Review of Consilience
When E.O. Wilson was a small boy in Florida, he put one eye out on a fish. It was a spiney Caribbean minnow and he managed to swing it into his right eye while trying to unhook it. The way he tells the story in his autobiography is entirely characteristic: "The pain was excruciating and I suffered for hours. But, being anxious to stay outdoors, I didn't complain very much. I continued fishing." The pain from his eye went away after a few days, but a couple of months later he developed a traumatic cataract in that eye and had to have the lens removed in a barbarous operation, held down on a table by grown men and anaesthetised by ether dripping on his face.
The story encapsulates a great deal about the man. There is the extraordinary physical and mental toughness necessary to keep fishing under such circumstances, coupled with a deeply romantic conviction that it is better, perhaps less lonely, to be fishing on the beach than to return to human company. In passing, he lets us know that his parents were splitting up that summer. The outcome of the accident was also important. It meant that when his boyish interest in hunting animals began to tip over into a naturalists interest in finding and classifying them. He could not study birds. His hearing had always been too poor to distinguish their high notes; and now his eyesight would not let him spot them. He turned his fierce monocular vision onto insects instead, with huge success. By the age of 26 he was an associate professor at Harvard (younger even than James Watson, the co-discoverer of DNA, whom he loathed) and by 1971, when he was 52, he bestrode the world of ants like a colossus. In that year he published an enormous study synthesising everything important that was known about social insects.
Hardworking, ambitious, and filled with a peculiarly American romanticism, he decided to follow this feat with a synthesis of everything that was known about social behaviour in animals, including man. He was one of the first people to appreciate the significance of the Oxford zoologist W.D. Hamilton's theory of kin selection when the original, impenetrably mathematical, papers were published in 1964. These solved one particular problem of interest to entomologists: why worker ants and bees are always female. But they also laid out a solution to the general problem of how altruistic behaviour might spread in a Darwinian world. In a later paper, Hamilton also introduced the figure of the selfish gene, calculating how to spread itself; but he is a man of extraordinary diffidence and the fame of the phrase went elsewhere.
Wilson's 1975 book Sociobiology was the first really ambitious attempt to work out in public the Darwinian basis of social behaviour. As well as Hamilton's theories of kin selection, he used the game-theoretical approaches of Robert Trivers, who came up with the idea of "Reciprocal Altruism" - a formalisation of the idea that if you scratch my back, and I scratch yours, we both will benefit. These were the strong foundations of sociobiology, which are now commonplace in any discussion of human nature. But his book also epitomised the weaknesses of this style of reasoning. None of them are fatal but all have been very damaging.
There was the profoundly confusing use of "selfish". For example, he writes: "the allele [gene] spreads through the population via the increased success of its collateral relatives" -- which is a nice succinct statement of the central idea. Or it would be a nice succinct statement if he hadn't actually written: "The allele is always 'selfish'; it spreads through the population by promoting itself via the increased success of its collateral relatives." Talking in this context of "selfish" entities "promoting themselves" turns a mathematical insight into a piece of dodgy anthropomorphism. It is as if Kepler had discovered his three laws and then explained that planets would rather travel elliptically than go round in the same boring old circles.
The second characteristic vice was the wild use of animal analogy: even the second, shortened edition of Sociobiology, reissued after considerable criticism, contains the information that overcrowded laboratory rats displayed "hypertensive behaviour … hypersexuality and homosexuality and engaged in cannibalism. … such behaviour has its close parallels in certain of the more dreadful aspects of human behaviour. There are some clear similarities, for example, between the social life of Calhoun's rats and that of people in concentration and prisoner of war camps." Except that there aren't. He seems to have confused two different sorts of camp.
But the third and deepest weakness was its inability to deal with the things that make us distinctively human. One has to be careful here because such criticism is generally taken as denying that we are animals, with fairly well-defined animal natures; or else as an assertion that we have immortal souls which of course have nothing to do with Darwinism. This last, was, incidentally, the position of Alfred Russel Wallace, the co-discoverer of Darwin's theory.
That part of Wallace's beliefs has been abandoned. Atheists and materialists see no difficulty in principle with the idea that all our human faculties, from the ridiculous to the sublime, should have emerged in the same way as out animal ones. There has even been some progress made in working out the application of these principles by evolutionary psychology, the successor of sociobiology. But the matter remains hideously complicated, because culture works to different ends and on a different time scale to genetic evolution.
What we know about the ways that culture works has been discovered by historians, philosophers, anthropologists and all the other people whom sociobiology was going to put out of business. The intellectual history of the last twenty years is a two way traffic: at the same time as pop Darwinism has spread through the humanities as the explanation for everything, a certain understanding of the complexities of what they are trying to explain has been travelling in the other direction. The application of Darwinian principles to human behaviour has turned out to be, like the study of consciousness, one of those interesting areas where you can't do the science without getting the philosophy straight as well.
