1948 , 1959 ) and visible through satellite imagery. In spite of the very small size of
Haradum and Tell Harmal, both houses and objects at these sites do not differ from
those found at their larger cousins (Malko 2006 ). Indeed, a comparison between the
object classes recovered from Haradum with those from Mashkan-shapir indicated
that only difference between the two sites were the royal inscriptions found only at
Mashkan-shapir, even though the latter is some sixty-five times the size of Haradum.
Moreover, the high density of these two small Early Dynastic settlements is matched
by seven early second millennium sites seen in the high resolution satellite imagery
(Figure 8. 8 ).
The similarities between the late Early Dynastic cities and villages and those dating
to the early second millennium BCstrongly suggest that Akkadian and Ur III
settlements differed little from their immediate forebears. The only question which
remains open is when city walls were reconfigured such that they surrounded the entire
site and not just each individual mound.
CONCLUSIONS
Cities–settlements with populations too large for everyone to know each other and with
an institutional complex unknown earlier–were first seen in Sumer. This chapter has
attempted to chronicle the evolution in settlement organization that brought this about.
Although the word “city” conjures visions of large groups of people living together, the
Mesopotamian data suggest that it was the institution of the temple that came first, with
high-density populations developing much more gradually. The early temples at Eridu
(and perhaps also Tepe Gawra) may have served as pilgrimage centers (McCorriston
2011 )–certainly as a burial center–for the surrounding population. Although Eridu itself
never gained population, as the Protoliterate proceeded the evidence suggests that
similar temples–especially but not exclusively those at Uruk–attracted ever larger
populations. The organization of these small centers suggests that the temples remained
somewhat remote, located far to the edge of the settlement in a part of town separated
from the majority of the inhabitants. Moreover, if cattle and perhaps sheep were still
housed within the settlements, population densities may not have been great.
Nevertheless, these fourth millennium BCsettlements were larger, more institutionally
complex and more populous than any other settlement elsewhere in the world.
It is in the middle of the third millennium BCthat we see evidence of the develop-
ment of true cities–indeed at this time it appears that virtually all of the settled popu-
lation were urban dwellers. These cities were walled and continued to be dominated by
temples, the most important of which retained their somewhat separate, peripheral
position within the urban framework. Moreover, the architectural details of these
structures suggest that they became more remote, more protected, as they incorporated
ever more daunting gatehouses (Delougaz, Hill and Lloyd 1967 ). By this time the
urban fabric was a dense mass of houses built around courtyards located along a maze
of small streets and alleyways. No room was left within the cities for domestic animals.
It was within this environment that we witness the growth of the second major insti-
tution, the palace, the earliest manifestations of which were located as far from the
temple as possible. Written records indicate that a third institution, the assembly, must
have existed by this time (Jacobsen 1943 ), but this has left no architectural traces that
we can recognize.
–– The organisation of a Sumerian town ––