Chapter 8

THE SUBLIMATION OF THE YOGINĪ:

The Subordination of the Feminine in High Hindu Tantra

In the opening chapter of this book, I suggested that it was sexual practice and in particular the ritualized consumption of sexual fluids that gave medieval South Asian Tantra its specificity—in other words, that differentiated Tantra from all other forms of religious practice of the period. This, the “hard core” of South Asian Tantra, first appeared as a coherent ritual system—the Kaula—in about the eighth century in central India; and there have since been more recent revivals of the original Kaula impetus, in fourteenth- to sixteenth-century Bengal and Nepal in particular. However, throughout most of South Asia, a marginalization of Kaula practice occurred in elite brahmanic circles, from a very early time onward, which sublimated the “hard core” of Kaula practice into a body of ritual and meditative techniques that did not threaten the purity regulations that have always been the basis for high-caste social constructions of the self.

The sublime edifice of what I have been calling “high” Hindu Tantra in these pages has been, in the main, an internalization, an aestheticization, and a semanticization of Kaula practice. It has been the transformation “from a kind of doing to a kind of knowing,” a system of “overcoding” that has permitted householder practitioners to have it both ways and lead conventional lives while experimenting in secret with Tantric identities.1 This transformation, which was effected over a relatively brief period of time, between the tenth and the twelfth centuries, especially involved the subordination of the feminine—of the multiple Yoginīs, Mothers, and Śaktis (and their human counterparts) of Kaula traditions—to the person of the male practitioner, the male guru in particular. This subordination occurred on a number of levels that involved: (1) the internalization of the Yoginīs and their circles into the cakras of hathayogic practice; (2) the semanticization of the Yoginīs into seed mantras; (3) the masculinization of Tantric initiation; and (4) the introduction of ritual substitutes for the referents of the five M-words, including maithuna.

1. Prehistory of the Cakras

In his masterful book The Kāpālikas and Kālāmukhas, David Lorenzen makes the following cogent point concerning the goals of yogic practice:

In spite of abundant textual references to various siddhis [supernatural enjoyments] in classical Yoga texts, many modern Indian scholars, and like-minded western ones as well, have seized on a single sūtra of Patañjali (3.37) to prove that magical powers were regarded as subsidiary, and even hindrances, to final liberation and consequently not worthy of concentrated pursuits. This attitude may have been operative in Vedāntic and Buddhist circles and is now popular among practitioners imbued with the spirit of the Hindu reformist movements, but it was not the view of Patañjali and certainly not the view of mediaeval exponents of Haha Yoga.2

It suffices to cast a glance at the Yoga Sūtras to see that the acquisition of siddhis was at the forefront of yogic theory and practice in the first centuries of the common era: nearly all of the fifty-five sūtras of book 3 of this work are devoted to the siddhis, and the “disclaimer” in verse 37 of this book—that “these powers are impediments to samādhi, but are acquisitions in a normal fluctuating state of mind”—seems only to apply, in fact, to the siddhis enumerated in the two preceding verses. This is a view shared by P. V. Kane.3

One finds very little of yogic practice, in the sense of techniques involving fixed postures (āsanas) and breath control (prāāyāma), in the Yoga Sūtras. They are, of course, the third and fourth limbs of Patañjali’s eight-limbed yoga (2.29); however, in the grand total of seven sūtras (2.46–52) he devotes to them, Patañjali gives absolutely no detail on these matters, save perhaps a veiled reference to diaphragmatic retention, which he terms stambha-vtti (2.50). References to the subtle body, the channels (nāīs) and energy centers (cakras), are entirely absent from this work (although the bhāya does briefly describe a limited number of āsanas). It would appear in fact that the circa sixth-century B.C.E. Chāndogya Upaniad (8.6.6) had already gone far beyond Patañjali and his commentators when it stated: “There are a hundred and one channels of the heart. One of these passes up to the crown of the head. Going up by it, one goes to immortality. The others are for departing in various directions.”

Moreover, Patañjali’s “classical” definition of “yoga” notwithstanding,4 many if not most pre-twelfth-century accounts of the practice of “yoga,” going back to the MBh,5 describe it not as a form of meditative or physical practice, but rather as a battery of techniques for the attainment of siddhis, including out-of-body experience, entering the bodies of others as a means to escaping death or simply to feed on them, invisibility, the power of flight, transmutation, and so on.6 Similarly, the term “yogin” (or yogeśvara, “master of yoga”), like its feminine form yoginī (or yogeśvarī), most often means “sorcerer” or “magician” in pre-twelfth-century sources: thus, for example, Kāntaśila, the rogue ascetic of the frame story of the KSS, is called a yogin; and Bhaā, who makes a meal of King Baka in the Rājataragiī, is called a yogeśvarī.7 The “Tantric yoga” that is being marketed in places like Hollywood has elided several centuries from the history of the origins and development of yoga, and altered its content beyond recognition.

In this section I will trace the development of a number of elements specific to haha yoga as such emerged in a variety of Hindu and Buddhist sources between the eighth and twelfth centuries C.E. These sources are the eighth-century Buddhist Hevajra Tantra and the following Hindu sources: the eighth-century Bhāgavata Purāa (BhP) and Tantrasadbhāva Tantra; the ninth- to tenth-century KJñN; the tenth- to eleventh-century KM and Jayadrathayāmala; the eleventh-century TĀ; the eleventh- to twelfth-century Rudrayāmala Tantra; and the twelfth-century Śrīmatottara Tantra. In this historical analysis, I will discuss (1) the emergence of the subtle body system of the cakras; (2) the projection of powerful feminine figures from the external world of Tantric ritual onto the grid of the subtle body; and (3) the role of these now-internalized feminine energies, including that known as the kualinī, in the male practitioner’s attainment of siddhis.

One need not go back very far to find the principal source of the seemingly timeless system of the six plus one cakras: this is Arthur Avalon’s edition and translation of a late work, the Sacakranirūpaa, as the principal element of his seminal study, The Serpent Power.8 Perhaps due to the power of the illustrations of this configuration in Avalon’s work, many scholars have taken this to be an immutable, eternal system, as old as yoga itself, and grounded, perhaps, in the yogin’s actual experience of the subtle body. A case in point is a recent work by Rahul Peter Das, which, while it offers an encyclopedic account of subtle body systems in Bengal, is constantly plagued by the author’s frustration in the face of the inconsistencies and contradictions between those systems.9 In fact, there is no “standard” system of the cakras. Every school, sometimes every teacher within each school, has had their own cakra system. These have developed over time, and an “archaeology” of the various configurations is in order.

We have already noted that Hindus have been worshiping groups of Mothers (mātcakras) since at least the sixth century.10 These were circular arrays of goddesses “in the world,” that is, outside of the body, circles represented in mandalas of every sort, including the circular, hypaethral Yoginī temples. The gradual internalization of these powerful female entities was effected by internalizing their formations into the hierarchized cakras of the yogic body. Two early instances of this process may be found in the KJñN and the KM.

