The Onion Has More Than You
a read-aloud about a thing everybody says about you that turns out not to work — and what an onion has to do with it, with every source kept on the page for the grown-up
What people say about it
We have met the four letters. And we have met the words they make, three letters at a time.
Now: what is the whole thing?
You will hear grown-ups say it is the plan for you. Some of them say blueprint, which is an old word for a drawing of a house, made before anybody builds the house.
So: is it a drawing of you, made before you?
That is a real question, and it has a real answer, and we can go and check.
This closes a three-part arc. No. 105 gave the alphabet, No. 106 gave the code that reads it three letters at a time, and each deliberately stopped short of the whole-genome question. This one takes it.
The blueprint metaphor is not a simplification of the science; it is a claim, and it is wrong. The reference behind this zine is M. Pigliucci, “Genotype–phenotype mapping and the end of the ‘genes as blueprint’ metaphor,” Phil. Trans. R. Soc. B 365(1540), 557–566 (2010), read at full text, whose own verdict is that “old-fashioned metaphors like genetic blueprint and genetic programme are not only woefully inadequate but positively misleading.”
The zine checks rather than asserts, which is the only honest way to take something apart for a child. A plan is a drawing of a finished thing made in advance; so a more complicated thing must need a bigger drawing. That is a prediction, it can be tested, and the next six spreads test it twice.
This piece stays on organisms throughout. It never argues about people, and what the metaphor has been used to do to people is in the refusals, addressed to you.
A bigger house needs a bigger drawing
Think about a real plan for a real house.
A little hut with one room and one window — you could draw that on one small piece of paper.
A castle, with towers and stairs and a hundred rooms, needs a much bigger drawing. Lots of sheets. Because there is more to say.
That is how plans work. More complicated thing, bigger plan.
So if the stuff inside you really is a plan of you — then something more complicated than an onion should have more of it than an onion.
Let us go and look.
This spread sets up a falsifiable prediction, deliberately. Pigliucci lists it among the metaphor’s failures as misleading proportionality — blueprint thinking implies that sequence length should track organismal complexity. It does not, and the failure is not marginal.
“More complicated” is doing unexamined work here and a curious child may say so. There is no agreed measure of organismal complexity, and that is a genuine problem in biology rather than a gap in the zine. The argument survives it: on any ordering a person would offer — cell types, organ systems, behaviour — a human comes out above an onion and above a millimetre-long worm, and the DNA does not follow. You do not need a definition of complexity to notice the prediction failing this badly.
The onion has more than you
Here is how much DNA you have. And here is how much an onion has.
The onion has about five times more.
Not a little bit more. Five times. An ordinary onion, the kind in the kitchen, the kind that makes you cry.
If the stuff inside a living thing were a plan of it, that would mean an onion is five times more complicated than you are.
An onion is not five times more complicated than you.
The figures. The first onion genome assembly gives Allium cepa at about 16,400 Mb per 1C, against a human genome of roughly 3,200 Mb — so about five times. The same paper calls it “the largest of all cultivated diploid crops and of a size comparable to the allo-hexaploid bread wheat.”
There is an explanation, and withholding it would be cheating. At least 95 per cent of the onion genome is repetitive sequence, mostly LTR retrotransposons — long stretches of the same thing over and over, copied and re-copied. So the onion is not carrying five times more instructions. It is carrying an enormous amount of repeat.
That rescues the onion and not the metaphor. If quantity of DNA can balloon like that without the organism changing to match, then quantity of DNA is not measuring the organism — which is exactly the point. And the next check removes the escape route by counting instructions instead.
The comparison is a standing challenge in the field, posed by T. Ryan Gregory as the onion test: can you explain why an onion needs about five times more non-coding DNA than a human? Within the genus Allium alone, genome size runs from about 7 to 31.5 picograms — so onions disagree with other onions by more than onions differ from us.
Maybe we counted the wrong thing
Somebody could say: that is not fair. You counted letters.
A very long book can have a lot of letters and still not say much, if it says the same thing over and over. And that is exactly what an onion does — most of its letters are the same bits, copied again and again.
Fair enough. Let us count the instructions instead.
Not letters. Whole instructions. The bits that actually tell the body to make something.
They have a name you already half know. They are called genes.
This spread exists because the first check has a real objection and the zine should make it rather than wait to be caught. Raw genome size is a poor proxy, the onion’s bulk is largely repetitive, and an adult who knows that would be right to say so. Handing a child the counter-argument and then addressing it is the same move No. 96 makes about how you ask a question.
“Gene” is being used in its narrow, countable sense — a protein-coding gene — because that is the sense in which anybody has a number. The word has several other meanings in working biology and they do not all count the same things.
