ties – notably the Màppilla of Kerala, the people of
Lakshadweep (both in India), and of the Mal-
dives – continued to remain so even after adopting
Islam. Here, men have rights of usufruct over prop-
erty held in common by the matrilineal group,
while the transmission of individually owned prop-
erty is governed by the principles of the Shàf≠ì
school of SunnìIslam (Dube 1969, 1994, Kutty
1972). Traditional residence patterns were matrilo-
cal, with the system of visiting husbands. In recent
decades, with increasing interaction with dominant
non-Muslim communities in India and Muslims in
the Gulf, matrilocality here is giving way to patrilo-
cality among both Muslims and non-Muslims.
Simultaneously, elsewhere in South Asia, long-term
male labor migration within and beyond the sub-
continent is forcing many patrilineal communities
to become “matri-weighted” (Naveed-i-Rahat 1990)
in practice, though not in ideology.
While matri-centered socioeconomic units have
long existed in many Muslim and non-Muslim
agricultural and herding communities – for exam-
ple in Rajasthan (India) and Sindh (Pakistan) –
where men entrust the cash earned for safe-keeping
to mothers and wives, among most traditional
Muslims, Hindus, and Sikhs direct female par-
ticipation in production processes outside the
precincts of the household economy depends on
and is an indicator of family wealth and social sta-
tus and the concept of (male) honor. The less the
visibility of such participation, the higher the
status, especially in urban areas (Jeffery 1979,
Mandelbaum 1988, Vreede-de Stuers 1968), but
also in many rural settings (Kotalová 1993,
Lindholm 1982, Naveed-i-Rahat 1990).
Status among South Asian Muslims is also
closely related to concepts of “origin,” pedigree
and blood (nasab), caste (jàt/jàtì) and lineage (birà
deri, khàndàn, kaum/qoum, kul, nukh are among
terms used, depending on locality). This is despite
the fact that for several centuries, community
boundaries remained fairly flexible, because of
individual and group conversions and immigration
from surrounding regions. Only in the early six-
teenth century did the Màppilla, for instance,
evolve into an endogamous community. Almost
identical, language-specific kinship terminology
and (especially female) first names among Muslims
and non-Muslims, as well as the numerous com-
munities who blended Islam and non-Brahmanical
Hinduism – Bishnois, Guptis, Hussaini Brahmins,
Pranamis, and so forth – may indicate a degree of
intermarriage. This continued between Muslims
and Hindus into the late Mughal and early colonial
south asia 339period among the ruling elite (for example, the Raj-
puts) and certain lower status professional groups
(for example, classical musicians). In remoter regions
(for example, Ladakh) Muslims and Buddhists
intermarried till even the late twentieth century,
when general religious fanaticism encouraged by
state policies intervened. In Bangladesh no laws bar
inter-faith marriages, but social pressure to convert
to Islam is strong. Pakistani law prohibits a Muslim
woman from marrying a non-Muslim and a Mus-
lim man from marrying a Hindu, while in Nepal a
Hindu loses his/her legal (and social) identity on
marrying a Muslim (Gaborieau 1995). In India,
under the Special Marriage Act 1954, Muslims and
non-Muslims may marry and retain their respective
religions; to remain valid such a marriage must be
monogamous. In reality, however, when Hindu
girls marry Muslim boys, increasingly the couples,
their respective families, and their possessions are
targeted by gangs affiliated to various Hindutva
organizations; the couples are forcibly separated,
the boys charged with abduction and the girls put
away – all this usually with the connivance of the
local police.
Ritual sororal and sibling bonds across religious
boundaries played major roles in subcontinental
history and still unite individuals and groups in vil-
lage India. The cognitive flexibility regarding com-
munity boundaries did not clash in practice with
the primacy accorded to traits transmitted through
paternal blood and maternal milk (Rao 2000).
Female mediated “fictive kinship” across religious
and community boundaries was most clearly em-
bodied in the precolonial elite institution of the wet
nurse, which existed well into the early decades of
the twentieth century. Based on the principle that
through the milk, the child would imbibe the traits
of the community/caste of the wet nurse, it united
families, while also following the Islamic rule gov-
erning ri∂à≠a.^1
Well ensconced in the overarching hierarchy of
the local version of the caste (varna-jàtì) system –
with the partial exception of the (Indian adminis-
tered) Kashmir Valley, some Pakistani “tribal areas”
(but see Barth 1971), and parts of Bangladesh
(Bertocci 1995) – Muslims also observe the norms
of caste stratification (albeit often with less extreme
adherence to notions of purity and pollution) that
prevail in socioeconomic interaction within any
given micro-region (Ansari 1960, Eglar 1960).
Islamization of local communities and immigrant
Muslim groups may have introduced a certain
spirit of egalitarianism, but they also brought along
other principles of ranking, which had been well