however, the S ́u ̄dracategory has always included many of those dominant,
martial, and landed castes that would elsewhere in India have been reckoned in
the royal or Ks.atriya varn.a. This seems consonant with the fact that possession
worship received royal patronage in Can.kam times (Hart 1975), and that many
socioeconomically and politically dominant castes in south India have contin-
ued to patronize such practices into recent times, often in worship of their own
family, clan, and caste deities (kula-de ̄vata ̄). This, I believe, is historically linked
with a qualitatively different ethnographic patterning one finds in possession
cults across south India as compared with the north. There are often greater
institutional investment, social organizational stability, and caste-based heredi-
tary lines of possession-priesthoods across south India, whereas such cults tend
to be more idiosyncratically personalistic, popularly contingent, and institu-
tionally marginal across much of north India. At any rate we find that while
many (especially urban) sections of the dominant and upper castes have
Brahmanized in terms of embracing temple culture and Sanskritic norms of
worship, many landed, village-based dominants have also continued their ances-
tral investment in local temple and shrine establishments that give prominence
to possessed worship. This is all the more the case for the bulk of those lower
castes who were traditionally excluded from temple entry, with the combined
result that possession often retains a level of ritual formalization and socially
structured entitlement one would associate with higher-caste priestly office in
the north. Teyyam worship exemplifies these points with great clarity, and
further suggests how tantric doctrine and practice mediated between the
Brahmanical and more localized religious complexes in the south.
We have little by way of hard historical documentation for teyyamfrom the
premodern, but the internal references, traditions, and subject-matter of the
teyyamliturgies as oral literature, conjoined with the areal spread of its ritual
forms, particular gods, shrine networks, and social entitlements, would all indi-
cate that it existed in substantially its contemporary form from at least the period
of the late medieval kingdoms of northern Kerala, from perhaps the fifteenth
century. The deities and their shrines distribute with clear ritual and social
cleavages in conformity with boundaries of the old kingdoms of Ko ̄ lattuna ̄t.u,
Nı ̄le ̄s ́varam, and Kumbal.am (covering today’s Kannur and Kasargode Districts).
Major events recounted in the liturgies go back to probably the thirteenth
century, and the distribution of cognate dance and possession forms, both
further south, throughout Kerala, and further north, in the bhu ̄taworship of the
Tul.u country in Karnataka (Claus 1978, Nambiar 1996), suggest the much
older distribution of possession worship, as clearly fundamental to the religious
organization of medieval society.
The dominant kingly line in the teyyamregion of Kerala, that of the Ko ̄ lattiri
Ra ̄ja, still celebrates its own lineage goddess in the form of a teyyam, in concert
with an amalgam of other local teyyamdeities at its royal, family temple in a
popular regional festival. This is despite the fact that, of course, the Ko ̄ lattiris
also worshipped at and extended their patronage to a variety of exclusively
higher-caste, Brahmanical temples and establishments throughout their
the teyyam tradition of kerala 309