We begin with the presentation in the KJñN of six categories of Śaktis—the “Field-born,” “Mound-born,” and so on—that were outlined in chapter 6.11 Here, a comparison may be drawn with a slightly later source, Kemarāja’s eleventh-century commentary to Netra Tantra 19.71. Citing the Tantrasadbhāva, Kemarāja names these same six categories of Śaktis, specifying that unlike the Yoginīs, who dwell in the worlds of Brahmā, Viu, and Indra, these six types of Śaktis all dwell within the body. He then goes on to identify these with six powerful and terrible classes of female entities: the Yoginīs, Devatās, Rupiīs, Śākinīs, Śābarikās, and Śivās. Most of these are described as draining the human body of its “five nectars,” its vital fluids, but the language is ambiguous and seems to imply that they do so from without rather than from within. Following its division of the six Śaktis into internal and external groups of three, the KJñN continues with a description of a seventh type, called “Lowest-born”—that is, an out-caste woman—and then shifts to a description of the worship of a cakra comprised of the sixty-four Yoginīs and the fifty-eight Virile Heroes, “duly presided over by the Sons of the Clan.”12 Fifteen verses later, two sets of seed mantras—termed the “Clan Group of Eight” and the “Wisdom Group of Eight,” comprised of vowels and consonants, respectively—are presented. These are to be written out eight times, with Clan and Wisdom graphemes interspersed. This entire sixty-four-part arrangement is termed the “Yoginī Sequence.”13

It is at this point that the term cakra first comes to be employed in a systematic way in this chapter. One who is devoted to meditation upon and worship of the first cakra, named “Mingling with the Yoginīs” (yoginīmelakam), obtains the eight supernatural powers (siddhis); with the second cakra, one obtains the power of attraction; and with the third, entering into the body of another person; and so on to the eighth, which confers the power of realizing one’s desires and mastery of the six powers of Tantric sorcery. This Great Cakra (mahācakra), raised at its apex (i.e., conical), is ascended through devotion to the Clan. The chapter concludes with the promise that one who knows the sixty-four arrangements becomes perfected, and that the “Sequence of the Sixty-four Yoginīs” is the concealed true essence of these arrangements.14 This data is repeated with variations in chapter 10, with the practitioner meditating on eight cakras of eight petals each, with the total of sixty-four corresponding to eight sets of eight seed mantras.15

In these KJñN passages, the term cakra is being used in a nontechnical way, to simply denote a circle or grouping of divinities, identified with arrangements of the Sanskrit graphemes. A similar situation obtains in the KM. This work—whose five-“cakra” system comprises groups of devīs, dūtīs, māts, yoginīs, and khecarī deities aligned along the vertical axis of the yogic body—nearly never refers to these groupings as cakras.16 According to Dorothea Heilijger-Seelens, the meaning of the term cakra was, in the period in which this work was compiled, generally restricted to the groups of deities located in a mandala, which served as their base or support. The term did not denote a circular array, and even less so one located within the yogic body. Moreover, in those rare cases in which the KM did present the six energy centers by their “standard” names (this is the earliest source in which these are found), it only once referred to one of these—the anāhata—as a cakra.17 These conceptual connections would be made later.

While the KM nonetheless insists that these are internal centers or groupings, it betrays a macrocosmic model when it speaks of their dimensions. The lowest group, the Devīcakra, identified with the element earth, is said to be one hundred kois (of yojanas, according to the commentary18) in size, with the other, higher groups a thousand, hundred thousand, 10 million, and 1,000 million kois in diameter, respectively. These are the precise measurements and proportions given in the tenth chapter of the Svacchanda Tantra—a text that predates the KM by at least a century—of the cosmic egg (one hundred kois), and the surrounding spheres of water, fire, air, and ether.19 This understanding is already present in the KJñN, a text coeval with the Svacchanda Tantra, which gives a measure often kois “beyond the visible [world]” for “this Kaula,” that is, this embodied universe. Also according to the KJñN, when the practitioner reaches a certain threshold of practice, “he sees the threefold universe, with its mobile and immobile beings, inside of his body. . . . With [an extension of] one thousand kois, he is Śiva himself, the maker and destroyer [of the universe].”20 The clear implication here is that the various dimensions of the “outer space” of the universe are being directly projected onto the “inner space” of the human body. In these early references, the circles or spheres of the outer elements, even when they are identified with various groupings of female divinities, are still far removed from the later, “standard” notion of the six cakras of the yogic body.21

2. The Emergence of the Cakras as Components of the Yogic Body

The earliest accounts of the cakras as “circles” or “wheels” of subtle energy located within the yogic body are found in the Caryāgīti and the Hevajra Tantra, two circa eighth-century Buddhist Tantric works that locate four cakras within the human body at the levels of the navel, heart, throat, and head.22 These cakras are identified with four geographical sites (pīhas), which appear to correspond to points of contact between the Indian subcontinent and inner Asia: these are Kāmākhyā (Gauhati, Assam), Uiyāna (Swat Valley?),23 Pūragiri (Punjab?), and Jālandhara (upper Punjab). This tradition is repeated in numerous sources, including those of the Nāth Siddhas, whose twelfth-century founder Gorakanātha identifies the same set of four pīhas with sites aligned along the spinal column within the yogic body.24 The TĀ offers a slightly longer list of pīhas “in the world,” before locating the same within the yogic body, a few verses later.25 The Hevajra Tantra26 also homologizes these four centers with a rich array of scholastic tetradic categories, including Buddha bodies, seed mantras, goddesses, truths, realities, schools, et cetera.27 Their locations in the yogic body appear to correspond as well to the mystic locations of the mind in its four states as described in a number of late Upanishadic traditions, which declare that while one is in a waking state, the mind dwells in the navel; during dreamless sleep, it dwells in the heart; during dream sleep, it resides in the throat; and when in the “fourth state” only attainable by the yogin, it resides in the head.28 Later sources locate ten and, still later, fifty-one pīhas (identified with the Sanskrit phonemes) within the subtle body.29

The vertical configuration of the six plus one cakras that many identify with Hindu subtle body mapping emerges slowly, in the course of the latter half of the first millennium C.E. Perhaps the earliest Hindu source on this system is the BhP, discussed in previous chapters.30 Here, the “six sites” (asu . . . sthāneu) named are the (1) navel (nābhī); (2) heart (hd); (3) breast (uras); (4) root of the palate (svatālumūlam); (5) place between the eyebrows (bhruvorantaram); and (6) cranial vault (mūrdhan), from which he “will then surge upward into the beyond (param).” What is the source of this enumeration in the BhP? A glance at the early medical literature indicates that these sites correspond quite exactly to anatomical notions of the vital points of the body (mahā-marmāi) or the supports of the vital breaths (prāāyatana). These are listed in the circa 100 C.E. Caraka Sahitā as follows: head (mūrdhan), throat (kaha), heart (hdaya), navel (nābhī), bladder (basti), and rectum (guda).31 Certain later sources add the frenum,32 the membrane that attaches the tongue to the lower jaw, to this list: this would correspond to the root of the palate listed in the BhP.

