ways in Northern Iraq (Wilkinson 1993 ), likely represent the routes taken by cattle on
their daily treks into the marshes. If so, then, like the smaller Early Dynastic I villages
where excavations have been carried out, even larger settlements like Umm al-‘Ajaj,
must have devoted considerable space to housing cattle overnight, a situation which
would also likely have prevailed at Uruk, given its position on a “bird’s foot” delta
between land and water (Pournelle 2007 : 43 ).
We must consider the possibility that large, medium and small Mesopotamian
settlements predating the late Early Dynastic period might have had a quite different
character from the later cities. The extraordinary high density of later Mesopotamian
cities–as is now made clear by our satellite imagery–leaves no space whatsoever within
the city for keeping herd animals like cattle and sheep. Indeed, beginning in the later
Early Dynastic period there are numerous textual sources pertaining to the manage-
ment of cattle and sheep by the state. There is sufficient space, not only at the West
Mound at Abu Salabikh and at Sagheri Sughir but also at Hububa Kabira, for sheep
and cattle to have been housed within the walls at night, an indicator perhaps that
population densities within settlements may have been much lower in these early
settlements. We can also contrast the fourth millennium sacred precincts at Uruk, with
their large open areas and complete absence of nearby domestic structures, with their
later Early Dynastic equivalents which were hemmed in by residential structures (see
below).
This discussion has been, by necessity, tentative, though some consistent strands
may be identified. The first is the importance of temples and the absence of any
evidence for large secular buildings that might have served as palaces or even high-
status residences for priests during the Protoliterate and Early Dynastic I periods. The
data do make clear that the division of at least the larger settlements into multiple
mounds had already taken place, and there is some evidence suggesting that the
temples and perhaps their dependents may not only have been located on their own
mounds, but that there may have been a directional preference for the northwestern
area of the site for these buildings.^8 As is to be expected, domestic structures dominated
these settements,^9 but these were made up not of the courtyard houses familiar from
the late Early Dynastic onward, but by a continuation of the tripartite houses known
from the ‘Ubaid period, probably grouped into larger compounds as has been docu-
mented at Habuba Kabira. All of these data suggest that southern Mesopotamian
settlement before the later Early Dynastic period was characterized by low density,
probable lack of public buildings other than temples and perhaps the only occasional
presence of fortification walls.^10 These sites would have been as reliant on marsh
products as on agricultural fields and orchards, and the data suggest that sheep and
especially cattle were likely housed within these settlements overnight.
LATER EARLY DYNASTIC SETTLEMENT
It is during the Early Dynastic period that the less dense, temple-focused sites of the
fourth millennium evolved the recognizable characteristics of the Mesopotamian city:
the institutional complexity of palace and temple and the dense urban fabric based on
courtyard houses.–– The organisation of a Sumerian town ––