The term Saivism here refers to a number of distinct but historically related systems comprising theology, ritual, observance and yoga, which have been propagated in India as the teachings of the Hindu deity Siva.

A Saiva is one who practices such a system. To understand the term to mean ‘a worshipper of 5>iva’ or ‘one whose deity is 5>iva’ is less precise; for a Saiva may well be a worshipper not of Siva but of the Goddess (Devi).

The scriptural revelations of the Saiva mainstream are called Tantras, and those that act in accordance with their prescriptions are consequently termed Tantrics (tantrika). The term tantra means simply a system of ritual or essential instruction; but when it is applied in this special context it serves to differentiate itself from the traditions that derive their authority from the Vedas (direct revelation: sruti) and a body of later texts that claim to be Veda-based (indirect revelation: smrti). This corpus ofsruti and smrti prescribes the rites, duties and beliefs that constitute the basic or orthodox order and soteriology of Hindu society. The Tantrics however saw their own texts as an additional and more specialised revelation (yis'esas'astra) which offers a more powerful soteriology to those who are bom into this exoteric order. The Tantric rituals of initiation (diksa) were held to destroy the rebirth-generating power of the individual’s past actions (karma) in the sphere of Veda-determined values, and to consubstantiate him with the deity in a transforming infusion of divine power.

The Saivas were not the only Tantrics. There were also the Vaisnava Tantrics of the Pancaratra system, whose Tantras, consid­ered by them to be the word of the deity Visnu, prescribed the rituals, duties and beliefs of the devotees of Vasudeva in his various aspects (vyuha) and emanations (vibhava, avatdra).

The production of Tantric revelations was not limited to those who accepted the supernatural authority of the Vedas. It went on, though on a much smaller scale, among the Jains, while the Buddhists added an enormous Tantric corpus to their canonical literature during the period c. 400-750 ce. By the end of this period the system of the Tantras, called ‘the Way of the Diamond’ (Vajrayana) or ‘of Mantras’ (Man­tray ana), was generally recognised among the followers of the Greater Way (Mahayana) as the highest and most direct means to liberation (nirvana), and its esoteric deities were enshrined in the monasteries as the high patrons of the faith. The Tibetans, who received Buddhism at this stage ofits development, preserve, in the Tantric section of their canon, translations from the Sanskrit of almost 500 revealed texts and over 2000 commentaries and explanatory works. Of these more than three quarters concern Tantras of the most radical kind, those of the Higher and Supreme Yogas (Yogottara-tantras and Yogdnuttara-tantras).

All these Tantrics were similarly related to the tradi­tional forms of religion, the Buddhists to the monastic discipline and the Vaisnavas and Saivas to Vedic orthopraxy. They were excluded by the traditionalists because they went beyond the boundaries of these systems of practice. But the Tantrics themselves, while excluding these exclusivists, included their systems as the outer level of a concentric hierarchy of ritual and discipline.

In those communities in which it was possible or desirable to add to the exoteric tradition this second, more esoteric level, there might be forms of the Tantric cult in which this transcendence entailed the infringement of the rules of conduct (dcdra) which bound the performer of ritual at the lower, more public, level of his practice.

The followers of these cults, even in their undomes­ticated form, should not be seen as rebels who rejected a ritualised social identity for a liberated cult of ecstasy. This popular view of Tantrism over­looks the highly-structured ritual contexts (Tantric and non-Tantric) of these un-Vedic practices. A person who underwent a Tantric initiation (diksd) was less an anti-ritualist than a super-ritualist. He was prepared to add more exacting and limiting ritual duties to those which already bound him. Indeed he has much in common with the most orthodox of Hindus, the srauta sacrificer, who transcended the simple and universal domestic rites pre­scribed in the secondary scripture (smrti) to undertake the great rituals of the primary and more ancient revelation (sruti). Though the srauta and the Tantric occupied the opposite ends of the spectrum of Hinduism they shared the character of being specialists of intensified ritual above the more relaxed middle ground of the smartas (the followers ofsmrti). This similarity is carried through into their doctrines of liberation from rebirth (moksa). Both thesrauta tradition articulated by the Bhatta Mimamsakas and the Tantric represented by the Saivas stood apart from the mainstream by holding that the mere performance of the rituals prescribed by their respective scriptures is a sufficient cause of final liberation (see pp.

