of a different tradition might be quoted with approval, it is generally only
regarded as a truth emanating from the level of the cosmos from which it derives.
Later, esoteric teachings transcend the previous revelation.
While these traditions maintain a hierarchical structure in the classifica-
tion of revelation, as we will see, we nevertheless have in the S ́aiva understand-
ing of tradition an example of the decentralizing strategies of what we call
“Hinduism,” which should make us skeptical of the usefulness of the category
in a historical context. While S ́aiva authors were keen to make totalizing claims
about the universal truth of their teachings, the model of tradition shared by all
S ́aiva schools is inherently pluralistic in the idea of the guru lineage, while simul-
taneously being hierarchical in its assumption of a graded cosmos or ontology
and a graded teaching. These initiatory lineages have been extremely important
in the history of S ́aivism and have mostly been associated with groups of texts
called Tantra. But there has also been a more general temple S ́aivism associated
with Sma ̄rta brahmanism. As Sanderson has shown, the term S ́aiva is techni-
cally restricted to an initiate into one of the S ́aiva systems, while the term
Mahes ́vara has been used for a brahmin worshipper of S ́iva within the Sma ̄rta
domain (Sanderson 1988b: 660–4). It is Sanderson’s general mapping of these
systems in the early medieval period that I follow here (Sanderson 1985; 1988;
1988) although a more complete mapping of the traditions by him, which will
revolutionize our understanding of S ́aivism and the history of Indian religions
more generally, will have to wait (Sanderson, forthcoming). But before we trace
this history a few remarks on the earliest indications of reverence for S ́iva and
the development of S ́aiva traditions are necessary.
Early and Pura ̄nic S ́aivism
Some scholars maintain that the worship of S ́iva goes back many thousands of
years in the subcontinent to the Indus valley civilization, where steatite seals
have been found suggestive of a deity akin to S ́iva. The famous “Pas ́upati” seal
shows a seated, perhaps ithyphallic, horned figure surrounded by animals. Sir
John Marshall has claimed that this is a prototype of S ́iva as the yogin and
Pas ́upati, the Lord of animals (Marshall 1931: 52). But is not clear from the seals
that this is a proto-S ́iva figure and Asko Parpola has convincingly suggested that
the seal is in fact a seated bull, almost identical to figures of seated bulls found
on early Elamite seals of ca. 3000–2750 bc(Parpola 1994: 248–50). It may be,
of course, that elements of S ́iva’s later iconography – such as the crescent moon
in his hair – can be traced to this period but unless the Indus Valley script is deci-
phered, these seals can only be suggestive. There are early textual references to
Rudra, arguably a forerunner of S ́iva, one of whose epithets is “auspicioius”
(s ́iva), in the R.g Veda. Here three hymns are addressed to Rudra, the “roarer’. He
is clothed in an animal skin, brown, with a black belly and a red back. Even
at this time he is an ambiguous deity who is like a ferocious beast destroying
204 gavin flood