household. A single entrance means access was shared by men and women, and most
of the rooms are accessible via each other and the central court. The width of rooms
was determined by the length of the timber available (which for domestic buildings was
limited to local palm or willow), so that rooms were often rectangular rather than
square. In many of the larger houses there was sometimes a long oblong chamber
adjoining the courtyard that perhaps represented the main reception room (Delougaz
et al. 1967 : 276 – 277 ). Internal windows were apparently rare but when present were
very small – an example discovered at Tell Asmar was only 25 cm across (Delougaz et
al. 1967 : 154 ) – and might be closed with grilles (see below) that were presumably
designed to exclude the sun, sand and dust in summer and conserve heat in winter. For
the same reasons, internal doors were also small and varied in size from 60 to 90 cm
(Delougaz et al. 1967 : 277 ). They were placed near one corner of a room rather than
in the middle of a wall.
Due to the lack of long roof beams, the rooms in homes were small and dimensions
of about 1 by 2 to 5 metres were not uncommon; a result of this may have been that
activities took place in the courtyard. It is because of this lack of domestic space for
productive activities that city dwellings may have differed from rural dwellings. In the
latter, space allowed for rooms to be ranged around a larger, open compound, as may
have been the case, for example, in the village site of Umm al-Jir (Crawford 2004 : 107 ).
Although houses with compounds are known from urban settings in northern Iraq,
such as at the late Early Dynastic III–Akkadian period site of Tell Taya (Reade 1973 ,
1982 ), this layout of buildings does not appear to be a feature of Sumerian towns. The
rooms of many rural houses were also probably given over to storage and shelter for
animals as well as for living; in this respect, many city dwellings may have been
different. Nonetheless, it is possible that courtyards in Sumerian city houses would
have been given over to some production (e.g., to house textile looms). Additional
living and storage space in the summer might have been provided by the roof which
ethnographic evidence from southern Iraq suggests would have been flat; access may
have been by stair from the courtyard. The reconstruction of second storeys of homes
at Ur by Leonard Woolley has been shown to be flawed (Crawford 2004 : 114 ).
Although the plan of domestic buildings remained essentially the same, there were
some developments from the Akkadian period onwards. Toilet rooms were introduced
into homes where, with the use of water in these spaces, baked bricks and bitumen
became more common. In addition, a damp course of baked bricks, which was only
used occasionally in homes in the Diyala, was increasingly utilised, becoming a
common feature in houses of the early second millennium BCat Ur (Delougaz et al.
1967 : 151 ). Baked brick was also used more widely for thresholds, pavements and door
arches. One significant new feature of the Ur III period houses at Ur was the presence
of shrines in prosperous houses; these rooms lay parallel with the main living space,
located in the most private area of the house, furthest from the door and work areas.
The shrines, which may have related to household deities and ancestors, were generally
associated with a baked brick burial vault (see below).
INTERNAL FITTINGS, FIXTURES AND FURNITURE
Only occasionally is it possible to explore the changing nature of a home over time
as well as its consistent features. One such example, among the finest and largest–– Paul Collins ––