The Temple Before the Farm

The standard account of how civilisation began has an order to it, and the order is the argument. Agriculture produces surplus. Surplus permits specialisation. Specialisation permits monumental construction, priesthoods and organised religion. Farming first, temples later. This sequence was taught with confidence for most of the twentieth century.

Göbekli Tepe is in southeastern Turkey, on a limestone ridge near Şanlıurfa. It was surveyed in the 1960s and dismissed as a medieval cemetery. Klaus Schmidt began excavating in 1995 and recognised what the T-shaped stones protruding from the hill actually were.

The site consists of circular and oval enclosures built from massive T-shaped limestone pillars, some over five metres tall and weighing several tonnes, arranged in rings with two larger pillars at the centre. Many are carved in relief with animals: foxes, boars, snakes, scorpions, vultures, cranes. Some of the central pillars have arms and hands carved along their sides, which suggests they are stylised figures rather than supports.

The radiocarbon dates place the earliest construction in the tenth millennium before the common era, around 9600 BCE. There is no pottery. There is no metal. There is no evidence of domesticated plants or animals at the site during its earliest phases, and the animal bones recovered are from wild species. The people who cut, moved and raised these stones were hunter-gatherers.

That reverses the sequence. Whatever gathered enough people to build this, and to keep building for something like fifteen hundred years, it was not agricultural surplus, because there was not yet any agriculture. Schmidt's proposal, which remains a proposal, was that the causation ran the other way: the need to feed large gatherings assembled for construction and ritual may have driven the domestication of grain. Some of the earliest known domesticated einkorn wheat has been identified within a few dozen kilometres of the site.

The other unusual fact is the ending. The enclosures were not abandoned to erosion. They were deliberately backfilled with rubble, debris and bone, and buried. Someone put considerable effort into covering the structures, which is why they survived so completely and which nobody has satisfactorily explained.

Care is needed here, because Göbekli Tepe attracts overclaiming faster than almost any site in archaeology. It is not a lost civilisation. There is no evidence of writing, of metallurgy, of advanced mathematics, or of the technologies that popular accounts routinely attach to it. Interpretations of the carvings as astronomical records are contested and rest on chains of assumption that the evidence does not currently support.

What the site does establish is narrower and more useful than any of that. Hunter-gatherers, without agriculture, pottery or metal, organised labour on a monumental scale for purposes that appear to be ritual. That single fact broke a sequence that had been treated as close to settled, and it broke it with stone that anyone can go and look at.

The revision has been absorbed by the field with relatively little drama, which is itself worth noting. The sequence was not defended to the death. It was a working assumption that turned out to be wrong, and the discipline changed it.

The question that leaves is which of the currently confident sequences are working assumptions of the same kind, sitting undisturbed because the site that would break them has not been dug yet.