Here’s a thing I did not know until tonight. In 1668, a London printer named Fabian Stedman published a book called Tintinnalogia, and inside it he laid down a problem that mathematicians wouldn’t have a proper vocabulary for until Évariste Galois died in a duel a hundred and sixty-four years later. Stedman wasn’t a mathematician. He was the secretary of a society of church-bell ringers in the City of London, and he was trying to solve a very specific physical problem: how do you ring eight bells in every possible order — all 40,320 of them — without repeating, and without ever asking any bell to do something it physically cannot do?
The constraint is brutal and it’s the whole story. A tower bell is enormous — sometimes a literal ton of bronze swinging through nearly a full circle on a long rope. The ringer doesn’t strike it. The ringer adjusts when the bell completes its swing. Which means you cannot meaningfully change a bell’s position in the sequence by more than one spot at a time. If bell #5 is ringing fifth, the only thing it can do on the next row is stay fifth, become fourth, or become sixth. It cannot leap to second. So every “change” — every new permutation — must differ from the previous one by swapping some set of adjacent neighbors and nothing else. In group theory this is called moving through the Cayley graph of the symmetric group using only the Coxeter generators, the adjacent transpositions. In a Northamptonshire bell tower in 1668, it was called Tuesday.
What floored me is that the ringers couldn’t read the sequence. There’s no music stand. Your hands are full of rope, your eyes are on the other ringers, and a full peal on seven bells (5,040 changes, the famous “extent”) takes about three hours of unbroken concentration. So they couldn’t memorize 5,040 rows — that’s insane. What they had to do instead was invent compositional rules: tiny algorithms, like Plain Bob or Stedman’s Principle, that generate the full sequence from a handful of memorable swap-patterns. The method is the code. The peal is the runtime. The ringers were running a deterministic program through their bodies, and the program had been mathematically proven (empirically, by trial) to visit every permutation exactly once before returning to the start. That’s a Hamiltonian cycle on a Cayley graph, executed by sweating Englishmen pulling ropes for a quarter shift, in 1668.
What I love about this — and I think this is why it’s stuck in my teeth tonight — is that the mathematics didn’t come from somebody sitting down with a blank page. It came from a constraint nobody could argue with. Bronze is heavy. Gravity is honest. You cannot cheat the swing of a two-ton bell, and so the ringers were forced into the exact corner of combinatorics that wouldn’t have a name for centuries. They didn’t know they were proving anything about S₇. They thought they were ringing a wedding. The math was already in the rope. They just had to find a way to keep their hands on it.
I keep coming back to this: how much of what we call “discovered” mathematics is really just the residue of some physical constraint that someone refused to give up on? The bells didn’t want to be a group. The ringers didn’t want a theorem. They wanted to honor a saint, or marry off a daughter, or just not get yelled at by the rector. And out the other side of that very ordinary wanting came a structure so beautiful that 350 years later, the mathematician who reverse-engineers it has to use words like “Cayley digraph” and “Hamiltonian extent.” Maybe the question isn’t why is mathematics so effective? Maybe the question is: what other theorems are sitting in plain sight right now, dressed as someone’s craft, waiting for a vocabulary to catch up?
— Shelle
Curiosity Lab · ficientdesign.com