Śaivasiddhānta sources give a slightly different account of the centers. These most commonly list five centers, which they call either sites (sthānas), knots (granthis), supports (ādhāras), or lotuses—but almost never cakras. These are the heart (ht); throat (kaha); palate (tālu); the place between the eyebrows (bhrūmadhya); and the fontanel (brahmarandhra). Quite often, the End of the Twelve (dvādaśānta)—because it is located at twelve finger-breadths above the fontanel—will also be mentioned in these sources, but not as a member of this set of five. So, too, Śaivasiddhānta works will sometimes evoke the root support (mūlādhāra) in its bipolar relationship to the brahmarandhra, but without mention of the intervening centers.33

The first Hindu source to list the locations found in the BhP, and perhaps the first to apply the term cakra to them as well, is the KJñN:

The various spokes [of the wheels] of divine maidens (divyakanyāra) are worshiped by the immortal host in (1) the secret place (genitals), (2) navel, (3) heart, (4) throat, (5) mouth, (6) forehead, and (7) crown of the head. [These maidens] are arrayed along the spine (pamadhye) [up] to the trident (tridaakam) [located at the level of] the fontanel (muasandhi). These cakras are of eleven sorts and comprised of thousands [of maidens?], O Goddess! [They are] five-spoked (pañcāram) and eight-leaved (aa-pattram), [as well as] ten- and twelve-leaved, sixteen- and one hundred–leaved, as well as one hundred thousand–leaved.34

This passage continues with a discussion of these divine maidens, through whom various siddhis are attained, each of whom is identified by the color of her garb (red, yellow, smoky, white, etc.). So it is that we find in this source a juxtaposition of (1) the locations of the cakras; (2) the use of the term cakra; (3) a description of the cakras as being composed of spokes and leaves (but not petals); and (4) a portrayal of color-coded divine maidens as dwelling in or on the spokes of these cakras. The problematic remark in this passage, that the cakras are in some way elevenfold, or of eleven sorts, appears to be explicated in the seventeenth chapter of the same source, which names eleven sites, of which six correspond to the six sites or cakras:

The (1) rectum, (2) secret place (genitals), along with the (3) navel [and] (4) the downturned lotus (padma) in the heart, (5) the cakra of breath and utterances (samīrastobhakam) [i.e., the throat], (6) the cooling knot (granthi) of the uvula, (7) the root (or tip) of the nose, and the (8) End of the Twelve;35 the (9) [site] located between the eyebrows; (10) the forehead; and the brilliant (11) cleft of brahman, located at the crown of the head: it is the stated doctrine that [this] elevenfold [system] is located in the midst of the body.36

In addition to using the term cakra, this passage also refers to the down-turned lotus (and not wheel) in the heart, as well as to a knot (granthi) located at the level of the uvula.37 It would appear that Matsyendra’s yogic body system contributed to the synthesis presented in the writings of Abhinavagupta. In TĀ 29.37 he names the End of the Twelve, the “upward kualinī (ūrdhvagakualinī), the place between the eyebrows (baindava), heart, umbilicus, and the “bulb” (kandam) as the six “secret places” (chommas) through which the kula is transmitted from teacher to disciple.38 Abhinavagupta’s system also features a trident (triśula), located at the level of the fontanel, and a thousand-spoked End of the Twelve. However, we must note that whereas the KJñN discusses these centers as wheels possessed of spokes or leaves, or as lotuses, the cakras of the subtle body in Trika Kaulas sources are whirling spoked wheels that, in the body of the nonpractitioner, become inextricable tangles of coils called knots (granthis) because they knot together spirit and matter.39

Another likely source of Abhinavagupta’s synthesis is the Netra Tantra, of which his disciple Kemarāja wrote an extensive commentary. The seventh chapter of this work, entitled the “Subtle Meditation on the ‘Death-Conquering’ [Mantra],” comprises a discussion of two subtle body systems, which Kemarāja qualifies as belonging to the “Kaula” and “Tantric” liturgies, respectively.40 Taken together, the two systems presented in the text and commentary appear to be more direct forerunners of the later haha yoga system of Gorakanātha than do the KJñN and other works attributed to Matsyendranātha, who was Gorakanātha’s guru, according to Nāth Siddha tradition. The Netra Tantra’s presentation of yogic practice combines breath control, meditation, “the piercing” of knots and the central channel, the raising of the “Śakti who is filled with one’s semen” the length of that channel,41 and the internal production of the nectar of immortality.42 At the same time, the Netra Tantra agrees with the KJñN on a number of subtle body locations; for example, the “Fire of Time” (kālāgni), which it locates at the tips of the toes; and “Fish-Belly,” which it locates at the level of the genitals.43 Such is not the case, however, for the Netra Tantra’s presentation of the six cakras, which is idiosyncratic with regard to every other yogic body system: “The nāīcakra is [located] in the ‘place of generation’; the [cakra] called māyā is in the navel; the yogicakra is placed in the heart; while the [cakra] known as bhedana is placed in the uvula. The dipticakra is placed in the ‘drop’ (bindu) and the [cakra] called śānta is in the ‘reverberation’ (nāda).”44 The sole source to mention any one of these cakras is the eighth-century Mālatī-Mādhava, in which it is the nāīcakra that powers Kapālakualā’s flight.45 A mention in the Jayadrathayāmala of “māyā [as] the mother of the phonemes . . . the kualinī may be a reference to the second of the Netra Tantra cakras.46

Returning to the KJñN, yet another discussion of subtle body mapping occurs in this source under the heading of sites (sthānas). Here, it describes eleven of these in terms of their spokes, leaves, and petals (dalas): in order, they are the four-leaved, eight-spoked, twelve-spoked, five-spoked, sixteen-spoked, sixty-four-petaled, one hundred–leaved, one thousand–petaled, 10 million–leaved, 5 million–leaved, and 30 million–leaved.47 It then goes on to discuss a number of other subtle sites (vyāpaka, vyāpinī, unmana, etc.), located in the upper cranial vault, that one finds in other Kaula sources, including the Svacchanda Tantra and Netra Tantra.48

A final KJñN evocation of the workings of the subtle body will serve to orient us, once again, toward the KM.49 This is the work’s fourteenth and longest chapter, much of which comprises a rambling account of supernatural powers realized by “working the mind” through a sequence (krama) of yogic body locations, variously called cakras and “kaulas” (“clans of internal Siddhas”?).50 Toward the end of this meditative ascent, the KJñN (14.92) evokes “this seal, which is called ‘Unnamed” (anāmā nama mudreyam), and states that “sealed with the five seals . . . one should pierce that door whose bolts are well-fitted.” One finds similar language in the KM, for which “Unnamed” is one of the names of the goddess Kubjikā.51 Here, the statement “applications of the bolts on the openings of the body,”52 occurs at the beginning of this work’s discussion of “upward progress” (utkrānti),53 which appears to be a type of hathayogic practice. The KM passage continues: “The rectum, penis, and navel, mouth, nose, ears and eyes: having fitted bolts in these places (i.e., the nine ‘doors’ or bodily orifices), one should impel the ‘crooked one’ (kuñcikā) upward.”54 Then follows a discussion of a number of yogic techniques—including the Cock Posture (kukuāsana)—which effect the piercing of the knot[s], confer numerous siddhis, and afford firmness of the self.55

Bhairava, the divine revealer of the KM, next states that he will provide a description of what he calls the “bolt-practices” of the knife (kurikādyargalābhyāsa), and so on, which effect upward progress (utkrānti-kāraam) in him who is empowered to use it, and great affliction in the unempowered. Having already discussed this ritual in earlier chapters, I will not go into a description of its details at this point.56 Here, the salient point of this passage concerns the names of the goddesses invoked and the bodily constituents offered to them. In order, their names are Kusumamālinī (“She Who Is Garlanded with Flowers”),57 Yakiī, Śakhinī, Kākinī, Lākinī, Rākiī, and ākinī. These Yoginīs are named in nearly identical order in the eighteenth chapter of the Śrīmatottara Tantra, a later text of the same Kubjikā tradition. Here, the names listed are ākinī, Rākiī, Lākinī, Kākinī, Śākinī, Hākinī, Yākinī, and Kusumā.58 They are listed in the same order in Agni Purāa 144.28b–29a. In this last case, their names are enumerated in instructions for the construction of the six-cornered Kubjikā mandala, with the ordering proceeding from the northwest corner.59 This mandala is identical to the Yoginīcakra, the fourth of the five cakras of the Kubjikā system, located at the level of the throat, as described in the fifteenth chapter of the KM itself.60 A shorter, variant list of these Yoginīs is found in two places in the KJñN, and chapter 4 of the KJñN, which is devoted to Tantric sorcery, appears to be a source for the data found in a number of later Kubjikā traditions.61

