Northern urbanisation then saw a discontinuity towards the end of the third
millennium (EJ IV–V); site size was reduced and site numbers decreased. This
disruption began in the early Akkadian period but was at least partially contemporary
with significant climate change (see below) in the Late Akkadian and Post-Akkadian
periods. However, the severity of that change and human reaction to it are debated, and
the north was not entirely abandoned in the Post-Akkadian period. Although
settlement size was reduced, occupation persisted at Brak, Mozan, Hamoukar, Tell al-
Hawa and Tell Taya, while construction of single large buildings atop relatively small
sites (Chagar Bazar, Arbid) implies the presence of a local political authority.
One of the most important aspects of northern Mesopotamian surveys has been
identification of ‘hollow ways’, linear features that are the remains of past paths of
traffic and goods transport. They connect settlements to agricultural hinterlands or to
other settlements (Ur 2003 ; Wilkinson 1993 , 2000 a, 2003 ); they originate in the third
millennium BCurban uplift. Hollow ways may appear as dark lines on satellite imagery
and, more rarely and seasonally dependent, as visible depressions with standing water
and/or denser vegetation from ground level. The identification of hollow ways has led
to development of a more nuanced treatment of Mesopotamian landscapes, shifting
from a model of a tell-dotted agricultural base to one brimming with lived experience
of resource clusters, quarries for flint or clay, work spaces and prominent natural or
man-made places. Studies of field manuring and other ‘off-site’ sites provide vital
balance to tell-focused reconstructions of space use; and these land-use aspects are
almost absent from our knowledge of Sumer’s landscape.
SITE-SIZE HIERARCHY AND CITY-STATES
Third millennium BCoccupation in the north exhibits a classic tiered pattern similar
to that of Sumer. It is centred on urban settlements of 50 to 155 hectares; for example,
Beydar, Brak, Chuera, Hamoukar, al-Hawa, Khoshi, Leilan, Mozan and Taya. There
is also an important story of mid-range settlements ( 10 – 40 ha) told by sites such as
Chagar Bazar, Tell Arbid, Tell Barri and Tell Billa. Smaller settlements ( 5 to less than 1
hectare), including Mohammed Arab, Tell Atij, Tell Bderi, Tell Raqa’i, Telul eth-
Thalathat V and Ziyadeh, add depth and detail to that story. However, there remain
sites that are difficult to classify; Ashur, for instance, has a major temple to an
important deity, but its third millennium BCsize and political importance are unclear.
The scarcity of texts in the north means that our reconstruction of these sites’ inter-
action is in some ways less detailed than our history of Sumer. We lack, for instance,
Sumer’s richly illustrated chapters on warfare and local dynasties. However, from
archaeology and settlement patterns alone, the political systems were similar: a network
of independent city-states comprising a central city and linked hierarchy of smaller sites
(Sallaberger and Ur 2004 ). Agricultural land was the primary local resource, but the
nomadic population was also a significant factor. Names of some city-states are known
from the few local and contemporary southern texts: Brak = Nagar, Beydar = Nabada,
Leilan = Shehna, Mozan = Urkesh.
–– North Mesopotamia ––