You'd expect "Tantra" to name a religion, the way "Buddhism" or "Christianity" does. It doesn't. Wallis argues Tantra is better understood as a mode of practice — a particular way of doing spiritual work, and of viewing reality — "that is not of any one religion, though it is found as the esoteric dimension of many religions."1 It's not a faith you belong to; it's an inner method that several different faiths grew. The same way mysticism shows up inside Christianity, Islam, and Judaism without being a religion itself, Tantra appears as the esoteric layer inside multiple Indian (and one non-Indian) traditions. This is why a single book can compare "the most Buddhistic school of Śaiva Tantra (the Krama)" with "the most Śaivist school of Buddhist Tantra (Dzogchen)" and find "little discernible difference" — they're the same mode surfacing in two religions.2
Two linked claims about Tantra's nature and its spread:3
Calling Tantra a mode rather than a religion solves a problem that calling it a religion creates. If Tantra were a religion, you'd have to explain why it shows up inside Shaivism, Buddhism, Vaiṣṇavism, Jainism, and Islam — five different religions — without any of them being "Tantra." A religion doesn't normally live inside five other religions. But a mode — a method, a way of doing the inner work — does exactly that: it's portable across host-traditions, taking on each host's deities and vocabulary while keeping its characteristic structure (deity-yoga, mantra, initiation, the subtle body, the recognition of reality as one Consciousness).
This is the same logic as the Dzogchen convergence writ large: if a Buddhist school and a Śaiva school describe ultimate reality identically, the thing they share isn't Buddhism or Shaivism — it's the trans-religious mode both adopted. And it explains the order-of-influence list: a mode spreads from where it was invented (Shaivism, Buddhism) to where it was adopted (Vaiṣṇavism, Jainism, Islam), the way a technology diffuses from its origin culture outward. Śaiva priority isn't a sectarian boast; it's the textual finding that the Buddhist Tantric texts borrowed from the Śaiva ones and not vice versa — the mode's point of origin, traceable by who-copied-whom.
This page is the conceptual frame behind several others: it grounds the Śaiva-priority origin history (the same priority claim, developed historically), explains why Dzogchen and NŚT converge (same mode, different religions), and licenses the feature-list definitions that must "encompass both Śaiva and Buddhist Tantra." It pairs with there-is-no-Hinduism — Tantra crossing between distinct pre-colonial religions is only thinkable once you stop treating them as one "Hinduism."
The claim that Shaivism has priority over Buddhist Tantra could sound like a Śaiva partisan flexing. Wallis grounds it in something harder than preference: the textual evidence of borrowing direction.8 Many Buddhist Tantric texts "drew directly or indirectly on Śaiva tantras, while the reverse was rarely if ever the case." When you find text A copying from text B but never B from A, you've established that B came first and was the source. That's a forensic argument, not a devotional one — and it's the same kind of evidence (later developed by Sanderson in detail) that the vault's critical-historical pages rely on.
What makes the case study sharp is that the priority finding sits inside the "Tantra is a trans-religious mode" thesis without contradicting it. You might think "if Tantra belongs to no religion, how can one religion have priority?" The resolution: the mode was largely invented within Shaivism (and Buddhism), then spread. Priority isn't ownership — Shaivism doesn't own Tantra — it's authorship of the innovations that the mode then carried into other hosts. Buddhism became "thoroughly Tantricized" and brilliantly developed the mode (carrying it to Tibet), but the textual record shows it received the core innovations from the Śaiva side. The case study models how a tradition can be both trans-sectarian (belonging to no one) and historically rooted (invented somewhere specific) — the mode floats free, but it has a birthplace, and the manuscripts record where.
