Tantric initiation required a dream. White draws on Obeyesekere and Nabokov: across South Asia, dreams and visions involve possession by a being, expelled through narrative interpretation and ritual exorcism by a cāmi — a low-caste specialist "recruited" by a goddess (himself possessed), whose prior possession empowers him to cure others.1 The cāmi "[is] similar — if not identical — to tāntrikas."2 Strickmann: the Tantric ritual master generates "a sustained state of (wakeful) dreaming," such that a properly performed rite is "a perfect dream" — and this is the Freudian "dream work," "the tāntrika's ritual processing of the terrible nightmare demonesses that possess his own mind and body."3 So both Kaula and high-Tantric initiation demand a dream-experience. But White's brilliant comparison shows two opposite kinds of dream — and the difference is the sublimation. He sets the c. 9th–10th-century Brahmayāmala (BY, Kaula) against the 1095 C.E. Somaśambhupaddhati (SP, Śaivasiddhānta): the BY summons the nightmare-demonesses; the SP walls them out. The Kaula dreams the Yoginīs in; high Tantra dreams them away.
The BY (Kaula): the nightmare embraced. The practitioner undertakes "the viewing of his own yoni" — meditating on "the vulvas of a circle of ritual consorts," the yoni being both his meditation-support and "the womb of the prior existence he is seeking to know" (he reads his past lives to find karmic impediments).4 In an underground chamber with eight śaktis (the "clan number"), he sexually arouses them, "effecting as many 'rebirths' as his energy allows," for up to six months.5 Then come "the great obstructors (mahāvighnāni), all of them very terrifying": a "fearsome serpent that seems to be devouring him," "a she-cat with sharp teeth," "a very terrifying she-rat," demonesses crying "Stand up! I devour you!", "buck-toothed śaktis licking him with their tongues… they lick the essence inside the practitioner for the sake of knowledge."6 He must not fear, must not halt the ritual — and at six months "there is the visible manifestation of the [great] Goddess." He "becomes a Virile Hero, surrounded by yonis."7 The BY dream is the full cremation-ground horror — the devouring Yoginīs summoned, survived, and mastered, the practitioner ending "at the center of a mandala of feminine entities he now controls" (the Kaula kulāmnāya: Bhairava enthroned among the eight Mothers).8
The SP (Śaivasiddhānta): the feminine walled out. The highest Śaivasiddhānta initiation (nirvāṇa-dīkṣā) "epitomizes this censoring of the female."9 The initiands sleep "protected by the astra [mantra]"; "a preliminary mantric protection is effected to ensure that no demonic power, no spirit of the dead, and especially no female entity invade the space of the initiation."10 The dreams are stimulated by the mantra of the "Little Dream Man" (svapnamāṇavaka) — "the Indian equivalent of the Sandman, here identified with Śiva."11 And the resulting dreams are tame: "not one of the dreams has either a sexual or a horrific content — they range from eating clotted milk to falling down an empty well" — "precisely because the sleeping initiands have been protected from demonic invasions… by a 'citadel of mantras.'"12 The SP dream is the defeminized, demilitarized dream — "no fearsome female entity troubles the sleep… The 'dream work' has been elided… with the total sublimation, or exclusion, of the terrifying feminine, in favor of an unthreatening male 'Sandman.'"13 The Yoginīs "have no place in this all-male universe."14
White's analytic payoff: the two traditions "share three common approaches."15 First, both require the dream (you must dream). Second, both work from the same worldview: "multitudes of dead and demonic beings at the periphery of a hierarchized mandala governed by a supreme deity." Third, both "give full value to the possession experience" — the goal is samāveśa, "the 'co-penetration' or 'mutual possession' of two beings — the human practitioner and the divine — until they are realized to be one."16 So the structure is identical (mandatory dream, mandala-worldview, possession-goal). The difference is the treatment of the terrifying feminine. The Kaula (BY) summons the devouring female nightmare-horde, transacts with it (the buck-toothed śaktis licking his essence), and masters it (becomes the Virile Hero surrounded by yonis). High Tantra (SP) walls out the feminine entirely (the citadel of mantras, the male Sandman, the tame milk-and-wells dreams). The masculinization (see the masculinization) is visible in the dream-content: the Kaula initiate dreams the Yoginīs; the high-Tantric initiate is protected from dreaming them. But — the crucial residue — the SP's elaborate protective measures (the citadel of mantras against the female entities, the blood-offerings to the bhūtas) are "ample evidence of their invasive power."17 You only build a fortress against an enemy that is real and dangerous — so the SP's walls prove the Yoginīs' continuing power even as they exclude them. The feminine is walled out, not abolished: the high tradition "denies the importance of the subconscious" (the terrifying feminine, the dream-work) by fortifying against it — which confirms that it is still there, pressing on the walls.
