trends. Social historians, frustrated by finding that
the records came from ruling households rather
than from peasants or commoners, felt free to
ignore them as unrepresentative. Only historians
interested in state formation studied the ruling
households of the region. Thus Blake (1979) study-
ing the Mughal Empire (1526–1858) found it nec-
essary to focus on the imperial household as a
microcosm of the “patrimonial-bureaucratic em-
pire.” Social anthropologists, on the other hand,
have written much on the household but have paid
little attention to religion (Vatuk 1989, 109). Since
neither the historians nor the anthropologists have
seen Islam as significant for the household, it is at
present only possible to outline a history of Muslim
households in the subcontinent.
Islam arrived in South Asia in the seventh cen-
tury. By the middle of the twentieth century, Sunnì
Islam had become the dominant school though
other variants of Islam maintained a strong pres-
ence within the Muslim community. This long his-
tory was marked by enormous diversity in the
ethnic-cultural practices that shaped Islamic insti-
tutions. The first Arab Muslim migrants were
joined by large numbers of Muslims from Central
and West Asia, and East Africa. The group and
individual conversion of South Asians to Islam
increased the diversity of belief and practice within
the community. Thus Muslims were linguistically,
ethnically, and socially diverse. Through the cen-
turies, connections developed between soldiers,
courtiers, scribes, and rural societies organized
around Islamic holy men, and class-rank distinc-
tions and hierarchies were elaborated within the
South Asian Muslim community.
Of these four main social groups, only women
from the courtly groups have left any significant lit-
erary traces. A comparative study of such records
allows us to simultaneously locate significant con-
tinuities in Muslim elite conceptions of “family,”
“ancestry,” and “kinship” as well as to glimpse the
tectonic shifts that had occurred by the twentieth
century. Structurally speaking, the multi-layered
household described by a sixteenth-century Mughal
princess in her History of Humàyùnresembles
those depicted in twentieth-century memoirs of
women from the landed gentry of Northern India.
But while the sixteenth-century household may
have contained the multiple wives of a Mughal
grandee, later elite households in the region were
rarely polygynous. In both periods, however, these
households contained multiple generations of con-
sanguineous kin, as well as large numbers of unre-
lated people. The wealthier households encompassed
large numbers of dependents who were either dis-256 household forms and composition
tant kindred or elderly kinswomen, or non-kin
recipients of charity. Since believing Muslims are
enjoined to give a portion of their income to char-
ity, the presence of unrelated or very minimally
related people within these households was itself a
public display of piety. But charitable intent not-
withstanding, such households were also organized
to extract, and differentially reward, labor services
of various kinds. It is always, therefore, difficult to
distinguish the dependent but free kinsman or
kinswoman from the dependent non-free servitor.
All accounts record an elaborate hierarchy of
female and male servants. Elderly supervisors and
wet-nurses were the most powerful and new young
slaves the least powerful. But the same accounts
also record change: thus wet-nurses often acquired
the status of foster-parents to children they had
suckled; and the children of wet-nurses were to be
regarded as foster-siblings by their charges. This
elevation would be registered during ceremonial
moments in the life cycle of the child and of the
household. The longer the record of service, the
greater the moral claim of such a servant upon the
attention and resources of the household.
With the spread of British colonialism in the nine-
teenth century, attention shifted to the gender-segre-
gated and secluded nature of women’s lives within
such households. While some accounts may be exag-
gerated, it is important to understand that seclusion
practices were being adopted by growing numbers of
Muslims as part of intensified social stratification.
While some ruling groups lost their wealth and fell in
status, families of scholars, artisans, and substantial
peasants began to claim higher rank. Since seclusion
had been an important symbol of social position,
upwardly mobile groups in the nineteenth century
also adopted this aristocratic practice. Purdah, or the
seclusion of women, accompanied claims to gene-
alogical depth, and/or “foreign” origins and served to
consolidate social prestige among Muslims.
Another social practice associated with a partic-
ular moment in the ascent of such a lineage was
increasing marital exclusivity. Anthropologists
refer to this as the adoption of endogamous mar-
riage for junior members. Endogamous marriages
were justified by the learned as required by kafà±a
(Arabic) or barabari(Urdu) (similarity of status of
bride and groom). The criteria for status, however,
differed between groups. Among the southern Indian
Muslim followers of the Shàfi≠ìlaw, the principal
criteria in the nineteenth century were descent, free-
dom, religion, and occupation. Some even required
a formal assent to a certain austerity in marriage
rites and celebrations. Simultaneously, the northern
Muslims following £anafìlaw applied the criteria