Consilience, his latest book, is an attempt to get everything straight. The epigraph comes from Francis Bacon: "Thus I made as it were a small globe of the intellectual world, as truly and faithfully as I could discover." And Bacon is one of his heroes. But here is the first doubts about the ambition of his project creep in. He writes that Bacon was born in 1561, and "during his lifetime, England, ruled successively by Elizabeth I and James I, passed tumultuously from a feudal society to a nation-state"; and one thinks, if he's wrong about that, why should he be right about anything else?
It's a trivial error, but of a significant sort. Difficulties of foreshortening and perspective are unavoidable in any attempt to place us in the universe, which is what Wilson is trying to do. But they are the besetting vice of Whig history, which is what he is writing here, even though he is vision of the world is not nearly as one-eyed as once it was. In fact, it is astonishing how far he has now retreated from the original territorial claims of sociobiology, that all the humanities would have to be melted down and reconstituted on a sound base of population biology. The joke is that his followers have gone ahead regardless: it is now taken for granted in smart glossy culture that the myth of DNA is well-established: a colour supplement piece on a jealous woman last year explained that "Having browsed 'The Selfish Gene', she is convinced that the only reason a man gets out of bed in the morning is to couple with creatures in high heels he sniffs at a bus stop."
This is funny and vivid, but it really isn't what Wilson (or Richard Dawkins) meant when they popularised these ideas in the Seventies. A large part of the popular success of sociobiology rests on the confusion, seldom resolved, between genes as necessary and sufficient explanations. Part of this is technical. A lot of the early successes of theoretical sociobiology were gained by using the word "gene" as if it meant simply "heritable variation" and ignoring the tricky questions of how many genes might be involved in producing the particular heritable variation we are interested in. But it is possible to be quite certain that genes are involved in something without having the faintest idea of which, how many, or where they are found. Wilson, after praising the "One Gene One Disease" or OGOD principle, which has led to the identification, he says, of 1,200 different diseases then goes on to admit that it is almost no use when dealing with the common diseases or conditions that interest us.
"For all its early success, however, the OGOD principle can be profoundly misleading when applied to human behavior. While it is true that a mutation in a single gene often causes a significant change in a trait, it does not at all follow that the gene determines the organ or process affected. Typically, many genes contribute to the prescription of each complex biological phenome-non. How many? For that kind of information it is necessary to turn from human beings to the house mouse, which, being a prime laboratory animal with a short life span, is genetically the best known of all the mammals. Even here knowledge is sketchy. In the mouse, genes contributing to the texture of the hairs and skin are known from no fewer than seventy-two chromosome sites. At least forty-one other genes have variants that cause defects in the organ of balance in the inner ear, resulting in abnormal head shaking and cir-cling behaviour."
And, as he goes on to say, mice are better understood genetically than humans.
"Whole organs and processes, as well as narrowly defined features within them, are commonly prescribed by ensembles of genes, each of which occupies a different array of positions on the chromosomes. The dif-ference in skin pigmentation between people of African and European ancestry is believed to be determined by three to six such 'polygenes.' The estimates for this and other such systems may be on the low side. In addition to the more potent genes, which are easier to detect, there can be many oth-ers that contribute small portions of the variation observed and thus remain undiscovered. [these factors supply] one reason that genes predisposing the development of chronic depression, manic-depressive syndrome, and other disorders have proven so elusive. Clinical depression in Ireland, for example, may have at least a partially different gene-based predisposition from clinical depression in Denmark."
This portion of Wilson's book is an excellent and conscientious corrective to the idea that the practical details will be trivial to fill in, now that we have a theoretical understanding of human genetics. He really does try to be a synthesiser as much as a populariser, and to give a fair account of what is known in the areas of science that interest him. His faith in the ultimate capacity of human beings to understand themselves and their world has the effect, common in firm believers, of making him unafraid of contrary evidence. He may be sure that sociology, for instance, will be able to give an account of the world "from the society down to the neuron", but he plays fair with the contrary evidence.
"Because human behavioural genetics is still in its infancy, there is a near-absence of direct links between particular genes and behaviour underlying universal cultural traits. The observed fit between theory and fact is based mostly on statistical correlation. … These shortcomings in behavioural genetics and development are conceptual, technical, and deep. But they are ultimately solvable. Unless new evidence commands otherwise, trust is wisely placed in the natural consilience of the disciplines now addressing the connection between heredity and culture, even if support for it is accumulating slowly and in bits and pieces."
This is so cautious and unfanatical it might be an Archbishop of Canterbury. Who could possibly be against the growth and interconnection of scientific knowledge "unless new evidence commands otherwise"? Something has gone dreadfully wrong with a revolution which within twenty years of its inception has gone from a promise to transform our understanding of everything to this kind of ringing banality, of which there are worse examples sprinkled throughout the book: "Science faces in ethics and religion its most interesting and humbling challenge while religion must somehow find the way to incorporate the discoveries of science in order to retain credibility."