What the Yoginīs are offered is of signal interest here: the first of these, Kusumamālinī, is urged to take or swallow (gha) the practitioner’s “prime bodily constituent,” that is, semen; the second, Yakiī, to crush his bones; the third, Śakhinī, to take his marrow; the fourth, Kākinī, to take his fat; the fifth, Lākinī, to eat his flesh; the sixth, Rākiī, to take his blood; and the seventh, ākinī, to take his skin.62 Clearly, the bodily constituents these goddesses are urged to consume constitute a hierarchy. These are, in fact, the standard series of the seven dhātus, the “bodily constituents” of Hindu medical tradition (with the sole exception being that skin has here replaced chyle [rasa]), which are serially burned in the fires of digestion, until semen, the “prime bodily constituent,” is produced.63 With each goddess invoked in this passage, the practitioner is offering the products of a series of refining processes.

To all appearances, this is a rudimentary form of the hathayogic raising of the kualinī. What is missing here is an identification between the goddesses to whom one’s hierarchized bodily constituents are offered and subtle body locations inside the practitioner. This connection is made, however, in another KM passage, which locates six Yoginīs, called the “regents of the six fortresses,” as follows: āmarī is located in the ādhāra, Rāmaī in the svādhihāna, Lambakarī in the maipura, Kākī in the anāhata, Sākinī in the viśuddhi, and Yakiī in the āā.64 In another chapter the KM lists two sequences of six goddesses as kulākula and kula, respectively. The first denotes the “northern course” of the six cakras, from the āā down to the ādhāra, and the latter the “southern course,” in reverse order. The former group is creative, and the latter—comprised of ākinī, Rākiī, Lākinī, Kākinī, Śākinī, and Hākinī—is destructive.65

A number of later sources,66 beginning with the Rudrayāmala Tantra, identify these goddesses, which they call Yoginīs, with the cakras as well as with the dhātus, the bodily constituents. The Rudrayāmala Tantra’s ordering identifies these Yoginīs with the following subtle body locations: ākinī is in the mūlādhāra; Rākiī in the svādhihāna; Lākinī in the maipura; Kākinī in the anāhata; Śākinī in the viśuddhi; and Hākinī in the āā.67 Kusumamālā, who is missing from this listing, is located in the feet in the Śrīmatottara Tantra;68 other works place a figure named Yākinī at the level of the sahasrāra.69 These Rudrayāmala Tantra locations correspond, of course, to the “standard” names of the six cakras of later hathayogic tradition. They are, in fact, first called by these names in the KM, which correlates the six standard yogic body locations with its Yoginīs of the “northern course.”

Mark Dyczkowski has argued that it was within the Kubjikā traditions that the six-cakra configuration was first developed into a fixed coherent system.70 The KM, the root Tantra of the Kubjikā tradition, locates the cakras and assigns each of them a number of “divisions” (bhedas) or “portions” (kalās), which approximates the number of “petals” assigned to each of these “lotuses” in later sources.71 We also encounter in the KM the notion of a process of yogic refinement or extraction of fluid bodily constituents, which is superimposed upon the vertical grid of the subtle body, along the spinal column, leading from the rectum to the cranial vault. Nonetheless, it would be incorrect to state that there is a hathayogic dynamic to the KM’s system of the cakras. What is lacking are the explicit application of the term cakra to these centers, the explicit identification of these centers with the elements,72 and the deification or hypostasization of the principle or dynamic of this refinement process: here I am referring to that commonplace of hathayogic theory, the female kualinī or serpent power—who has perhaps been evoked, albeit not by name, in the statement made in this source that one should, through utkrānti, “impel the crooked one upward” (KM 23.114a).

3. The Kualinī and the Channeling of Feminine Energies

The KM makes a number of other statements that appear to betray its familiarity with a notion of this serpentine feminine nexus of yogic energy.73 In KM 5.84 we read that “[śakti] having the form of a sleeping serpent [is located] at the End of the Twelve. . . . Nevertheless, she is also to be found dwelling in the navel. . . .”74 This serpentine (bhuja[]ga-ākārā) śakti is connected in this passage to mantras and subtle levels of speech, through which she is reunited with Śiva. A later passage (KM 12.60–67) describes the sexual “churning” (ma[n]thanam) of an inner phallus (ligam) and vulva (yoni) that occurs in the maipura cakra,75 that is, at the level of the navel. Here, however, the language is not phonematic, but rather fluid: this churning of Śiva and Śakti produces a flood of nectar.

This is not, however, the earliest mention of this indwelling female serpent to be found in Hindu literature. This distinction likely falls to the circa eighth-century C.E. Tantrasadbhāva Tantra,76 which similarly evokes her in a discussion of the phonematic energy that also uses the image of churning:

This energy is called supreme, subtle, transcending all norm or practice. . . . Enclosing within herself the fluid drop (bindu) of the heart, her aspect is that of a snake lying in deep sleep . . . she is awakened by the supreme sound whose nature is knowledge, being churned by the bindu resting in her womb. . . . Awakened by this [luminous throbbing], the subtle force (kalā), Kualī is aroused. The sovereign bindu [Śiva], who is in the womb of Śakti, is possessed of a fourfold force (kalā). By the union of the Churner and of She that is Being Churned, this [Kualī] becomes straight. This [Śakti], when she abides between the two bindus, is called Jyehā. . . . In the heart, she is said to be of one atom. In the throat, she is of two atoms. She is known as being of three atoms when permanently abiding on the tip of the tongue. . . .77

In this passage we may be in the presence of the earliest mention of a coiled “serpent energy”; however, the term that is used here is kualī, which simply means “she who is ring-shaped.” This is also the term that one encounters in the KJñN, which evokes the following goddesses in succession as the Mothers (mātkās) who are identified with the “mass of sound” (śabdarāśi) located in “all of the knots” (sarvagrantheu) of the subtle body: Vāmā, Kualī, Jyehā, Manonmanī, Rudra-śakti, and Kāmākhyā.78 Also mentioned in this passage are the “female” phonemes called the Mātkās (“Little Mothers”) and the “male” phonemes called the Śabdarāśi (“Mass of Sounds”). Here we already detect the process of the semanticization of the Goddess and her energies, a process that becomes predominant in later Tantric traditions.79 In another passage the KJñN describes Vāmā as having an annular or serpentine form (kualākti) and extending from the feet to the crown of the head: the raising of this goddess from the rectum culminates with her absorption at the End of the Twelve.80 Once again the kualinī serpent appears to be present here in everything but precise name.

