Ask five scholars what makes a practice "Tantric" and you get five lists. Wallis combines them — Tribe, Hodge, Goudriaan, Brooks, Lopez — into one master list of about thirty features: new revelation, centrality of ritual, proliferation of deities, visualization and self-identification with the deity, mantras installed on the body, the ontological identity of mantra and deity, initiation and secrecy, maṇḍalas, the subtle body and kuṇḍalinī, linguistic mysticism, the guru, worldly aims by magical means, nondualism, revaluation of the body and of "negative" mental states, the centrality of śakti, the status of women, transgressive acts, sexual yoga, bliss, spontaneity (sahaja), and more.1 It's a sprawling inventory. But buried in it is a single organizing key — deity-yoga — that turns the checklist from a grab-bag into a system.2
Two scholarly tools for recognizing Tantra, stacked. First, the master list — the composite of ~30 features that characterize Tantra as a spiritual movement.3 Crucially, Wallis flags that the last ten features (revaluation of the body and negative states, śakti's centrality, the role of women, transgression, sexual yoga, bliss, spontaneity, the deity-transformation meditations) apply only to nondualist or "left-current" Tantra — they're not universal.4
Second, the six universal features that constitute the most prominent traits of classical Tantra: yogic meditation, mantras, maṇḍalas, guru, initiation, and ritual worship of the divine.5 But Wallis adds a sharp qualification: these six "cannot completely define Tantra by themselves," because all of them except maṇḍalas are also found in non-Tantric Indian religions.6 Only the maṇḍala is unique. So the six don't do the defining work alone — they're "organized and given their specific form by the seventh universal feature": deity-yoga.7
Here's the move that rescues the definition from being a list. Deity-yoga isn't a thirty-first feature; it's the principle that arranges the other six.8 It means working with the power of a particular aspect of the Divine to which you've been linked through initiation — through that deity's mantra, yantra, visualization, and/or consecrated image. So meditation, mantra, maṇḍala, guru, initiation, and ritual worship aren't six separate boxes to tick; they're six aspects of one activity — relating to your initiatory deity. The mantra is that deity's mantra; the maṇḍala is that deity's diagram; the initiation is into that deity; the worship is of that deity; the guru is the one who linked you to it. Deity-yoga is the thread that strings the beads.
This delivers the actual criterion. What makes a practice "indisputably Tantric," Wallis concludes, is that it involves deities and their mantras revealed in explicitly Tantrik scriptures.9 You can call any teaching "Tantric" if it appears in an original Tantric source — but if you want the word to keep a specific meaning, that scriptural-deity-mantra criterion is the line. This is why a "Tantric" practice with no initiatory deity, no scripture-revealed mantra (a weekend "Tantric sex" workshop, say), isn't Tantra in any precise sense — it's missing the organizing principle the whole system is built around.
This page is the scholarly-definitional companion to the tradition's own indigenous definition (Rāma Kaṇṭha) — Wallis pairs the insider definition with these outsider lists. Deity-yoga is operationalized in the mantra-deity (the deity's four bodies) and the initiation that links you to it. The "last ten features only apply to the left current" point feeds directly into the left-right spectrum, and the maṇḍala-as-uniquely-Tantric claim connects to the maṇḍala as deity-substrate.
The most instructive thing in Wallis's treatment is his refusal to let the obvious definition stand. The six universal features — meditation, mantra, maṇḍala, guru, initiation, ritual worship — look like a perfectly good definition of Tantra. Most popular accounts stop there. Wallis points out the problem: five of the six are also found in non-Tantric Indian religions.10 Vedic religion has ritual worship, fire offerings, mantras. Pre-Tantric yoga has meditation and the guru. Initiation exists across Indian traditions. So if you define Tantra by those five, you've defined something far broader than Tantra — you've described much of Indian religion.
Only the maṇḍala is genuinely unique to Tantra among the six. But a definition resting on one feature is thin. So Wallis does something more sophisticated: he argues the six are defined not by their presence but by their organization — and the organizing principle is deity-yoga, working with a specific initiatory deity through all six channels at once. This is why the same act (reciting a mantra, say) is Vedic in one context and Tantric in another: in Tantra it's this deity's mantra, received through initiation into this deity, part of a unified relationship with this aspect of the Divine. The case study shows that defining a tradition isn't about finding features no one else has — it's about finding the principle that arranges shared features into something distinct. Tantra didn't invent meditation or mantra; it invented a way of organizing them around the initiatory deity, and that organization is the tradition.
