What did we build before we had cities?

Göbekli Tepe in south-eastern Türkiye contains massive T-shaped limestone pillars, some over five metres tall and carved with animals, arranged in circular enclosures. The site dates to roughly the tenth to ninth millennium BCE, built by people who were not yet farming.

It was inscribed as a UNESCO World Heritage Site on 1 July 2018. Geophysical survey found enclosures across the whole mound and put the probable total at twenty or more. By 2012 eight had been excavated, two of them down to the floor.

Nevalı Çori, now under a reservoir, had a building with T-shaped pillars of the same kind, and Jerf el Ahmar and Tell Abr 3 in Syria, which had communal buildings of their own, have produced carved objects whose animals closely echo Göbekli Tepe’s. It is a tradition, not a single anomaly.

Çatalhöyük, first settled around 7100 BCE and some 540 kilometres to the west, reached somewhere between three and a half and eight thousand people, with painted walls, figurines and the dead buried under the house floors. Ancient genomes show that in its early centuries the people buried in one house were mostly close kin, and that later they mostly were not. There is no sign of a ruling class.

The Antikythera mechanism, recovered from a shipwreck in 1901, is a geared bronze device for predicting astronomical positions and eclipses. Imaging published in 2006 reconstructed how its gears worked and doubled the number of its inscriptions that could be read. Nothing of comparable mechanical complexity is known from the following thousand years.

In 2012 the excavators called Göbekli Tepe a hill sanctuary and not a settlement, on the grounds that no domestic buildings had been found and the hill had no reliable water. Work since then has turned up round houses, an activity area with hearths and bead making, a burial cut under a floor, and a channel carved in the bedrock near a large rock-cut pit that may have held rainwater. The team now reads the rectangular buildings known since the 1990s as houses too.

One of the project’s two heads now writes that those finds have “exposed the weaknesses of the temple-narrative”, and that the archaeologists themselves are the ones who put the “World’s First Temples” into circulation. The interpretation is being argued about by the people digging it.

Klaus Schmidt, who excavated the site, proposed that feeding the workforce for construction on this scale put a strain on hunting and gathering that domestication relieved: that monuments produced agriculture rather than the other way round. The ranges of the wild founder crops overlap in this part of Upper Mesopotamia.

The site has produced grinding tools by the thousand and no large storage, which fits food made for immediate use at seasonal work feasts. The hypothesis is serious and it is not established, and the reading of the site it grew out of is the one now under revision.

The pillars carry foxes, snakes, scorpions, boars, aurochs, gazelle, wild ass and birds, alongside arms and hands that make the pillars themselves stylised human figures. There is no key. Interpretations are constrained by context and plausibility and little else, which is why incompatible ones coexist.

The enclosures are full of limestone rubble, flint, and animal bone smashed for marrow. For two decades the excavators read that fill as deliberate, ritual burial at the end of each building’s life.

A review of the evidence by the current team turned the question over: the fill in the main excavation area looks like settlement debris that slid in from the higher slopes of the mound. The team’s own paper of 2019 had still read the layers in one building as rapid backfilling, with two carved heads set into them on purpose. How much was rite and how much was landslide is not settled.

Lidar and geophysical survey keep finding structures beneath forest and soil. Airborne laser scanning of more than two thousand square kilometres of lowland Guatemala mapped a dense, interconnected Maya landscape under the canopy, and the geophysics at Göbekli Tepe found enclosures across the whole mound.

These sites may show that large coordinated projects do not require permanent elites or states. Çatalhöyük was thousands of people with no ruling class the excavators can find, and the gazelle bones at Göbekli Tepe point to mass hunts between midsummer and autumn, which reads as people gathering to work and then dispersing.

The dispersal half of that is the part now in doubt, since houses have been found on the mound. A reading can be right about the hierarchy and wrong about the coming and going.

The Antikythera mechanism suggests a tradition of precision gearing that left almost no other trace. Whether that tradition was narrow and fragile, or broad and simply unpreserved, cannot currently be decided, which is a more interesting situation than either answer.