Let us dwell for a moment on the names of the Mother goddesses evoked in the KJñN. In Śaivasiddhānta metaphysics, the goddess Jyehā(devī), mentioned in the KJñN and Tantrasadbhāva passages, is described as assuming eight forms, by which she represents eight tattvas: these are Vāmā (earth), Jyehā herself (water), Raudrī (fire), Kālī (air), Kalavikaraī (ether), Balavikaraī (moon), Balapramathinī (sun), and Manonmanī (Śiva-hood). This group of eight are said to be the śaktis of the eight male Vidyeśvaras of the Śaivasiddhānta system, the deifications of the eight categories of being that separate the “pure” worlds from the “impure.”81 With this enumeration, we may surmise that Matsyendranātha was drawing on the same source as the Saiddhāntika metaphysicians.82 In addition, we once more see a hierarchization of internalized goddesses, identified here with the five elements (and a number of their subtler evolutes), as well as with the ordering of phonemes within the yogic body. That these are projected upon the grid of the yogic body is made clear by the fact that they are said to be located “in all the knots.” Finally, this list of deities from the Saiddhāntika system is complemented by the Mother named Kualī whom the KJñN locates between Vāmā (earth) and Jyehā (water).83 It is a commonplace of later subtle body mapping to identify the five lower cakras with the five elements: Kualī would thus be located, according to this schema, between the rectal mūlādhāra (earth) and the genital svādhihāna (water).

Jyehā (“Eldest”) is a goddess whose cult goes back to the time of the fifth- to second-century B.C.E. Baudhāyana Ghya Sūtra.84 As was indicated in chapter 2, she is a dread goddess who is mentioned together or identified with such terrible Mothers as Harītī, Pūtanā, and Jarā,85 and inauspicious (alakmī) astrological configurations: in the Indian calendar, the month of Jyaia, falling as it does in the deadly heat of the premonsoon season, is the cruelest month. Jyehā’s names and epithets are all dire—“Ass-Rider,” “Crow-Bannered,” and “Bad Woman” (Alakmī)—and she is depicted in her iconography with a sweeping broom, the symbolic homologue of the winnowing fan carried by the smallpox goddess Śītalā.86 Jyehā belongs to an early triad of goddesses—the other two being Vāmā and Raudrī—who would later become identified with the three Śaktis (Icchā-, Jñānā- and Kriyā-), the three phonemes A, Ā, and Ī, as well as the goddesses Parā, Aparā, and Parāparā of the Trika pantheon. References to Parā, Aparā, and Parāparā in the Mālinīvijayottara Tantra (3.30–33) indicate that this triad was an appropriation of an earlier threefold division of classes of Mothers: those that liberate souls (aghorā), those that impede souls (ghorā), and those that drag souls downward (ghorātarā).87

Both the KJñN account of the raising of the ring-shaped goddess Vāmā from the level of the rectum to the End of the Twelve and the statement in KM 5.84 that Śakti dwells in the form of a sleeping serpent in both the cranial vault and the navel are precursors of the dynamic role of the kualinī in later hathayogic sources. In the KJñN passage, the goddess’s ring shape evokes the circles of Yoginīs that rise into the air at the conclusion of their cremation-ground rites—and it should be recalled here that the cakras themselves are referred to as cremation grounds in the later hathayogic literature.88 In the KM passage, it is the upward motion of feminine energy that is stressed.

Perhaps the earliest occurrence of the term kualinī (as opposed to kualī) is found in the third hexad (aka) of the tenth- to eleventh-century Jayadrathayāmala,89 which, in a discussion of the origin of mantras from the supreme god Bhairava, relates the kualinī to phonemes as well as to the kalās, to which we will return:

Māyā is the mother of the phonemes and is known as the fire-stick of the mantras. She is the kualinī Śakti, and is to be known as the supreme kalā. From that spring forth the mantras as well as the separate clans, and likewise the Tantras. . . .90

Abhinavagupta, who likely took his inspiration from all of the sources we have been reviewing, develops this principle in his discussion of the upper and lower kualinīs, which are two phases of the same energy, in expansion and contraction, that effects the descent of transcendent consciousness into the human microcosm, and the return of human consciousness toward its transcendent source. Often he portrays these as spoked wheels that, aligned along a central axis or axle, rise and descend to whirl in harmony with one another. In spite of the highly evocative sexual language he employs, Abhinavagupta’s model is nonetheless one of phone-matic, rather than fluid, expansion and contraction.91

It not until the Rudrayāmala Tantra and the later hathayogic classics attributed to Gorakanātha that the kualinī becomes the vehicle for fluid, rather than phonematic transactions and transfers. This role of the kualinī in the dynamics of yogic body fluid transfer is brought to the fore in a portion of the Tantric practice of the five M-words, which Agehananda Bharati describes:

When the practitioner is poised to drink the liquor, he says “I sacrifice”; and as he does so, he mentally draws the coiled energy of the Clan (kulakualinī) from her seat in the base cakra. This time, however, he does not draw her up into the thousand-petaled sahasrāra in the cranial vault, but instead he brings her to the tip of his tongue and seats her there. At this moment he drinks the beverage from its bowl, and as he drinks she impresses the thought on his mind that it is not he himself who is drinking, but the kula-kualinī now seated on the tip of his tongue, to whom he is offering the liquid as a libation. In the same manner he now empties all the other bowls as he visualizes that he feeds their contents as oblations to the Goddess—for the kula-kualinī is the microcosmic aspect of the universal Śakti.92

Here, the coiled energy at the tip of the practitioner’s tongue is not spitting phonemes, as in the Tantrasadbhāva Tantra passage quoted above, but rather drinking ritual fluids, which are so many substitutes for, or actual instantiations of, vital bodily fluids. One may speculate as to why it is that the feminine principle of yogic energy comes to be represented as a serpent, now coiled, and now straightened. Of course, there seems to be some sort of elective affinity between the kualinī’s function and form—however, the avian gander (hasa), which doubles for the kualinī in a number of sources, appears to fulfill the same function of raising energy from the lower to the upper body.93 The KJñN’s discussion of the “goddess named Vāmā” is framed, tellingly, by a disquisition on the hasa:

From below to above the gander sports, until it is absorbed at the End of the Twelve. Seated in the heart it remains motionless, like water inside a pot. Having the appearance of a lotus fiber, it partakes neither of being nor of nonbeing. Neither supporting nor supported, it is omniscient, rising in every direction. Spontaneously, it moves upward, and spontaneously it returns downward. . . . Knowing its essence, one [is freed] from the bonds of existence. . . . In the ear [orally] and in the heart, the description of the gander is to be made known. [Its] call becomes manifest in the throat, [audible] near and far. From the base of the feet to the highest height, the [goddess] named Vāmā has the form of a ring (kualāktim). It is she who, seated in the anus, rises upward until she is absorbed at the End of the Twelve. Thus indeed the gander sports in the midst of a body that is both auspicious and inauspicious.94

Lilian Silburn suggests that it is the serpent’s coiling and straightening that explain its projection upon the subtle body: a venomous serpent, when coiled, is dangerous; straightened, it is no longer threatening. This would be of a piece with the characterization of the kualinī as “poison” when she lies coiled in the lower body and “nectar” when she is extended upward into the cranial vault. Or, Silburn suggests, the image of the kualinī is one that borrows from the Vedic creatures Ahir Budhnya and Aja Ekapāda, or the Puranic Śea and Ananta. In fact, the KJñN describes the Goddess’s body as being “enveloped in fire and having the form of Ekapāda (i.e., of a serpent).”95 I am more inclined to see the kualinī’s origins in the role of the serpent in Indian iconography. Temples and other buildings are symbolically supported by a serpent that coils around their foundations, an image represented graphically by a certain number of Hindu temples in Indonesia. Similarly, images of the Buddha and later of Viu are figured with a serpent support and canopy. Finally, the phallic emblem of Śiva, the ligam, is often sculpted with a coiled serpent around its base, whose spread hood serves as its canopy. This is a particularly evocative image when one recalls that the kualinī is figured in the classical hathayogic sources as sleeping coiled three and a half times around an internal ligam, with her hood or mouth covering its tip. When the yogin awakens her through his practice of postures and breath control, she pierces the lower door to the medial suumā channel and “flies” upward to the place of Śiva in the cranial vault.

