
The cooperation, indeed, breaks down whenever the chance arises, as in so-called segregation distorter' genes. There is a gene in mice known as the t gene. In double dose t causes sterility or death, and there must be strong natural selection against it. But in single dose in males, it has a very odd effect. Normally, each copy of a gene should find itself in 50 per cent of the sperms made by a male. I have brown eyes like my mother, but my father has blue, so I know that I carry one copy of the gene for blue eyes and 50 per cent of my sperms carry the blue-eyed gene. In male mice, t doesn't behave in this orderly way. More than 90 per cent of an affected male's sperms contain t. Distorting sperm production is what the t gene does. It is its equivalent of making brown eyes or curly hair. And you can see that, in spite of lethality in double dose, once t arises in a population of mice, it will tend to spread because of its huge success in getting itself into sperms. It has been suggested that outbreaks of t arise in wild populations of mice, spreading like a sort of population cancer
and eventually driving the local population extinct, t is an illustration of what can happen when cooperation among genes breaks down. 'The exception that proves the rule' is often a rather silly expression, but this is a rare occasion when it is appropriate.
To repeat, the main suites of cooperating genes are the whole gene pools of species. Cheetah genes cooperate with cheetah genes but not with camel genes, and vice versa. This is not because cheetah genes, even in the most poetic sense, see any virtue in the preservation of the cheetah species. They are not working to save the cheetah from extinction like some molecular World Wildlife Fund. They are simply surviving in their environment, and their environment largely consists of other genes from the cheetah gene pool. Therefore, abilities to cooperate with other cheetah genes (but not with camel genes or codfish genes) are among the main qualities favoured in the struggle between rival cheetah genes. Just as, in arctic climates, genes for withstanding the cold come to predominate, so, in cheetah gene pools, do genes that are equipped to flourish in the climate of other cheetah genes. As far as each gene is concerned, the other genes in its gene pool are just another aspect of the weather.
The level at which the genes constitute 'weather' for each other is mostly buried in cellular chemistry. Genes code the production of enzymes, protein molecules which work as machine tools churning out one particular component in a chemical production line. There are alternative chemical pathways to the same end, which means alternative production lines. It may not matter greatly which of two production lines is adopted, so long as the cell doesn't attempt both at once. Either of the two production lines might be equally good, but intermediate products yielded by production line A can't be used by production line B, and vice versa. Once again, it is tempting to say that the entire production line is naturally selected, as a unit. This is wrong. What is naturally selected is each individual gene, against the background or climate provided by all the other genes. If the population happens to be dominated by genes for all but one of the steps in production line A, this constitutes a chemical climate in which the gene for the missing A step is favoured. Conversely, a pre-existing climate of B genes favours B genes over A genes. We aren't talking about which is 'better', as though there were some kind of contest between production line A and production line B. What we are saying is that either of the two is fine, but a mixture is unstable. The population has two alternative stable climates of mutually cooperating genes and natural selection will tend to steer the population towards whichever of the two stable states it is already closest to.
But we don't have to talk of biochemistry. We can use the metaphor of genetic climate at the level of organs and behaviour. A cheetah is a
beautifully integrated killing machine, equipped with long, muscled legs and a sinuously sprung backbone for outrunning prey, powerful jaws
and dagger teeth for stabbing them, forward-focused eyes for aiming at them, short guts with appropriate enzymes for digesting them, a brain pre-loaded with carnivorous behaviour software, and collections of other features that make it a typical hunter. On the other side of the arms race, antelopes are equivalently well equipped to eat plants and avoid being caught by predators. Long guts, complicated by blind alleys stuffed with cellulose-digesting bacteria, go with flat grinding teeth, go with brains pre-programmed to alarm and rapid escape, go with exquisitely camouflaged dappling of the pelt. These are two alternative ways of making a living. Neither is obviously better than the other, but either is better than an uneasy compromise: carnivorous guts combined with herbivorous teeth, say, or carnivorous pursuit instincts combined with herbivorous digestive enzymes.
Yet again, it is tempting to speak of the 'whole cheetah' or the 'whole antelope' as being selected 'as a unit'. Tempting, but superficial. Also lazy. It requires some extra thinking work to see what is really going on. Genes that programme the development of carnivorous guts flourish in a
genetic climate that is already dominated by genes programming carnivorous brains. And vice versa. Genes that programme defensive camouflage flourish in a genetic climate that is already dominated by genes programming herbivorous teeth. And vice versa. There are lots and lots of ways of making a living. To mention only a few mammal examples, there is the cheetah way, the impala way, the mole way, the baboon way, the koala way. There is no need to say that one way is better than any other. All of them work. What is bad is to be caught with half your adaptations aimed at one way of life, half aimed at another.
This kind of argument is best expressed at the level of the separate genes. At each genetic locus, the gene most likely to be favoured is the one that is compatible with the genetic climate afforded by the others, the one
that survives in that climate through repeated generations. Since this applies to each one of the genes that constitute the climate - since every gene is potentially part of the climate of every other - the result is that a species gene pool tends to coalesce into a gang of mutually compatible partners. Sorry to go on about this, but some of my respected colleagues refuse to get the point, obstinately insisting that the 'individual' is the 'true' unit of natural selection!
More widely, the environment in which a gene has to survive includes the other species with which it comes into contact. The DNA of any one species doesn't literally come into direct contact with the DNA molecules of its predators, competitors or mutualistic partners. 'Climate' has to be understood less intimately than when the arena of gene cooperation is
the interior of cells, as it is for genes within one species. In the larger arena, it is the consequences of genes in other species - their 'phenotypic effects' - that constitute an important part of the environment in which the natural selection of genes within neighbouring species goes on. A rainforest is a special kind of environment, fashioned and defined by the plants and animals that live in it. Every one of the species in a tropical rainforest consists of a gene pool, isolated from all other gene pools as far as sexual mixing is concerned, but in contact with their bodily effects.
