The Selfish Gene

The Selfish Gene - Richard Dawkins

I bought this book because I'm fascinated by the idea of evolution - I mean, at first glance it appears utterly preposterous, right? So I wanted to take a closer look. I started by reading The Origin of Species (Darwin, of course). That was well worth-while but clearly his theory was wrong, for many reasons, most of which are given in the book, by Darwin himself. The key problem for Darwin was that whilst he knew there had to be some kind of inheritance of characteristics, he had no idea what the mechanism was. Genetics came to the rescue of evolutionary theories by providing such a mechanism. OK - so now I had to find out what a modern theory of evolution looked like. I read Niles Eldredge's Re-inventing Darwin, which turned out to be a book making a counter-case to ideas proposed by Dawkins. I found it pretty convincing, but then I hadn't read any Dawkins. It didn't really provide what I was looking for, anyway; the book doesn't set out a complete theory of evolution, instead it takes an academic debate into the popular science arena. Much time goes by and I end up with Gould's last collection of essays. It turns out that he was strongly opposed to many of Dawkins' ideas, too - but that book only gave a sketch of a theory, in two essays. Time to actually read some Dawkins, then and give the guy a fair hearing. I picked up his most famous book and found that I'd made another blunder; The Selfish Gene is about the evolution of altruism! Basically it's about animal behaviour. For goodness' sake; I might have to start reading the blurb before I buy books on evolution! That said, the underpinnings for a general theory of evolution is laid out in this book. Phew!

So Dawkins' basic thesis is that altruistic behaviour amongst animals can be understood as arising from natural selection at the level of individual genes or gene-complexes (I'm just gonna say "genes" henceforth). You share a proportion of your genes with your relatives - hence the probability of survival of those genes could be increased by altruistic behaviour towards them. Dawkins makes a convincing case for this, but along the way he makes some surprising blunders and wanders down some unexpected roads. (Unexpected if you don't know anything about Dawkins, that is.) For example, he has a bash at the concept of the Virgin Birth (justifiably from a theological perspective but what's it doing in a book about animal behaviour?) Also at Catholicism for prohibiting contraception - I agree, but what's it doing in a book about...and eventually an outright attack on religion in general that is only vaguely linked to the topic at hand.

The blunders? Well, Dawkins never properly defines some of his terms. There are neo-Darwinians, Darwinians and group-selectionists; I think the former two are supposed to be the same but the latter is meant to be different. Then what is a group-selectionist? The idea is that groups of animals within a species are selected for survival or extinction. Dawkins spends an age bashing this idea - then gives examples of it actually occurring! I think he intended merely to attack the idea that the individuals of a group or species act in the best interests of the species, not the possibility that groups are sometimes selected for or against. (Mainly against, in my view - groups get wiped out or they don't. Getting wiped out is much more important in evolution than carrying on much the same as before, I suspect.) Another blunder refers to "man" (grrrr! - but Dawkins does show regret for using sexist pronouns to describe homo sapiens sapiens, in a preface to this edition) "out-witting" his genes by using contraception. According to his own theory it is logically impossible for genes to be outwitted - the influence of one gene might be stronger than that of another, however. Dawkins talks about copulation, but he has not really thought carefully enough - there's the urge to have an orgasm, the urge to copulate and the urge to become a parent. Contraception is irrelevant to the discussion when the species can and does masturbate - and masturbation is not the sole preserve of humans, is it, Rover? No - now stop that! Rover - stop right now! OK - you're going outside!


The book is full of over-estimates of the uniqueness of the human species, such as this one. The chapter on memes is an ill-considered disaster, though there seems to be a good idea at the bottom of it.

One key concept is that of the evolutionarily stable strategy (ESS). This idea came out of the application of game theory to animal behaviour. The idea is that over time one "strategy" regarding some important aspect of animal life, e.g. defending a territory, comes to dominate within a population at the expense of all others. More-over it becomes impossible for any other strategy to "invade" and come to dominate the behaviour of the population. The thing that struck me about this concept is that it supports a key contention of Gould and Eldredge - that under normal circumstances natural selection tends to preserve species unchanged. This idea arose because the fossil record does not show gradual change amongst species (as Darwin predicted) - instead the abrupt ceasing of species and equally abrupt arrival of new ones. Gould and Eldredge put forward the theory of Punctuated Equilibrium to explain this: Species are stable until a major disruption to the eco-system occurs - then some species go extinct and others evolve rapidly (on a geological time-scale). Dawkins in the book gives support to the PE idea but seems to have mis-understood it; he suggests a "little bit" of evolution occurs when the equilibrium is disrupted - in fact the theory is that lots of evolution occurs - this is when speciation happens and not at any other time. In a footnote to this edition Dawkins withdraws his support for this idea.

