since the resolution of our imagery is not sufficient for the identification of doorways.
At both Lagash and at Khafajah, houses in very different parts of these cities have the
same range of room widths, providing a strong suggestion that the physical divisions
between residential areas occupied by people of different classes so prevalent in today’s
cities did not exist in ancient Mesopotamia.
Although room widths may be consistent within individual cities, they differ
considerably between cities. Four Early Dynastic urban centers have sufficient areas of
domestic housing for it to be possible to arrive at some understanding of overall house
size: excavated domestic areas at Tell Asmar (Delougaz, Hill and Lloyd 1967 ), surface
scraping and excavations at Abu Salabikh (Postgate 1990 , 1994 ), both excavations and
high-resolution satellite data for Khafajah and satellite data alone for Lagash. These
data show that the houses in the Diyala region (Khafajah and Asmar) were all built at
a significantly smaller scale than those elsewhere, with Abu Salabikh having the largest
houses of all. Since these are all walled settlements, it is likely that these differences
reflect different degrees of crowding (Postgate 1994 : 63 ). Khafajah is the site where we
have evidence of the most extreme crowding, a pattern which is seen throughout the
site when we include the traces visible in the high-resolution satellite imagery. Indeed,
if the fairly small areas of excavated housing were typical of the site as a whole, it would
appear that the smallest houses were associated with the earlier phases of settlement and
that the slightly more expansive houses–though still very small when compared with
other sites–were built at the same time as the Temple Oval (Delougaz, Hill and Lloyd
1967 : pl. 2 – 14 ). One possibility is that the construction of the new temple–which is
located at the edge of the site–might have been accompanied by an expansion in the
overall size of the settlement as a whole.
Fortification walls were typical of Early Dynastic cities and doubtless were a
contributing factor to the degree of crowding outlined above. At both Abu Salabikh
and Lagash, walls circumvallate the individual component mounds that made up the
ancient city, but no single wall surrounds the whole. At Abu Salabikh, part of the wall
around the main mound has been recovered through excavation and surface scraping
(Postgate 1990 : 96 , 98 ) and is further visible in the satellite imagery. This separates it
from the contemporary South Mound, located across a watercourse and which most
likely contained the palace (Postgate 1990 : 104 – 106 ). We also see traces of a fortification
wall all the way around the northern sector of Lagash in the satellite imagery, even
though there was much more to this very large site. The data at hand are insufficient
to indicate whether these walls imply real political divisions within these early
cities–between a temple city and a palace city for example–or if instead they were
designed to protect the residential parts of these early cities, built on still quite low
mounds, from flooding.
At Ur, the only Early Dynastic evidence comes from the cemetery. Although best
known for the important group of royal tombs, these were located within a much wider
area given over to the burial of the less exalted, whose grave goods were few (Woolley
1934 : 135 – 146 ). A similar cemetery was also found at Kish, where once again quite
elaborate graves were found only partially separated from more modest examples
(Moorey 1978 : 61 – 75 ). The discovery of graves beneath houses at sites like Khafajah
(Delougaz, Hill and Lloyd 1967 ) and Abu Salabikh (Postgate 1990 : 99 – 102 ) suggest that
either some sites at some times had cemeteries whereas at other times people were
buried beneath their houses, or perhaps that some people were buried in cemeteries
–– The organisation of a Sumerian town ––