family and household 243
half of the nineteenth century, as an effect of Mohammed Ali’s attempt to introduce
a modern, centralized form of government, women were excluded from a whole
range of activities which had previously been significant sources of income. A new and
sterner patriarchal system made them socially and economically peripheral and increas-
ingly dependent on male breadwinners. Elite women were the ones mostly affected,
but all other women are likely to have suffered the same fate in different degrees.
As far as household structure is concerned, however, generalizing from the better
documented upper strata to the whole of the population is an arduous task. A still
different question is whether significant differences existed between rural and urban
settings. Some answers are provided by the population census of Egypt in 1848, the
oldest comprehensive statistical source covering all the inhabitants of this or indeed of
any other Ottoman country. The analysis of a large sample of 20 163 individuals,
representing the 256 679 inhabitants of the city, reveals that in Cairo “complex”
households accounted for only 30% of the total. The remaining 70% consisted either
of individuals who lived alone or of nuclear families, which explains a mean household
size of only 3.54 persons and “stands in contrast to villages and rural areas still domi-
nated by large, multinuclear households” (Fargues, 2003: 38).
A comparison of these figures with the largely-similar ones estimated by Duben
and Behar (1991) for Istanbul in 1907 would seem to point to the existence of an
urban pattern where the share of complex households was not negligible (of the order
of 30%), and yet far from being dominant in an urban society where “every man,
provided that he survived long enough, founded a household and established inde-
pendent living arrangements” (Fargues, 2003: 38). However, the need to resist hasty
generalizations is demonstrated by the very different pattern detectable in Damascus
in the early twentieth century, where “solitaries” were roughly as numerous as in
Cairo and Istanbul but the proportional weight of “complex” households was much
greater, and so was the mean household size (Okawara, 2003). Interestingly, in 1991
complex residential arrangements still accounted for about two-thirds of all house-
holds in the Muslim population of Damascus. By contrast, a 1980 survey of Cairo
revealed that complex households comprised only 18% of the sample. This points to
persisting or even increasing contrasts between different urban contexts in the same
region. It also confirms, on the other hand, that living alone or in small families was
and still is a common experience in cities and towns.
How do these small urban families cope in times of crisis, when illness incapacitates
men and women from fulfilling their responsibilities in a household, or old age makes
elderly people frail and vulnerable? These are some of the questions tackled by
Singerman in an ethnographic study she started in Cairo in that very year, 1980. She
found that a pervasive familial ethos not only prescribed that parents provide for
their children but also that the children provide for their parents when they
become elderly (Singerman, 1995: 83–84). Yet, large complex households structur-
ally capable of ensuring support to the elderly and other needy people were not to
be found in great numbers. Singerman discovered that along with, or instead of,
co-residence with close relatives, other sources of support could be resorted to, the
most important one being represented by “reciprocity networks” between kin and
neighbors. The very boundary between kin and neighbors was itself blurred. Rugh,
doing research on family in Cairo in the same years, was struck by the expandable
meaning of qarayib, a term which designates both neighbors and relatives stressing