The definition gives a usable test, which Wallis essentially supplies for evaluating modern "Tantra." Ask: is there an initiatory deity? Genuine Tantra works with a specific aspect of the Divine you've been linked to through initiation — not a generic spirituality. No initiatory deity, no deity-yoga, no Tantra in the precise sense. Ask: which tantra (scripture) does it follow? Wallis's own field test for a teacher claiming "Tantra" is to ask which tantra — which revealed scripture — they follow; genuine lineages can answer.11 Don't be fooled by the universal six alone. Meditation, mantra, a guru, initiation, ritual — these appear in many traditions; their mere presence doesn't make something Tantric. Watch for the left-current markers if "nondual" is claimed. The revaluation of the body, of women, of "negative" states, the centrality of śakti, transgression, bliss, spontaneity — these belong specifically to nondual/left-current Tantra, so a tradition claiming that lineage should carry them. The whole test reduces to: a scripture-revealed deity and its mantra, received through initiation, worked through the six channels. Everything else is necessary but not sufficient.
Feature-checklist literalism. You treat the presence of some Tantric features (a mantra here, a maṇḍala image there, talk of kuṇḍalinī) as proof that something is Tantra, ignoring that the features must be organized by deity-yoga and rooted in scripture-revealed deities. The tell: "this is Tantric because it has mantras / chakras / energy," with no initiatory deity, no scripture, no lineage behind it.
Over-broadening (defining Tantra as all of Indian spirituality). You define Tantra by the universal six alone and end up with a category so wide it includes Vedic ritual and Patañjali's yoga. The tell: calling nearly any meditative, mantra-using Indian practice "Tantric," losing the specific meaning the deity-yoga-plus-scripture criterion preserves.
A careful scholarly synthesis (five named scholars' lists) plus Wallis's own organizing argument.12 The genuine tension is internal to the definitional project: the master list is descriptive (here's what tends to show up) while the deity-yoga criterion is prescriptive (here's what's required to call it Tantra), and the two don't perfectly align — some scripture-revealed Tantric material is light on deity-yoga, and some deity-focused practice isn't scripturally Tantric. Wallis's escape hatch ("call anything Tantric if it appears in an original Tantric source") is honest but loosens the criterion he just tightened. The open question is whether "Tantra" names a coherent thing with a real essence (deity-yoga + scripture) or a family resemblance across a sprawling feature-set with no single necessary trait — a question that matters for whether the word can do any precise work at all.
The deep methodological convergence-and-split here is between the indigenous definition (Rāma Kaṇṭha) and the Western scholarly feature-lists, which Wallis deliberately uses together. Rāma Kaṇṭha defines from inside (initiation, deity-worship, twofold goal, divine revelation) — concise, essence-oriented, but assuming a reader who already inhabits the tradition. The Western scholars define from outside (a long checklist of observable features) — comprehensive, recognizable to a newcomer, but with no organizing center until Wallis supplies deity-yoga. They converge on the content (both foreground deity-worship, initiation, scripture) and split on method: the insider names the essence and trusts you to fill in the features; the outsider catalogs the features and struggles to find the essence. What the pairing reveals is that neither definition is complete alone — the insider's is too compressed for a newcomer to apply, the outsiders' is too diffuse to mean anything until organized. Wallis's move is to use the outsider list to make the insider essence recognizable: deity-yoga (close to Rāma Kaṇṭha's "worship of a form of the Divine") is the principle that turns the scholars' checklist into the tradition's actual shape. The definition that works is the one that puts the insider's essence and the outsider's catalog in the same frame.
The plain link: this is a worked problem in how you define a category that has no single defining feature — which connects to the psychology of concept-formation and the behavioral mechanics of how movements mark their boundaries.
The Sharpest Implication If five of the six "universal features" of Tantra are also found in non-Tantric religions, then nearly everything people point to as proof that something is "Tantric" — it has mantras, meditation, a guru, initiation, ritual — proves almost nothing, because those features are everywhere in Indian spirituality. The destabilizing part: the markers most people use to identify Tantra are the markers least able to identify it. What actually distinguishes Tantra isn't any exotic feature but a principle of organization — working with a specific scripture-revealed deity through all the channels at once — which is invisible to anyone counting features. This means the popular game of "is this Tantric?" played by checking for chakras, kuṇḍalinī, and mantras is the wrong game entirely; the real question is whether there's an initiatory deity and a scripture behind it, and that question the marketplace almost never asks because it's the one most modern "Tantra" can't survive.
Generative Questions