Picture a librarian handed a shelf of esoteric Buddhist manuals and told to sort them. One axis is easy: left-handed or right-handed — does the tradition use female deities and powers, or doesn't it. But a second axis cuts across the first, and it has nothing to do with iconography. It asks: did this school keep to the tenets its own lineage teachers set down, or did it borrow from everywhere? Pure, or mixed. Lay the two axes over each other and you get a grid, and onto that grid Dukes drops every esoteric school — including, pointedly, the Hindu ones, which land in a single box: left-handed and mixed.1
Dukes's section sorts the "Thunderbolt Path" — Vajrayana, the schools that dealt with powerful primitive energies — using two independent classifications that modern scholars apply to esoteric Buddhism (Chinese Mi Chiao, Japanese Mikkyo).2
The first axis is the familiar left/right-hand split. Tibetan female Vajrayana, using "images and symbols which included representation of female deities and powers," is termed left-handed; the Chinese male Vajrayana is right-handed.3 The second axis is purity of lineage: "Pure Mikkyo keeps to the tenets formulated by its lineage teachers, while the mixed incorporates tenets from many other sources."4 These two axes are independent — that is the structural point. A school's gender-iconography tells you nothing about its doctrinal purity, and vice versa. "Esoteric sects can thus be formed of all the various combinations of these."4
Then the filing decision that gives the page its edge: "Non-Buddhist (i.e., Hindu) esoteric teachings can thus be classified as a 'left-handed tantra,' of the 'mixed' type."5 Hindu esoterica is not granted its own category. It is slotted into the same two-axis grid built for Buddhist schools, and it lands in the corner Dukes treats as furthest from the pure Chinese center.
The page carries a second load: a refutation. It is "often said that the King of Udiyana, Indrabhuti (678–717), Padmasambhava's father, founded the Vajrayana, but this is unlikely, as esoteric Buddhism in general, and the title of Vajrayana, was already widely known in both India and China."6 Dukes denies Indrabhuti the founder's crown — the tradition predates him.
Walk why the grid matters. A one-axis taxonomy — just left/right-hand — would force you to read iconography as the master variable: female deities mean one kind of school, male another, and that's the story. Dukes refuses that. By adding the purity axis, he separates what symbols a school uses from how faithfully it transmits its own lineage. These can vary independently. You can have a right-handed school that has gone mixed, or a left-handed one that is doctrinally pure to its teachers.
The Tibetan tantric followers, he notes, "used noncanonical collections of religious writings, which they termed 'tantras,'" and "in Tibet, these often supplanted the Sutras as sources of doctrine."7 That is a vivid marker of the mixed pole at work: scripture itself getting displaced. The pure pole, by contrast, holds the line at the lineage teachers' tenets. So the second axis is really a measure of boundary discipline — how porous a school's doctrinal membrane is — and it operates regardless of which hand the school dances with.
The Indrabhuti refutation is the same instinct applied to chronology. A founder-myth compresses a diffuse tradition into one originating king. Dukes decompresses it: the title and the practice were "already widely known" before Indrabhuti's lifetime, so the crown is honorific, not historical.6 The man's actual distinctive contribution, Dukes allows, was iconographic — his Vajrayana "used images and symbols which included representation of female deities and powers, unlike the Chinese or other esoteric schools."3 He shaped the left-handed pole. He did not found the path.
This page is the classification spine the rest of the Vajrayana cluster hangs on. Tantric Left-Hand/Right-Hand Path develops the first axis in depth; this page shows the second axis crossing it and why a single axis underdetermines a school. Cham: Lama Dances as Nata Survival needs the sect-frame this page provides — the Nyingma-Pa that preserved the dances sits at a definite spot on the grid, descending from Padmasambhava's left-handed line. Chuan Fa Transmission Lineage is about lineage fidelity in the martial arts; the pure/mixed axis is the doctrinal version of the same concern — does the line hold its source, or leak. And the refutation hands Wandering Acarya Transmission a caution: founder-attributions in a mobile, diffuse tradition are usually retrofits.
Read the filing sentence as a power move, because it is one.5 Dukes has built a two-axis grid. He could have placed Hindu esoteric teaching anywhere on it, or off it. He places it in one specific corner: left-handed and mixed. Both labels are, in his frame, the marked, peripheral values — left-handed against the Chinese right-handed center, mixed against the pure lineage-faithful ideal. The Hindu material thus lands as far from the pure Chinese center as the grid allows.
Notice what this does. It folds a non-Buddhist tradition into a Buddhist classificatory system rather than meeting it on its own terms — Hindu tantra becomes a variety of mixed left-handed Mikkyo, defined by its distance from a Buddhist pure pole. From inside Dukes's practitioner-lineage standpoint this is coherent: he is sorting by fidelity to Buddhist lineage teachers, and Hindu material is by definition outside that line, hence mixed. But the move quietly installs a hierarchy with the Chinese pure school at the apex and Hindu tantra at the base. The vault's White/Wallis corpus would read the same Hindu tantra as a sovereign tradition with its own logic, not a degenerate dialect of someone else's grid. Both readings are kept. Dukes's is a classification from a position, and the position is the practitioner-scholar defending his own lineage's center.
