SITE MORPHOLOGY AND ARCHITECTUREThree main site types provide a cross-section of third millennium BCnorthern
Mesopotamian settlements: small tells, large complex tells, and tells comprising a dis-
tinct Upper Town and Lower Town.
Small tells are the most variable and unpredictable in function and yet the most
useful for reconstruction of economic dynamics. Tell Raqa’i on the Middle Khabur, for
instance, was not a simple village despite its size of less than 1 hectare; it had grain
storage facilities, a shrine and a tight group of houses (Curvers and Schwartz 1990 ;
Schwartz 2000 ; Schwartz and Curvers 1992 ). The associated cluster of similar sites of
EJ I–II date (Tell Atij, Tell Gudeda, Kerma, Kneidig, Ziyadeh) presents a model of
directed agricultural production, surplus management and processing. Questions
remain about the grain’s source (small-scale irrigated fields along the Middle Khabur
or larger fields further north), its intended destination (north to Brak, south to Mari,
or storage by and for a local nomadic population), and its mode of management,
whether corporate or elite (Hole 1999 ; Riehl 2006 ). In the better-watered upper Tigris,
excavation at Telul eth-Thalathat V also revealed a contemporary grain storage facility
(Fukai et al. 1974 ), perhaps connected to Tell al-Hawa or another urban centre nearby.
This northern agricultural management picture is almost completely missing from the
archaeological record of contemporary Sumer, where administration is assumed from
texts but is inadequately represented in architecture.
Tell Brak provides a stark contrast to Raqa’i: it is a multi-period massive mound with
complex morphology that has close Sumerian cousins, such as Nippur. Mozan, Tell al-
Hawa and Nineveh’s Kuyunjik are parallels in the north. In EJ I–II, with Brak only
approximately 15 – 20 hectares, known public buildings are limited to a small single-
room shrine (Area HS; Matthews 2003 ) with thick walls, benches and a niche. Its small
scale and relatively easy accessibility match shrines at Raqa’i, Arbid and Kashkashok
and argue against full-time specialised priests or significant power over people or
resources. The one- to two-room rectangular plans of these shrines are at odds with the
large temple complexes of Sumer. Other EJ I–II architecture at Brak is irregular and
incompletely excavated (a similar situation pertains at many sites; Roaf 2003 ). By EJ
IIIB, the centre of the then 50 – 60 hectare site housed a massive institutional building
(Area TC; Emberling et al. 1999 ; Emberling and McDonald 2001 , 2003 ). The excavated
portion of this structure may be a provisioning area, with ovens, storage and facilities
for grinding grain and possibly brewing beer (Emberling and McDonald 2001 ). The
building’s main function may be secular or religious; the core rooms have not been
revealed. Nagar is known to be the primary Upper Khabur kingdom in the mid-third
millennium BC, with power over Nabada/Beydar (see below); this seems to be secular
and economic control, rather than religious.
In EJ IV, Brak was taken over by Akkadian kings, who built two extensive
religious–administrative complexes at the south and north city edges (Areas SS and FS;
Oates et al. 2001 ) and a ‘palace’, also at the south (Mallowan 1947 ). Mud-bricks
stamped with Naram-Sin’s name in the latter leave no doubt it was a southern-imposed
and commissioned construction. Its large courtyards and narrow rooms suggest a
storehouse or barracks rather than a palace. Contemporary houses and a scribal school
were exposed to its east (Area ER; Oates et al. 2001 ). The highly visible placement of
these public buildings made a power statement both within the urban landscape and
–– Augusta McMahon ––