Sumerian

The oldest pantheon the atlas records is organised as a council. The Anunna gather, and what they do when gathered is fix destinies, which is a decision rather than a prophecy. Sumerian gods have jurisdiction, and jurisdiction is allocated.

At the head stands An, the sky, who is senior and increasingly inactive. The god who actually governs is Enlil, whose temple at Nippur, the Ekur, is where the assembly meets and whose approval legitimates kingship across a landscape of competing cities. The pattern is withdrawal at the top and delegation below it, rather than a father overthrown by a son.

Every city has its own god and its own temple, and the god is understood to live there. Enlil at Nippur, Enki at Eridu, Inanna at Uruk, Nanna at Ur: the divine map and the political map are the same map, and when a city fell, the literature says the god had left. This is a pantheon assembled from a landscape of city cults rather than imposed on one.

Human beings were made for the labour the gods no longer wished to do, and that low estimate of the human position is stated flatly in the literature. Nonetheless the line was crossed at the top: kings of the Third Dynasty of Ur were written with the sign that marks a divine name, and hymns were composed to them as gods. Deification of the living ruler is a Sumerian development, and it did not survive the period that produced it.

Sumerian is a language with no known relatives, and it was recovered because the Babylonians and Assyrians who inherited the writing system treated it as a classical tongue and produced bilingual material for their own students. Modern understanding of the oldest pantheon in the atlas rests, in large part, on ancient scribal homework.

The literary corpus comes overwhelmingly from one place and one moment: the scribal schools of Nippur in the early second millennium, where students copied a curriculum of hymns, disputations, laments and narratives about the gods. This means the Sumerian literature that survives is what was being taught several centuries after the culture that produced it was at its height, selected by teachers for what a scribe ought to know.

Alongside it stands administrative material in enormous volume, tens of thousands of tablets recording offerings, rations, personnel and temple estates. These are dry and they are decisive, because they show which gods actually received support, at what scale, and where, independent of what any hymn claims.

God lists, which pair deities with explanations and equivalences, add a third layer: the pantheon as the scribes themselves systematised it.

Two limits are worth stating. Much of the corpus is still being edited, and readings change. And the ordinary religion of people outside the temple economy is largely invisible, because it was not the kind of thing anyone wrote on a tablet and filed.