Two possible explanations for this bathos present themselves. The obvious explanation is that sociobiology explains too little and has not remotely caught up with its ambitions. We may know little known about the workings of developmental genetics, and the ways that bodies are constructed by the interplay of genes and the environment, but we know far less about the construction of culture from genetically coded preferences. It's not really controversial - at least in our culture - that certain forms of culture fit human nature better than others (if they didn't, how could we talk coherently about human rights?). But the fit is baggy and variable, even at the level of body language. To quote the anthropologist Maurice Bloch. "There may well be foods that human beings find universally repulsive. But what has that got to do with food taboos?" Religions, after all, ban things which are good to eat. So far as I know, no religion has ever taught that you will go to hell for eating dogshit.
But Wilson's project is more ambitious than simply establishing that there are in built human predispositions towards certain cultural patterns. He wants to analyse the ways in which cultural and genetic change influence each other: again, the general principle is not controversial. There is no doubt that cultural developments can modify the environment in which genes are selected in ways which are both obvious and less so. The invention or taming of fire must have diminished the selection pressure for the sort of chewing apparatus that can cope with raw meat. Much later, the invention of agriculture makes possible a huge growth in population densities. So far, so ungenetic. But once people live together in villages and later cities, they can support a breeding population of all kinds of bugs and bacteria, which will in turn set up selection pressures for resistance to them at a genetic level. We may not notice these pressures, but when societies which have not been exposed to them meet new diseases, the effects are often devastating. Conversely, further cultural developments, like antibiotics and sewerage, could remove the pressures, a prospect which greatly worried the Victorians. There's not much new in the idea of gene-culture co-evolution.
Wilson, though he speaks of co-evolution, still tends to see the genes as more fundamental than culture in this process, rather than it being the interplay of equally important factors. In his sketch of the problem in Consilience he still sees cultural innovation as arising from genetic changes which in turn change "epigenetic rules", his term for such mental skills as colour vision or certain phobias.
One of the most interesting epigenetic rules discovered of sociobiology is that we appear to make innate distinctions between living and non-living things. The category of living, purposive beings, to whom teleological ascriptions are appropriate, seems to be hard-wired into us. This makes sense, both causally and teleologically: the animals who pay more attention to the intentions of predators than to trees and rocks will leave more descendants: the primates who understand their fellows' motivations best will also flourish. This insight is at the root of one of the main sociobiological theories of how human intelligence evolved: it holds in essence that we got smart in order to outsmart each other.
But what all the plausible candidates for epigenetic rules have in common is that they are ways of seeing or understanding the world, not rules of conduct. Seeing the world in particular ways may lead people to behave in certain ways: that is the insight that has led to the sociobiological study of our emotional constitution. The difficulty, however, lies in the fact that these epigenetic rules or biasses are deep and implicit. Explicit rules of conduct are always variable and culturally conditioned.
It is in the details of the link between genes and culture that his project really runs into trouble. For a start, he believes that both genetic and cultural evolution involves the same kind of mechanism and that there is a unit of cultural change, or meme: a kind of brain structure which will perform in cultural change the same function as a gene performs in biology. It's not a tremendous objection to this project to say that we don't know what such a thing would look like: the foundations of genetics were laid by people who didn't know what chemical genes were made of. The real problem, however, is that in biological evolution there is a fundamental distinction between genes and the bodies that transmit them or between replicators and interactors; and this distinction is nowhere to be found in culture. Everything that is proposed as a meme is much more like a body than a gene.
As John Maynard Smith, one of the founders of sociobiology, has pointed out, memes fails both the tests one would like an explanatory theory to pass: it cannot tell us what might happen, and, more importantly, it doesn't tell us what can't happen. But without such a working theory, Wilson rapidly falls back on platitudes when he discusses culture; and when he is not explaining too little, he finds himself explaining too much.
The peculiar beauty and power of natural selection is the way in which it allows causal explanations to act as if they were teleological. In Richard Dawkins' phrase, it acts as the Blind Watchmaker. Huge advances in understanding result from this convertibility. Two particularly elegant examples are the luminescent bellies of some deep-sea fishes (which makes them less visible against the paler water above) and the development of morning sickness in humans, which seems to make the mother unable to bear possibly toxic foods at the time in pregnancy when the baby is most developmentally sensitive. Natural selection is not the only way to bind these explanations together. There is also the tradition of enlightenment Christianity, exemplified by Archdeacon Paley, who argued that the exquisite fit of cause and purpose in the complexity of the world proved the existence of an over-riding purpose, which was God.
Paley was wrong in the conclusion that he drew; but finishing Wilson's book, the melancholy thought occurs to me that he might have been wrong in thinking the premises well-established, too. Paleyites, whether Christian like the Archdeacon himself, or atheistic like Wilson, agree that the conjunction of causal and teleological explanations is the most interesting thing about biology. They look at the world and see a unifying principle behind it. This assumption of fundamental rationality is what links the Enlightenment to the form of Christianity from which it grew and against which it rebelled.
None the less, it is only an assumption. Wilson himself has a really profound love of nature: think of him as a the small boy fishing all day, even though his eye was on fire. He has a profound love of understanding. But nowhere in all his labours at summarising what is known about the sort of animals we are, does he manage to synthesise these two great passions convincingly. And if a unifying principle is nowhere to be found in humans, how can we be so confident of its existence outside us?