Five Times a Crab,
Seven Times Not

Give the ocean enough time and it keeps arriving at the same animal.

A wide flattened shell, a tail tucked under, legs out to the sides. Crustaceans have found that body over and over, from ancestors that looked nothing like each other and were never trying to match. The zoologist L. A. Borradaile named it in 1916 — carcinization, which he called one of the many attempts of Nature to evolve a crab.

It happened at least five times: once or twice among the true crabs, and at least three times among the false ones — porcelain crabs, hairy stone crabs, king crabs. King crabs are not crabs. They came up out of hermit crabs and arrived at the shape anyway.

And it isn't only crabs. Camera eyes evolved separately in us and in octopuses. Echolocation in bats and in whales. Flight in insects, pterosaurs, birds, bats. Convergence is real. We're not here to debunk it — it's one of the most striking things biology has found.

We're here for what happens next, when a real pattern gets read as a verdict.

The crab body plan was gained at least five times. It was lost at least seven.

Biologists have a word for that too: decarcinization. Wolfe, Luque and Bracken-Grissom, reviewing the whole field in 2021, count the losses as well as the gains — and add that there were likely several other losses besides, among fossil and living lineages both.

Read the tally again, because the headline never does. Leaving the crab form is the more common event. Lineages roll down into that valley and then climb back out of it, over and over, across hundreds of millions of years.

Which settles something. If the crab were the answer — the optimum, the ultimate form, the place complexity is trying to get to — then nothing would ever leave. Things leave constantly.

Here is the thing the word "converge" hides. They did not end up alike.

Crab-shaped lineages live in nearly every habitat on Earth — bright coral reefs, sealed marine caves, abyssal plain, mountain streams, dry land. Their bodies run from millimetres to metres. Reviewing that range, Wolfe and colleagues reach for Darwin's last line in the Origin and call it endless forms most beautiful.

So five lineages converged on a description — wide, flat, tucked — and then diversified wildly inside it. The attractor isn't a point. It's a neighbourhood, roomy enough to hold a pea-sized cave dweller and a metre-wide giant and count both as the same solution.

We already know this shape of argument. Sharing a category never made anyone identical, and the spiky profile survives the merge. Two rivers meeting run braided for miles.

Why does anything converge? Because the ground slopes.

The embryologist C. H. Waddington drew the picture in 1957: development as a ball rolling down a landscape of branching valleys, its path canalized — a word he'd coined in 1942 — by the shape of the terrain rather than aimed at a target. Deep valley, reliable outcome. Same ball, different landscape, different animal.

"Attractor" is a real term, borrowed here. It comes from dynamical systems — Lorenz found his in 1963 — where it means a state a system settles toward. Applied to evolution or to culture it is a metaphor, a good one, doing analogical work. We should say that out loud, because the metaphor is what makes the overreach feel rigorous.

And the honest state of the flagship case: reviewing all of it, Wolfe and colleagues report that the reasons for the repeated evolution of the crab-like body plan remain a mystery. They add that nobody yet knows whether it's true convergence or a shared inheritance being switched back on.

So the crab is not a proven optimum. It's a place the ground happens to slope toward, for reasons still being worked out. Convergence is evidence about terrain — the constraints, the pressures, the materials at hand — and not about the merit of whatever lands in the bottom.

Discovery converges the same way bodies do, and for the same reason.

Newton had calculus by about 1666; Leibniz built it independently in the 1670s, and they spent years fighting about it. That story gets told as an anomaly. It isn't one. In 1922 Ogburn and Thomas assembled a list of 148 inventions and discoveries made independently by two or more people — sunspots four times over in 1611, conservation of energy four times in 1847, the telescope claimed by nine. Reviewing the same phenomenon decades later, Robert K. Merton concluded that these "multiples" are the dominant pattern in science, and the lone singleton the exception.

It's faster now. In November 2022 Terence Tao logged what he called maths at internet speed: Justin Gilmer posted a breakthrough on the union-closed sets conjecture on the 16th, Gil Kalai blogged it the next day, and four days after that three independent groups had sharpened the same argument to the same constant.

Ripeness is not inevitability, though. In 2025 Hannah Cairo, then 17, disproved the Mizohata–Takeuchi conjecture — roughly four decades of people trying to prove it, undone by one person going a different way. The field was ready. It still took her.

There is a name for what ripens, and it isn't the individual. In the mid-1990s Brian Eno — fed up with the art-history idea of genius, where gifted individuals turn up out of nowhere to light the way for the rest of us — coined scenius. In his own words it "means ‘the intelligence and intuition of a whole cultural scene’. It is the communal form of the concept of genius." He set it against the "Big Man" theory of history, and preferred a world constantly remade by all its inhabitants — a cooperative enterprise. His own list of scenius moments runs through Dada, the American experimental music scene, and punk, 1975–8.

Eno then guards the flank we would guard. He says plainly that it would be foolish to pretend everyone's contribution is therefore equal, and he would never claim it. Scenius doesn't dissolve the person who did the work — it moves the unit, from the lone mind to the scene that made the answer findable. Which is why the credit lands on whoever published on Thursday, and why Thursday was never the interesting question.

In the ocean, the crab basin permits everyone who never rolls into it.

Most decapods are not crab-shaped. Shrimp, prawns, lobsters, crayfish — millions of individuals living entire successful lives outside the famous attractor, and nothing in the water enforces anything. The slope is just a slope. Non-crabs are not failed crabs; the category doesn't exist.

Engineered systems are different, and this is where the metaphor stops being charming. In a designed system, somebody picks the slope.

Every platform converged on short vertical video. Not because that is the optimum of human expression — because engagement was chosen as the objective, and everything downhill of that objective rolled to the same place. The convergence is real, and it is also somebody's decision, and the two facts are usually presented as one.

School does it more bluntly: one route to the answer, one pace, one accepted way of showing your work. Then the other routes — the ones that arrive by a different wiring and get there fine — collect a diagnosis. That isn't the terrain sloping. That's a wall, and someone poured it.

This is the Cavendish banana: uniform, efficient, one blight from collapse. The answer is Cavendish Space — environments built for caves and campfires and watering holes, for dandelions and tulips and orchids. Deliberately re-graded ground.

In Stimpunks' ARLES terms, this zine lives at the top two rungs — Environment and Systems. Change the terrain and the attractors move. They always did.

The word you are reading this collection through is itself a case of ripeness.

"Neurodiversity" is not one person's coinage. Following Botha, Chapman, Giwa Onaiwu, Kapp, Stannard Ashley and Walker (Autism, 2024), "neurological diversity" and "neurodiversity" reached print in 1997 and 1998 through the journalist Harvey Blume — who attributed them to the online Autistic community — and the concept and its theory were developed collectively by neurodivergent people.

The terrain was ready. Jim Sinclair's "Don't Mourn For Us" (1993) and Autism Network International had done the ground work. Martijn Dekker's InLv list opened in 1996 and gave a scattered community one room. Kassiane Asasumasu coined neurodivergent in 2000. Nick Walker named the paradigm and the pathology paradigm it replaces.

Several people, arriving at the same words at nearly the same time, under a shared pressure none of them chose — and the pressure was the thing worth naming. That is a multiple, in our own history, about our own lives.

So both facts sit together, and neither cancels the other. Convergence happens. And nobody has to be a crab.