Göbekli Tepe: monuments before pottery

Scene · reconstructed from the paperworkPicture it. A limestone ridge in south-east Türkiye, 1994. A survey team is walking the ground: Klaus Schmidt, M. Akman and M. Morsch. The hill has been noted before, but this time they recognise what it is. According to the German Archaeological Institute[dai-project], that survey is the moment Göbekli Tepe was rediscovered. The next year, 1995, the excavations began, and they have continued[dai-project] ever since.

There is no night of in this file in the usual sense. There is no witness, no radar, no flash in the sky. The first witness is the ground itself, and the people who dig it. That makes this an unusual folder for the Archive: the testimony is archaeology, and the documents are excavation reports and peer-reviewed papers.

What came out of the hill was a set of enclosures built from T-shaped stone pillars. The tallest stand up to 5.5 metres. Some are plain; some are richly carved. The DAI[dai-project] places them in the Pre-Pottery Neolithic A and B, from about 9500 or 9250 to 8000 or 7750 BC. That label matters. It means the people who quarried, carved and raised these pillars had not yet made pottery.

The man who led the work for its first two decades was Klaus Schmidt. He remained head of excavations at Göbekli Tepe until his unexpected death on 20 July 2014, as the DAI recorded[dai-schmidt-2014]. The project continues under the institute, with Lee Clare named as project lead on its project page[dai-project].

In 2018 the site was inscribed on the UNESCO World Heritage List[unesco-1572]. That is recognition, not interpretation. UNESCO does not settle what the pillars were for. It records that the place matters.

The first reading of the site was bold. For its first twenty years Göbekli Tepe was read as a primarily ritual site, a mountain sanctuary, built by people who came together to raise pillars rather than to live. Newer finds of domestic activity changed that. The DAI[dai-project] now describes it as a settlement with a pronounced ritual component.

So the first night of this file is quieter than most. A survey team on a hill. A date that came from digging and radiocarbon, not from a story. And a question that has followed the site ever since: who built monuments this large, this early, and why? The date is the solid part of that question. The why is where the arguments begin, and the rest of this file follows them.

The paper trail of Göbekli Tepe is unusually good for an ancient site, because the excavators keep publishing. The base document is the DAI project page[dai-project]: rediscovery in 1994, excavation since 1995, Pre-Pottery Neolithic dates, pillars up to 5.5 metres, and the move from sanctuary to settlement with a ritual component.

The excavators also write for the public. Their blog, The Tepe Telegrams[tepe-telegrams-2017], carries news and notes from the research staff. It is where they give plain-language versions of their published papers, and it is where they state numbers that matter for any theory about the carvings: more than sixty T-pillars are known.

In 2020 the paper trail gained a geometry study. Gil Haklay and Avi Gopher, in the Cambridge Archaeological Journal[haklay-gopher-2020], analysed the floor plans of the enclosures. They found an underlying geometric pattern and concluded that three of the large stone enclosures were planned and initially built as a single project. The university's release[tau-press-2020] put it as a knowledge-based, top-down process carried out by specialists, in a complex built 11,500 to 11,000 years ago.

That is a strong claim, and it is a mainstream one. It does not need anything unknown. It says that early Neolithic communities could plan large architecture before they made pottery. The Archive marks it as a published analysis, not the last word: other readings of the enclosures differ, and the question of whether they were built as one plan or added over time stays open.

The other half of the paper trail is the argument about the carvings. In March 2017 Martin Sweatman and Dimitrios Tsikritsis published a peer-reviewed proposal[sweatman-2017] that animal carvings on Pillar 43, the so-called Vulture Stone, match star asterisms and form a date stamp for 10950 BC, plus or minus 250 years. In May the excavators replied in the same journal[notroff-2017].

In 2024 the University of Edinburgh[edinburgh-2024] announced a follow-up study by Sweatman, reading V-shaped marks on a pillar as a possible solar calendar recording a swarm of comet fragments around 10,850 BC. The paper itself[sweatman-2024] is listed in this file as a reference only: the publisher blocked our fetch, and we have not read it.