Note what this spread does not do. It does not say the repetitive DNA is useless, and the zine takes no position in that argument, which is live and unresolved. All that is needed here is that raw length is not the right ruler; what the rest of the length is doing is a separate question.
People guessed first, and wrote it down
Before anybody had counted a single one, people guessed how many instructions a person would have.
They guessed big. Some said a hundred thousand.
You can see why. People are complicated. A hundred thousand felt about right for something as complicated as a person.
Then it took years and years of work, and they counted.
The real number is about twenty thousand.
Not a hundred thousand. About a fifth of that.
The range of pre-Human-Genome-Project estimates ran roughly 40,000 to 100,000, and the zine uses the top of it because that is the figure that circulated. The published count settled near 20,000 protein-coding genes, with later curation pushing it slightly below; the exact figure still moves a little with annotation method, which is why the page says “about”.
This is the part of the story worth a child hearing, and it is not the number. A large, expensive, well-run scientific effort published a prediction and the prediction was wrong by a factor of five, and everyone said so, and the number was corrected. That is the process working rather than failing.
Why the guess ran high is the blueprint metaphor doing its work in plain view: if the genome is a specification of the organism, a complicated organism needs a great many specifications. The guess was a prediction of the metaphor, and it is the metaphor the result falsified.
And the worm has about the same
Now here is a worm.
It is about as long as this line: —
You could hold a hundred of them in your hand and not be sure they were there. It has no eyes. It has about a thousand cells in the whole of it. You have trillions.
It has about twenty thousand instructions.
The same as you. Near enough the same number as a whole person.
And all of that fits in a space thirty-two times smaller than yours.
The worm is Caenorhabditis elegans, the nematode used in laboratories everywhere, about 1 mm long. Its genome is 100,291,840 bases against the human 3,200 million — hence thirty-two times smaller — and it carries roughly 20,000 protein-coding genes. The figure moves between WormBase releases with ongoing curation (19,735 in one, 20,512 in a later one), which is why the page says “about”.
“About a thousand cells” is exact and worth knowing: an adult hermaphrodite has 959 somatic cells, and every one of them has been traced and named. That is why the animal is in every laboratory.
So the prediction fails on both rulers. By quantity of DNA an onion beats us five to one; by number of instructions a millimetre of worm draws level. Neither count tracks anything anybody would call complexity, and the metaphor needed them to.
What actually differs is not in these numbers at all — it is in how genes are regulated, when and where they are switched on, how their products are combined, and what the developing body is doing while all that happens. Which is the next spread.
So it is not a plan of you
We checked it two ways. Both ways said the same thing.
We counted the letters: an onion has five times more than you.
We counted the instructions: a worm you can hardly see has about as many as you.
A plan of a more complicated thing is bigger. This is not.
So whatever the stuff inside you is, it is not a drawing of you made beforehand.
Nobody drew you.
“Nobody drew you” is the sentence this zine is built to reach, and it is a claim about a metaphor rather than about anything metaphysical. There is no advance representation of the finished organism anywhere in a genome. That is not a poetic reading of the biology; it is what the failure of proportionality on two independent measures actually means.
Pigliucci lists two further failures the zine does not have room for, and both point the same way. The mapping is not one-to-one — very different genotypes can give the same phenotype — and blueprint thinking predicts brittleness, whereas real developmental systems are strikingly robust to perturbation. Four failures, one conclusion.
What replaces it is not “a recipe”, which is the usual substitute and carries some of the same trouble. Pigliucci’s term is developmental encoding: the genome specifies not an outcome but a set of generative rules, whose results depend on the developing system and its environment at every step.
The plan and the building are the same thing
Here is the difference, and it is the whole answer.
A drawing of a house sits still. It does nothing. Somebody else has to come along and build the house, and the drawing just lies there being looked at.
What is inside you does not lie there being looked at.
It is doing the building. It is the plan and the builders, at the same time, and it never stops.
Which means it is not a picture of a finished you that somebody is working towards.
There is no finished you in there to be a picture of.
There is just the building, happening, all the way along.
This answer is older than the metaphor it replaces, which is the best thing in this zine. Erwin Schrödinger introduced code-script in What Is Life? (1944) — the book Watson, Crick and Wilkins all credited — and then, in the very same paragraph, said it was not enough:
“But the term code-script is, of course, too narrow. The chromosome structures are at the same time instrumental in bringing about the development they foreshadow. They are law-code and executive power — or, to use another simile, they are architect’s plan and builder’s craft — in one.”
