THE SELFISH GENE is a book to read on paper, with a pencil or, better yet, a ballpoint pen in your hand for scribbling in the margins. It demands engagement in a way that few others can. It has an enormous amount to teach, and, if you have learned its lessons thoroughly, you will understand that it is all true, right down to the title — provided you understand that the “Selfish” doesn’t mean selfish, “Gene” doesn’t mean gene, and even the definite article is a bit dodgy.
That “Selfish” doesn’t mean selfish is a point made repeatedly by Richard Dawkins himself. “At no time does this evolution of the ‘co-operative gene’ violate the fundamental principle of the selfish gene,” he writes, before going on to suggest that the book could just as well have been called The Altruistic Gene. But, if it had been, no one would have taken any notice. The Selfish Gene is a shocking title and so an effective one. It is shocking because we are co-operative animals, who have evolved to take selfishness and manipulation seriously; but it also illuminates two really important aspects of natural selection.
The first is death. Genes are copied (they cannot, of course, copy themselves), and bodies are not. This is not a new point, but Dawkins made it more vividly than anyone since Tennyson, who wrote that nature was “So careful of the type. . . So careless of the single life.”
If your criterion of value is longevity, then genes are obviously more valuable and more important than their bearers. It is very easy to slip from there to the view that we only serve their purposes, or that, as Dawkins wrote, “DNA neither knows nor cares. DNA just is, and we dance to its tune.”
If it neither knows nor cares, how can it be said to be “selfish”? Dawkins’s answer is two-fold. It is “selfish” because it can cause “altruistic” behaviour in organisms. The gene that causes a worker bee to sting and die in defence of the queen means that the queen will live on to lay more eggs that share that gene. In that sense, the gene spreads “selfishly” at the expense of some of the bees that carry it. The mathematics of this were worked out in the 1960s, but it was Dawkins who popularised them and made them seem comprehensible and urgent to a huge readership.
THE second “selfishness” of genes is, I think, original to Dawkins. To quote my own book The Darwin Wars (Simon & Schuster, 1999), he uses it “to mean ‘having a distinct identity on which selective processes can act’. To be selfish in this sense implies nothing whatever about psychology, motivation, or even behaviour. It simply means to be visible to Darwinian selection. This is an extremely important quality and Dawkins’s originality consisted in being the first person to stress its importance in all sorts of contexts. If he’d thought of an original name for it as well, he might have saved a lot of confusion, not all of it other people’s.”
This form of “selfishness” is integral to the idea of a gene as a distinct, heritable cause of bodily form or behaviour. It is a theoretical concept that need have nothing to do with DNA. The word “gene” itself comes from the early 20th century, and the mathematical foundations of altruism were worked out in the 1960s, 15 years before a single gene was sequenced.
Dawkins does explain that the genes he talks about are mathematical abstractions. To a theoretical biologist, he says, “Genes are things you can count as the generations go by and it doesn’t matter what their molecular nature is . . . [To] put the point another way, The Selfish Gene is quite likely a valid account of life on other planets even if the genes on those other planets have no connection with DNA.”
In the epilogue to the 50th-anniversary edition, he writes: “Genomics is not genetics. It’s the study of what genes actually are,” and, immediately afterwards, he explains that geneticists, like him, are quite uninterested in “what genes actually are”. Instead, “Genetics is the study of heredity, of how inherited traits reassort and pass down the generations.”
The mere details of DNA are unimportant, in his view, compared with the patterns of Darwinian evolution. The maths is more important than the territory. This understanding of universal Darwinism was hugely fashionable in the early years of the century, articulated and developed by the philosopher Daniel Dennett and the circle around the literary agent whom he shared with Dawkins, John Brockman.
Central to this picture of evolution was the distinction that Dawkins makes between “replicators” and “vehicles”. “The gene is the unit in the sense of replicator. The organism is the unit in the sense of vehicle. Both are important. . . [T]hey represent two completely distinct kinds of unit and we shall be hopelessly confused unless we recognise the distinction.”
THIS gnostic divide, corresponding in some ways to the division between matter and spirit, or software and hardware, is central to the argument of The Selfish Gene and is also its most important mistake.
This was pointed out by the philosopher David Hull, whose book Science as a Process (University of Chicago Press, 1990) is an attempt, inspired by Dawkins, to analyse science itself as a process of Darwinian evolution whose product is knowledge. Hull used the term “interactor” in place of Dawkins’s “vehicle”, because it stresses the independent and unpredictable nature of an organism’s behaviour.
Natural selection, or Darwinian evolution, is a two-stage process, endlessly repeated: some things are copied, stuff happens, and, as a result, not all the same things are copied next time round, and so the population of things being copied changes over time. That is the “selection” part of natural selection.
Hull argues that replication and selection, or copying and winnowing, are roles, not ontological categories. Darwinian evolution requires replication and differential selection, but it does not require genes or anything like them.
This also is clear from Dawkins, if you read him far enough. When he discusses the origin of replicators and the likelihood that DNA itself evolved from RNA, he is talking about molecules that were, in his terminology, replicators that were also their own vehicles.
So, if you want to know what is wrong with The Selfish Gene, and with the fashions of thought which followed it, just read it carefully, and then read The Extended Phenotype. Neither, however, can explain how he produced an account of natural selection which denied all agency to the agents who actually do the selecting. It hardly needs adding that everything he writes about religion is nonsense, and that anything about “memes” is nonsense on stilts.