The inventory, then: an excavating institute's record, a World Heritage listing, an excavators' blog, a geometry paper, and a published exchange about comets. That is a better paper trail than most ancient mysteries get.

The turn in this file came in the spring of 2017, when the comet reading met the excavators.

Sweatman and Tsikritsis made a specific, testable claim. In their paper[sweatman-2017] they argued that the Vulture Stone is a date stamp for 10950 BC, plus or minus 250 years, and that the site may be the smoking gun for a cometary encounter at the start of the Younger Dryas cold period. That gave the idea a number, and a number can be checked.

The Göbekli Tepe research team checked it. Their reply[notroff-2017] in the same journal, and its plain-language version[tepe-telegrams-2017], made two points. First, the proposed date is 700 to 1000 years older than the oldest radiocarbon date for Enclosure D, the enclosure that holds Pillar 43. Second, with more than sixty T-pillars known, choosing a few carvings to match star patterns is arbitrary. Their sharpest line is short: any available date for Göbekli Tepe rather marks the end than the beginning of the Younger Dryas.

That is the turn. A radiocarbon record that can be resampled met an astronomical reading that depends on which carvings are chosen. The Archive prefers the record that can be resampled, and says so in the margin.

The argument did not stop there. In 2024 Edinburgh[edinburgh-2024] reported Sweatman's new study on a possible solar calendar and a comet swarm around 10,850 BC. The release is hedged throughout, with may and could. The Archive records it in the timeline and the Other Side box, and keeps the paper in to-verify until it can be read.

There is a second, larger argument that uses Göbekli Tepe as evidence: that a lost advanced civilisation, destroyed at the end of the Ice Age, taught its builders. The Society for American Archaeology[saa-letter-2022] answered that idea in 2022, in an open letter about a television series. A century of fieldwork, it wrote, has produced no evidence for an advanced global Ice Age civilisation. The letter does not discuss Göbekli Tepe by name, and the Archive says so.

So the turn leaves the file in a clear shape. The date: confirmed, from excavation and radiocarbon. The planning: a published analysis says three enclosures were laid out together. The comet: a peer-reviewed proposal, rejected by the excavators on dates. The lost civilisation: no evidence reported. Our Other Side[sweatman-2017] box gives the comet case at full strength.

What is still open at Göbekli Tepe is mostly the ordinary work of archaeology, and that is the good news.

First, how people lived there. The DAI[dai-project] moved from reading the site as a sanctuary to reading it as a settlement with a pronounced ritual component. How permanent the occupation was, and how ritual and daily life shared the hill, is still being dug. Each season can change the answer.

Second, how the enclosures were built. The Haklay and Gopher study[haklay-gopher-2020] argues for one planned project behind three large enclosures. Earlier readings saw them added over time. The geometry is published; whether it reflects a single plan is a question for more dates and more digging.

Third, what the carvings meant. The animals and symbols on the pillars are real and well recorded. What they meant to the people who carved them is interpretation. The comet reading[sweatman-2017] is one interpretation; the excavators[tepe-telegrams-2017] reject it on dates and on the selection of pillars. Other readings exist. None can be checked against a written record, because there is none.

Fourth, the 2024 calendar paper. The Edinburgh release[edinburgh-2024] describes it; the paper[sweatman-2024] is not yet read by the Archive. Until it is, its argument stays in the timeline and in to-verify, not in the analysis.

Fifth, the popular books. Graham Hancock's treatment of Göbekli Tepe in Magicians of the Gods (2015) is widely read and widely cited. We have not opened it, so it stays in to-verify. The archaeologists' letter[saa-letter-2022] answers the general idea, not the book.

The honest position: Confirmed, monumental enclosures with T-pillars up to 5.5 metres, built in the Pre-Pottery Neolithic from about 9500 BC, as the excavating institute[dai-project] records. Published, a geometric plan behind three enclosures. Proposed and rejected by the excavators, a comet date on the Vulture Stone. Not supported by evidence, a lost Ice Age civilisation.

• How permanent was occupation?
• Were the enclosures one plan or many?
• What did the carvings mean?

The folder stays open, and the file stays a draft, because the digging has not stopped.