later. Instead there was a patchwork of urban-centred zones of varying size along
connective but often narrow canal or river-branch corridors, and zones of irrigated
agriculture, apparently interspersed with marsh and steppe deemed difficult or
unsuitable (sometimes too saline) to justify irrigating.
(Adams 2008 : 15 , 1978 )This differentiated pattern of irrigation also implies that there were limitations to
the extent of either provincial or royal authority which may not have spread into the
more remote areas of the countryside (Adams 2008 : 16 ).
WATER SUPPLY AND AGRICULTURE
The terminology from mathematical texts, school texts and the ‘Farmers Instructions’,
when combined with knowledge of modern irrigation techniques and ethnography,
enable plausible reconstructions to be made of ancient hydraulic works (Pemberton et
al. 1988 ; Postgate 1990 ; Hunt 1988 ). It is necessary to be wary, however, not only
because some features of the Sumerian landscape will have been significantly different
from those of the recent past (Zettler 2003 ), but also because the term ‘water works’
rather than ‘irrigation canals’ may be more appropriate for much of the terminology
because many activities were focused on flood control, not just irrigation (Powell 1988 :
162 ; Civil 1994 ).
Powell ( 1988 ) subdivides Sumerian water works into three broad groups:- Canals and irrigation ditches
- Dams and barrages
- Wells, cisterns and reservoirs.
Channels, canals, banks and irrigation ditches
The term pa 4 /pa 5 –(r) refers to an irrigation ditch (Civil 1994 : 109 ), but because the state
authorities were more concerned with labour expended on digging canals, the texts
refer to the volume of soil up-cast and the size of the resultant levees. Therefore
evidence of canals derives mainly from references to the mounds of up-cast alongside
(ég), rather than the ditch itself (Civil 1994 : 109 ; Pemberton et al. 1988 : 212 – 213 ). For
example, at Girsu, texts refer to small embankments with cross-sections ranging from
0. 5 x 0. 5 to 1. 25 x 2. 5 m (Ibid.: 118 ), whereas around Umma cross-section areas ranged
from 6 x 0. 5 m to 9 x 2. 5 m (Ibid.: 118 ). The latter represents a canal of substantial size,
especially because there would probably be two banks, one on each side. In addition,
those parts of the levees that were vulnerable to erosion or to rising flood waters were
reinforced with reeds or rushes (ú-sag 11 : Ibid.: 121 ).
Canal construction was often commemorated, and the king Rim-Sin refers to the
excavation of the canal known as Simat-Erra as follows: ‘its two embankments/levees
are like mountains’. This is an apt description for anyone who has seen the landscape
associated with major canals which dominate the topography, often rising above the
ancient city mounds. Other texts give a flavour of the landscape itself, by referring to
levees with tamarisks, poplars or ash trees (Ibid.: 113 ), a phenomenon also mentioned
by Guest ( 1933 ) for twentieth-century Iraq.
–– Tony J. Wilkinson ––