Evolution and Arithmetic

Chapter 13: Shape

← What the Robots Didn't Have to DoHundreds of Genes for a Face →

What the Field Can Say

You might reasonably assume that somewhere in developmental biology someone already has a working answer, and that this book just hasn't got round to explaining it. That isn't so, and it's better to say so directly. Nobody has anything close to a complete account of how hundreds of millions of cells, each aware of little beyond its immediate neighbors, reliably produce a specific curving structure like the rim of an ear. This isn't a gap the book is choosing to skip. It's an acknowledged frontier in the field itself.

It's also a young field. The idea that DNA has much to do with shape at all, beyond specifying proteins, is only a few decades old, and the harder question of how many separately shaped parts get organized into one coherent body is younger still. Less has been mapped and fewer questions are settled. In places, this chapter will say plainly that nobody knows, rather than cover a gap with confident-sounding language.

That's not to say nothing is being done. Several serious lines of research are chipping away at pieces of the problem.

The oldest idea is that cells read their position from chemical signals that spread outward from fixed points. A cell can gauge roughly where it sits by how strong a signal is where it is, the way you might judge your distance from a campfire by how warm the air feels. This explains some coarse patterns well, such as which end of a developing limb becomes the thumb side and how many fingers form. It wasn't built to explain how to trace a particular curving edge, point by point.

A second idea goes back to a 1952 paper by Alan Turing, the mathematician best remembered for breaking wartime codes. He showed that two chemicals spreading and reacting at different speeds can produce regular repeating patterns, stripes and spots and evenly spaced structures, with no blueprint at all. It's a powerful idea, and it does explain some patterns in nature. But it describes chemicals, not the behavior of individual cells, and that gap has kept it from winning full acceptance among biologists working with real tissue.

A third line treats a developing tissue as a mechanical problem as much as a chemical one: cells pushing and pulling on their neighbors, and the resulting stresses helping to guide what happens next. This has started to explain how sheets of cells fold, bend, and buckle. By its own researchers' account, it's only beginning to be joined up with the genetic and chemical side of the story.

The newest approach borrows from computer science: searching by computer for simple cell-by-cell rules that could, in principle, produce a given shape, and checking the results against real tissue. Its practitioners describe programming cells toward a chosen outcome as extremely difficult, for exactly the reason this chapter has been circling: real development emerges from countless local decisions with no central coordination.

Put plainly, there is no agreed account of how a shape like an ear gets built, cell by cell, curve by curve. There are several serious partial approaches, each acknowledged by the people pursuing it to be incomplete. The question this chapter opened with isn't answered anywhere in the current scientific literature.

Last revised 2026-11-19

← What the Robots Didn't Have to DoHundreds of Genes for a Face →