Within each of those separate gene pools, natural selection favours those genes that cooperate within their own gene pool, as we have seen. But it also favours those genes that are good at surviving alongside the consequences of the other gene pools in the rainforest - the trees, vines, monkeys, dung beetles, wood lice and soil bacteria. In the long run this may make the whole forest look like a single harmonious whole, with each unit pulling for the benefit of all, every tree and every soil mite, even every predator and every parasite, playing its part in one big, happy family. Once again, this is a tempting way of looking at it. Once again, it is lazy - bad poetic science. A much truer vision, still poetic science but (it is the purpose of this chapter to persuade you) good poetic science, sees the forest as an anarchistic federation of selfish genes, each selected as being good at surviving within its own gene pool against the background of the environment provided by all the others.
Yes, there is a wishy-washy sense in which organisms in a rainforest perform a valuable service for other species, and even for the maintenance of the whole forest community. Certainly, if you removed all the soil bacteria, the consequences for the trees and ultimately for most of the life of the forest, would be dire. But that is not why the soil bacteria are there. Yes, of course they do break down the dead leaves, dead animals and manure into compost which is useful for the continued prosperity of the whole forest. But they aren't doing it for the sake of making compost. They are using the dead leaves and dead animals as food for themselves, for the good of the genes that programme their compost-making activities. It is an incidental consequence of this self- interested activity that the soil is improved from the point of view of the plants, and therefore the herbivores that eat them, and therefore the carnivores that eat the herbivores. Species in a rainforest community flourish in the presence of the other species in that community because the community is the environment in which their ancestors survived. Maybe there are plants that flourish in the absence of a rich culture of soil bacteria, but those are not the plants we find in a rainforest. We are more likely to find them in a desert.
This is the right way to handle the temptation of 'Gaia': the overrated romantic fancy of the whole world as an organism; of each species doing
its bit for the welfare of the whole; of bacteria, for example, working to improve the gas content of the earth's atmosphere for the good of all life. The most extreme example I know of this kind of bad poetic science comes from a famous and senior 'ecologist' (the quotation marks denote an activist for green politics, rather than a genuine scholar of the academic subject of ecology). It was reported to me by Professor John Maynard Smith, who was attending a conference sponsored by the Open University in Britain. The conversation turned to the mass extinction of the dinosaurs and whether this catastrophe was caused by a cometary collision. The bearded ecologist was in no doubt. 'Of course not,' he said decisively, Gaia would not have permitted it! '
Gaia was the Greek earth goddess whose name has been adopted by James Lovelock, an English atmospheric chemist and inventor, to personify his poetic notion that the whole planet should be regarded as a single living thing. All living creatures are Gaia's body parts and they work together as a well-adjusted thermostat, reacting to perturbations so as to preserve all life. Lovelock is avowedly embarrassed by those, like the ecologist I have just quoted, who take his idea right over the top. Gaia has become a cult, almost a religion, and Lovelock now understandably wants to distance himself from this. But some of his own early suggestions, when you think them through, are only slightly more realistic. He proposed, for instance, that bacteria produce methane gas because of the valuable role it plays in regulating the chemistry of the earth's atmosphere. The problem with this is that individual bacteria are asked to be nicer than natural selection can explain. The bacteria are supposed to produce methane beyond their own needs. They are expected to produce enough methane to benefit the planet in general. It is no good pleading that this is in their own long-term interests because if the planet goes extinct so will they. Natural selection is never aware of the long-term future. It is not aware of anything. Improvements come about not through foresight but by genes coming to outnumber their rivals in gene pools. Unfortunately, genes that make rebel bacteria sit back and enjoy the benefits of their rivals' altruistic production of methane are bound to prosper at the expense of the altruists. So the world will become relatively more full of selfish bacteria. This will continue even if, because of their selfishness, the total number of bacteria (and of everything else) is going down. It will continue even to the point of extinction. How should it not? There is no foresight.
If Lovelock were to retort that the bacteria produce methane as a by- product of something else that they do for their own good, and it is only incidentally useful for the world, I should agree wholeheartedly. But in that case the whole rhetoric of Gaia is superfluous and misleading. You don't need to talk about bacteria working for the good of anything other than their own short-term genetic good. We are left with the conclusion
that individuals work for Gaia only when it suits them to do so - so why bother to bring Gaia into the discussion? We are better off thinking about genes, which are the real, self-replicating units of natural selection, flourishing in an environment which includes the genetic climate furnished by the other genes. I am quite happy to generalize the notion of the genetic climate to include all the genes in the whole world. But that is not Gaia. Gaia falsely focuses attention on planetary life as a single unit. Planetary life is a shifting pattern of genetic weather.
Lovelock's main comrade-in-arms as a champion of Gaia is the American bacteriologist Lynn Margulis. Despite her pugnacious disposition, she places herself firmly on the gentle side of the continuum which I am attacking as bad poetic science. Here she is, writing with her son Dorion Sagan:
Next, the view of evolution as chronic bloody competition among individuals and species, a popular distortion of Darwin's notion of 'survival of the fittest,' dissolves before a new view of continual cooperation, strong interaction, and mutual dependence among life forms. Life did not take over the globe by combat, but by networking. Life forms multiplied and complexified by co-opting others, not just by killing them. Microcosmos: Four Billion Years
of Microbial Evolution (1987)
Margulis and Sagan are in a superficial sense not too far from being right here. But they are misled by bad poetic science into expressing it wrongly. As I emphasized at the beginning of this chapter, the opposition 'combat versus cooperation' is the wrong dichotomy to stress. There is fundamental conflict at the level of the genes. But since the
environments of genes are dominated by each other, cooperation and 'networking' arise automatically as a favoured manifestation of that conflict.
Where Lovelock is a student of the world's atmosphere, Margulis approaches from the other direction, as a specialist in bacteria. She rightly grants bacteria centre stage among life forms on our planet. At the level of biochemistry, there is a range of fundamental ways of making a living. These are practised by one or another kind of bacteria. One of these basic life recipes has been adopted by eukaryotes (that's everybody except bacteria), and we get it from bacteria. Margulis has successfully argued over many years that most of our biochemistry is carried out for us by what were once free bacteria now living within our cells. Here's another quotation from the same book by Margulis and Sagan.
Bacteria, by contrast, exhibit a far wider range of metabolic variations than eukaryotes.
