Evolution and Arithmetic

Chapter 12: RNA World to the Rescue?

← Is RNA a Great Savior?When Could Any of This Have Happened? →

Codes and Handoffs

There are theories, seriously pursued by researchers, for how all of this might have happened. The most direct attempt to answer the specific question just raised — how RNA would ever get, and pass on, the information needed to build a particular protein — is called the stereochemical theory: the idea that certain RNA sequences are chemically shaped to directly grab onto certain amino acids, the way a key fits a lock, with no code needed at all. It's a clever idea.

However, researchers testing this idea — RNA molecules directly grabbing an amino acid onto themselves, without any enzyme's help — found only weak, inconsistent preference for one amino acid over another, not the sharp, dependable specificity the stereochemical theory actually needs to work. One of the researchers who has spent decades studying exactly this question reviewed the evidence and stated that this kind of chemical attraction does not seem to account for the origin of the genetic code. The best answer on offer, tested against the evidence, doesn't hold up.

In any case, grabbing was never the real issue. What's actually needed is a specific sequence — amino acids linked together in a particular order that happens to be useful, and a way to have that same sequence produced again, consistently, rather than as a one-off accident. Grabbing alone, however well it works, doesn't touch that problem — because a chain is built by joining, not by grabbing, and nothing about being grabbed determines where in a chain an amino acid ends up.

Some researchers have gone further and tested this directly — whether an RNA molecule's own sequence could actually influence which amino acid gets joined onto it, a real, if modest, attempt to show the sequence problem being solved rather than sidestepped. One recent, well-designed study did find a small effect: the sequence at one end of an RNA molecule made a measurable difference to which amino acid attached there. But by the researchers' own description, the effect was "not particularly high" — weak enough that they characterize what it could plausibly produce as only "loosely coded" short peptides, not the sharp, dependable process needed to build a specific, functional protein. In other words, the best demonstrated result in this exact line of research still can't get past producing short, effectively useless strings. The mechanism has been tried and tested, and it hasn't delivered; this isn't a gap nobody has bothered to look into. The sequence problem hasn't been sidestepped in this corner of the research. It's been tested directly, and it's still standing.

Underneath all of this sits the broader question Chapter 5 already raised: every version of this story involves handoffs — RNA doing the job alone, then RNA and protein sharing it, then DNA taking over storage while protein takes over the chemistry — and nobody has offered a working mechanism for any of them. Each is assumed to have happened. How, specifically, is left unanswered every time.

The same trouble shows up again in a different disguise. Some researchers propose that early life used only a handful of amino acids, not the full twenty, and gradually expanded its toolkit over time. That's a reasonable-sounding way to make the starting problem smaller. But it just relocates the transition problem instead of solving it: at some point, a system already fully built around a few amino acids has to switch over to a system built around many more — every protein, every pairing, every piece of the reading machinery has to somehow accommodate the new arrivals without destroying what already works. That either means quietly building an entire second, expanded system alongside the first and eventually swapping over, or running some far more tangled arrangement that manages to hold the original system in place while continuously altering it out from under itself. Either way, a proposal that starts simple to make one part of the puzzle easier tends to hand you a new, comparably hard puzzle of its own, a little further down the road.

Put all of this together, and the honest answer to "can RNA make protein" isn't simply "not yet demonstrated." It's that every step in the chain — sorting the right amino acids from a chemical mess, joining them without any known machinery to do so, producing anything but a random sequence, and remembering a lucky result well enough to produce it again — is missing at once, and each missing piece is exactly the kind of elaborate, coordinated system that would itself need explaining before it could be used to explain anything else.

Last revised 2026-11-18

← Is RNA a Great Savior?When Could Any of This Have Happened? →