Evolution and Arithmetic

Chapter 14: Instinct

← A Journey No Single Butterfly MakesSteering by the Sun →

Not Just More of the Same

Everything this book has looked at so far, however complicated, has belonged to one family of problem. DNA builds a protein. The protein folds into a shape, joins a structure, or switches another gene on, and somewhere down the line a body part comes into being: a working ear, a branching network of blood vessels. It's staggering, and this book has spent chapters showing how staggering. But at bottom it's one kind of task, repeated on an enormous scale: build the right physical thing, in the right place, out of the right materials.

Some of what makes migration work is exactly that familiar story in different clothes. A hormone is held back, and a monarch stops breeding and lives many times longer than its mother did. A bird's body reads the shortening days and responds by packing on fat, in some species doubling its own weight. And a migratory direction can be inherited like any other trait. The best-known example is the blackcap, a small European songbird. Until the 1950s blackcaps hardly ever wintered in Britain; most of those from Germany headed southwest, toward Spain. Then some began flying west-northwest instead, to Britain. Researchers bred these birds in captivity and found that their chicks, never having flown anywhere, chose the same new direction. The new route had evolved, and spread, within about thirty years.

That deserves to be taken at face value: the direction setting, the dial that says "lean west-northwest, not southwest," can change quickly. But a dial is only useful if it's connected to something. The blackcap's new heading worked because it plugged into equipment the bird already had: a compass to hold the heading, a clock, the urge to leave at the right time, the fat to fly on. Turning the dial is the easy part.

Now look at the cuckoo chick again. It has to sense, by touch alone, in a dark nest, that something else is in there with it. It has to work its body underneath that object without seeing it, lift it, brace, and heave it over the rim, in a particular sequence, using muscles it has never used before. Nobody shows it how, and it doesn't have long to learn: the urge comes on within hours of hatching and fades after a few days, before the host's own chicks grow too big to move. It can try again when an egg rolls back in, but a chick that hasn't finished by then grows up beside a rival that takes part of its food. That isn't a dial. It's a program, carried out against the unpredictable layout of one particular nest, correctly, in cuckoo chick after cuckoo chick.

Or take the compass. A gene can set a value, the way a gene sets eye color. But something in the bird also has to take the position of the sun, correct it against a running internal clock, check the result against the turning night sky, and check that against the Earth's magnetic field: three different physical phenomena, from three different branches of physics, none of them anything to do with proteins or chemical bonds. DNA has never seen the sun. It has no way of knowing that stars exist, or that they seem to wheel around a fixed point overhead, or that the Earth behaves like a weak magnet. Yet generation after generation, it builds creatures that use all three, in combination, to reach a place they've never seen.

That's the shift this chapter asks you to notice. Building the right protein in the right place is hard enough. But a built-in system that arrives already matched to the sun, the stars, the Earth's magnetism, or the layout of a nest, and responds correctly on the first attempt, is a different kind of achievement. It's an internal system fitted to features of an outside world that nothing inside the organism could have sensed or checked in advance.

The standard answer is that nothing needed checking in advance. Given enough time and enough generations, blind variation can stumble onto a mechanism that happens to fit an unknown reality, not because anything was aiming at the sun or the stars, but because whatever happened to work survived, and whatever didn't, died out with the animals carrying it. That's a coherent answer, and this chapter takes it seriously. The question is whether the number of blind attempts it needs, and the time available to make them, still look reasonable once we see what has to be found.

Last revised 2026-11-18

← A Journey No Single Butterfly MakesSteering by the Sun →