The extreme Gouldian view - certainly the view inspired by his rhetoric, though it is hard to tell from his own words whether he literally holds it himself - is radically different from and utterly incompatible with the standard neo-Darwinian model. It also, as I shall show, has implications which, once they are spelled out, anybody can see are absurd. It is very clearly expressed - betrayed might be a better word - in asides in Stuart Kauffman's At Home in the Universe (1995):
One might imagine that the first multicellular creatures would all he very similar, only later diversifying, from the bottom up, into different genera, families, orders, classes, and so on. That, indeed, would be the expectation of the strictest conventional Darwinist. Darwin, profoundly influenced by the emerging view of geologic gradualism, proposed that all evolution occurred by the very gradual accumulation of useful variations. Thus the earliest multicellular creatures themselves ought to have diverged gradually from one another.
So far, this is a fine summary of the orthodox neo-Darwinian view. Now, in a bizarre passage, Kauffman goes on:
But this appears to be false. One of the wonderful and puzzling features of the Cambrian explosion is that the chart was filled in from the top
down. Nature suddenly sprang forth with many wildly different body plans - the phyla - elaborating on these basic designs to form the classes, orders, families, and genera . . . In his book about the Cambrian explosion, Wonderful Life: The Burgess Shale and the Nature of History, Stephen Jay Gould remarks on this top-down quality of the Cambrian with wonder.
As well he might! You only have to think for one moment about what 'top down' filling in would have to mean for the animals on the ground and you immediately see how preposterous it is, 'Body plans' like the mollusc body plan, or the echinoderm body plan, are not ideal essences hanging in the sky, waiting, like designer dresses, to be adopted by real animals. Real animals is all there ever was: living, breathing, walking, eating, excreting, fighting, copulating real animals, who had to survive and who can't have been dramatically different from their real parents and grandparents. For a new body plan - a new phylum - to spring into existence, what actually has to happen on the ground is that a child is born which suddenly, out of the blue, is as different from its parents as a snail is from an earthworm. No zoologist who thinks through the implications, not even the most ardent saltationist, has ever supported any such notion. Ardent saltationists have been content to postulate the sudden bursting into existence of new species, and even that relatively modest idea has been highly controversial. When you spell out the Gouldian rhetoric into real-life practicalities, it stands revealed as the purest of bad poetic science.
Kauffman is even more explicit in a later chapter. In discussing some of his ingenious mathematical models of evolution on 'rugged fitness landscapes', Kauffman notes a pattern that he thinks sounds a lot like the Cambrian explosion. Early on in the branching process, we find a variety of long-jump mutations that differ from the stem and from one another quite dramatically. These species have sufficient morphological differences to be categorized as founders of distinct phyla. These founders also branch, but do so via slightly closer long-jump variants, yielding branches from, each founder of a phylum to dissimilar daughter species, the founders of classes. As the process continues, fitter variants are found in progressively more nearby neighborhoods, so founders of orders, families, and genera emerge in succession.
Kauffman's earlier, more technical book, The Origins of Order (1993), says something similar about life in the Cambrian:
Not only did a very large number of novel body forms originate rapidly, but the Cambrian explosion exhibited another novelty: Species which founded taxa appear to have built up the higher taxa from the top down.
That is, exemplars of major phyla were present first, followed by progressive filling in at class, order, and lower taxonomic levels. . .
Now, one way of reading this is inoffensive to the point of obviousness. On our 'converging backwards' model it would have to be true that species splittings that are eventually going to become phylum divides would normally precede those that are destined to become divides between orders and lower taxonomic levels. But Kauffman clearly doesn't think he is saying something ordinary and obvious. This is apparent from his statement that 'the Cambrian explosion exhibited another novelty', and from his phrase long-jump mutations'. He thinks he is attributing to the Cambrian something revolutionary. He really does seem genuinely to intend the alternative reading, in which 'long-jump mutations' give rise, on the instant, to brand new phyla.
I hasten to emphasize that these passages of Kauffman's are embedded in a pair of books that are for the most part interesting, creative and uninfluenced by Gould. The same is true of Richard Leakey and Roger Lewin's The Sixth Extinction (1996), another recent book, admirable in most of its chapters, but sadly marred by one, 'The Mainspring of Evolution', which is explicitly and avowedly influenced by Gould. Here are a couple of relevant passages:
It was as if the facility for making evolutionary leaps that produced major functional novelties - the basis of new phyla - had somehow been lost when the Cambrian period came to an end. It was as if the mainspring of evolution had lost some of its power.
Hence, evolution in Cambrian organisms could take bigger leaps, including phylum-level leaps, while later on it would be more constrained, making only modest jumps, up to the class level.
As I have written before, it is as though a gardener looked at an old oak tree and remarked, wonderingly: 'Isn't it strange that no major new boughs have appeared on this tree for many years. These days, all the new growth appears to be at the twig level! ' Just think once again what a 'phylum-level leap' or even a 'modest' {modest? ) class level leap would have to mean. Animals of different phyla, remember, are animals with different fundamental body plans, like molluscs and vertebrates. Or like starfish and insects. A long-jump, phylum level mutation would have to mean that a couple of parents belonging to one phylum mated and gave birth to a child belonging to a different phylum. The difference between parent and offspring would have to be on the same scale as the difference between a snail and a lobster, or a starfish and a codfish. A class level leap would be equivalent to a pair of birds giving birth to a mammal.
Picture the parents gazing wonderingly into the nest at what they have produced, and the full comedy of the notion becomes apparent.
My assurance in ridiculing these ideas is not based simply upon knowledge of the facts of modern animals. Obviously if it were just that, one could retort that things were different in the Cambrian. No, the argument against Kaufman's long jumps, or Leakey and Lewin's phylum level leaps, is a theoretical one, and an extremely strong one. It is this. Even if mutations on this gigantic scale occurred, the products would not have survived. This is fundamentally because, as I have said before, however many ways there may be of being alive, there are almost infinitely more ways of being dead. A small mutation, representing a minor step away from a parent which has proved its ability to survive by virtue of being a parent, has a good chance of surviving for the same reason, and it may even be an improvement. A gigantic, phylum level mutation is a leap into the wild blue yonder. I said that the long-jump mutation we are talking about would be of the same magnitude as a mutation from a mollusc to an insect. But it would never, of course, have been a jump from a mollusc to an insect. An insect is a highly tuned piece of survival machinery. If a mollusc parent gave birth to a new phylum, the leap would have been a random leap, like any other mutation. And the chance that a random leap of that magnitude would produce an insect, or anything with the faintest chance of surviving, is small enough to be discounted totally. The chance of its being viable is impossibly small, no matter how empty the ecosystem, how wide open the niches. A phylum level leap would be a mess.