I give the book four stars because I enjoyed the process of reading it. Thinking about evolution is something I enjoy, because it is so awkward to deal with; it has to work from molecular bio-chemistry right up to geological history - and the thing it has to explain is The Origin of Species! How species arise is a tremendously difficult problem. Some of the ideas presented here look as if they might be part of a genuine (i.e. predictive) theory of evolution - others are obviously wrong. Some things are badly explained, or contradict earlier statements. Other things are irrelevant. Hence the book itself is only worth two stars.

Where next? Dawkins' The Extended Phenotype and Gould's The Structure of Evolutionary Theory.

*************Notes made whilst reading are below*********************

It would seem that Dawkins has been oft criticised for personifying genes - saying that they have motivations and these motivations are selfish. He answers that this is merely metaphor and neither he nor anybody with any wit genuinely believes genes are conscious of anything and that this is made clear in the text. Fair enough. My chemistry teacher in school used to do the same thing, talking about electrons "wanting" to be in the least energy state. He made the same warning about metaphor, too. The question is, does this metaphor clarify or confuse matters? I rather think it can cause errors. Discussions of evolution are full of Just-So stories about how attributes and behaviours evolved. Personification can lead to just this kind of useless hypothesis masquerading as theory.

Dawkins makes a mistake when he says humans have evolved to a point where they can subvert their genes e.g. every time they use contraception. This comes after discussing how a runt might be serving the survival of its genes by allowing itself to be eaten by its siblings. According Dawkins' theory, use of contraception might allow parents to focus time effort and money on giving a smaller number of children the best care. This might be a better strategy than having many children, in a society where there is good hygiene, health care and food supply. Hence use of contraception may not be a "subversion" of genetic "will" at all. Also, the gene complexes that allow invention and use of contraception are almost certainly at least overlapping with those that cause the development of a "big brain" which just about everybody suggests is a positive survival trait even though it is the cause of many difficult births. It is simply not possible to "subvert" one's genes, according to Dawkins' theory. Even a childless suicide hasn't; he's just a victim of "bad" (anti-survival) gene complexes.

Dawkins adds notes to the text for this anniversary edition, discussing criticisms of certain passages and comments about things that are now outdated or he has altered his view on. He and I are both amused by the discussion of 1976 computing power.

Dawkins talks about brains = computers and talks about chess programs (also a dated discussion). He suggests that computers that can beat grand masters would teach us humility; rather the opposite I think. He also talks about how the programs are personified by players and chess writers, hinting that the program has a rudimentary conciousness. But apply Occam's Razor: 1. Computer is conscious; 1 additional entity.
2. Humans are "projecting"; no additional entities. Therefore it is more likely that chess programs are not conscious.

There is something of a tacit assumption behind Dawkins' theory; evolution occurs by natural selection. This is not obvious or proven and therefore should not be treated as axiomatic. The fossil record does not back the idea of continuous evolution, rather the opposite. This is why Eldredge and Gould developed the Punctuated Equilibrium Theory. At one point Dawkins mentions this theory and backs it, but seems to have mis-understood it slightly; he suggests that "a little bit" of evolution occurs when something alters the species' environment. The theory is actually that natural selection tends to preserve species unchanged until a radical change in environment occurs and then radical evolution (speciation) occurs on a time scale short in comparison to the fossil record. In a footnote Dawkins suggests that he now believes Punctuated Equilibrium has been accorded too much significance. I think he now recognises that his theory and that of PE are not very compatible.

Early on Dawkins presents a theory of the origin of life - the primordial soup idea. He tells us that there are other theories and does not claim to be wedded to any particular one. He explains the theory but glosses over the two major weaknesses in it without pointing them out. In fact I've never seen these weaknesses pointed out, so I'm going to do it here: It's known that amino-acids can occur naturally in a number of environments (including interplanetary space). Amino-acids are the building blocks of proteins and proteins are the building blocks of many components of living organisms. It is not known how to get from naturally occurring amino-acids to proteins without the intervention of living organisms. This is the first thing that is glossed over. The second thing is cell membranes. If you have RNA already it can make proteins (but you need proteins in the first place to make RNA - see above) and cover itself in a protein shield. Many viruses are exactly this - RNA strands surrounded by proteins. From here the theory leaps straight to single-cell organisms without explaining how a giant molecule can spontaneously generate a cell membrane round itself, whilst floating in a small puddle. Any theory of the origin of life worth the name must tackle these two problems. I caught the end of a Radio 4 programme recently in which it was suggested (I think) that membranes could be created if the puddles underwent a freeze-thaw process that concentrates the large molecules in the puddle of water. I wish I had heard the whole item as it sounded like a major breakthrough.