At Home in the Universe

To make the intuition concrete, Kauffman builds a thought experiment he calls autocatalytic sets, and it rewards a moment of patience. Picture a population of molecules where some of them can catalyze — speed up — the formation of others. A catalyzes the making of B, B helps make C, and somewhere in the tangle a molecule loops back to help make A. Now imagine adding more and more types of molecules to the mix. Each new kind brings new possible reactions, and crucially, new possible catalysts for reactions already present.

Here is the part that genuinely surprises. As the diversity of molecules climbs, the number of reactions between them climbs faster. Past a certain ratio, it becomes almost unavoidable that somewhere in the web you find a closed loop — a set of molecules that, taken together, catalyzes its own production from raw materials in the surroundings. The set makes itself. Nothing in it is special; no single molecule is the miracle. The self-maintaining quality belongs to the network as a whole, and it switches on suddenly, the way water suddenly freezes, once the connections are dense enough.

The same idea travels well beyond chemistry, and Kauffman's most striking demonstration comes from genetics. Every cell in your body carries the same genome, yet a liver cell and a neuron behave nothing alike. The difference is which genes are switched on. Genes regulate one another — one gene's product turns another on or off — so the genome is really a vast network of switches, each governed by the others. The puzzle is how such a tangled, jostling network ever settles into the stable, repeatable patterns we call cell types.

To probe this, Kauffman built abstract models of networks made of simple on-off elements, each wired to a handful of neighbors, each following a little rule about when to flip. Think of thousands of lightbulbs, each watching two others and turning on or off accordingly. You would expect such a system, left running, to flicker forever in chaos. Most wirings do exactly that. But Kauffman found that when each bulb listens to only a couple of others, the whole network does something remarkable: it falls, on its own, into a small number of repeating cycles. From astronomically many possible states, it settles into a handful of stable patterns.

Step back and the shape of Kauffman's wager comes clear. Darwin gave biology one engine of order: natural selection, the patient sieve that keeps useful variations and discards the rest. It is real, and Kauffman never denies it. His claim is that selection has been doing the work of two engines and getting credit for both. Alongside it, mostly unnoticed, sits a second source of order — the spontaneous self-organization that complex systems fall into on their own. Autocatalytic sets, the few hundred stable cell types, the edge of chaos: in each case order appears before selection has a chance to act. Kauffman calls it order for free.

If that is true, the relationship between the two engines flips from what we assumed. We have imagined selection as the sculptor and raw matter as formless clay, with every living structure owing its shape entirely to the chisel. Kauffman's picture is different. The clay arrives already half-organized, predisposed toward certain forms, and selection works with that grain rather than against a blank. Evolution becomes a collaboration between chance, selection, and the deep tendencies of complex systems to organize themselves. The order we see is partly carved and partly, as he puts it, the order we get for nothing.

Return to the warm pond and the crushing odds. Kauffman never disputes the arithmetic; he disputes the picture that gave it meaning. A lone molecule against the universe is impossibly unlikely. A rich web of molecules helping make one another is something complex chemistry tends to do, the way buttons and threads tend to snap into a single connected whole once the connections grow dense enough. Move from the hero to the network and the miracle dissolves into a tendency. That move, repeated from chemistry to genomes to economies, is the whole of his argument.