This page makes the masculinization visible in the dream-content: the Kaula (BY) summons the devouring Yoginīs (the nightmare embraced and mastered); high Tantra (SP) walls them out (the citadel of mantras, the male Sandman, the tame dreams). It supplies the shared structure (mandatory dream / mandala-worldview / samāveśa possession-goal) that makes the difference (transacting-with vs walling-out the feminine) precisely the sublimation. It delivers the fortress-proves-the-enemy point (the SP's protective measures confirm the Yoginīs' continuing power). And it connects the Tantric dream-work to the cāmi/possession ethnography (Obeyesekere/Nabokov) and the Freudian dream-work.
The decisive comparison is the content of the two prescribed dreams, set side by side.18 The BY initiate dreams: a devouring serpent, a sharp-toothed she-cat, a terrifying she-rat, demonesses screaming "I devour you!", and "buck-toothed śaktis licking him with their tongues… lick[ing] the essence inside the practitioner for the sake of knowledge." The SP initiate dreams: "eating clotted milk" or "falling down an empty well." The case study is the gulf between them. Both are prescribed initiatory dreams; both are required; both lead (supposedly) to the same goal (samāveśa, union with the divine). But the BY dream is a cremation-ground horror (the practitioner licked and devoured by buck-toothed goddesses) and the SP dream is a bowl of clotted milk. What changed between the 10th-century BY and the 1095 SP is not the structure (both require a dream) but the feminine: the Kaula initiate must survive being licked by devouring śaktis; the Śaivasiddhānta initiate is protected so thoroughly that he dreams of dairy. The case study reveals the sublimation as a taming of the dream: the terrifying feminine (the buck-toothed śaktis, the devouring serpent) has been walled out and replaced by the innocuous (clotted milk, an empty well). And the cost is registered in the SP's own anxiety: the elaborate "citadel of mantras," the astra-protection, the blood-offerings to keep the bhūtas away — the high tradition builds a fortress to ensure the initiate dreams of milk rather than being licked by goddesses, and the fortress itself proves the goddesses are still out there, pressing to get in. The bowl of clotted milk is the defeminized dream — and the walls around it are the measure of what was excluded. The Kaula dreamed the Yoginī devouring him; the high tradition dreams of dairy, behind a wall built against her.
You encounter two versions of a prescribed initiatory dream/vision and want to read the sublimation between them. The discipline: hold the structure constant and read the content's treatment of the threatening other. First, confirm the shared structure: do both require the dream, share a worldview (the mandala of beings), and aim at the same goal (union/possession)? If the structure is constant, the difference is significant. Second, read the content's treatment of the threatening party: does one version summon the threatening other (the devouring feminine, the demonic) and transact with/master it, while the other walls it out (protective measures, a tame substitute)? Transacting-with vs walling-out is the sublimation. Third, find the fortress: does the sublimated version build elaborate protections (a citadel of mantras, blood-offerings) against the excluded party? The fortress proves the enemy — the protections confirm the excluded party's continuing power. Fourth, read the substitute: what replaces the threatening content (the male Sandman for the devouring goddesses; clotted milk for the buck-toothed śaktis)? The substitute's innocuousness measures the exclusion's thoroughness. Fifth, note the denial: does the sublimated version deny the importance of the excluded realm (the subconscious, the feminine) precisely by fortifying against it? The workflow is to read the two prescribed dreams as the same rite with the threatening other summoned-and-mastered (original) vs walled-out-and-substituted (sublimated) — and to recognize that the walls and the tameness of the sublimated dream are the evidence of the power it excludes.
The failure is reading the tame high-Tantric dream (the Sandman, the clotted milk) as the original/normal initiatory experience, missing that it is the defeminized, walled-out version of a Kaula dream that summoned the devouring Yoginīs (the buck-toothed śaktis licking the initiate). You can tell you've done it when your Tantric dream has no terrifying feminine, no nightmare, no cremation-ground horror — only milk and wells. A second failure: missing the shared structure (both require the dream, the mandala-worldview, the samāveśa goal) that makes the difference (summon vs wall-out) the sublimation rather than two unrelated systems. A third sign: missing that the SP's protective fortress (the citadel of mantras against the female entities) proves the Yoginīs' continuing power — reading the walls as evidence of the feminine's absence rather than as evidence of her excluded presence. A fourth: missing the denial-by-fortification (the high tradition denies the subconscious/feminine precisely by building walls against it). If your prescribed dream is tame, all-male, and unfortified, you've taken the walled-out dream for an original and missed the devouring goddesses on the other side of the wall.
