Beautiful metaphors, bad science

The reaction against Darwin in this century has taken two religious forms. One is to reject his theory. The other is to ransack it for objects of worship. It seems almost impossible to hold clearly in our minds the idea that we may not matter-as if it were logically impossible that anything capable of thinking and feeling unimportant should in fact be unimportant. The clearest illustration of this is to be found in the work of Richard Dawkins.

The Selfish Gene was a book about genes, but almost everyone read it as a book about people. What it said about people is not scientific. It is a religious message: we are unimportant wisps of matter, at the mercy of an omnipotent power. This has no necessary connection to biology or genetics. One of the finest expressions of this message is found in a writer whose style could not be further from Richard Dawkins': in Russell Hoban's great mud-soaked post-apocalyptic novel Riddley Walker, civilisation, technology and even English have all been smashed to bits:

You know Riddley theres something in us it don't have no name... Its looking out thru our eye hoals. You dont take no noatis of it only some times. Say you get woak up suddn in the middl of the nite. 1 minim youre a sleap and the nex youre on your feet with a spear in your han. Wel it wernt you put that spear in your han. It puts us on like we put on our cloes. Some times we dont fit. Some times it cant fynd the arm hoals and it tears us a part...



Hoban does not name the something that controls us. Dawkins calls it DNA. "DNA neither knows nor cares. DNA just is. And we dance to its music," he wrote in his worst book, River Out of Eden. "We are survival machines... blindly programmed to preserve the selfish molecules known as genes."

It may seem unfair to concentrate on this sort of rhetoric against the bracingly clear scientific exposition that Dawkins does so well. But it is, I think, the rhetoric and not the science that has made him famous. Dawkins did not reach his present eminence as God's own atheist by preaching evolution. There are plenty of practising Christian Darwinians, such as Sam Berry, professor of genetics at University College London. What makes Dawkins unique is not merely the vigour of his style but the fact that his metaphors, like Virginia Woolf's imagination, are fitted with an accelerator but no brakes.

With the exception of The Extended Phenotype, his books have been arranged around a single, dazzling phrase: The Blind Watchmaker, River Out of Eden, The Selfish Gene. Some of these have entered the language, and deservedly so. Climbing Mount Improbable, his latest book, is unlikely to do so. It is a distillation of old themes: there is a bit of computer modelling; a bit of God-bashing (though not until page 209); and a great deal about the way in which the accumulation of very small changes can lead to very large ones. This is the central argument about the origin of species.

Evolution on a small scale has been known and accepted ever since man began to breed plants and animals. No one doubts that the ancestors of dachshunds were like wolves, and that the change from wolf to dachshund was made by choosing the most dachshund-like dogs in each generation, and then choosing the most dachshund-like in their progeny.

Darwin's first great insight was that the process which differentiates dachshunds from wolves over a few thousand generations could account for all differences between living things if prolonged over the immensities of time. This is the first thing creationists deny; their denial seems credible enough at first when you try to imagine the common ancestor of a cow and a lemur, let alone of a fruit fly and a banana. One of Dawkins' gifts is to show how unreliable a guide common sense must be when judging questions of probability over the generations.

In Climbing Mount Improbable, the central metaphor is of an escarpment of probability. At its summit stands something hugely complicated, like the human eye. The approach straight up the cliff from where we are seems quite improbable without supernatural assistance. Yet if you go around the back of the escarpment, where the slope is gentle, it is possible to reach enormous heights without wings and almost without exertion. All you need is time; and time has been available for evolution in quantities to drown the human imagination. With the aid of computer simulations, Dawkins reckons that the eye could have evolved 40 different times quite separately, and gives cogent arguments for supposing how it might have been done, with each successive refinement slightly improving its bearer's chances of survival.

The same sort of point was made in The Blind Watchmaker, by his invention of Biomorphs-a computer simulation. Biomorphs breed on screen from a tiny and plain stick-insect stock into extraordinary branching shapes like crawling candelabra. They are huge fun. They make very clearly and quite realistically the point they were designed to make: that the ratcheting accumulation of small changes can amount over time to huge irreversible changes. But as a device for explaining evolution, they suffer from the drawback that they require a God. Someone must sit at the keyboard and select among them.

But Darwin's second great insight, drawn from economics, was that evolution doesn't need a God. The complexities of an ecological system, like those of a market, will always supply selective pressures. There will always be competition for resources-be they food, territory, or sexual success; there will always be competition not to become something else's resources. Providing that some of the elements of success in these competitions are heritable, these pressures will have selective effects. Order, of a sort, emerges from markets: order and complexity emerge from evolution. Neither was designed in, or designed towards. Neither entails any idea of progress. In evolutionary terms, the E. coli bacteria in our entrails are far more successful than we are: more widespread, more stable, less vulnerable to catastrophe.

Curiously enough, a distrust and misunderstanding of evolution is more common among atheists and agnostics than among the happily religious. I suspect that the main reason is that the cruelty and contingency of the world are easier to look at straight if you believe there is another dimension in which they somehow balance out. If you know that God loves you personally, you are unlikely to find the success of E. coli in his world too upsetting.