I do not believe the authors I am quoting really believe what their printed words undoubtedly appear to be saying. I think they were simply intoxicated by Gould's rhetoric and didn't think it through. The whole point of quoting them in this chapter is to illustrate the power to mislead that a skilled poet can unwittingly exert, especially if he has first misled himself And the poetry of the Cambrian as a blissful dawn of innovation is undoubtedly beguiling. Kauffman gets completely carried away by it:
Soon after multicelled forms were invented, a grand burst of evolutionary novelty thrust itself outward. One almost gets the sense of multicellular life gleefully trying out all its possible ramifications, in a kind of wild dance of heedless exploration
At Home in the Universe (1995)
Yes. One does get exactly that sense. But one gets it from Gould's rhetoric, not from the facts of Cambrian fossils nor from sober reasoning about evolutionary principles.
If scientists of the calibre of Kauffman, Leakey and Lewin can be seduced by bad poetic science, what chance has the non-specialist?
Daniel Dennett has told me of a conversation with a philosopher colleague who had read Wonderful Lifers arguing that the Cambrian phyla did not have a common ancestor - that they had sprung up as independent origins of life! When Dennett assured him that this was not Gould's intention, his colleague's response was, 'Well then, what is all the fuss about? '
Excellence in writing is a double-edged sword, as the distinguished evolutionary scientist John Maynard Smith has noted, in the New York Review of Books, November 1995:
Gould occupies a rather curious position, particularly on his side of the Atlantic. Because of the excellence of his essays, he has come to be seen by non-biologists as the pre-eminent evolutionary theorist. In contrast, the evolutionary biologists with whom I have discussed his work tend to see him as a man whose ideas are so confused as to be hardly worth bothering with, but as one who should not be publicly criticized because he is at least on our side against the creationists. All this would not matter, were it not that he is giving non-biologists a largely false picture of the state of evolutionary theory.
Maynard Smith was reviewing Dennett's book Darwin's Dangerous Idea (1995), which contains a devastating and, one might hope, terminal critique of Gould's influence on evolutionary thinking.
What really happened in the Cambrian? Simon Conway Morris of Cambridge University is, as Gould fulsomely acknowledges, one of the three leading modern investigators of the Burgess Shale, the Cambrian fossil bed which is the subject of Wonderful Life. Conway Morris has recently published his own fascinating book on the subject. The Crucible of Creation (1998), which is critical of almost every aspect of Gould's view. Like Conway Morris, I don't think there's any good reason to think that the process of evolution was different in the Cambrian from the way it is today. But there is no doubt that a large number of major animal groups are seen in the fossil record for the first time in the Cambrian. The obvious hypothesis has occurred to many people. Perhaps several groups of animals evolved hard, fossilizable skeletons around the same time and perhaps for the same reason. One possibility is an evolutionary arms
race between predators and prey, but there are other ideas like a dramatic change in the chemistry of the atmosphere. Conway Morris finds no support at all for the poetic idea of an exuberant and
extravagant flowering of life in a wild dance of Cambrian diversity and disparity, subsequently pruned to today's more limited repertoire of
animal types. If anything, the reverse seems to be true, as most evolutionists would expect.
Where does that leave the question of the timing of the branch points of the major phyla? Recall that this is a separate question from the undoubted Cambrian explosion of fossil availability. The controversial matter is whether the branch points in the divergence of all the major phyla are concentrated in the Cambrian - the branch point explosion hypothesis. I said standard neo-Darwinism was compatible with this hypothesis. But I still don't think it is at all likely.
One possible way to tackle the question is by looking at molecular clocks. 'Molecular clock' refers to the observation that certain biological molecules change at a rather fixed rate over the millions of years. If you accept this, you can take blood from any two modern animals and calculate how long ago their common ancestor lived. Some recent molecular clock studies have pushed the branch points of various pairs of phyla deep into the Precambrian era. If these studies are right, the whole rhetoric of an evolutionary explosion becomes superfluous. But there is controversy over the interpretation of molecular clock results so far back in deep time, and we should wait for more evidence.
Meanwhile, there is a logical argument which I can assert with more confidence. The only evidence in favour of the branch point explosion hypothesis is negative: there aren't any fossils of many of the phyla before the Cambrian. But those fossil animals that have no fossil ancestors must have had ancestors of some kind. They can't have sprung from nothing. Therefore there must have been ancestors that didn't fossilize, absence of fossils does not mean absence of animals. The only question that remains is whether the missing ancestors going back to the branch points, who must have existed, were all compressed into the Cambrian, or whether they were strung out through the previous hundreds of millions of years. Since the only reason to suppose that they were compressed into the Cambrian is the absence of fossils, and since we have just proved logically the irrelevance of that absence, I conclude that there is no good reason at all to favour the branch point explosion hypothesis. Doubtless it has great poetic appeal.
9
THE SELFISH COOPERATOR
Wonder. . . and not any expectation of advantage from its discoveries, is the first principle which prompts mankind to the study of Philosophy, of that science which pretends to lay open the concealed connections that unite the various appearances of nature.
ADAM SMITH, The History of Astronomy' (1795)
The medieval bestiaries continued an earlier tradition of hijacking nature as a source of moral tales. In its modern form, in the development of evolutionary ideas, the same tradition underlies one of the most egregious forms of bad poetic science. I refer to the illusion that there is a simple opposition between nasty and nice, social and antisocial, selfish and altruistic, tough and gentle; that these pairs of binary opposites all correspond to the other pairs, and that the history of evolutionary controversy about society is described by a pendulum swinging back and forth along a continuum between these opposites. I am not denying that there are interesting issues to be discussed hereabouts. What I am criticizing is the 'poetic' idea that there is a single continuum and that worthwhile arguments are to be had between vantage points along its length. To invoke the rainmakers yet again, there is no more connection between a selfish gene and a selfish human than there is between a rock and a rain cloud.