Un-Vedic though it was, the Tantric tradition was destined to have a far greater influence than the srauta on the middle ground. While the srauta tradition all but died out, the Tantric came to pervade almost all areas of Indian religion. The distinction between the Vedic and the Tantric in refigion continued to be crucial, and it was drawn in such a way that the Tantric continued to be the tradition of a minority; but what was called Vedic here was essentially Tantric in its range of deities and liturgical forms. It differed from the properly Tantric principally in its mantras. This became the chief formal criterion: in Vedic worship (puja) the actions that compose the liturgy were empowered by the recitation of Vedic mantras drawn from the Rgveda and Yajurveda rather than by that of the heterodox mantras of the Tantras. At the same time these de-Tantricised reflexes of Tantric worship were non-sectarian. While properly Tantric worship was more or less exclu­sive, being emphatically centred in a particular deity, Vedic domestic wor­ship was inclusive. Its most typical form is thepahcayatanapujd, the worship of the five shrines, in which offerings are made to the five principal deities, Siva, Visnu, Surya, Ganapati and the Goddess (Devi). The scriptural author­ity for these neo-Vedic rituals was provided by the indirect revelation (smrti) and in particular by the ever-expanding Puranas, though the exact text of worship is generally a matter of unascribed tradition. Such worship is there­fore called smarta (‘s/Mrti-based’) orpauranika (‘Purana-based’), or, where the properly Vedic or s'rauta tradition is completely absent, simply vaidika (‘Veda-based’). Its form in a particular community is largely the product of the history of the various Tantric traditions within that community.

What follows falls into two parts. In the first I pre­sent Saivism on the evidence of what survives, mostly unpublished, of its earliest scriptural sources. In most of this material the domestication of which I have spoken above has yet to begin. We shall therefore be looking in the main at traditions of Saiva asceticism. Here the married man (grhastha) is altogether absent or at best subordinate. In the second part (pp. 690-704) we shall consider what happened to these traditions when they were domesti­cated. Our evidence here is the literature of the Saiva community of Kashmir from the ninth century onwards.

This second body of texts is our earliest datable and locatable evidence for the Saiva traditions. How much earlier than these Kashmiri works the scriptures to which they refer were composed cannot be decided yet with any precision. At best we can say that the main body of these early Tantras must have been composed between about 400 and 800 ce. To this we can add some relative chronology. We know even less about where these Tantras were composed or about the areas within which they were followed. However I incline to the view that when these traditions became the object of sophisticated Kashmiri exegesis between the ninth and thir­teenth centuries they were widely represented throughout India. It is certain that the Kashmiri and the Newars of the Kathmandu valley looked out on much the same distribution and interrelation of Saiva Tantric cults at this time, and it is highly probable that each community inherited these traditions independently by participating in a more widespread system, which may have included even the Tamil-speaking regions of the far south of the subcontinent.

The Kashmiri exegesis considered in the second part is a local tradition of much more than local impact. In a very short time it was acknowledged as the standard both in its theological metaphysics and in its liturgical prescription among the Saivas of the Tamil south. This was the case in the Saiva Siddhanta (see pp. 691-2), the Trika (see pp. 692-6), the Krama (see pp. 696-9), and the cults of Tripurasundari (see pp. 688-9) and Kubjika (see pp. 686-8). Consequently, while the Hindu culture of Kashmir declined in influence and vitality after the thirteenth century with large-scale conver­sion to Islam and periodic persecutions, the Tantrics of the far south con­tinued the classical tradition, and through their many and outstanding con­tributions to Tantric literature guaranteed it a pan-Indian influence down to modern times. These southern and subsequent developments are unfortu­nately beyond the scope of this essay.