The evidence is the Obeyesekere/Nabokov possession-ethnography (the cāmi ≈ tāntrika); Strickmann's "perfect dream" and the Freudian dream-work; the BY "viewing one's own yoni" (the underground chamber, the eight śaktis, the buck-toothed devouring goddesses, the Virile Hero surrounded by yonis); the SP nirvāṇa-dīkṣā (the citadel of mantras, the male Sandman, the tame milk-and-wells dreams); the three shared approaches; and the fortress-proves-the-enemy residue.19 The tension is whether the BY/SP contrast is representative or cherry-picked: White selects one Kaula text (the BY's most elaborate envisioning) and one Śaivasiddhānta text (the SP's nirvāṇa-dīkṣā) as "opposite ends of the spectrum" — but the actual range of Tantric dream-practice is wider and messier, and the clean opposition (devouring-Yoginīs vs clotted-milk) may sharpen a contrast that was, in practice, more graded. The open question is the psychological status of the difference: White (via Freud) reads the SP's walling-out of the terrifying feminine as a denial of the subconscious — the high tradition repressing the dream-work the Kaula did. But one could read it differently: perhaps the high tradition genuinely achieved a less-terrifying initiatory experience (the samāveśa without the cremation-ground horror), in which case the SP is not repressing the feminine but transcending the need to confront her. The fortress-proves-the-enemy argument favors White (you wall out what is real and dangerous) — but a defender of the high tradition would say the walls are prudent protection for the unready (the nirvāṇa-dīkṣā is for renouncers seeking liberation, not Virile Heroes seeking the siddhis), not repression. The deepest open question is whether the Kaula's nightmare-confrontation (dreaming and mastering the devouring goddesses) is psychologically superior (integrating the terrifying feminine, the Freudian dream-work done) or more dangerous (courting the horror), and whether the high tradition's walls are repression (the dream-work elided, the feminine denied) or protection (the unready shielded) — and White's Freudian reading takes the former (the Kaula integrates, the high tradition represses), which is a psychological judgment the sources alone cannot settle.
White synthesizes Gananath Obeyesekere and Isabelle Nabokov (the dream/possession ethnography), Michel Strickmann (the "perfect dream"), and Sigmund Freud (the "dream work"), reading the BY/SP contrast through a psychoanalytic lens. The convergence of ethnography (Obeyesekere/Nabokov on living possession-cults), history-of-religions (Strickmann), and depth psychology (Freud) on the same phenomenon (the Tantric dream as the processing of terrifying feminine/demonic content) is the page's interpretive power. What White's psychoanalytic reading adds is the evaluative frame: the Kaula (BY) does the dream-work (confronts and integrates the terrifying feminine), while high Tantra (SP) elides it (walls out the feminine, denies the subconscious) — so the masculinization is, psychologically, a repression. This is White at his most interpretively bold (and most contestable): the Freudian frame lets him read the SP's walls as repression and the BY's nightmare as integration, valuing the Kaula's confrontation over the high tradition's protection. The tension is that the Freudian judgment (confrontation = healthy integration; walling-out = unhealthy repression) is a modern psychological value imported onto a medieval religious difference — a Śaivasiddhānta theologian would not accept that his nirvāṇa-dīkṣā is "repression"; he would call it protection and transcendence. So the BY/SP contrast is real (the texts genuinely differ in their treatment of the feminine), but White's evaluation of it (Kaula-integration-good, high-Tantra-repression-bad) is a psychoanalytic reading that the traditions themselves would dispute. What the convergence-and-tension reveals is that the masculinization (the walling-out of the feminine) can be read two ways — as repression (White/Freud: the feminine denied, the dream-work elided) or as transcendence/protection (the tradition's self-understanding: the unready shielded, the samāveśa achieved without the horror) — and the fortress (the protective measures) is the ambiguous evidence both readings claim: for White it proves the repressed enemy; for the tradition it is prudent defense.
Plainly: one tradition summoned the terrifying feminine in dreams and mastered her; another walled her out with protective mantras and dreamed of clotted milk instead — and the walls themselves prove she was still out there, pressing to get in. That is a problem about how traditions handle the threatening unconscious, and whether confrontation or defense is the deeper path.
The Sharpest Implication If the high-Tantric initiate is protected so thoroughly from the terrifying feminine that he dreams of clotted milk and empty wells — behind a citadel of mantras built specifically against female entities — then the walls themselves confess what the official doctrine denies: the Yoginīs are still out there, real and dangerous, pressing on the fortifications, and the all-male tranquility of the high tradition is not the absence of the terrifying feminine but the successful exclusion of a power that remains. Take that seriously and the masculinization is visible in the dream-content as a repression: where the Kaula initiate dreamed the devouring goddesses, survived being licked by them, and mastered them (the shadow confronted and integrated), the high-Tantric initiate is walled off from them and dreams of dairy (the shadow repressed, displaced into the innocuous). The discomfort is that the peace of the high tradition (the tame dreams, the all-male universe, the Sandman) is purchased by exclusion — and the fortress is the receipt: you do not build a citadel of mantras against female entities unless those entities are real and invasive, so the very serenity of the nirvāṇa-dīkṣā is the evidence of the devouring Yoginīs it has walled out. The bowl of clotted milk is the sublimation's dream — and somewhere outside the wall, the buck-toothed śaktis are still licking, waiting for the citadel to fail.
Generative Questions