To explain the poetic continuum I am criticizing, I might as well borrow a line from a real poet, Tennyson's 'Nature, red in tooth and claw', from In Memoriam (1850), widely assumed to be inspired by On the Origin of Species but actually published nine years earlier. At one end of the poetic continuum are supposed to stand Thomas Hobbes, Adam Smith, Charles Darwin, T. H. Huxley and all those, such as the distinguished American evolutionist George C. Williams and today's advocates of 'the selfish gene', who emphasize that nature really is red in tooth and claw. At the other end of the continuum are Prince Peter Kropotkin the Russian anarchist and author of Mutual Aid (1902), the gullible but immensely influential American anthropologist Margaret Mead, and today a spate of authors reacting indignantly to the idea that nature is genetically selfish, of
whom Frans de Waal, author of Good Natured (1996), is representative.
De Waal, a chimpanzee expert who understandably loves his animals, is distressed at what he mistakenly sees as a neo-Darwinian tendency to emphasize the 'nastiness of our apish past'. Some of those who share his romantic fancy have recently become fond of the pygmy chimpanzee or bonobo as a yet more benign role model. Where common chimpanzees often resort to violence, and even cannibalism, bonobos say it with sex. They seem to copulate in all possible combinations at every conceivable opportunity. Where we might shake hands, they copulate. Make love not war is their watchword. Margaret Mead would have warmed to them. But the very idea of taking animals to be role models, as in the bestiaries, is a piece of bad poetic science. Animals are not there to be role models, they are there to survive and reproduce.
Moralistic devotees of the bonobo are apt to compound this error with an outright evolutionary falsehood. Probably because of their powerful 'feelgood factor', bonobos are often claimed as more closely related to us than common chimpanzees are. But this cannot be, so long as we accept, as everybody does, that bonobos and common chimpanzees are more closely related to each other than either is to humans. You need no more than that simple and uncontroversial premise to conclude that bonobos and common chimpanzees are exactly equally closely related to us. They are connected to us via the common ancestor that they share and we don't. Certainly we may resemble one of the two species more than the other in some respects (and very probably resemble the other in other respects), but such comparative judgements absolutely cannot be reflections of differential evolutionary closeness.
* I should explain that Margaret Mead is 'gullible but influential' because a large section of American academic culture enthusiastically adopted her rose-tinted environmentalist theory of human nature which, it later transpired, she had built on a somewhat insecure foundation: systematic misinformation fed her as a joke by two mischievous Samoan girls, during her brief period of fieldwork in their island. She didn't stay in Samoa long enough to learn the language well, unlike her professional nemesis, the Australian anthropologist Derek Freeman, who uncovered the whole story years later in the course of a more detailed study of Samoan life.
De Waal's book is full of anecdotal demonstrations (which should surprise nobody) that animals are sometimes kind to each other, cooperate for mutual good, care for each other's welfare, console each other in distress, share food and do other heartwarmingly good things. The position I have always adopted is that much of animal nature is indeed altruistic, cooperative and even attended by benevolent subjective emotions, but that this follows from, rather than contradicts, selfishness at the genetic level. Animals are sometimes nice and sometimes nasty, since either can suit the self-interest of genes at different times. That is precisely the reason for speaking of 'the selfish gene' rather than, say, 'the selfish chimpanzee'. The opposition that de Waal and others have erected, between biologists who believe human and animal nature is fundamentally selfish, and those who believe it is fundamentally 'good- natured', is a false opposition - bad poetry.
It is now widely understood that altruism at the level of the individual organism can be a means by which the underlying genes maximize their self-interest. However, I don't want to dwell on what I have expounded in earlier books such as The Selfish Gene. What I would now re-emphasize from that book - it has been overlooked by critics who appear to have read it by title only - is the important sense in which genes, though in
one way purely selfish, at the same time enter into cooperative cartels with each other. This is poetic science, if you like, but I hope to show that it is good poetic science which aids understanding rather than impedes it. I shall do the same with other examples in the remaining chapters. The key insight of Darwinism can be expressed in genetic terms. The genes that exist in many copies in the population are the ones that are good at making copies, which also means good at surviving. Surviving where? Surviving in individual bodies in ancestral environments. That means surviving in the environment typical of the species: in a desert for camels, up trees for monkeys, in the deep sea for giant squids, and so on. The reason individual bodies are so good at surviving in their environments is mainly that they have been built by genes that have survived in the same environment for many generations, in the form of copies.
But never mind deserts and ice floes, seas and forests; they are only part of the story. A far more prominent aspect of the ancestral environment in which genes have survived is the other genes with which they have had to share a succession of individual bodies. The genes that survive in camels will, to be sure, include some that are particularly good at surviving in deserts, and they may even be shared with desert rats and desert foxes. But more importantly, successful genes will be those that are good at surviving in an environment consisting of the other genes that are typically found in the species. So, the genes of a species become selected to be good at cooperating with each other. Genetic cooperation, which is good scientific poetry whereas universal cooperation is not, will be the subject of this chapter.
The following fact is often misunderstood. It is not the genes of any given individual that cooperate particularly well together. They have never been together before in that combination, for every genome in a sexually reproducing species is unique (with the usual exception of identical twins). It is the genes of a species at large that cooperate, because they have met before, often, and in the intimately shared environment of the cell, though always in different combinations. What they cooperate at is the business of making individuals of the same general type as the present one. There is no particular reason to expect the genes of any particular individual to be especially good at cooperating with one another when compared with any other genes of the same species. It is largely a matter of accident which particular companions the lottery of sexual reproduction has drawn for them from the gene pool of the species. Individuals with unfavourable combinations of genes tend to die. Individuals with favourable combinations tend to pass those genes on to the future. But it is not the favourable combinations themselves that are passed on in the long term. Sexual reshuffling sees to that. Instead, what are passed on are the genes that tend to be good at forming favourable
combinations with the other genes that the species gene pool has to offer. Over the generations, whatever else the surviving genes may be good at, they will be good at working together with other genes of the species.
