Why evolution is more ruthless than any human engineer.

This article is an excerpt from Flights of Fancy: Defying Gravity by Design and Evolution (2021), where Richard Dawkins explores how nature and human innovation have mastered the challenge of flight

When we face a problem, like how to avoid stalling in a flying machine, it’s convenient to think, “How would I set about solving the problem?” In the case of man-made planes, design engineers really do think like that. They notice a problem. They imagine possible solutions to it, such as wing slats. They sketch ideas on drawing boards, maybe they get together in brainstorming sessions with a whiteboard, or a computer with graphics software. Perhaps they build prototypes, or scale models which they test in a wind tunnel. And the solution that emerges is finally put into production. The whole research and development process (R & D) takes only a few years or even less.

With animals the process is different, and much slower. The R & D, if you can call it that, takes many generations spread over millions of years. No thought goes into it, no clever ideas, no deliberate ingenuity, no creative inventiveness. There are no drawing boards, no brainstorming engineers, no wind tunnels to test prototypes or scaled-down models. All that happens is that some individuals in the population just happen, by random genetic luck (mutation and sexual shuffling of genes), to be a little bit better than average at flying. Maybe a mutant gene gives a falcon a slight edge in speed. Individual falcons bearing this gene are a little bit more likely to catch prey. Or perhaps a mutant starling is a bit more manoeuvrable than rivals in the flock, which makes all the difference between dodging a predator and being eaten. When a starling is eaten because of a “slow flying gene” the gene is eaten too, and is not passed on to the next generation. Or another genetic type might be a little bit less likely than others to stall, because of a subtle difference in wing shape. So they are a bit more likely to survive and therefore reproduce, passing on the genes that made them slightly better flyers than their contemporaries. Slowly slowly, gradually gradually, generation by generation, the good-at-flying genes become more numerous in the population. The bad-at-flying genes become less numerous, as animals that possess them are a little bit more likely to die or fail to reproduce.

The same thing is going on all the time with lots of different genes in the population, each influencing flying in its own way. So, after lots of generations, after millions of years of accumulating good flying genes in the population, what do we see? We see a population of very good flyers. Good in all sorts of subtle details including anti-stalling devices; including sensitive nervous control of the muscles that adjust wing shape to every little detail of wind eddies and up-draughts; including more efficient wing muscles that get just that little bit less tired. Wings and tails have evolved to become just the right shape and size, just right in every detail, just right as if a human engineer had perfected the design on a drawing board and tested it in a wind tunnel. The end products of human design and evolutionary design are both so good, both fly so well, that we find it convenient to forget how different are the two processes of improvement. This forgetting shows itself in the language we use. You will have noticed in this book that I have used a kind of short-hand language. I’ve written as though birds and bats, pterosaurs and insects set about solving the problems of flight in the same kind of way as human engineers: as though birds themselves solve the problems rather than Darwinian natural selection. The short-hand is convenient partly because it really is short: it takes fewer words than spelling out how natural selection works every time. But it’s also convenient because you and I are human and we know as humans what it’s like to be faced with a problem, and what it’s like to imagine solutions to the problem.

It’s tempting to suggest that the similarity between evolution and human design goes further. We might suspect that engineers’ new ideas, say for an anti-stalling device, are rather like genetic mutations. These “idea mutations” are then subject to something like natural selection. An idea may die immediately, when the inventor quickly realises it won’t work. Or it may die as a prototype which is seen to fail in a preliminary test, or perhaps in a computer simulation or a wind tunnel, and is consequently rejected. Failing in a wind tunnel is relatively harmless: nobody dies. Natural selection of animal flyers is more cruel: failure really does mean death. Not necessarily a fatal crash, but perhaps the defective design is slower to escape a predator. Or less adept at catching prey on the wing, which increases the chance of starvation. Evolution doesn’t have a gentle substitute for death, like trial by wind tunnel. Failure really does mean failure: death or at least failure to reproduce.

Richard Dawkins

---

To support The Poetry of Reality with Richard Dawkins, shop the latest merch on: thepoetryofreality-shop.fourthwall.com