Chapter 5: A Universe in a Hurry
The RNA World
So what does modern science actually say happened? The leading answer is a clever piece of reasoning. But being clever isn't the same thing as being demonstrated.
The problem it's trying to solve is a serious one for any theory of origins: a modern cell needs DNA to store its instructions, but DNA can't be copied without protein machines to do the copying — and those protein machines can't be built without DNA to specify them. It's a classic chicken-and-egg problem, dressed up in molecules.
The proposed way out is that neither came first. Instead, a third kind of molecule — RNA, DNA's close chemical cousin, which we met in Chapter 3 — supposedly did both jobs at once, early on. RNA can, in principle, store information the way DNA does, and unlike DNA, it can also fold up and act as a crude little machine, the way a protein does. RNA machines of this kind are called ribozymes. If that's right, the earliest life wouldn't have needed the chicken or the egg — just RNA, wearing both hats until something better came along.
In its standard form, the story then runs in three stages. First, the RNA World proper: RNA does everything, storing the information and doing the work. Second, an "RNA-protein world": RNA invents a way of building proteins — an early version of the protein-making machine every cell has today — and begins directing the construction of proteins, while still serving as the genetic material itself. Third, the handoff to DNA: DNA arrives as a more stable place to store the information, which gets copied over into it from RNA.
If you go looking for the case in favor of the RNA World yourself, you'll run into one piece of evidence again and again, presented as the strongest card in the hand. It goes like this.
Deep inside every living cell, at the heart of the machine that builds every protein the cell will ever make, is a giant molecular assembly called the ribosome. It's a combination machine — built from both RNA and protein, bundled together — and every known form of life has one, bacteria included. The protein parts of a ribosome mostly serve as scaffolding, holding everything in the right shape. But here's the striking part: the actual chemical step where two amino acids get stitched together — the one job you'd most expect a protein to handle, since proteins are supposed to be the molecules that do the work — isn't done by the protein parts at all. It's done by the RNA.
Defenders of the RNA World call this a "molecular fossil" — a leftover trace of an earlier, RNA-run version of life, still sitting at the center of modern biology because it was never fully replaced. On first hearing, it's a compelling image: like finding an ancient stone tool still doing useful work inside a modern factory.
Here's the criticism. The word "fossil" is borrowed from an entirely different field, and it's borrowing more than the people using it may realize. A real fossil — a bone, a shell, an imprint in rock — is a physical object you can dig up, date, and examine on its own terms, independent of any story about how it got there. What actually exists here is nothing like that. Nobody dug up an ancient ribosome. The entire observation is just this: modern cells use RNA, not protein, for this one job, today. That's it.
Whether that's because RNA is a leftover from deep history, or simply because RNA happens to be the best tool for this particular chemical task regardless of history, is not something the observation itself can tell you. Both stories fit the same fact. Calling it a "fossil" quietly picks one of the two and hands it a label that makes it sound already settled — before any actual evidence has settled it.
And even setting the naming problem aside, the "fossil" argument, at its very best, could only tell you that RNA machinery is old. It says nothing at all about whether that old RNA machinery was ever capable of doing the one thing the RNA World needs it to have done: copy itself, without outside help, so that a purely RNA-based life could exist and reproduce on its own before proteins entered the picture.
There has been real progress on that front, but no one has yet achieved what the RNA World actually needs. Researchers have built pairs of ribozymes that assemble each other from ready-made halves, multiplying rapidly and even evolving. In 2024, a lab-evolved ribozyme was shown to copy other RNA molecules accurately enough for those molecules to evolve. In 2026, a ribozyme only forty-five letters long was shown to build its own complementary strand, and, separately, a copy of itself — though in about two and a half months, only a fraction of one percent of the possible copies got made. Every one of these is a hard-won result. And every one depends on a chemist first supplying carefully prepared molecular pieces that nothing on the early Earth would have handed over for free; none of them puts all the steps together into one self-sustaining cycle running on the simple raw materials a real early ocean would have offered. As one review of the field put it, robust, continuous RNA self-replication from realistic raw materials "has not yet been possible" to demonstrate — and the newest results haven't changed that. Genuine, unassisted self-copying, the one ability this whole idea requires, is still a goal, not a finding. The "fossil," even granting the word every benefit of the doubt, is a fossil of the wrong thing.
So what we have is an observation, a clever hypothesis built on top of it, and a piece of terminology doing more persuading than the evidence underneath it can support.
There's another problem that gets far less attention outside specialist journals. RNA is a fragile molecule. Left sitting in ordinary water, it doesn't wait patiently to be useful — its backbone breaks down over time, and individual letters in its code decay chemically on their own, with no help needed. That's manageable inside a modern cell, which spends a great deal of protein machinery protecting and repairing its RNA. But the RNA World story needs RNA to have done its earliest work with none of that protection in place — no cell, no proteins, nothing but the molecule and its raw surroundings.