For all we know, particular camel genes might be good at cooperating with particular cheetah genes. But they are never called upon to do so. Presumably mammal genes are better at cooperating with other mammal genes than with bird genes. But the speculation must remain hypothetical, because one of the characteristics of life on our planet is that, genetic engineering aside, genes are mixed only within species. We can test watered-down versions of such speculations by looking at hybrids. Hybrids between different species, when they exist at all, often survive less well, or are less fertile, than pure-bred individuals. At least part of the reason for this is incompatibilities between their genes. Species A genes that work well against a genetic background or 'climate' of other Species A genes do not work when transplanted into Species B, and vice versa. Similar effects are sometimes seen when varieties or races within one species hybridize.
I first understood this while listening to lectures by the late E. B. Ford, legendary Oxford aesthete and eccentric founder of the now neglected school of Ecological Geneticists. Most of Ford's research was on wild populations of butterflies and moths.
Among these was the Lesser Yellow Underwing moth, Triphaena comes. This moth is normally yellowish brown, but there is a variant called curtisii which is blackish. Curtisii is not found in England at all; however, in Scotland and the Isles curtisii coexists with the normal comes. The curtisii dark colour pattern is nearly completely dominant to the normal comes pattern. 'Dominant to' is a technical term, which is why I can't just say 'dominates'. It means that hybrids between the two look like curtisii even though they bear the genes of both. Ford caught specimens from Barra in the Outer Hebrides, west of Scotland, and from one of the Orkney islands, north of Scotland, as well as from the Scottish mainland itself. Each of the two island forms looks exactly like its opposite number at the other island site, and the dark curtisii gene is dominant on the two islands, as well as on the mainland. Other evidence shows that the curtisii gene is the very same gene in all localities. In view of this you'd expect that, when you cross- bred specimens from different islands, the normal dominance pattern would hold up. But it doesn't, and this is the point of the story. Ford caught individuals from Barra and mated them with individuals from Orkney. And the dominance of curtisii completely disappeared. A full range of intermediates turned up in the hybrid families, just as if there was no dominance. What seems to be going on is this. The curtisii gene does not, in itself, encode the formula for the coloured pigment by which we distinguish the moths, nor is dominance ever a property of a gene on its own. Instead, like any other gene, the curtisii gene should be thought
of as having its effects only in the context of a suite of other genes, some of which it 'switches on'. This suite of other genes is part of what I mean by 'genetic background' or 'genetic climate'. In theory, any gene could therefore exert radically different effects on different islands, in the presence of different suites of other genes. In the case of Ford's Yellow Underwings, things are a little more complicated, and very illuminating. The curtisii gene is a 'switch gene' which has what looks like the same effect on both Barra and Orkney, but it achieves it by switching on different suites of genes on the different islands. We notice this only when the two populations are cross-bred. The curtisii switch gene finds itself in a genetic climate which is neither one thing nor the other. It is a mixture of Barra genes and Orkney genes, and the colour pattern that either suite could produce, on its own, breaks down.
What is interesting about this is that either the Barra mixture or the Orkney mixture can put together the colour pattern. There is more than one way of achieving the same result. Both of them involve cooperating suites of genes, but they are two different suites and the members of each suite don't cooperate well with the other. I take this to be a model for what often goes on among working genes within any gene pool. In The Selfish Gene, I used a rowing analogy. A crew of eight oarsmen needs to be well coordinated. Eight men who have trained together can expect to work well together. But if you mix four men from one crew with four from another equally good crew, they don't gel: their rowing fails apart. This is analogous to the mixing of two suites of genes which worked well when each was with its previous companions, but whose coordination breaks down when each is pushed into the foreign genetic climate provided by the other.
Now at this point many biologists get carried away and say that natural selection must work at the level of the whole crew as a unit, the whole suite of genes, or the whole individual organism. They are right that the individual organism is a very important unit in the hierarchy of life. And it really does display unitary qualities. (This is less true of plants than of animals, who tend to have a fixed set of parts, all neatly parcelled inside a skin with a discrete, unitary shape. Individual plants are often harder to delimit as they straggle and vegetatively propagate themselves through meadows and undergrowth. ) But, however unitary and discrete an individual wolf or buffalo, say, may be, the package is temporary and it is unique. Successful buffaloes don't duplicate themselves around the world in the form of multiple copies, they duplicate their genes. The true unit of natural selection has to be a unit of which you can say it has a frequency. It has a frequency which goes up when its type is successful, down when it fails. This is exactly what you can say of genes in gene pools. But you can't say it of individual buffaloes. Successful buffaloes don't become more frequent. Each buffalo is unique. It has a frequency of
one. You can define a buffalo as successful if its genes increase in frequency in future populations.
Field Marshal Montgomery, never the humblest of men, was once heard to remark, 'Now God said (and I agree with Him) . . . ' I feel a bit like that when I read of God's covenant with Abraham. He didn't promise Abraham eternal life as an individual (though Abraham was only 99 at the time, a spring chicken by Genesis standards). But he did promise him something else.
And I will make my covenant between me and thee, and will multiply thee exceedingly . . . and thou shalt be a father of many nations . . . And I will make thee exceeding fruitful, and I will make nations of thee, and kings shall come out of thee. (Genesis 17
Abraham was left in no doubt that the future lay with his seed, not his individuality. God knew his Darwinism.
To resume, the point I am making is that genes, for all that they are the separate units naturally selected in the Darwinian process, are highly cooperative. Selection favours or disfavours single genes for their capacity to survive in their environment, but the most important part of that environment is the genetic climate furnished by other genes. The consequence is that cooperating suites of genes come together in gene pools. Individual bodies are as unitary and coherent as they are, not because natural selection chooses them as units, but because they are built by genes that have been selected to cooperate with other members of the gene pool. They cooperate specifically in the enterprise of building individual bodies. But it is an anarchistic, 'each gene for itself kind of cooperation.
