You sit before two paintings. The first is a Pāñcarātra Vaiṣṇava maṇḍala from the early centuries CE: Nārāyaṇa enthroned at the centre, his four vyūha-emanations radiating outward — Vāsudeva, Saṃkarṣaṇa, Pradyumna, Aniruddha — each calm, regal, clothed in the dignified iconography of legitimate divine sovereignty. The deities are dressed for the audience-hall. Their gestures express benediction. The maṇḍala fits neatly inside the dharmic moral order.
The second painting is a Buddhist Cakrasaṃvara maṇḍala from the eleventh century. Cakrasaṃvara stands at the centre with his consort Vajravārāhī in sexual union, both naked except for bone ornaments, both with three eyes, his blue body and her red body clasped in a posture that the modern observer registers as transgressive before any iconographic decoding has even begun. Around them in the four directions are four wrathful goddesses — Ḍākinī, Lāmā, Khaṇḍarohā, Rūpiṇī — also naked, with hair flying upward, holding skull-cups and chopping-knives. Around them in turn are sixty-four more figures, each at a cremation-ground with corpses and ash and human-bone offerings. The maṇḍala is built to make a viewer who has not been initiated deeply uncomfortable.
Both are maṇḍalas. The architectural structure is the same — centre with directional emanations, layered out from the centre. But the content could not be more different. Samuel's distinction in Chapter 9 names what divides them: the first is a pre-Tantric maṇḍala, "respectable, aligned with moral order"; the second is a Tantric maṇḍala, "wild, aggressive, antinomian, and deliberately shocking."1 Both deploy the same architectonic form. Only the second carries what Samuel and other recent scholarship use the term "Tantric" to mean.
Samuel writes: "While these kinds of schemes have conventionally been called 'tantric', we are in a different world here from the kind of wild, aggressive, antinomian, and deliberately shocking imagery that is often associated with Tantra. Tantra in the sense of the Pāñcarātra or of the Buddhist Tantric texts we are discussing here is about the ritual worship of divine beings who — at least once the new state has been established — are very much aligned with the respectable moral order."2
The point is methodological. The word "Tantric" gets used in modern scholarship to cover several different things. It can mean texts called Tantra (which in Brahmanical contexts can mean little more than "non-Vedic"). It can mean practices using the maṇḍala-architecture (the broader sense). It can mean practices that are wild, transgressive, antinomian, sexually-explicit, cremation-ground-oriented (the narrower sense). The three senses do not pick out the same body of material.
Samuel's distinction is to separate the broad sense from the narrow sense. The pre-Tantric maṇḍala in the broad sense — the Pāñcarātra vyūha-architecture, the Suvarṇaprabhāsa four-Buddha scheme, the Bṛhadāraṇyaka proto-maṇḍala — uses the centred-and-emanated form but for respectable deity-cult work that aligns with the dharmic moral order. The Tantric maṇḍala in the narrow sense — Cakrasaṃvara, Hevajra, the developed Bhairava-Tantras — uses the same architectonic form but populated by transgressive deities and operating with rituals that the dharmic moral order explicitly disallows.
The distinction matters because it lets us see the historical development clearly. The maṇḍala-architecture itself is not a Tantric innovation. It is a pre-Tantric structural form that the Tantric movement absorbed and re-populated. The early Vedic-Upaniṣadic and Mahāyāna-Buddhist maṇḍala-architecture was already developed. When the Tantric movement emerges in the 5th-8th centuries CE on the Śaiva side and the 7th-12th centuries on the Buddhist side, it adopts the existing maṇḍala-form and fills it with transgressive content.
This is structurally what enables Tantra to graft onto existing religious institutions. The new wild-content fits inside the inherited architectural form. A Buddhist monastery already had maṇḍala-practice for the Mahāyāna sūtras (Suvarṇaprabhāsa, Avataṃsaka). When the Tantric texts emerge with their transgressive deities, the monastery does not have to invent a new architecture — it has to re-populate the existing architecture with new content. The integration is smoother than it would otherwise be.
The distinction also clarifies what Tantra adds to the inherited maṇḍala-architecture. The wild deity at the centre, the cremation-ground setting at the periphery, the transgressive substances used in ritual, the antinomian practices — these are the Tantric innovations. The architectural form is inherited. The transgressive content is new.
