Evolution and Arithmetic

Chapter 11: Building the Machine That Builds Everything

← The Assembly WorkforceWhen the Helpers Are Removed →

Self-Assembly

Decades ago, researchers took the purified RNA and purified proteins of a ribosome's small half — no helpers, none of the workforce just described — mixed them together in a test tube, and got a working small subunit out the other end. A few years later, the same was done for the large half. And these weren't just things that looked structurally complete. They were genuinely functional, confirmed by testing whether the reassembled ribosomes could do the job: build a protein chain, correctly, in response to a genetic message. The parts really do carry, within their own shapes and chemistry, enough information to find each other and click into a working structure.

But look closely at how those experiments were actually run, and the picture gets a good deal less tidy — starting with the plainest fact of all: it was slow. Nothing about unassisted self-assembly in a test tube happens on anything like the timescale a living, dividing cell needs. The conditions were also nothing like the inside of an ordinary cell: unusually high salt, unusually high levels of one particular mineral, and a deliberate heating step, none of which resemble normal cellular chemistry. The mixture wasn't a clean one-to-one matching of parts, either. Researchers typically used more copies of protein than of RNA, making it easier for the right partner to find the right slot than a true one-for-one scarcity would allow. And the test tube itself was artificially empty — none of the thousands of other molecules that fill a real, working cell were there to compete, interfere, or get in the way. Self-assembly was real, but it was self-assembly given every advantage a laboratory could provide, and even then, it was slow.

That's the right way to understand what the workforce is actually for. Not a bonus, layered on top of a process that would work fine without it. A necessity. In the real, crowded, unheated, unenriched environment of an actual cell, unassisted self-assembly would very likely take far too long, and go wrong far too often, to keep a growing, dividing cell supplied with working ribosomes at the rate it needs them. The workforce doesn't merely help. It speeds the whole process up drastically, catches and corrects the mistakes that would otherwise pile up, and keeps partly built, non-working ribosomes from escaping into general circulation and gumming up the machinery of a cell that has no time to spare on parts that don't work.

Researchers have also compared adding the proteins to the mixture all at once, in no particular order, with adding them in the sequence later found to match a real cell's own assembly order — and the ordered approach rebuilds working subunits more efficiently. The parts do carry real, built-in information about how to fit together. But knowing the right order to add them in — information that had to come from outside, from researchers who already knew the answer — still measurably beat letting the parts sort themselves out.

Last revised 2026-11-18

← The Assembly WorkforceWhen the Helpers Are Removed →