4. Transformations in the Art of Love

The theoreticians of post-tenth-century C.E. high Hindu Tantra were especially innovative in their integration of aesthetic and linguistic theory into their reinterpretation of earlier theory and practice. As such, the acoustic and photic registers lie at the forefront of their metaphysical systems, according to which the absolute godhead, which is effulgent pure consciousness, communicates itself to the world and especially to the human microcosm as a stream or wave of phosphorescent light, and as a “garland” of the vibrating phonemes of the Sanskrit language. And because the universe is brought into being by a divine outpouring of light and sound, the Tantric practitioner may return to and identify himself with this pure consciousness by meditatively recondensing those same photemes of light and phonemes of sound into their higher principles.

This is, in the main, a gnoseological process, in which knowing takes priority over doing. In fact, as Alexis Sanderson has argued, one may see in the high Hindu Tantra of the later Trika and Śrīvidyā the end of ritual: “since [the] Impurity [that is the sole impediment to liberation] has been dematerialized, ritual must work on ignorance itself; and to do this it must be a kind of knowing.”96 Of course, a similar transformation had already occurred over two millennia earlier in India, in what Jan Heesterman has termed the transformation of sacrifice into ritual:

The “science of ritual” . . . should be rated as a paradigm of what Max Weber called “formal rationality.” Its rational bent becomes apparent when we notice that it is not just to be done but is required to be “known.” What has to be known are the equivalences, the keystone of ritualistic thought, to which the ubiquitous phrase “he who knows thus” refers.97

In a sense, high Hindu Tantra ritualizes—that is, “gnoseologizes”—Kaula sacrifice in the same way that the Brāhmaas did the sacrificial system of the Vedic Sahitās; and it is worth recalling here that the term “Tantra” originally applied to the auxiliary acts of the ritual complex of a given sacrifice. It is the general and largely unchangeable part of the complex and the same for all sacrifices of the same type.98

With this, we return to the practice of the kāmakalā, introduced in chapter 4.99 In the high Hindu Tantric context, the ritual component of the kāmakalā—that is, rajapāna, the drinking of female discharge—becomes abstracted into a program of meditation whose goal is a nondiscursive realization of the enlightened nondual consciousness that had there-to-fore been one’s object of knowledge. Through the meditative practice of mantras (phonematic, acoustic manifestations of the absolute) and of mandalas or yantras (photemes, i.e., luminous, graphic, visual representations of the same), the consciousness of the practitioner is uplifted and transformed to gradually become god-consciousness. But what is the nature of the “practice” involved here? It is reduced to knowing, as the most significant Śrīvidyā work on the kāmakalā, aptly entitled Kāmakalāvilāsa (“The Love-Play of the Particle of Desire”),100 makes clear (verse 8): “Now this is the vidyā of Kāma-kalā, which deals with the sequence of the cakras [the nine triangles of the Śrīcakra] of the Devī. He by whom this is known becomes liberated and [the supreme Goddess] Mahātripurāsundarī Herself.”

Yet even as the acoustic and the photic, phosphorescing drops of sound lay at the forefront of high Hindu Tantric practice, there was a substratum that persisted from other traditions, a substratum that was neither acoustic nor photic but, rather, fluid, with the fluid in question being sexual fluid. As we have seen in these earlier or parallel traditions, it was via a sexually transmitted stream or flow of sexual fluids that the practitioner tapped into the source of that stream, usually the male Śiva, who has been represented iconographically, since at least the second century B.C.E., as a phallic image, a ligam. Śiva does not, however, stand alone in this flow of sexual fluids. In most Tantric contexts, his self-manifestation is effected through his female hypostasis, the Goddess, whose own sexual fluid carries his divine germ plasm through the lineages or transmissions of the Tantric clans, clans in which the Yoginīs play a crucial role. In the earlier Kaula practice, it was via this flow of the clan fluid through the wombs of Yoginīs that the male practitioner was empowered to return to and identify himself with the godhead. It was this that lay at the root of the original practice of the kāmakalā, the Art of Love.

5. Śrīvidyā Practice of the Kāmakalā

Here I present a detailed account of the multileveled symbolism of the kāmakalā, as it is found in the primary Śrīvidyā sources, in order to demonstrate how the description itself of the kāmakalā diagram represents a semanticization or overcoding of the Kaula ritual upon which it is based. A word on the meanings and usages of this term is in order, composed as it is of two extremely common nouns, both of which are possessed of a wide semantic field. The simplest translation of the term might well be “The Art (kalā) of Love (kāma).” Two other important senses of the term kalā yield the additional meanings of “Love’s Lunar Digit” or “Love’s Sixteenth Portion.” Earlier, we also saw the use of the term kalā in early yogic body descriptions as a subtle force, synonymous with the kualinī, and the mother of phonemes.101 Commenting on Abhinavagupta’s Tantrāloka (TĀ), Jayaratha (fl. ca. 1225–1275) refers to the kāmakalā or kāmatattva as the “Particle (or Essence) of Love,” a gloss to which I will return.102

Nowhere in the history of these medieval traditions is the kāmakalā accorded greater importance than in Śrīvidyā, which, likely born in Kashmir in the eleventh century, came to know its greatest success in south India, where it has remained the mainstream form of Śākta Tantra in Tamil Nadu, down to the present day.103 The kāmakalā is of central importance to Śrīvidyā because it is this diagram that grounds and animates the Śrīcakra or Śrīyantra, the primary diagrammatic representation of the godhead in that tradition. Thus verse 8 of the thirteenth-century Kāmakalāvilāsa [KKV] of Puyānandanātha states that “the Vidyā of the Kāmakalā . . . deals with the sequence of the Cakra [of the Śrīcakra] of the Devī. . . .”104

The Śrīcakra is portrayed as a “drop” (bindu) located at the center of an elaborate diagram of nine nesting and interlocking triangles (called cakras), surrounded by two circles of lotus petals, with the whole encased within the standard gated frame, called the “earth citadel” (bhūpura). The principal ritual practice of Śrīvidyā is meditation on this cosmogram, which stands as an abstract depiction of the interactions of the male and female forces that generate, animate, and ultimately cause to re-implode the phenomenal universe-as-consciousness. The practitioner’s meditative absorption into the heart of this diagram effects a gnoseological implosion of the manifest universe back into its nonmanifest divine source, and of mundane human consciousness back into supermundane god-consciousness, the vanishing point at the heart of the diagram. In the Śrīvidyā system, these male and female principles are named Kāmeśvara and Kāmeśvarī, “Lord of Love” and “Our Lady of Love,” a pair we have already encountered in an Indonesian ritual of royal consecration.105

