Map

The Twenty-Eight Āgamas

The full traditional catalogue of the Siddhānta canon — ten Śiva-bheda and eighteen Rudra-bheda — with what actually survives of each, how the upāgamas attach, which texts govern South Indian temple worship today, and where a reader can consult the editions.

Twenty-eight is the number the Siddhānta gives for its own revelation: ten Āgamas uttered from Śiva's higher nature, the Śiva-bheda, and eighteen from his Rudra nature, the Rudra-bheda. This page lays the whole catalogue on the table, name by name — together with the uncomfortable truth every honest catalogue of this canon must carry: for a good number of these titles, nothing survives but the name itself.

Where the list comes from — and why it variesतन्त्रावतार

There is no single master document called "the list of the twenty-eight". The catalogue is transmitted inside the scriptures themselves — in the tantrāvatāra or scripture-descent chapters that many Āgamas open with, where the text narrates its own transmission from Śiva downward and situates itself among its siblings — and in the paddhatis, the ritual manuals of the exegetes, which rehearse the canon when they establish their own authority. These witnesses agree on the frame: ten plus eighteen, twenty-eight in all. They do not agree on every name. Titles swap in and out between recensions, a text counted as a primary Āgama in one list appears as a subsidiary in another, and orthography drifts (Niḥśvāsa/Niśvāsa, Sarvokta/Śarvokta, Āgneya/Anala). The table below gives the form of the list most widely rehearsed in the South Indian transmission, and flags divergence where it matters. Treat the catalogue as the tradition's self-portrait, not as a shelf inventory.

The ten Śiva-bhedaशिवभेद

The eighteen Rudra-bhedaरुद्रभेद

Notice the shape of the evidence. The texts that survive with the strongest early credentials — Niśvāsa, Pārameśvara, Kiraṇa, the Svāyambhuva and Raurava sūtrasaṅgrahas — survive largely because Nepal's climate and scribal culture preserved early palm-leaf copies that the plains destroyed. The texts that survive as living documents — Kāmika, Kāraṇa, Suprabheda — survive because Tamil temple priests never stopped using them, copying them, and expanding them. These are two different kinds of survival, and a reader should not expect the living ritual compilations to be verbatim witnesses of the first millennium.

The upāgamasउपागम

Each of the twenty-eight is said to carry a retinue of subsidiary scriptures, upāgamas, traditionally totalled at two hundred and seven — a number, like twenty-eight itself, that describes a canon rather than a count. The relationship is one of formal affiliation: an upāgama presents itself as an elaboration or specialisation under the authority of its primary. What makes the class impossible to ignore is that some of the most intellectually serious works of the whole Siddhānta are technically upāgamas:

Mṛgendraमृगेन्द्र · under the Kāmika

A doctrinally sophisticated scripture in four sections (vidyā, yoga, kriyā, caryā), with the commentary of Bhaṭṭa Nārāyaṇakaṇṭha. Its ritual section was edited with translation in the Pondicherry series (1962). For philosophy it outranks most of the primary twenty-eight

Mataṅgapārameśvaraमतङ्गपारमेश्वर · under the Pārameśvara

Long, learned, and equipped with Rāmakaṇṭha's commentary; critically edited by N. R. Bhatt at Pondicherry (1977 and 1982). With the Kiraṇa and Mṛgendra, one of the three scriptures a student of Siddhānta doctrine actually reads

The Kālottara recensionsकालोत्तर

A family of related redactions in different verse-counts (the Sārdhatriśatikālottara, "the 350-verse Kālottara", edited at Pondicherry in 1979, is the best known; Rāmakaṇṭha commented on it). Traditionally affiliated to the Vātula; enormously influential on the paddhati literature

Sarvajñānottaraसर्वज्ञानोत्तर

An early scripture with a notably non-dualist colouring inside the Siddhānta transmission — a standing reminder that the dualist reading of these texts was an achievement of the exegetes, not a given. Early Nepalese manuscript evidence exists; its formal affiliation is variously stated

The upāgama class is also where the catalogue leaks. Lists disagree about which text hangs under which primary, and a title counted among the twenty-eight in one witness may be filed as an upāgama in another. The tradition itself treated the architecture as flexible scaffolding; modern scholarship treats each affiliation as a claim to be tested against the text's own statements.

In the temples todayदक्षिणदेश

The twenty-eight are not an antiquarian matter. Āgamic ritual is performed daily across the Tamil country and wherever Tamil Śaiva priesthood has travelled, and the working canon of that priesthood is drawn from this catalogue — above all the Kāmika and Kāraṇa, whose prescriptions, mediated through the twelfth-century paddhati of Aghoraśiva (the Kriyākramadyotikā), govern the daily and festival liturgy of a great many major temples. Priestly families identify their practice by Āgama, and temple disputes over ritual form are still argued, in the twenty-first century, by citation of these texts.

Attaching one named Āgama to one named temple is harder than the guidebooks suggest, and this catalogue will not pretend otherwise. Such attributions rest on the temple's own priestly tradition, and traditions have been known to shift or to be asserted retrospectively. Cidambaram is the standing illustration of the complexity: the great Naṭarāja temple is central to the history of the Siddhānta — Aghoraśiva worked there — yet its hereditary Dīkṣitar priesthood maintains, by its own account, a distinct ritual usage of its own rather than a straightforward performance of one of the twenty-eight. The honest general statement is this: the Kāmika and Kāraṇa dominate the living tradition; specific temple-to-Āgama attributions should be taken from each temple's own priesthood, case by case.

Using this catalogueसाधनसामग्री

What can a reader actually open? The centre of gravity is Pondicherry. The Institut Français de Pondichéry and the École française d'Extrême-Orient built, from the 1950s onward, both the great manuscript collection of the Saiddhāntika corpus — thousands of bundles gathered from South Indian collections, now internationally recognised as a documentary heritage of the first importance — and the edition series founded on it, above all the critical editions of N. R. Bhatt: the Rauravāgama (from 1961), the Ajitāgama (from 1964), the Mataṅgapārameśvarāgama (1977, 1982), with the Mṛgendra ritual section and the Sārdhatriśatikālottara alongside.

• For the early Siddhānta: Dominic Goodall's editions — the Kiraṇa with Rāmakaṇṭha's commentary (1998) and the earliest books of the Niśvāsatattvasaṃhitā (2015, with collaborators) — are the places where the oldest recoverable stratum of the canon can be read with a full scholarly apparatus.
• For the shape of the whole religion: Alexis Sanderson's "Śaivism and the Tantric Traditions" (1988) remains the map most readers start from, and his "The Śaiva Age" (2009) reconstructs the historical world in which this canon ruled.
• For the living ritual texts: the South Indian printed editions of the Kāmika, Kāraṇa and Suprabheda serve the priesthood rather than the philologist — indispensable for what is practised, unreliable as witnesses to the first millennium.
• For the manuscripts themselves: the IFP collection in Pondicherry for the southern transmission; the Nepalese archives, microfilmed over three decades by the Nepal-German Manuscript Preservation Project, for the early northern witnesses.

And a final orientation: this page is the dualist half of a two-part canon. The same tradition that counted twenty-eight Āgamas of the mild stream counted sixty-four tantras of the fierce — for that half, with its own table of survivals and losses, see the sixty-four Bhairava tantras.