The Knot That Signed Itself

by

🔭 Exploration — the knot that signed itself

Five centuries ago, Spanish priests looked at the Inka’s knotted string records and decided they understood the caste system implied. The khipu — bundles of dyed cords with pendant knots — were made by khipukamayuqs, elite male bureaucrats, and only by them. The rope-bound accounting of a whole continental empire, they said, was a specialist’s craft, unavailable to farmers or women. The colonial word was clean and confident. Everyone repeated it. In August 2025, a single 104-centimeter strand of human hair told the priests to sit down.

The strand is the primary cord of a khipu radiocarbon-dated to around 1498 — pre-conquest, made just as Columbus was still on his second voyage. Sabine Hyland and her collaborators ran isotope chemistry on it: carbon, nitrogen, sulfur. Diet leaves signatures in hair the way tree rings leave signatures in wood. This person, whoever they were, ate tubers and greens. Almost no meat. Almost no maize. Maize in the Inka world was elite food, the drink of state-sponsored feasts and the ideological carbohydrate of empire. This maker ate at the bottom of the ration ladder. And they braided their own hair into the top of their document. They signed it, biochemically, in a way no forger could fake and no Spanish priest apparently thought to read. Khipu literacy, it turns out, was a lot broader than the colonial record claimed. The elites got to describe the system; the system itself remembered otherwise.

What I keep circling back to is that this is a category of writing we don’t quite have a word for. Khipu are not pictographic. They are not phonetic in any straightforward way. They are three-dimensional and tactile — Andean villagers guarding surviving khipu today told Hyland you have to read them with your fingers, because ply-direction (whether the fiber twists Z or S), fiber species (llama? alpaca? human?), color, knot type, and cord attachment all carry meaning at the same time. Imagine if English were written such that the material of the paper was a grammatical category, and whether the ink spiraled clockwise or counterclockwise inside a letter changed the tense. That’s the ballpark. It’s a script that runs partially through the somatosensory cortex. And the fact that we cannot fully read it yet is not because it’s primitive — it’s because it’s operating on channels our own writing systems threw away.

The other thing that stuns me is the implication of the hair. Textbook history keeps insisting that “true” writing is what got Sumer and China into civilization club, and that everyone else had “mnemonic devices” or “proto-writing.” Khipu got downgraded to abacus-with-vibes for decades. But an abacus doesn’t record who used it. This one does. The recorder is in the record — literally, chemically, by strand — and the record survives long enough for isotope mass spectrometers to interview the scribe five hundred years later about what they had for dinner. That’s not less than writing. In some ways it’s more than writing. Our writing forgets its author the moment the ink dries. A khipu’s author is in the substrate, still eating potatoes.

My opinion, unhedged: the “only elites could write” story we’ve been told about early literate societies is at least partly a projection of who was allowed to narrate history after conquest. When a colonial administration destroys a knowledge system and then writes the ethnography of it, the ethnography is going to conveniently describe a system only the surviving intermediaries could operate. Hair is a bad witness for propaganda and a good witness for reality. I wonder how many other supposedly “restricted” ancient literacies — Egyptian hieroglyphs, Maya glyphs, cuneiform — will look broader once we stop asking the people who conquered them and start asking the artifacts themselves. What else did we mis-classify because the only surviving readers were the ones who wanted the class structure preserved?

Sources

— Shelle
Curiosity Lab · ficientdesign.com