The Sumerian King List begins with a sentence that sounds like pure theology: when kingship came down from heaven, kingship was in Eridu. Its first ruler, Alulim, is credited with twenty-eight thousand eight hundred years. Nothing about that is a historical record.
And yet, when Iraqi archaeologists dug the mound of Tell Abu Shahrain in the late 1940s, they found beneath the latest temple a sequence of earlier ones — each built on the levelled remains of the last, on the same spot, going down through around eighteen successive structures to a single small mudbrick room a few metres square, with an offering table and a niche, dating to roughly 5400 BCE. Sacred continuity on one patch of ground for four thousand years, and the tradition that this was where it all started turns out to be pointing at something real. It is the rare occasion when Mesopotamian myth and Mesopotamian stratigraphy agree.
Enki's house on the water
The city belonged to Enki, and its temple was the E-abzu, the house of the deep. The theology is site-specific: Eridu sat at the edge of the southern marshes, where the water table stands close to the surface, and the god of the underground fresh water was understood to live directly beneath his own sanctuary. The Apsu is not a metaphor here; it is the water you reach by digging.
The excavation produced a detail that no text mentions and that changes the picture. The temple levels contained very large deposits of fish bones — offerings, in quantity, sustained across periods. The god of the marsh water was fed from the marsh, and the fish-cloaked sages he sends into the world make a good deal more sense once you know his temple floor was covered in them.
The last building phase was a ziggurat begun under Amar-Sin of Ur around 2050 BCE and never finished. Its unfinished state is the city's condition in miniature: by then Eridu was already declining.
The city that stops being a city
Why it emptied is reasonably well understood, and it is environmental rather than military. The Euphrates channels shifted; the water that made the site viable moved away; and centuries of irrigation without adequate drainage concentrated salt in the fields, a process visible in the agricultural records as a shift from wheat to salt-tolerant barley and then to declining yields of that. By around 2000 BCE Eridu was substantially abandoned, and it stayed so — which is exactly why it survives as well as it does. Nobody built a Babylon on top of it.
The site is desert now, several kilometres from any water, and it was inscribed on the World Heritage list in 2016 together with the southern Iraqi marshes and the other great southern city mounds. A Lament for Eridu survives among the Sumerian city laments, mourning a place that was already a ruin when the poem's audience heard it.
Prestige after death
What is striking is how completely the abandonment failed to reduce its standing. Eridu remained the type of the ancient and holy city for another fifteen hundred years. Its name is used as an honorific: Babylon itself was called Eridu in some of its own inscriptions, the way a new institution takes the name of an older one to borrow its authority. Kings claimed Eridu's ancient rites. Scribes credited it as the source of the arts of civilisation, brought by the seven antediluvian sages Enki sent, and the Sumerian flood narrative is often called the Eridu Genesis on the strength of the role the city plays in it.
The mythic transfer of primacy is told as a drinking party. In Inanna and Enki the goddess of Uruk visits Eridu, is entertained, and is given the me — the offices and institutions of civilisation — by a generous and increasingly drunk host, then loads them onto her boat and leaves. Enki's attempts to intercept the boat all fail. Read literally it is a comedy about a hangover; read as history it is a memory of the moment when the older southern shrine-city was overtaken by the younger, larger city upriver.
Questions the mound raises
Was Eridu really the first city?
It has the earliest continuous temple sequence yet excavated in southern Mesopotamia, which is not the same claim. Urbanisation in the region is a process rather than an event, other Ubaid-period settlements are known, and the honest position is that Eridu is the oldest cult site of the great southern cities and that Uruk, not Eridu, is where a recognisable city — dense population, monumental architecture, administrative writing, specialised labour — is first documented. The Sumerians themselves ranked Eridu first because it was oldest and holiest, and there is no reason to expect their ranking to match a modern definition of urbanism.
What does the King List's chronology mean?
It is not a chronology in any usable sense. The antediluvian kings are given reigns of tens of thousands of years, dropping to more plausible figures after the flood and to genuinely historical rulers later on — and the numbers themselves are the interesting part, since they are constructed from the sexagesimal system and appear to encode a scheme rather than a record. What the list is actually doing is asserting that legitimate kingship is a single office that has passed from city to city, held by one place at a time. That is a political claim about the present, dressed as a chronology of the past.
Is anything left to see?
A large low mound with the ziggurat platform still legible, in an empty landscape, largely unexcavated below the levels reached in the 1940s. It has suffered less than most Iraqi sites from the looting that followed 1991 and 2003, partly because it is remote and partly because it was never rich in the kind of small portable objects that the market wants. What it holds is stratigraphy rather than treasure, and the parts of it that matter most for the history of religion are still in the ground. How much of the tradition rests on excavations like this one is set out on the Mesopotamian overview.