Connects to Maṇḍala as Galactic Polity — the architectural form that both pre-Tantric and Tantric maṇḍalas deploy. Connects to Transgression as Religious Resource — the broader Indic principle the Tantric content embodies. Connects to Śiva-Rudra as Outsider Soteriological Deity — the deity-tradition that supplies the transgressive content. Connects to Kāpālika-Style Ascetics: Bhairava Mahāvrata Doctrine — the historical proto-Tantric population that develops the transgressive content. Connects to Wild Goddesses: Disease-Demonesses Tantric Origin — the female-deity component of the Tantric content.
The Suvarṇaprabhāsa Sūtra (Sutra of Golden Light) gives us one of the earliest Buddhist proto-maṇḍala scenes (probably before 421 CE). In Chapter 2, the Bodhisattva Ruciraketu has a visionary experience in which his house becomes a palace of beryl with four Buddhas seated on lotus-thrones in the four directions. The four Buddhas are peaceful. They wear monastic robes (or the iconographic equivalent of monastic dignity). They give him teaching. The Sūtra is read at courts and recited in monasteries as state-protective ritual.
The Cakrasaṃvara Tantra, composed in the eighth or ninth century, gives us one of the most elaborated transgressive Buddhist maṇḍalas. Cakrasaṃvara and his consort Vajravārāhī occupy the centre in sexual union. The maṇḍala-periphery is composed of sixty-four ḍākinīs and yoginīs at twenty-four cremation-grounds. The deities are wrathful, naked, holding cutting instruments and skull-cups. The Sūtra is read at initiated gatherings, often at cremation-grounds, with ritual substances that the dharmic order disallows.
Both are Buddhist. Both deploy the maṇḍala-architecture. Both are technically called "Tantric" in modern usage (the Suvarṇaprabhāsa is sometimes catalogued in Buddhist Tantric canon, especially in Tibetan classification). But the world they inhabit is different. The Suvarṇaprabhāsa is operating in the respectable Mahāyāna sūtra world. The Cakrasaṃvara is operating in the transgressive Tantric world. Samuel's distinction is to separate them analytically while acknowledging the architectural form they share.
You encounter a maṇḍala. You ask three questions to locate it.
First, who is at the centre? A peaceful Buddha or peaceful supreme deity (Vairocana, Amitābha, peaceful Viṣṇu, peaceful Śiva-form) places the maṇḍala on the pre-Tantric or respectable-Tantric side. A wrathful figure in union with a consort (Cakrasaṃvara, Hevajra, Heruka, Bhairava-with-Kālī) places it on the transgressive-Tantric side.
Second, what is at the periphery? Bodhisattvas, peaceful directional Buddhas, dignified attendant deities — pre-Tantric or respectable side. Wrathful goddesses, cremation-grounds, skeletal protectors, ash-smeared yogins — transgressive side.
Third, what ritual context does the maṇḍala belong to? Public temple ritual, state-protective ceremony, mainstream monastic visualisation — respectable side. Initiation-restricted practice, cremation-ground ritual, substance-using ritual, sexual ritual — transgressive side.
The three questions usually resolve the placement. Some maṇḍalas occupy intermediate positions — peaceful in the central figure but with wrathful protectors at the periphery, used in both public and initiated contexts. The pre-Tantric/Tantric distinction is a spectrum with clear endpoints rather than a binary, but the spectrum is real and reading it precisely matters.
You use "Tantric" to cover any maṇḍala-using practice, and you flatten the wild-transgressive content into the respectable architectural form.
You read modern decontextualised maṇḍalas (the kind on the wall of the yoga studio) as if they were always transgressive Tantric maṇḍalas, when actually they are more often modern decontextualised versions of pre-Tantric architectural forms.
You assume that because Tantra "uses the maṇḍala" the maṇḍala itself is a Tantric invention, missing the centuries of pre-Tantric maṇḍala-development that the Tantric tradition inherited and re-populated.
You assume the Suvarṇaprabhāsa and the Cakrasaṃvara are operating in the same religious world because both can be called "Buddhist Tantric texts" in the broad sense. They are not in the same religious world. The first is dharmic-state-ritual; the second is initiated-transgressive practice.
You read modern Hindu temple-ritual maṇḍalas (which are largely pre-Tantric or respectable in Samuel's distinction) and conclude that Hindu Tantra has been domesticated. Some of it has, but the transgressive lineages also continue to operate, often in contexts the public temple-ritual does not show.
Strongest evidence: the textual and iconographic record clearly distinguishes the two kinds of content. The Cakrasaṃvara-Tantra's wild-content can be precisely placed against the Suvarṇaprabhāsa's peaceful-content. The historical trajectory is documentable.