To maintain the image of the drop, as the Śrīvidyā sources do, it is appropriate to conceive the entire diagram, with its many “stress lines” of intersecting flows of energy and consciousness, as a diffraction pattern of the wave action initiated when the energy of a single drop, falling into a square recipient of calm water, sends out a set of ripples that interfere constructively and destructively with one another. This, too, appears to be the image the Śrīvidyā theoreticians had in mind when they described the relationship of the nonmanifest male and manifest female aspects of the godhead in terms of water and waves. In his commentary on Yoginīhdaya (YH) 1.55, the thirteenth- to fourteenth-century Amtānanda (whose teacher, Puyānandanātha, was the author of the KKV)106 states:

The waves are the amassing, the multitude of the constituent parts of Kāmeśvara and Kāmeśvarī. It [the heart of the Śrīcakra] is surrounded by these waves and ripples as they heave [together]. . . . Here, the word “wave” (ūrmi) means that Parameśvara [here, a synonym of Kāmeśvara], who is light, is the ocean; and Kāmeśvarī, who is conscious awareness, is its flowing waters, with the waves being the multitude of energies into which they [Parameśvara and Kāmeśvarī] amass themselves. Just as waves arise on the [surface of the] ocean and are reabsorbed into it, so too the [Śrī]cakra, composed of the thirty-six tattvas . . . arises from and goes [back to Parameśvara].107

It is, then, a phosphorescing (sphurad) drop of sound (bindu) that animates this cosmogram and the universe, and into which the mind of the person who meditates upon it is resorbed. This drop is the point located at the center of the Śrīcakra, and the kāmakalā is a “close-up,” as it were, of this drop. When one zooms in on it meditatively, one sees that it is composed of three or four elements whose interplay constitutes the first moment of the transition, within the godhead, from pure interiority to external manifestation, from the pure light of effulgent consciousness (prakāśa) to conscious awareness (vimarśa). I now give an account of these constituent elements of the kāmakalā and the means and ends of meditation upon them, as described in the Śrīvidyā and the broader Tantric literature.108

Dirk Jan Hoens has translated kāmakalā as the “Divine Principle (kalā) [manifesting itself as] Desire (kāma).” In this context,

the triad of Śiva-Śakti-Nāda [are given] the name Kāmakalā. . . . Śiva and Śakti are called Kāmeśvara and Kāmeśvarī. The kāmakalā symbolizes the creative union of the primeval parental pair; a pulsating, cosmic atom with two nuclei graphically represented by a white and red dot which automatically produce a third point of gravity. This situation is often represented in graphical form as a triangle. This can be done in two ways: with the point upwards or downwards. . . . A final step is taken when this triad is enriched with a fourth element so as to constitute the graphic representation of the most potent parts of Devī’s mystical body (also in this context she is called Kāmakalā or Tripurāsundarī): her face, two breasts (the white and red bindus) and womb [yoni]. They are represented by the letter Ī written in an older form akin to the Newari [or Brahmi] sign, or by the ha (the “womb” is often called hārdhakalā, “the particle consisting of half the ha,” i.e., its lower part). . . .109

image

Figure 8.a. Kāmakalā yantra according to the Yoginīhdaya Tantra. Adobe Photoshop image.

In this yantra (fig. 8.a),110 the upturned triangle represents Śiva, and the downturned triangle, his consort Śakti.111 At the apex of the upturned Śiva triangle, we find the Sanskrit grapheme A, which is also the sun and the mouth of the maiden who is the support for this meditation. This is also termed the “medial bindu.” The two bindus or points that form the visarga (the surd Sanskrit phoneme represented by two bindus) are the two base angles of this triangle: they are identified with fire and the moon. They are also the breasts of the maiden. Located between these two and pointing downward is the apex of the downturned Śakti triangle, which is the yoni of the maiden and the locus of the grapheme HA. Natānandanātha, the commentator of the KKV, explains that these elements, taken together, constitute a phonematic rendering of the kāmakalā, since Kāma is Paramaśiva (whose desire to create gives rise to the universe), pure effulgence, and the first phoneme, which is A; and Kalā signifies reflective consciousness and the last phoneme, which is HA.112

Located in the heart of the hexagon formed by the two intersecting triangles is the kualinī, the coiled serpent who here takes the form of the Sanskrit grapheme Ī (which, together with the bindu—the graphic dot over a Sanskrit character that represents the nasalization of a sound—becomes Ī) However, Ī is also the special grapheme of the supreme Śrīvidyā goddess, Tripurāsundarī. Termed the trikhaā (“having three parts”), it is meditatively viewed as the body of the goddess, composed of head, breasts, and yoni.113 As such, it constitutes a redoubling of the symbolism of the intersecting Śiva and Śakti triangles. It is in the form of the Ī grapheme then, that energy, in the coiled form of the kualinī serpent, dwells between the bindu and the visarga, that is, between the first and last phonemes and graphemes of the Sanskrit “garland of letters.” Lastly, the kualinī is represented in the form of the serpentine grapheme Ī because it is a commonplace of the Hindu yogic tradition that the female Śakti, which dwells in a tightly coiled form in the lower abdomen of humans, can be awakened through yogic practice to uncoil and rise upward, along the spinal column, to the cranial vault. Here then, the grapheme Ī also represents a yogic process that extends from the base to the apex of the yogic body. Later commentators would find additional correlates to this configuration, identifying the four components of face, breasts, and yoni with four goddesses, four stages of speech, and four cakras within the subtle body.114

There are no less than six levels of overcoding in the Tantric interpretation of this diagram, which reflect so many bipolar oppositions mediated by a third dynamic or transformative element. These oppositions are (1) Śiva and Śakti, the male and female principles of the universe in essence and manifestation; (2) the phonemes A and HA, the primal and final utterances of the phonematic continuum that is the Sanskrit alphabet; (3) the effulgent graphemes or photemes representing the phonemes A and HA, here the bindu (a single point or drop) and the visarga (a double point or drop);115 (4) two subtle or yogic “drops,” the one red and female and the other white and male, which combine to form a third “great drop”;116 (5) male and female sexual emissions; and (6) the corporeal mouth and vulva of the maiden upon whom this diagram is projected in Kaula-based practice.

These bipolarities are mediated by the serpentine nexus of female energy, the kualinī, who in her yogic rise from the base to the apex of the system is described as telescoping the lower phonemes and graphemes of the Sanskrit garland of letters back into their higher evolutes, until all are absorbed in the bindu, the dimensionless point at which all manifest sound and image dissolve into silence and emptiness, in the cranial vault. Also bearing a yogic valence in this diagram and its interpretation are the elements sun, moon, and fire. Identified here with the upper bindu and lower visarga, respectively, these also represent the three primary channels of yogic energy, the right, left, and central channels, respectively.

