Here's a clean way to feel the distinction White draws. A Tantra is like a terse recipe card: do this, then this, with these ingredients, on this night. No explanation, no philosophy, just the procedure. A Tantraśāstra is the glossy cookbook that came later — written in beautiful prose by a famous chef, explaining what the recipe means, why it works, what state of mind it expresses. White's point: nearly all Western scholarship has been reading the glossy cookbook and calling it "Tantra," when the actual Tantra is the bare recipe card the cookbook was written about.1 And the cookbook layer has a name with a sting in it: when you translate Tantraśāstra honestly, you get "Tantrism" — the exact Orientalist word scholars have been trying to banish.
White separates three things people mush together. Tantra is ritual practice. The Tantras are "unembellished compendia of ritual practice, with a minimum of the explanatory material and erudite exegesis found in later works."2 Tantraśāstra — literally "Theoretical Treatises on Tantra" — is the commentary corpus: signed works in excellent Sanskrit by brahmin intellectuals, appearing later than the Tantras they comment on. The flagship is Abhinavagupta's eleventh-century Tantrāloka, whose very title means "Elucidation of the Tantras."3 The name admits the dependency: it's a light shone on the Tantras, not the Tantras themselves.
White's translation move is deliberate and a little mischievous: "I can think of no better translation for the term 'Tantraśāstra' than 'Tantrism.'"4 "Tantrism" — the -ism suffix that turns a practice into a doctrine — is precisely the Orientalist bogey recent scholars have tried to exorcise from their writing. The irony White lands: even as they exorcise the word, their scholarship has stayed "more squarely based on Tantraśāstra than on Tantra."5 They renamed the thing while continuing to study the thing.
The etymology backs the recipe-card reading. The term "tantra" goes back to the pre-fifth-century-BCE Śrauta Sūtras, the Vedic ritual manuals, where it means "ritual framework" or "interweaving of rites."6 That's the loom sense of the root — warp threads, a framework you weave rites onto. So a Tantra is, at root, a ritual framework, and the Tantras are exactly that: frameworks of practice.
White even offers a tautological definition that's useful up to a point: Tantra is the subject matter of works that call themselves "Tantras."7 But he immediately complicates it — not every "Tantra" is a ritual text (the famous Pañcatantra is a book of animal fables; grammar, astronomy, and alchemy works also took the title), and not every ritual text is called a "Tantra" (the Kubjikāmata is a Mata, the Brahmayāmala a Yāmala, the Kaulajñānanirṇaya a Nirṇaya; coeval Śaivasiddhānta and Vaiṣṇava Pāñcarātra texts call themselves Āgamas or Saṃhitās).8 The naming is messy; the practice-vs-exegesis distinction is what's clean and load-bearing.
This page is a filing rule for the whole corpus. Whenever the vault cites a "Tantric" source, this distinction asks: are we reading practice or commentary? It protects the hard-core thesis (see Hard-Core vs. Soft-Core Tantra) by explaining why the field looks so mystical — scholars sampled the commentary, which is where the mysticism lives. It also primes the sublimation thesis: Tantraśāstra isn't neutral explanation; it's the layer that transforms practice into metaphysics, often distorting the original rite (see The Sublimation of the Yoginī).
White names the smoking gun: the exegetes don't just explain the Tantras, they bend them. His prime examples are Kṣemarāja's eleventh-century Trika commentary on the Netra Tantra, and Abhinavagupta's "forced reading of the dualist principles of the Mālinīvijayottara Tantra into a nondualist Trika mold" in the Tantrāloka.9 Look at what that means. The Mālinīvijayottara is a dualist text. Abhinavagupta needed it to be nondualist, because his system was nondualist. So his "elucidation" reads a nondualism into a source that didn't have it. That's not commentary clarifying a text; that's commentary overwriting it. The case study shows the danger of mistaking Tantraśāstra for Tantra in its purest form: if you read Abhinavagupta to learn what the Mālinīvijayottara says, you'll learn what Abhinavagupta needed it to say. The cookbook isn't just explaining the recipe; it's quietly changing the ingredients to fit the chef's signature style.
You've got a Tantric citation in front of you and you want to use it honestly. Here's the sort.
You ask first: is this a bare instruction or an explanation of an instruction? If it tells you what to do — terse, procedural, unembarrassed — you're probably in a Tantra, reading practice. If it tells you what the doing means, in elegant Sanskrit, with a named author and a metaphysical payoff, you're in Tantraśāstra, reading commentary. Then you ask the harder question: does this commentary have a system to defend? If the commentator is a known nondualist and the text he's elucidating is dualist, you treat his reading as his, not the source's. You quote the commentary for what the commentator believed and the Tantra for what was practiced — and you never let the eloquence of the former stand in for the content of the latter.
You've mistaken the cookbook for the kitchen when your account of "what Tantra is" is built from elegant signed treatises and reads like philosophy of consciousness. The diagnostic tell: your sources all have famous authors and beautiful prose. Real practice texts tend to be anonymous, terse, and a little embarrassing. A second sign: you find the tradition is "really about" a refined inner state — because that refined inner state is exactly what the exegetical layer was built to articulate, and you've been reading only that layer. The whole field caught this disease, which is why White has to diagnose it: a century of scholarship on "Tantra" that was actually scholarship on "Tantrism," dressed in a word-swap.
The evidence is philological (the Śrauta-Sūtra pedigree of "tantra" as ritual-framework), textual (the -āloka "elucidation" title structure), and demonstrative (the Kṣemarāja and Abhinavagupta over-readings).10 The tension is internal to the field: White is accusing the mainstream of Tantric studies — Woodroffe, Silburn, Padoux, and others, whose work he calls "nothing short of brilliant" — of studying the wrong corpus.11 He's not saying their work is bad; he's saying it's about the commentary, mislabeled as being about the practice. Open question: is the practice even recoverable without the commentary? The Tantras are terse precisely because they assumed a living context now lost — so reading them "raw" may import a different distortion (the modern scholar's guesswork) in place of the exegete's. White wants the recipe card, but recipe cards assume you already know how to cook.
White and Padoux stand on the same fault line from opposite sides. Padoux is one of the scholars whose work White praises and whose focus he diagnoses — Padoux's field-defining definition of Tantra (see Hard-Core vs. Soft-Core Tantra, where it's folded in) is built on the goals of practice as articulated by the exegetes, which is to say it's a definition of Tantraśāstra. They converge completely on the value of the commentary corpus — neither thinks Abhinavagupta is worthless; White calls the syntheses "arguably the greatest metaphysical writings of the entire medieval period." They split on whether that corpus is the subject or a secondary development about the subject. The split reveals a trap that runs through all textual scholarship: the most sophisticated layer of a tradition is also the most available — best written, best preserved, most quotable — and availability quietly masquerades as centrality. Padoux studied what was there to be studied beautifully; White's correction is that what's most studiable is rarely what's most foundational.
The plain version: there's a difference between what people do and the polished theory they later write about what they do — and scholars across fields keep mistaking the second for the first because the theory is easier to read.
The Sharpest Implication If the commentary can overwrite the text it claims to explain — if Abhinavagupta can read nondualism into a dualist source and call it "elucidation" — then every authoritative interpretation is a potential act of quiet authorship. Take that seriously and you can't treat any great commentator as a transparent window onto an earlier text. The more brilliant the exegesis, the more it warps. The implication for your own reading: when a master explains a tradition, you're learning the master's system at least as much as the tradition — and the better the master, the harder that is to see.
Generative Questions