Tension on terminology: the term "Tantra" has been used in many ways. Some traditions use it broadly to include the pre-Tantric maṇḍala material; some use it narrowly for the transgressive material only. Samuel's distinction is one effort to clarify; other scholars use other terminological strategies (some prefer "esoteric Buddhism" to "Tantra" for the broader sense).
Tension with practitioner-lineage warrants: the Tibetan Vajrayāna lineage warrant treats the entire spectrum from peaceful kriyā-tantras to wrathful yoginī-tantras as one continuous Vajrayāna tradition, organised by tier of practice but not split into "pre-Tantric" and "Tantric" categories. Samuel's distinction is a critical-historical analytical category that the practitioner-lineage may not endorse. The vault holds both vantages.
Open question: where does the transition between pre-Tantric and Tantric maṇḍalas happen historically? Samuel locates it around the 5th-8th centuries CE on the Śaiva side, perhaps slightly later on the Buddhist side. But the transition was not sharp — there is intermediate material. What does the intermediate material show us about the mechanism of the shift?
Samuel, Davidson, Sanderson, and White converge on the basic distinction between earlier respectable maṇḍala-architecture and later transgressive Tantric content. The terminology varies (Samuel uses pre-Tantric/Tantric; Sanderson sometimes uses Mantra-mārga vs Mantra-mārga-yoginī-Tantra; White uses pollution-as-power as the marker of the genuinely-Tantric). The substantive distinction is consistent.
The interesting tension is with practitioner-lineage warrants that treat the full spectrum as one continuous tradition. Tibetan Vajrayāna and Hindu Tantric Śaiva-Siddhānta both have organising frameworks that include both the peaceful and wrathful material as parts of a single architecture. The Samuel-style critical-historical distinction is analytically clarifying but does not necessarily correspond to how practitioners organise the material from within their own lineages. Both vantages are real.
The distinction between respectable and transgressive deployments of the same architectural form has parallels across the vault's domains.
The handshake to behavioral-mechanics sits in the broader principle that the same operational architecture can be deployed for respectable or transgressive purposes. A persuasion-architecture can be deployed by a politician giving a public speech or by a manipulator running an exploitative private operation. The architecture is the same; the content and ethical position differ radically. The vault's behavioral-mechanics corpus has documented this for influence-architecture. The maṇḍala-as-architecture / content-as-variable distinction is structurally analogous. The insight: any sufficiently general architectural form is content-neutral and can be deployed for the full range of moral-orientation that the host culture supports — the maṇḍala-architecture supports the dharmic-respectable deployment and the antinomian-transgressive deployment with equal facility.
The handshake to history sits in the comparative case. Christian liturgical architecture (the church-as-building, the liturgical-calendar-as-temporal-architecture) can be deployed in mainstream-respectable mode or in heterodox-transgressive mode (witchcraft as inverted-mass, the various antinomian Christian movements). Sufi ṭarīqa architecture can be deployed in respectable orthodox Sufi mode or in malāmatī-transgressive mode (the Qalandar dervishes). The same pattern of architectural-form-with-variable-content recurs. The insight: religious-cultural architectures generally seem to permit both respectable and transgressive deployments, and the transgressive deployment often does work the respectable deployment cannot — granting the host culture access to forces and operations that the respectable mode cannot acknowledge.
The handshake to psychology is in what the transgressive maṇḍala lets the practitioner engage psychologically. The respectable maṇḍala places the practitioner inside a moral-ordered cosmos. The transgressive maṇḍala places the practitioner inside a cosmos that has encompassed-the-bad-mother-and-bad-father (in Samuel's earlier phrase) — that has integrated rather than suppressed the dangerous, the sexual, the violent, the polluted. The practitioner doing transgressive maṇḍala-yoga is engaging psychologically with material that the respectable maṇḍala does not let access. The vault's psychology corpus on shadow-integration can absorb this case. The insight: the Tantric transgressive maṇḍala is a technical instrument for accessing and integrating psychologically-difficult material under structured-architectural conditions that the respectable practice does not provide.
Sharpest implication: modern decontextualised Tantric-coded practices in the West largely deploy the architectural form (centred meditation diagrams, deity visualisation, mantra recitation) without the transgressive content that historically gave Tantra its distinctive operational character. The result is pre-Tantric-style practice carrying the Tantric label. This is not necessarily a bad thing — pre-Tantric practice has its own value — but the marketing of the practice as historically-continuous-Tantra is misleading.
Generative questions: