The Tantric Corpus
The scriptures themselves: what the Agamas and Tantras are, how the canon was counted and organised, and the individual works that survive.
The Tantric canon is not one book, and the tradition never agreed on how many books it had. Lists of twenty-eight Agamas and sixty-four Tantras are repeated everywhere, and the moment they are checked against the manuscripts the counts stop matching, the titles drift, and whole streams turn out to survive only as names quoted by somebody else.
These pages take the corpus as it actually is. What was written, what was lost, what each surviving work contains, and who counted it. The numbers in the titles are claims the tradition makes about itself, and they are treated as evidence rather than as facts.
Where an edition and a manuscript disagree, the disagreement is reported rather than resolved, and where a printed text and a commentator quoting it do not match, that is reported too. A canon is a claim about what counts, and the claims are worth reading as carefully as the books.
Guides in this topic 13 readings
Sanatana Dharma carries two bodies of revelation that both claim to be revealed: the Veda and the Āgama. How they differ, how the orthodox denied the second, how the tantrics defended it, and how most worship today quietly reconciles them.
It is cited as the wine-and-sex Tantra. It is in fact the text that says: if drinking wine brought liberation, every drunkard would be liberated. The Kularnava is the most quoted Kaula manual, and its real subject is the guru, discipline, and the exposure of tantric frauds.
The Śaiva Siddhānta names a fixed canon of 28 revealed Āgamas, ten from Shiva and eighteen from Rudra. What is in them, what they govern, and why the tidy number hides a messier truth.
The tradition speaks of sixty-four Tantras of the Goddess. The number is famous, quoted in the Saundaryalaharī, and it is not a fixed list. What the Śākta scriptures are, and why no two sources agree on them.
The Vaiṣṇava stream of the tantric revelation is the Pāñcarātra, the scriptures that shaped much of Vishnu temple worship. Their theology, their three jewels, and the long fight over whether they count as Vedic.
Every classification of the tantric canon lists a Saura stream. A Shaiva text names eighty-five Saura scriptures, and not one manuscript of any of them has ever been found. What survives is one Sun-tantra that had already gone over to the Shaivas, a hymn that reached us inside a Shaiva commentary, and a daily rite still performed in Bali.
Everyone says there are seventy-seven Shiva Sutras, revealed to Vasugupta on a rock that turned at his touch. One of the four surviving commentaries prints seventy-nine, calls the divisions by a different name, and says the sutras came by a Siddha's command with no rock at all. It has been in print since 1916.
A goddess asks God the oldest question: what is reality, really? And instead of a doctrine he gives her one hundred and twelve ways to find out for herself. It may be the most practical scripture ever written: a manual of doorways into the infinite, most of them hidden inside ordinary life.
Alongside the Veda stands a second body of revealed scripture: the Āgamas, Tantras and Saṃhitās. What that corpus is, how it is organised, and why it insists it came down from the mouth of a god.
One ritual instruction says the guest-offering is poured with fragrant liquor. In Kashmir around the year 1000 a commentator said water was going in instead, and he did not stop at philology. He said they were seized by the demon of caste. His is, so far as anyone knows, the only Sanskrit commentary on this scripture that survives entire, so the argument reaches us almost entirely in the winner's version.
A man in a South Indian temple city sat up at night with a vision of a goddess, and wrote a tantra in the wrong language. Prakrit for the verses, then a lifetime of Sanskrit explaining his own verses. His last verse salutes the goddess and calls what has gone before the seventy Prakrit sutras, counting from outside the count. The Kashmir edition ends its gathas at seventy; the two southern editions at seventy-one.
The earliest surviving Shaiva tantra keeps coming up short. Four divisions where the system wants five, thirty-two principles where it wants thirty-six, two channels where it needs three. The shortfall is the argument for its age, and the full count arrives later inside the same corpus.
Every survey of tantra prints the same branching diagram of the Bhairava canon. No single Sanskrit text draws it. The sources give four seats rather than two, the clean two-seat formula survives only as a citation of a lost scripture, and the earliest text to classify the Svacchandatantra puts it on the opposite side from where the handbooks put it.