Dawkins is such a vigorous and inexhaustible controversialist against religion that it is easy to overlook the extent to which he is controversial even among scientists. These controversies are difficult to represent, partly for technical reasons; and partly because the two sides, like all true believers, suppose that only those who agree with them deserve the name of Darwinians. Both sides agree that species arise as a result of a multiplication of small changes in DNA. They disagree over whether the species, once they have arisen, exist as entities. For Dawkins, the unit of selection is the individual gene, since it is only changes in the gene which are preserved down generations. From this perspective neither individuals nor species exist in any interesting way. (It puts us on like we put on our cloes. Some times we dont fit.)

This is an odd conclusion to reach. Part of the problem is that the gene can be defined as a unit of inheritance; and from this definition, all the desired qualities follow. But a gene can also be defined as a sequence of DNA; and although these two definitions overlap, they are not the same: the distinction between them separates science from metaphysics.

When you look for the unit of heredity in a string of DNA, things grow extremely complicated. Talk of the DNA "code" suggests that each fragment has a rigid meaning; but genes do not speak a code susceptible of machine translation, they speak something more like a language. A particular sequence of DNA can have differing effects on the organism that contains it, depending on where in the chromosome it is found, when the sequence is being expressed, and so on. It is like a syllable, the meaning of which varies according to the word, the context and the language.

Dawkins admits this, in all his books, and in conversation. But the importance of this admission tends to be lost in all the talk about "genes" as if they were discrete, dependable entities. Indeed, his commitment to gene-level selectionism is very close to describing language only at the level of syllables. His opponents in this dispute, Stephen Jay Gould and Niles Eldredge among others, see evolution as taking place in interacting hierarchies, so that natural selection works on genes, on bodies and on species; and all these differing levels of selection interact with each other. At all levels this selection is blind: genes are not selected for their usefulness to individuals; nor individuals for their usefulness to species. But the different levels do affect and constrain one another. For example, new species often arise as a result of geographical isolation, which is difficult to understand at the level of DNA, yet is a perfectly clear and simple idea if you are looking higher up the hierarchy.

None the less, Dawkins is entirely serious in his commitment to the primacy of DNA. The houses built by caddis-flies he sees as an effect of caddis genes: given the degree to which insect behaviour is hard-wired, this seems reasonable. Then, he goes on to say, a beaver dam is an effect that beaver DNA has on the landscape. "Natural selection consists of molecules which have a causal influence on something; which vary; and which are inherited." This concentration on molecules, to the extent of building them into the definition of natural selection, is odd. Why should molecules be regarded as somehow more essential-in fact, more worthy of religious consideration-than their constituent atoms, than individuals, or species? His answer is that molecules are unchanging. In the beginning was the DNA and it abideth forever. Individuals, species, whole phyla appear and vanish. Only the DNA remains what it was.

Yet DNA in itself is not very interesting. The effects, the meanings, if you like, of DNA are not intrinsic to it. Information is an arbitrary property. It is not contained in DNA as a substance: it appears only when DNA is found in the immensely complicated chemical bath of a cell. To identify DNA with some kind of platonic essence of life-see River Out of Eden-is another example of the superimposition of two definitions of gene. It makes for beautiful metaphors and lousy science.

This is a lot of argument about a small thing. But Dawkins is a popular phenomenon. He is the Darwinian everyone knows; and his metaphors and cast of mind have been very influential outside his field, especially among the computerate. His new Oxford Chair of the Public Understanding of Science was endowed for him by Charles Simonyi, a Hungarian who made his fortune at Microsoft. Daniel Dennett, a philosopher whose career has been built on the prospect of artificial intelligence, and who shares Dawkins' talent for lucid destruction, is another fan. So is Douglas Adams.

To fit alongside the gene as a unit of heredity, Dawkins has proposed the "meme" as a unit of culture: memes are ideas, tunes, religious beliefs-anything found in our minds as opposed to our brains. This idea has spread round the world as a, well, meme should. But he himself has partially disowned it, under a storm of criticism.

Memes have something in common with his perspective on DNA. Both ways of looking at the world tend finally to eliminate the individual. First my body becomes no more than a caddis house: a shelter for a swarm of genes; and then my mind provides shelter and sustenance for a swarm of memes. It takes a remarkable ego to contemplate a self like that with equanimity. This hostility to the individual may explain the ferocity of his attacks on religion. I take as much pleasure as the next man in watching a bishop humiliated and confused. But there is something unnerving about the relish Dawkins brings to the task. Many scientists share his faith in reason without sharing-or at least showing-his scorn for outsiders. Dawkins' peculiar fury is directed at the argument from design, but this argument plays no part in most people's religious imaginations. What they seem to work from instead is an intuition of personal significance. They do not ask that the whole universe be designed so they can fit into it; only that their own experience should be compatible with their existence as significant beings. This is a fairly humble, commonsensical idea of the individual. It would be easy enough to construct a just-so story of how it might have evolved among social primates. That is not the sort of small, empirical step that appeals to Dawkins. For him it seems clear that since natural selection has produced Richard Dawkins, there is no further need for the hypothesis of God.