Göbekli Tepe is a Neolithic site on a limestone ridge in Şanlıurfa Province, southeastern Turkey, where people built great stone enclosures between about 9500 and 8000 BCE, more than 6,000 years before the first stage of Stonehenge. Its T-shaped limestone pillars, carved with foxes, boars, snakes and vultures, are the oldest known megaliths in the world. The builders lived before farming had fully taken hold, which is why the site rewrote the story of how monuments, religion and settled life began. Excavated from 1995 by the German archaeologist Klaus Schmidt, it joined UNESCO's World Heritage List in 2018, and only about a tenth of it had been dug as of 2021.
The name means "Potbelly Hill" in Turkish; local Kurdish speakers call it Girê Mirazan, "Wish Hill", and treated it as a sacred place long before archaeologists arrived. The mound, or tell, is about 15 metres high and covers 8 hectares near the village of Örencik, in the foothills of the Taurus Mountains, overlooking the Harran plain and the headwaters of the Balikh River, a tributary of the Euphrates.
Klaus Schmidt's 1994 Visit to a Field Near Örencik
The site had been seen before, and misread. A 1963 survey led by Halet Çambel of Istanbul University and Robert Braidwood of the University of Chicago recorded the mound and collected flint tools from its surface. The American archaeologist Peter Benedict identified them as Neolithic, but took the tops of the buried pillars sticking out of the soil for medieval grave markers. Farmers had been ploughing the hill for generations, piling up stones from their fields, and some had tried to break up the pillars as ordinary obstacles.
In October 1994 Klaus Schmidt, who had already worked at the nearby Neolithic site of Nevalı Çori, went looking for similar places. Asking in local villages about hills with flint, he was led to the mound by Mahmut Yıldız, whose family owned the land and had already reported finds turned up by the plough. Schmidt recognised the flat slabs from Nevalı Çori: these were not gravestones but the heads of prehistoric megaliths. He began digging the following year, and the first large T-shaped pillar soon came out of the ground. Yıldız joined the excavation and later became the site's guard.
Schmidt directed the work for the Şanlıurfa Museum and the German Archaeological Institute (DAI) until his death in 2014. Since then the project has been run jointly by Istanbul University, the Şanlıurfa Museum and the DAI under the Turkish prehistorian Necmi Karul, with Lee Clare coordinating the German side. The newer campaigns dig less and document more, concentrating on conserving what is already exposed.
Dating Enclosures A to D, 9500–8000 BCE
Getting a firm date took two decades. The first radiocarbon results, published in 1998, came out 500 to 1,000 years later than Schmidt expected. His team argued that the soil filling the enclosures had been brought in from elsewhere when they were deliberately buried, and turned instead to organic material trapped in wall plaster. Later research dropped both ideas: the deliberate-burial theory has fallen out of favour, with landslides now blamed for much of the fill, and plaster dates turned out to be skewed by old wood. The revised chronology, built on new samples, places the whole sequence between 9500 and 8000 BCE, spanning the late Pre-Pottery Neolithic A and early Pre-Pottery Neolithic B.
The new sequence has eight building phases over at least 1,500 years. In Phase 1, in the second half of the 10th millennium BCE, the first versions of the four excavated enclosures, labelled A, B, C and D, went up alongside round and oval houses. In Phase 2, early in the 9th millennium, new walls were built inside them, and the first monolithic T-shaped pillars were set into those walls. Later phases added benches, inner walls and rectangular buildings. The enclosures sat more than 10 metres below the highest part of the settlement, and landslides kept burying them. One at the end of the early PPNB filled Enclosure D with household rubble; terrace walls built to hold the slope failed, and a second slide probably ended its use late in the 9th millennium BCE.
T-Shaped Pillars, Vultures and Foxes in Stone
The enclosures are circles 10 to 30 metres across. Thick walls of unworked stone hold a ring of T-shaped limestone pillars, and two taller pillars face each other in the centre. Builders cut the slabs from bedrock about 100 metres away, using flint points to work the soft limestone. Geophysical surveys show at least 16 more enclosures still buried, each with up to eight pillars, implying nearly 200 pillars in all.