You're sorting the shelf yourself now. Pick up a manual. First question, fast: does it use female deities and powers? Yes — set it to the left. No — set it to the right. Now the harder question, the one that takes reading: does this school hold to what its own lineage teachers laid down, or has it pulled in tenets from everywhere? Feel for the tell — are the canonical Sutras still the doctrinal floor, or have noncanonical "tantras" supplanted them? If scripture has been displaced, it's mixed; slide it that way. Two questions, two slides, and the manual has a coordinate. Now you pick up a Hindu tantric text and your hand hesitates — it's not Buddhist at all. But the grid takes it anyway: left, and mixed, the far corner. And right there you feel the grid stop being neutral. You are not just sorting. You are ranking, with your own lineage at the center.
The source is Dukes reporting how "modern scholars differentiate" the esoteric schools, layered with his own practitioner standpoint.2 The two-axis structure is presented as scholarly consensus; the placement of Hindu esoterica in the far corner is his application of it. The Indrabhuti refutation rests on a general claim — that the title was "already widely known" before him — rather than a dated citation, so it is [PLAUSIBLE — needs corroboration], though the reasoning is sound.
The live tension is the filing of Hindu tantra. Classifying a sovereign tradition as a peripheral variety of one's own is a defensible move within a lineage frame and a contestable one across frames. Preserved, not resolved.
Open: do the "modern scholars" Dukes invokes actually use this exact two-axis grid, or has he combined the pure/mixed Mikkyo distinction (a real scholarly category) with the left/right-hand distinction into a composite of his own? And does any school actually occupy the right-handed-and-mixed box, or is that corner empty in practice — which would quietly reduce the grid back toward a single axis?
Dukes writes as a practitioner-scholar whose center is the Chinese/Japanese pure-lineage esoteric stream his own art descends from, and the grid tilts toward that center by construction — pure and right-handed sit at the apex, and that happens to be home. Set this against the vault's Samuel and White material and the friction is clean. White would not classify Hindu tantra as mixed left-handed Mikkyo; he would treat it as a tradition with its own operative logic, and would likely read Dukes's grid as a sectarian instrument rather than a neutral taxonomy. Samuel would ask whether the "modern scholars" consensus is as settled as Dukes implies. The convergence: all three would agree the Indrabhuti founder-myth deserves the skepticism Dukes gives it — a diffuse tradition rarely has a single king at its source. On the Hindu-filing, the page holds the tension open. From inside the lineage, the placement is rigorous; from outside, it is a value-judgment wearing a coordinate. [CONTESTED] across frames; kept as found.
Category as Control — Behavioral-mechanics treats the power to name and sort as a power move in itself: whoever defines the categories defines the hierarchy, because every taxonomy smuggles in a center and a periphery. Dukes's grid is a precise instance. By sorting all esoteric teaching on his two axes, with pure-right at the apex, he positions every rival tradition relative to his own — Hindu tantra not as a peer but as a far-corner variety. The tactical literature knows that controlling the frame controls the ranking; this page supplies a religious-historical case where the framing is done so smoothly it reads as scholarship. The insight neither domain produces alone: behavioral-mechanics would spot the move but treat it as manipulation; the religious-studies frame would read it as classification; together they show that classification IS positioning — that the most durable way to subordinate a rival tradition is not to attack it but to grant it a coordinate in your grid. The Hindu tradition is not denounced. It is filed. Filing is the softer, more permanent move.
Founder Myth as Compression — Historians repeatedly find that diffuse, multi-source traditions get retrofitted with a single founding figure, because a named originator is easier to teach, defend, and venerate than a slow institutional emergence. Dukes's refutation of Indrabhuti is the historian's instinct applied in real time: the title and practice were "already widely known," so the founder-crown is compression, not record.6 The historical frame supplies the general pattern; this page supplies a case where a practitioner-scholar resists the compression of his own tradition's lore — refusing a founder-myth that would actually flatter the lineage (Padmasambhava's own father as founder of the whole path). The joint insight: the willingness to dismantle a flattering founder-myth is itself a credibility signal. A source that keeps the convenient origin story is doing PR; a source that demolishes it, as Dukes does here, is doing history — which makes his other hedges (Cham, the Natyasastra chain) more trustworthy by association.
Sharpest implication: a classification system is never neutral furniture — it always has a center, and the center is wherever the classifier is standing; Dukes filing Hindu tantra as "mixed left-handed Mikkyo" is the whole politics of the esoteric world compressed into one coordinate.
Generative questions: