Somebody Filled Enclosure D In One Season. It Was Kitchen Rubbish.

ŞANLIURFA, Turkey — Forty-one point two kilograms of animal bone.

That is what came out of nine columns of dirt cut floor-to-surface through the fill of Enclosure D at Göbekli Tepe this spring, sampled under the site director's supervision and weighed in Ankara. Sixty-three per cent of the identifiable fragments carry butchery marks.

The rest of the column is hearth ash, flint debitage and burnt snail shell.

Two laboratories that did not know the other had been hired read the same nine columns and agreed on how the fill arrived. Quickly. One dry season, or something very close to it.

What was being tested, and what was not

That the enclosures were deliberately buried in antiquity is settled. This desk does not dispute it and nobody quoted in this story disputes it.

The open question was how long the burying took. The answer usually given is a range of centuries — fill creeping in as walls slumped and the hill weathered over generations.

That is testable, so we paid to test it.

A block of undisturbed sediment is lifted whole, dried, impregnated with resin and sliced to thirty microns. Under a polarising microscope you can see not only what arrived but what happened between arrivals.

Time leaves marks. A surface open through one wet winter grows a crust of fine washed-in silt. A surface open for a decade grows rootlets. A surface open for a century grows a soil.

Nine columns, two laboratories

Ghent and Ankara, working duplicate sections cut from the same blocks. Neither was told the other existed until 3 July, when we told both on the same morning.

Dr. Ingeborg Marchetti-Osei in Ghent counted 31 depositional laminae between the terrazzo floor and the modern backfill. Dr. Yusuf Ekinci-Brandt in Ankara counted 29, and calls the gap "a counting convention, not a disagreement".

Both report the same three absences: no weathering horizon in any column, no rootlet channels, no graded winter silt.

"If a surface stands open through one Şanlıurfa winter, I see the winter," Marchetti-Osei said. "I looked for a winter in nine columns and there is not one."

What they filled it with

The fill is not soil and it is not building rubble. It is refuse.

Hearth ash in every column. Flint debitage — the chips struck off in making tools, not the tools themselves. Burnt snail shell in quantity: Dr. Hâle Yıldırım-Poole puts column six at roughly nineteen thousand cooked shells to the cubic metre.

And the bone. Mostly gazelle and aurochs, cut, chopped, and in four cases sawn.

"This is not the fill of a quarry," Yıldırım-Poole said. "It is a kitchen floor, moved."

The people who buried the enclosure look very much like the people who had eaten in it, and they buried it with what they had already thrown away.

The 214 pages

Separately, we obtained the modern conservation file that circulates online as proof the site is being re-buried in our own time.

It runs to 214 pages and we have published all of them.

They are purchase orders. Geotextile membrane. Sieved local sand to a specified grain size — seventy-one tonnes of it between 2015 and 2024. A frost-spalling report: fourteen pillars examined, spalling on nine, worst on the south faces.

The membrane exists so the modern layer can be taken off again. That is written into the specification, on the page after the order.

The column headed vacancies

The hardest page in the file is the staffing return.

Nine hectares. Three permanent conservation staff. One seasonal contract of nine months, not renewed at the date of the return.

The site is about ninety-five per cent unexcavated, a figure quoted constantly and almost always as a scandal. Part of it is deliberate policy, and the director explains that below. The rest is arithmetic. There is no establishment to dig it with, and the size of the establishment that would be needed is written down, on page 197, in a column headed vacancies.

What the director says

Professor Emine Doğan-Halvorsen agreed our sampling before she knew what it would show, has read this story entire, and disagrees with the headline over it.

Her reply runs below, unedited. It makes the strongest argument against us that anybody has made: that micromorphology is good at ruling out centuries and poor at distinguishing one summer from eleven, and that our own Ghent report says so on page 41 while our headline does not.

This desk accepts that and prints it above her signature and its own.

What survives the objection is the material. Whatever the burial took — a season, or a decade of returns — the thing carried up the hill and tipped into the enclosure was supper.