Mesopotamian settlement organization was established by the later Early Dynastic and
remained quite stable after that time.
Moreover, this chapter could not be written based on excavated data alone since no
small or even medium-sized Sumerian sites have been excavated. However, the high
resolution of modern satellite imagery means that we can recover sub-surface archi-
tectural traces over much larger areas and for a much more varied inventory of sites
than is the case with architectural data. The project to analyze these sources of
settlement data is still in its infancy and here too, there are little data for the early stages
of urban growth or for the latter part of the third millennium,^1 and the process of
digitizing the traces we see has barely begun.
URBAN BEGINNINGS
Unlike the later Early Dynastic period, the available excavated data from the earliest
stages of Mesopotamian urbanism remain limited, and the only fourth millennium BC
buildings excavated in Iraq are temples. Fortunately, data relating to the Uruk period
from beyond Iraq, as well as data from high-resolution satellite imagery, can be used
to flesh out a view of Mesopotamia’s earliest centers.
The key Mesopotamian institution of the temple goes back to Neolithic times. At
Eridu a small shrine grew into a large temple built on a platform during the course of
the fifth millennium. By the end of this sequence, the Eridu temple did not differ from
later temples found at other cities, and although no associated settlement has been
found at Eridu, there is an extensive cemetery, suggesting perhaps that the residents
of neighboring villages chose to be buried close to the temple (Safar, Mustafa and Lloyd
1981 ). Beyond Eridu, other ‘Ubaid sites tend to be very small and made up of loosely
spaced tripartite houses with rows of rooms on each side of a large central space (Roaf
1989 ), though some have more complicated T-shaped structures (Jasim 1985 , 1989 ).
For the crucial Protoliterate and Early Dynastic I periods, the data are more
equivocal. We do know quite a bit about Protoliterate temples, but have no excavated
houses.^2 The temples were found in two dense clusters at Uruk itself, and another such
temple has been uncovered at Tell ‘Uqair (Lloyd and Safar 1943 ). All Protoliterate
temples are broadly comparable in plan with the early example from Eridu, but differ
in that they were associated with larger settlements. The temple at Uqair is part of a
double-mounded site which now measures some 11 ha. (see below), whereas the
protoliterate sherd scatters at Uruk (Finkbeiner 1991 ) cover some 225 hectares.^3
There is one largely excavated Uruk period town, Habuba Kabira, but it is located
in northern Syria (Strommenger 1980 ), not southern Iraq. It is associated with the so-
called Uruk expansion whereby Uruk-style material is found beside and within local
settlements far from the Mesopotamian heartland. These data are generally understood
as reflecting an effort by the population of the city of Uruk to control trade routes up
the Euphrates and, to a lesser extent, the Tigris rivers (Algaze 2008 ). Habuba Kabira
(Strommenger 1980 ) is the one urban-sized excavated settlement associated with this
phenomenon and its material culture is overwhelmingly Uruk in nature.^4
Although a cursory examination of the plan of Habuba Kabira might suggest the
same crowded spaces known from later Mesopotamian cities, the analysis by Vallet
( 1996 ) has shown that this city was both less dense than its later counterparts and that
much of the architecture was likely not used for residence.^5 Habuba Kabira households
–– The organisation of a Sumerian town ––