The cooperation, indeed, breaks down whenever the chance arises, as in so-called segregation distorter' genes. There is a gene in mice known as the t gene. In double dose t causes sterility or death, and there must be strong natural selection against it. But in single dose in males, it has a very odd effect. Normally, each copy of a gene should find itself in 50 per cent of the sperms made by a male. I have brown eyes like my mother, but my father has blue, so I know that I carry one copy of the gene for blue eyes and 50 per cent of my sperms carry the blue-eyed gene. In male mice, t doesn't behave in this orderly way. More than 90 per cent of an affected male's sperms contain t. Distorting sperm production is what the t gene does. It is its equivalent of making brown eyes or curly hair. And you can see that, in spite of lethality in double dose, once t arises in a population of mice, it will tend to spread because of its huge success in getting itself into sperms. It has been suggested that outbreaks of t arise in wild populations of mice, spreading like a sort of population cancer
and eventually driving the local population extinct, t is an illustration of what can happen when cooperation among genes breaks down. 'The exception that proves the rule' is often a rather silly expression, but this is a rare occasion when it is appropriate.
To repeat, the main suites of cooperating genes are the whole gene pools of species. Cheetah genes cooperate with cheetah genes but not with camel genes, and vice versa. This is not because cheetah genes, even in the most poetic sense, see any virtue in the preservation of the cheetah species. They are not working to save the cheetah from extinction like some molecular World Wildlife Fund. They are simply surviving in their environment, and their environment largely consists of other genes from the cheetah gene pool. Therefore, abilities to cooperate with other cheetah genes (but not with camel genes or codfish genes) are among the main qualities favoured in the struggle between rival cheetah genes. Just as, in arctic climates, genes for withstanding the cold come to predominate, so, in cheetah gene pools, do genes that are equipped to flourish in the climate of other cheetah genes. As far as each gene is concerned, the other genes in its gene pool are just another aspect of the weather.
The level at which the genes constitute 'weather' for each other is mostly buried in cellular chemistry. Genes code the production of enzymes, protein molecules which work as machine tools churning out one particular component in a chemical production line. There are alternative chemical pathways to the same end, which means alternative production lines. It may not matter greatly which of two production lines is adopted, so long as the cell doesn't attempt both at once. Either of the two production lines might be equally good, but intermediate products yielded by production line A can't be used by production line B, and vice versa. Once again, it is tempting to say that the entire production line is naturally selected, as a unit. This is wrong. What is naturally selected is each individual gene, against the background or climate provided by all the other genes. If the population happens to be dominated by genes for all but one of the steps in production line A, this constitutes a chemical climate in which the gene for the missing A step is favoured. Conversely, a pre-existing climate of B genes favours B genes over A genes. We aren't talking about which is 'better', as though there were some kind of contest between production line A and production line B. What we are saying is that either of the two is fine, but a mixture is unstable. The population has two alternative stable climates of mutually cooperating genes and natural selection will tend to steer the population towards whichever of the two stable states it is already closest to.
But we don't have to talk of biochemistry. We can use the metaphor of genetic climate at the level of organs and behaviour. A cheetah is a
beautifully integrated killing machine, equipped with long, muscled legs and a sinuously sprung backbone for outrunning prey, powerful jaws
and dagger teeth for stabbing them, forward-focused eyes for aiming at them, short guts with appropriate enzymes for digesting them, a brain pre-loaded with carnivorous behaviour software, and collections of other features that make it a typical hunter. On the other side of the arms race, antelopes are equivalently well equipped to eat plants and avoid being caught by predators. Long guts, complicated by blind alleys stuffed with cellulose-digesting bacteria, go with flat grinding teeth, go with brains pre-programmed to alarm and rapid escape, go with exquisitely camouflaged dappling of the pelt. These are two alternative ways of making a living. Neither is obviously better than the other, but either is better than an uneasy compromise: carnivorous guts combined with herbivorous teeth, say, or carnivorous pursuit instincts combined with herbivorous digestive enzymes.
Yet again, it is tempting to speak of the 'whole cheetah' or the 'whole antelope' as being selected 'as a unit'. Tempting, but superficial. Also lazy. It requires some extra thinking work to see what is really going on. Genes that programme the development of carnivorous guts flourish in a
genetic climate that is already dominated by genes programming carnivorous brains. And vice versa. Genes that programme defensive camouflage flourish in a genetic climate that is already dominated by genes programming herbivorous teeth. And vice versa. There are lots and lots of ways of making a living. To mention only a few mammal examples, there is the cheetah way, the impala way, the mole way, the baboon way, the koala way. There is no need to say that one way is better than any other. All of them work. What is bad is to be caught with half your adaptations aimed at one way of life, half aimed at another.
This kind of argument is best expressed at the level of the separate genes. At each genetic locus, the gene most likely to be favoured is the one that is compatible with the genetic climate afforded by the others, the one
that survives in that climate through repeated generations. Since this applies to each one of the genes that constitute the climate - since every gene is potentially part of the climate of every other - the result is that a species gene pool tends to coalesce into a gang of mutually compatible partners. Sorry to go on about this, but some of my respected colleagues refuse to get the point, obstinately insisting that the 'individual' is the 'true' unit of natural selection!
More widely, the environment in which a gene has to survive includes the other species with which it comes into contact. The DNA of any one species doesn't literally come into direct contact with the DNA molecules of its predators, competitors or mutualistic partners. 'Climate' has to be understood less intimately than when the arena of gene cooperation is
the interior of cells, as it is for genes within one species. In the larger arena, it is the consequences of genes in other species - their 'phenotypic effects' - that constitute an important part of the environment in which the natural selection of genes within neighbouring species goes on. A rainforest is a special kind of environment, fashioned and defined by the plants and animals that live in it. Every one of the species in a tropical rainforest consists of a gene pool, isolated from all other gene pools as far as sexual mixing is concerned, but in contact with their bodily effects.
