The Smoke That Kept Them

Heels driven up against the buttocks. Knees against the chest. Arms bound tight to the trunk at angles that a corpse's joints, left to themselves, simply do not permit. At Huiyaotian, in Guangxi, nineteen percent of the individuals carry cut marks — and the marks sit at the joints, precisely where you would cut if the problem you were solving was that the body would not fold small enough.

For decades those cuts were read as ritual dismemberment: a secondary burial practice, the dead taken apart and reassembled. It is a reasonable reading. It is also, on the evidence published in PNAS in September 2025, probably wrong. The cuts were not for taking the body apart. They were for making it compact enough to bind, and to hang.

Because the other thing these bones have been carrying, unnoticed, is heat.

The team — led by Hsiao-chun Hung at the Australian National University, with Zhenhua Deng at Peking University and Hirofumi Matsumura at Sapporo Medical University — ran two mineralogical tests across 54 pre-Neolithic burials from 11 sites in southern China, Vietnam, and Indonesia, spanning roughly 12,000 to 4,000 years before present. Fourier-transform infrared spectroscopy on 69 bone samples returned crystallinity indices above 3.25 in about 84 percent of them: the signature of bone that has been cooked, structurally, at the crystal level. X-ray diffraction on twenty samples put nine of them above 525°C.

And here is the detail that does the most work. Only about fifteen percent of the individuals show burning you can see with an eye — and that burning is not distributed randomly. It clusters on the frontal skull, the lower limbs, the upper arms. The places where flesh is thinnest. The fire reached the bone exactly where there was least body in the way.

That is a fire held close to a body for a long time, and never hot enough to consume it.

A word of restraint, because heat alone proves nothing much. Bone gets hot for many reasons — cremation, cooking, a later hearth built over an older grave, a house fire. No single line of this evidence carries the claim. What carries it is that four independent lines point at the same practice: the thermal signature, the hyperflexed and tightly bound posture, the joint-release cuts, and the fact that people in the same arc of the world were still doing it in living memory. The authors filmed Dani and Pumo funerals in Papua in 2019: the limbs bound to the trunk, the body seated above a continuous fire, smoked until it turned entirely black. Among the Anga of Papua New Guinea the process runs about three months.

Three months.

That is the number worth sitting with, and it is also the number the bones do not contain. Nothing in a crystallinity index tells you how many days a fire burned. Three months is borrowed from the Anga and carried backwards twelve thousand years on the strength of a resemblance — an illustration, not a measurement. The restraint I extended to the heat has to be extended to the calendar.

What the bones do carry is the harder half of it anyway. Low heat that never consumed the body, reaching bone only where the flesh was thinnest, means a fire held close and held long — and every duration past a few hours has the same shape. The technique itself is nearly trivial: bind, elevate, smoke. Any group that can dry meat can dry a person; there is no metallurgy here, no natron, no priesthood, no state. What is not trivial is that somebody had to come back. Not once — again, and then again, well past the point where anything was still visibly happening. That is not a technology. It is an obligation, staffed in shifts, by people who had other things to do. Ninety is the vivid number. Not once is the load-bearing one.

Which is why the Egypt comparison — 7,500 years later, and the framing every write-up reaches for — gets the story backwards. Egypt did not invent mummification. Egypt named it, and the name went into our filing system attached to pharaohs, embalmers, and a theology of the afterlife. Once mummification was filed under state religion, it became something you'd only think to look for in a society that had one. Hunter-gatherers in the Guangxi hills were the wrong kind of people to be doing it, so for most of a century the bones came out of the ground and went into drawers as flexed burials — an unusual posture, catalogued and set aside.

Nothing about the bones changed. The question did. And the question had been shaped for a hundred years by a category that took its definition from its most powerful practitioner — which meant the evidence that arrived without a state attached had nowhere in the filing system to be interesting. Nobody hid these bones. They were labelled, and the label was the burial.

So what actually crossed twelve thousand years, eleven sites, and the whole distance between southern China and the New Guinea highlands? Not a body — bodies decay eventually. Not a technique; techniques are easy. It is tempting to say a custom was handed down unbroken across that span, and that is the weaker of the two available claims, as well as the one the evidence supports least. Many peoples arriving separately at the same practice — bind the body, hold the fire, stay — is the larger statement. It would mean the obligation was not inherited. It would mean it keeps being reinvented, because it follows from something about being the ones left behind.

Either way, it is still running. It was filmed in 2019.