Whether the actual conditions of the early Earth — the temperature, the acidity, the unfiltered ultraviolet light pouring down with no ozone layer yet to block it — would have let RNA survive long enough to do anything useful is an unresolved argument among the scientists who study it for a living. Some point to particular pockets of environment — colder spots, certain mineral surfaces, patches of more acidic water — where the chemistry looks more forgiving. Others argue the obstacles are severe enough to cast real doubt on the whole scenario. Nobody has settled it either way.
So the molecule at the center of the story comes with its own well-documented survival problem, stacked on top of the self-copying problem.
There's a deeper problem with the RNA World, though — the one this book cares about most. Grant it every benefit of the doubt: grant that RNA could survive, grant that it somehow learned to copy itself, grant all of it. The arithmetic problem doesn't go away. It just moves.
A working ribozyme, one actually capable of doing something useful, still has to be a specific working sequence out of an almost unimaginably larger number of non-working ones — the same needle-in-a-haystack search we've been dealing with all along, with a different molecule holding the needle. Swapping "protein" for "RNA" doesn't touch the underlying problem. It relocates it.
Moreover, if RNA's chemistry problems can only be solved by having it form and thrive in some special, limited environment, the arithmetic gets substantially worse. A small, special place offers far fewer tries than a whole ocean — and the tries are exactly what a search like this needs.
Set all of that aside, and there's a further gap nobody has closed: how, exactly, did the world get from RNA doing double duty to the DNA-and-protein system every living thing uses today — while the RNA-based version, presumably, had to keep working the whole time, since there was nothing else around to keep the show running? The standard three-stage story gives an answer, but look at what the answer requires. The way out of the RNA World isn't one new invention. It's several, and each raises the same question in a new form.
Start with protein. In the standard story, RNA itself invents protein-making. But that isn't one invention. It's a whole system: a machine to join amino acids together in order, a code linking stretches of RNA to particular amino acids, and adapter molecules to fetch the right amino acid for each stretch and bring it to the machine. None of these parts is any use without the others. A code with nothing to read it is useless; a machine with no code to follow makes random chains; adapters with no machine to deliver to are just more molecules in the water.
And even once the whole system works, it only translates. Nothing in a code tells RNA which sequence to write. A useful protein still has to be found in the same vast search space as before, by the same blind trial and error — now through RNA genes changing at random — with the same arithmetic waiting at the end of it. Nor is finding it enough. A useful protein that turned up loose in the surrounding water, with nothing connecting it to the RNA system, would be just one more molecule drifting off as chemical noise. To matter, it has to be made inside the RNA system, be useful to that system, and be kept by it — three separate hurdles, not one.
Then there's DNA, which faces the same problem from a different angle. Storing the instructions in DNA takes at least two more specific inventions: one machine to make DNA's building blocks out of RNA's, by removing a single oxygen atom from each, and another to copy the RNA's information into DNA, letter by letter. Researchers have shown, in the laboratory, that a ribozyme evolved to copy RNA can also make short strands of DNA from an RNA template — up to thirty-two letters long, using DNA building blocks the chemists supplied. The first of the two machines is harder still: the chemistry needed to make DNA's building blocks is widely thought to be beyond anything RNA could do on its own, which is why one camp argues that proteins must have come before DNA. Another camp argues the reverse: that the protein-making system needs more stored instructions than an RNA genome could reliably hold, so DNA must have come first. The question is open. Either way, each new stage needs its own specific, working molecules, found by the same blind search.
So, essentially: does starting from an RNA-only world make the eventual DNA-and-protein world any easier to reach than if that world had to arise on its own, from scratch? The RNA World might buy some extra time for the necessary molecules to be found, if it really existed. But time is only half the equation. The other half, the number of tries available, gets worse, not better, once the search moves inside a primitive, space-limited system instead of playing out across a whole ocean — fewer places for a lucky sequence to turn up, for the protein and for the DNA that has to store its blueprint, with no mechanism supplied, at any point, for finding either one, or for recognizing it once found.
Nobody has a demonstrated answer to any of these questions.
And then there's the question: what happened to the RNA World? Why is there no trace of it? What exists is an argument from absence: the newer DNA-and-protein system was so much better that it completely replaced the older one everywhere, leaving nothing behind to find. That may even be true. But notice what kind of claim it is. It isn't evidence for what happened. It's an explanation of why we shouldn't expect to find any.
So the RNA World, even taken entirely at face value, isn't an escape from the arithmetic problem. It's an attempt to solve one version of it by quietly introducing several more — and it offers no working answer for any of them.
Last revised 2026-11-17