Within each of those separate gene pools, natural selection favours those genes that cooperate within their own gene pool, as we have seen. But it also favours those genes that are good at surviving alongside the consequences of the other gene pools in the rainforest - the trees, vines, monkeys, dung beetles, wood lice and soil bacteria. In the long run this may make the whole forest look like a single harmonious whole, with each unit pulling for the benefit of all, every tree and every soil mite, even every predator and every parasite, playing its part in one big, happy family. Once again, this is a tempting way of looking at it. Once again, it is lazy - bad poetic science. A much truer vision, still poetic science but (it is the purpose of this chapter to persuade you) good poetic science, sees the forest as an anarchistic federation of selfish genes, each selected as being good at surviving within its own gene pool against the background of the environment provided by all the others.
Yes, there is a wishy-washy sense in which organisms in a rainforest perform a valuable service for other species, and even for the maintenance of the whole forest community. Certainly, if you removed all the soil bacteria, the consequences for the trees and ultimately for most of the life of the forest, would be dire. But that is not why the soil bacteria are there. Yes, of course they do break down the dead leaves, dead animals and manure into compost which is useful for the continued prosperity of the whole forest. But they aren't doing it for the sake of making compost. They are using the dead leaves and dead animals as food for themselves, for the good of the genes that programme their compost-making activities. It is an incidental consequence of this self- interested activity that the soil is improved from the point of view of the plants, and therefore the herbivores that eat them, and therefore the carnivores that eat the herbivores. Species in a rainforest community flourish in the presence of the other species in that community because the community is the environment in which their ancestors survived. Maybe there are plants that flourish in the absence of a rich culture of soil bacteria, but those are not the plants we find in a rainforest. We are more likely to find them in a desert.
This is the right way to handle the temptation of 'Gaia': the overrated romantic fancy of the whole world as an organism; of each species doing
its bit for the welfare of the whole; of bacteria, for example, working to improve the gas content of the earth's atmosphere for the good of all life. The most extreme example I know of this kind of bad poetic science comes from a famous and senior 'ecologist' (the quotation marks denote an activist for green politics, rather than a genuine scholar of the academic subject of ecology). It was reported to me by Professor John Maynard Smith, who was attending a conference sponsored by the Open University in Britain. The conversation turned to the mass extinction of the dinosaurs and whether this catastrophe was caused by a cometary collision. The bearded ecologist was in no doubt. 'Of course not,' he said decisively, Gaia would not have permitted it! '
Gaia was the Greek earth goddess whose name has been adopted by James Lovelock, an English atmospheric chemist and inventor, to personify his poetic notion that the whole planet should be regarded as a single living thing. All living creatures are Gaia's body parts and they work together as a well-adjusted thermostat, reacting to perturbations so as to preserve all life. Lovelock is avowedly embarrassed by those, like the ecologist I have just quoted, who take his idea right over the top. Gaia has become a cult, almost a religion, and Lovelock now understandably wants to distance himself from this. But some of his own early suggestions, when you think them through, are only slightly more realistic. He proposed, for instance, that bacteria produce methane gas because of the valuable role it plays in regulating the chemistry of the earth's atmosphere. The problem with this is that individual bacteria are asked to be nicer than natural selection can explain. The bacteria are supposed to produce methane beyond their own needs. They are expected to produce enough methane to benefit the planet in general. It is no good pleading that this is in their own long-term interests because if the planet goes extinct so will they. Natural selection is never aware of the long-term future. It is not aware of anything. Improvements come about not through foresight but by genes coming to outnumber their rivals in gene pools. Unfortunately, genes that make rebel bacteria sit back and enjoy the benefits of their rivals' altruistic production of methane are bound to prosper at the expense of the altruists. So the world will become relatively more full of selfish bacteria. This will continue even if, because of their selfishness, the total number of bacteria (and of everything else) is going down. It will continue even to the point of extinction. How should it not? There is no foresight.
If Lovelock were to retort that the bacteria produce methane as a by- product of something else that they do for their own good, and it is only incidentally useful for the world, I should agree wholeheartedly. But in that case the whole rhetoric of Gaia is superfluous and misleading. You don't need to talk about bacteria working for the good of anything other than their own short-term genetic good. We are left with the conclusion
that individuals work for Gaia only when it suits them to do so - so why bother to bring Gaia into the discussion? We are better off thinking about genes, which are the real, self-replicating units of natural selection, flourishing in an environment which includes the genetic climate furnished by the other genes. I am quite happy to generalize the notion of the genetic climate to include all the genes in the whole world. But that is not Gaia. Gaia falsely focuses attention on planetary life as a single unit. Planetary life is a shifting pattern of genetic weather.
Lovelock's main comrade-in-arms as a champion of Gaia is the American bacteriologist Lynn Margulis. Despite her pugnacious disposition, she places herself firmly on the gentle side of the continuum which I am attacking as bad poetic science. Here she is, writing with her son Dorion Sagan:
Next, the view of evolution as chronic bloody competition among individuals and species, a popular distortion of Darwin's notion of 'survival of the fittest,' dissolves before a new view of continual cooperation, strong interaction, and mutual dependence among life forms. Life did not take over the globe by combat, but by networking. Life forms multiplied and complexified by co-opting others, not just by killing them. Microcosmos: Four Billion Years
of Microbial Evolution (1987)
Margulis and Sagan are in a superficial sense not too far from being right here. But they are misled by bad poetic science into expressing it wrongly. As I emphasized at the beginning of this chapter, the opposition 'combat versus cooperation' is the wrong dichotomy to stress. There is fundamental conflict at the level of the genes. But since the
environments of genes are dominated by each other, cooperation and 'networking' arise automatically as a favoured manifestation of that conflict.
Where Lovelock is a student of the world's atmosphere, Margulis approaches from the other direction, as a specialist in bacteria. She rightly grants bacteria centre stage among life forms on our planet. At the level of biochemistry, there is a range of fundamental ways of making a living. These are practised by one or another kind of bacteria. One of these basic life recipes has been adopted by eukaryotes (that's everybody except bacteria), and we get it from bacteria. Margulis has successfully argued over many years that most of our biochemistry is carried out for us by what were once free bacteria now living within our cells. Here's another quotation from the same book by Margulis and Sagan.
Bacteria, by contrast, exhibit a far wider range of metabolic variations than eukaryotes.