The pillars are bodies. Several have arms carved down their sides, with hands, a belt and a loincloth, so the shaft reads as a torso and the crossbar as a head. The two central pillars of Enclosure D are the clearest example. Most of the other carving is animals: snakes, foxes and wild boar above all, with gazelle, wild sheep, wild asses, ducks, scorpions, spiders and vultures. Where the sex can be seen they are male, and many are shown in aggressive poses. Pillar 43 in Enclosure D includes a headless male figure. Human figures are rare, and there is only one certain depiction of a woman, on a slab.
What is missing is just as telling. The people who built Göbekli Tepe hunted gazelle and other game, but the carvings generally ignore the animals the community depended on, such as deer, and show formidable creatures instead, with no hunting scenes and no wounded prey. Schmidt and the zooarchaeologist Joris Peters suggested the images were protective or totemic rather than a single mythology. A reassembled "totem pole" 192 cm tall, carved with three stacked figures and snakes, came from one of the structures and dates to the early PPNB. In 2017 the team also found fragments of human skulls with deliberate cut marks, a sign of the skull cult known from other Neolithic sites such as Jericho and Çatalhöyük.
The Temple-Before-Cities Debate
Schmidt read the site as a ritual centre, the reading behind its popular label as the "world's first temple." In his view mobile hunter-gatherer bands from across the region gathered on the hilltop at certain seasons, built and rebuilt the enclosures, feasted on gazelle and scattered again, and their seasonal visits were timed to when game was most plentiful. He saw the building of Göbekli Tepe as a step towards the later rise of urban civilisation, and popular accounts boiled this down to a temple before cities. The researchers Haklay and Gopher added a further twist, arguing that Enclosures B, C and D were planned together as one complex laid out on an equilateral triangle.
Since his death, the evidence has pulled the other way. Findings published in 2019 identified houses, a water supply and heavy cereal processing. More than 7,000 grinding stones have been found, and rain-fed channels filled cisterns cut into the bedrock that held at least 150 cubic metres of water. People seem to have lived here for at least part of the year, eating cereals that may have been wild or early cultivated forms. The landscape was wetter than today: open steppe grassland full of wild einkorn, wheat and barley, crossed by wild sheep, goats, gazelle and wild asses, with pistachio and almond trees supplying about 90 percent of the charcoal found on site. The enclosures may still have been special places, but many researchers now see them as communal buildings inside a village, where ritual and daily life mixed, as at other large Neolithic sites.
The labour argument has shifted too. Schmidt, comparing the pillars with Easter Island's moai, thought hundreds of workers were needed, perhaps drawn in by a religious elite. Other estimates suggest seven to fourteen people could move a pillar with ropes and a lubricant, and experiments on site suggest a group of 12 to 24 could have built all the exposed later structures in under four months, within reach of one extended family or village. Claims that the site was an astronomical observatory have been largely rejected by the excavation team.
Karahan Tepe, the Taş Tepeler and UNESCO 2018
Göbekli Tepe is no longer alone. It belongs to a cluster of about a dozen Neolithic hilltop sites in the Şanlıurfa region known as the Taş Tepeler, the "Stone Hills", many of them with large stone buildings and T-shaped pillars. The best known is Karahan Tepe, about 46 km to the east inside the Tek Tek Mountains National Park. Discovered in 1997 by Bahattin Çelik of Harran University, it covers almost 10 hectares, had more than 266 visible pillars recorded by 2000, and may date from 10,000 to 9500 BCE, slightly earlier than Göbekli Tepe. In September 2023 excavators found a 2.3-metre statue of a seated man there, echoing the 1.8-metre Urfa Man found in Şanlıurfa in 1993. Older still is Körtik Tepe, dated to 10,700–9250 BCE, whose art may anticipate the whole tradition.
UNESCO added Göbekli Tepe to the World Heritage List in 2018, citing it as "one of the first manifestations of human-made monumental architecture." Turning a working dig into a visitor site brought its own row. That same year Çiğdem Köksal Schmidt, an archaeologist and Klaus Schmidt's widow, said concrete and heavy machinery used to build a new walkway had damaged the site; the Ministry of Culture and Tourism said no concrete had been used and no harm done. Meanwhile geophysical surveys show at least 20 large enclosures in the mound, most of them still unexcavated.
Göbekli Tepe sits at the northern edge of the arc of early farming lands traced in our guide to the Fertile Crescent. For a sense of how much later Europe's best-known megaliths came, compare it with Stonehenge, raised from about 3100 BCE on Salisbury Plain.