For a researcher or cross-traditional practitioner, the mode-not-religion frame is a working lens. Look for the structure, not the labels. When comparing a Śaiva and a Buddhist Tantric practice, set aside the differing deity-names and cosmologies and look at the mode: is there deity-yoga, mantra, initiation, a subtle-body map, the recognition of reality as Consciousness? The shared structure is the Tantra; the host supplies the costume. Use priority to trace, not to rank. Śaiva priority tells you the direction of historical flow (innovations Shaiva → Buddhist → others), useful for dating and tracing influence — not that Śaiva Tantra is "better." Expect convergence at the deep end. The further into the esoteric core of any host-tradition's Tantra you go, the more it converges with the others (Krama ≈ Dzogchen), because you're approaching the shared mode beneath the host-specific surface. Don't mistake the host for the thing. A practice isn't "Buddhist" or "Hindu" at its Tantric core; it's the trans-religious mode wearing a Buddhist or Śaiva face — which is the practical key to reading across the two traditions.
Religion-category error. You treat Tantra as a bounded religion with members and orthodoxy, and get confused when it shows up inside five different faiths or when a "Buddhist" and "Hindu" Tantra look nearly identical at their core. The tell: trying to locate Tantra's "real" religious home, or arguing about whether it's "really" Hindu or Buddhist, when the answer is "neither — it's a mode both adopted."
Priority-as-ownership. You read Śaiva priority as Shaivism owning Tantra, and either inflate Śaiva claims or dismiss the priority finding as sectarian bias. The tell: a debate over who Tantra "belongs to," missing that priority is a forensic claim about borrowing-direction (who-copied-whom), not a property deed.
A combination of a conceptual claim (Tantra as trans-sectarian mode) and an empirical one (Śaiva priority via textual borrowing), both well-grounded.9 The genuine tension is between the trans-sectarian framing (Tantra belongs to no religion) and the priority framing (it was invented within Shaivism) — Wallis reconciles them (the mode was authored in Shaivism then spread), but the reconciliation has soft edges: how much did Buddhist Tantra independently develop the mode versus merely receive it, and at what point does brilliant development amount to co-authorship rather than borrowing? The open question is whether "Śaiva priority" fully holds across all the innovations, or whether some Tantric developments (certain Buddhist yoginī-tantra material, say) have a more genuinely two-way or independent history than the clean priority-claim suggests — a live debate in the scholarship Wallis is summarizing.
Wallis here converges with the critical-historical scholarship (Sanderson's lineage, which is also Wallis's own training) on Śaiva priority — and the convergence is straightforward because this is a Part-2-history claim, where Wallis writes as a rigorous Sanderson-trained scholar rather than a devotee. It's worth contrasting with how he handles the Buddhist side. He could, as a Śaiva partisan, press Śaiva priority into Śaiva superiority. He doesn't — he calls Buddhism "thoroughly Tantricized," credits it with carrying the mode to Tibet, and elsewhere treats Dzogchen as describing the same reality as NŚT with "no significant difference." So the priority claim and the convergence claim sit together without rivalry: Shaivism authored the innovations, Buddhism developed them magnificently, and at the deep end they meet. What that reveals is a discipline the partisan version lacks — historical priority (who came first, traceable by manuscripts) is held entirely separately from spiritual ranking (which tradition is "higher"), and conflating the two is exactly the sectarian move Wallis refuses. The mode has a birthplace and no owner; both facts are allowed to stand.
The plain link: this is a model of a method or technology that diffuses across host-cultures while keeping its structure — which connects to the psychology of how practices transfer across contexts and the behavioral mechanics of innovation diffusion.
The Sharpest Implication If Tantra is a mode that lives inside five different religions rather than a religion of its own, then the boundaries between those religions are far more porous at the esoteric level than at the surface — a Śaiva and a Buddhist tāntrika, doctrinally "different religions," may be doing nearly the same inner work, while a Śaiva tāntrika and an ordinary temple-going Śaiva, nominally the "same religion," may be doing utterly different things. The destabilizing part: religious identity at the surface (which faith you belong to) can be a poor guide to what you're actually doing at the depth, where the trans-religious modes cut across the sectarian map. This means the real kinships in spiritual life may not run along the lines we use to sort religions — the deep structure groups people by method (everyone doing deity-yoga toward the recognition of one Consciousness) rather than by creed, and those two groupings can be almost orthogonal. The map of religions and the map of what people actually practice are different maps.
Generative Questions