The man who coined the metaphor flagged its limit on the spot, and the limit is what got dropped. Plan and craft, in one — and what travelled onward into blueprints and programmes and popular science was the plan, without the craft. Read at full text.
That loss is a whole piece of its own and is not this one. How a qualifier falls off between a founding text and a schoolroom is a chain, and it belongs in How We Got Here rather than in a read-aloud. This zine takes the answer and leaves the history.
Even onions disagree with onions
One last thing, because it is the best bit and it is true.
We said an onion has about five times more than you.
But onions do not all have the same amount either.
Onions and their close relatives — garlic, leeks, all that family — go from a little to more than four times that little.
So onions disagree with other onions by more than onions disagree with us.
They are all still onions. Every one of them.
Within the single genus Allium — onions, garlic, leeks, chives and their relatives — genome size runs from about 7 to 31.5 picograms, a range of roughly four and a half times, inside one genus of closely related plants that are all recognisably the same sort of thing.
This is the strongest form of the argument and it is why the spread exists. The first two checks compared distant organisms, where somebody might still want to argue about how you measure complexity. This one does not need any such measure: these are all onions and their cousins, nobody thinks one of them is four times the plant another is, and the DNA quantity varies four and a half fold anyway.
What it is not. It is not an argument that genome size is meaningless — it varies for real and interesting reasons, mostly transposable-element activity. It is an argument that genome size is not a measure of the organism, which is the only thing the blueprint metaphor needed it to be.
What this one refuses
This page is for you rather than for the child. The zine above stops at onions and worms, deliberately, and this is where the rest of it goes.
A plan of a more complicated thing is bigger. This one is not. Nobody drew you — there is no finished you in there to be a drawing of.
If it is not a drawing of you, what do you think it is more like?
Why do you think people guessed a hundred thousand?
The worm has the same number as you. What is different, then?
Can you think of something else that is the plan and the builder at once?
Where all of this comes from
The argument, and the four failures of the metaphor. M. Pigliucci, “Genotype–phenotype mapping and the end of the ‘genes as blueprint’ metaphor,” Phil. Trans. R. Soc. B 365(1540), 557–566 (2010), doi:10.1098/rstb.2009.0241. Read at full text. Source of the misleading-proportionality failure the zine tests, and of developmental encoding as the replacement. This zine exists because No. 106 cited that paper as a lead it had not read. It has now been read.
The onion. First genome assembly of onion, G3 (2021) — Allium cepa at about 16,400 Mb/1C, “the largest of all cultivated diploid crops,” at least 95 per cent repetitive, mostly LTR retrotransposons. The onion test is T. Ryan Gregory’s, posed as a challenge to universal-function claims about non-coding DNA. The Allium range of about 7 to 31.5 picograms is reported at second hand and not traced to the measurement papers.
The worm. C. elegans genome 100,291,840 bases; roughly 20,000 protein-coding genes, the figure moving between WormBase releases with curation (19,735 and 20,512 both appear); 959 somatic cells in the adult hermaphrodite. Pre-HGP human gene estimates of 40,000 to 100,000 against a settled count near 20,000.
The answer on spread nine. E. Schrödinger, What Is Life? (1944), the “Hereditary Code-Script” section. Read at full text, and the passage quoted verbatim rather than from any secondary account — including the sentence immediately before it, “But the term code-script is, of course, too narrow,” which is the half that matters most and the half nobody quotes.
How this one is built
Two layers on every spread. Large plain type for the child, first in the document, then a bordered box for the adult holding the figures, the citations and every place the evidence runs out. There is no demonstration in this one, and that is not a gap: you cannot weigh a genome at a table, and a forced demonstration is worse than none. What it offers instead is a check — a claim, a prediction, and two goes at testing it, which is the shape of the thing rather than a decoration on it.
The objection is handed to the child before it is answered. Spread five gives away that counting letters is unfair to the idea, because the onion’s bulk is mostly repeats — and then counts instructions instead. An argument that only presents its own best evidence is not a check, it is a sales pitch.
What we left out, and what is still open
The history is deliberately absent, and it is the better piece. Schrödinger named the limit of his own metaphor in 1944 and the limit was dropped; how that happened, through the Modern Synthesis and the genome project and out into ordinary speech, is a chain with graded joints and it belongs in How We Got Here. It is planned and it is not this. A read-aloud that tried to carry it would carry neither.
Open, and said on the page. Whether the onion’s repetitive DNA has function is a live argument this zine takes no side in, and spread five says so. “More complicated” has no agreed measure, which spread three admits rather than hides. And the Allium range is second-hand.
For grown-ups who want the rest of it. No. 105 and No. 106 are the first two parts: the alphabet, then the code. This is the third, and the three are meant to be read in that order.