waterfowl (ibid.: 302 , 303 ). ED III texts found in Girsu of the é-mihousehold, headed
by the wife of the ruler of the state of Lagash, detail produce from a number of
dependent laborers, including fishermen. Renamed the é-Bauunder Urukagina of
Lagash, of the approximately 1 , 200 members of the productive household, 100 were
listed as fishermen, and another 125 as oarsmen, pilots, longshoremen, and sailors
(Maekawa 1974 ; Van De Mieroop 1987 ). Economic activities included fresh- and salt-
water fishing, and fish and dried fish brought into the household were both deposited
in the store-room and issued as purchase goods to merchants acting on behalf of the
household (Postgate 1994 : 114 , 202 ).
This accumulated evidence suggests that, while palm groves, gardens, temples, kilns,
and other institutions were consolidated on turtlebacks and levees away from seasonal
inundation (by peoples well accustomed to thorough exploitation of wetlands),
concomitant with intensified agricultural production, reed and other marsh products
were also intensively harvested to underwrite urbanizing consumption. The multiple
canal off-takes cutting through the relict levee abutting site ES 156 clearly directed water
flow into the alluvial basin adjacent to the site – but notinto any apparent field
irrigation system. Instead, the water flow seems designed to augment catchment into
what is now a desiccated wetland. The surface morphology of the area is directly com-
parable to desiccated habitation areas on the seasonally inundated edges of massive,
permanent marsh reed beds such as those of the al-Khuraib marsh south of Amara
(Figure 1. 6 ).Separating the twins: UR III and beyond
By painstakingly reconstructing late third millennium BCtravel and shipping itiner-
aries from the city of Umma, Piotr Steinkeller has shown that a watercourse thought
by textual scholars to represent “the eastern branch of the Euphrates” (Jacobsen 1960 )
was known at Umma as the Tigris (Idigna) – but the major channels of the twin rivers
still flowed so closely together that direct interconnection was maintained, possibly just
south of Mashkan-shapir (Steinkeller 2001 ).^7 At Umma itself, texts show clearly the
continuing importance of marshland resources like reeds (Steinkeller 1987 ), fish
(Englund 1990 ), fowl, pigs, and even trees (Heimpel 2011 ). They also record quotas
for production of marsh-based products like reed, bitumen, boats, mats, and stan-
dardized fish baskets (de Genouillac 1920 : 603 – 606 ). However, upstream, the primacy
of Isin, Kish, and Babylon during the early to mid-second millennium BCdemarcates
progressive westward (Euphrates) and eastward (Tigris) channel succession. As the
climate dried and became more seasonalized, that succession privileged the (to those
cities, proximate) Euphrates as a source of irrigation water, fostering the pearls-strung-
through-the-desert view handed down through later historical periods (Figure 1. 7 a).
CONCLUSION
A hallmark of nineteenth-century beliefs about non-urban landscapes was that marshes
are inherently diseased, sodden wastelands, and that the appropriate effort of good
government was to transform them into cultivated agricultural land. This emphasis
on the importance of transforming “waste” marshes for “useful” agricultural endeavor
was especially operative during the formative period of Mesopotamian archaeology.–– Physical geography ––