Finally, we also detect a sexual substrate to this diagram. First of all, the first member of the compound is, after all, “kāma,” erotic love, and the name of the Indian Eros or Love, whose arts are described in works like the Kāma Sūtra. Second, the ritual support of this meditation is a maiden’s naked body. Of course, in high Hindu Tantra, the flesh-and-blood maiden substrate is done away with, with the abstract schematic visualization sufficing for the refined practitioner. Yet she remains present, just beneath the surface of her geometric and semantic abstraction, as such was effected in these later cosmeticized traditions. In a discussion of the kāmakalā, the Yoginīhdaya describes the two bindus that make up the corners of the base of the Śiva triangle and the breasts of the maiden as red and white in color. Here, the white and red drops are “Śiva and Śakti absorbed in their movement of expansion and contraction.”117 Clearly, the bindus so described are not abstract points but rather subtle drops of sexual fluids, that is, male semen and female uterine blood.118 Thus, the bindu as photic grapheme (dimensionless point of light) and the bindu as acoustic phoneme (dimensionless vibration, particle of speech) are overcodings of the abstract red and white bindus of the subtle body physiology of yogic practice, which are in turn overcodings of concrete drops of male and female sexual fluids (particles of love). These unite, in the upper bindu at the apex of the triangle, in the mouth (mukham) of the maiden, into a mahābindu, a “great drop.” We are reminded, however, that her mouth, the apex of the upturned Śiva triangle, is “reflected” in her vulva, the apex of the downturned Śakti triangle.119 Furthermore, as we have noted, a woman’s oral cavity is reflected, redoubled in her vulva, her “nether mouth.”

The fact that these divine principles were transacting in something more concrete than graphemes and phonemes is made abundantly clear even in these scholasticist, semanticizing sources. On the basis of terminology alone, we can see that the conceptual matrix is sexual. The absolute flashes forth, in phosphorescent effulgence (sphurattā; ullāsa). It expands as a phosphorescent wave, a welling, a swelling (sphurad-ūrmi)120 . . . thereby manifesting the cosmos made up of the thirty-six metaphysical categories (tattvas), from Śiva down to the element earth. . . . The Goddess is luminously conscious (prakāśāmarśana). . . . She is “throbbing incarnate” (spandarūpinī), being immersed in bliss (ānanda). . . . The cosmos is her manifest form, but, though shining as the “essence of divine loveplay” (divyakrīārasollāsa), the Absolute is pure undivided light and bliss.121

The subliminal sexual referents of this abstract image of the “Art of Love” were not entirely lost on the Śrīvidyā theoreticians. That they were aware of such is made clear from a debate that raged within the school concerning the relative legitimacy of conventional (samaya) meditation on the kāmakalā as opposed to the Kaula form of the same. It was in this latter (and earlier) case that a maiden’s naked body was used as the meditational substrate.122 A number of Śrīvidyā commentators, led by the venerable seventeenth-century master Bhāskararāya, insisted on the literal use of this meditation support, together with the referents of the five M-words, all of which smacked of the Kaula practices.123 Finally, the names of “Our Lord and Lady of Love,” in addition to their associations evoked above, are also identified, in the pre-fourteenth-century Kālikā Purāa, with the pīha of Kāmākhyā, whose sexual associations are legion in Tantric traditions.124

Elsewhere the worship of the sixteen Nityā goddesses who constitute the Goddess’s retinue, and which Śrīvidyā tradition identifies with the sixteen lunar tithis,125 includes offerings of meat and alcohol. It is especially the names of these sixteen Nityā goddesses that constitute the most obvious bridge between this and the earlier Kaula version of the same, given that these sixteen names are identical to those of the sixteen kalā-śaktis of the Śilpa Prakāśa.126 In Śrīvidyā sources these sixteen form the immediate entourage (āvaraa) of the Goddess, to whom sacrifices are to be offered, either in the central triangle or between the sixteen-petaled lotus and the square of the Śrīcakra. In other words, they occupy the same place in these sources as the sculpted images of the “eightfold practice of the kāmakalā” occupied on the early Tantric temples of Orissa.127 The sole variation between the two lists lies in the name of the first Śakti: she is Kāmeśvarī in Śrīvidyā sources and Kāmeśī in the Kaula diagram.128

6. Mantric Decoding and Kāmatattva in the Later Trika

It was the Kashmiri theoreticians, specifically Abhinavagupta and his disciple Kemarāja, who were most responsible for the semanticization of Kaula ritual into a form acceptable to the Hindu “mainstream,” to married householders, seekers of liberation, for whom the antinomian practices of the former were untenable. Here, in the socioreligious context of eleventh-century Kashmir, these reformers of the Trika sought to win the hearts and minds of a conformist populace by presenting a cleansed version of Kaula theory and practice, while continuing to observe the original Kaula rites in secret, among the initiated virtuosi. This trend of the progressive refinement of antinomian practice into a gnoseological system grounded in the aesthetics of vision and audition culminates in the Śrīvidyā tradition. Quite significantly, it is the image of a drop (bindu) that recurs, across the entire gamut of Tantric theory and practice, as the form that encapsulates the being, energy, and pure consciousness of the divine; and so it is that we encounter a multiplicity of references to drops of fluid, drops of light, drops of sound, and drops of gnosis. The language of phonemes and photemes, of mantras and yantras, make it possible for practitioners of high Hindu Tantra to discuss, in abstract, asexual terms, what was and remains, at bottom, a sexual body of practice. Through it, particles of love become transformed into particles of speech.

This is the explicit teaching of the (twelfth-century?) Vātūlanātha Sūtra (VNS) and its commentary by the sixteenth-century Anantaśaktipāda,129 according to which the mystic is effortlessly initiated, without the aid of external gurus or masters, by his own divinized powers of cognition, called “Yoginīs.”130 In the sixth verse to his commentarial introduction, Anantaśaktipāda evokes the “sūtras emitted from the mouth of the Yoginī,” and, in fact, each of the aphoristic teachings of this text is, according to him, presented by an internal Yoginī. It is in this way that that the overtly sexual language of the fifth sūtra, “From the sexual union of the Siddha and the Yoginī the great mingling (mahāmelāpa) arises,”131 is entirely sublimated and semanticized by Anantaśaktipāda:

The expression “Siddha-Yoginī” designates those who are Yoginīs and Siddhas, that is, the divinities of the senses and the objects of the senses. Their close contact is the “sexual union” of the two: the coming together of object [what is grasped] and subject [grasper]; or, again, their mutual and perfect embrace. By virtue of this embrace, an uninterrupted “great mystic union” (mahāmelāpa) occurs; that is, a sudden awakening or fluid equilibrium (mahāsāmarasya) which takes place constantly and everywhere in the ether of transcendent consciousness, when the duality of subjectivity and objectivity has melted away.

Here, the ritualized and sexualized Kaula “minglings” (melāpas) of flesh-and-blood Yoginīs and Siddhas that once took place on isolated hilltops on new moon nights now occur at all times within the “heart” of the enlightened Tantric practitioner, where they form the entourage of Bhairava-as-pure-consciousness and are characterized by their “extremely subtle vibrational activity.”132 In the context of these semanticized renderings, it is mantras that render one’s practice effective, containing in their very sound structure a mystic gnosis that, in a gnoseological system, is liberating. In every Tantric tradition, mantras are phonematic embodiments of deities and their energies, such that to know the mantra, and to be able to pronounce and wield it correctly, becomes the sine qua non of Tantric practice.133 These mantras, nondiscursive agglomerations of syllables, are entirely meaningless to an outsider; yet knowledge of their arcane meaning and, perhaps more importantly, the very divine energies embedded in their phonematic configuration render them incalculably powerful in transforming the practitioner into a “second Śiva” and affording him unlimited power in the world.

It is for this reason that mantras are themselves a matter of great secrecy and thereby subject to a wide array of security measures in their use and transmission.134 First of all, a mantra will generally be pronounced silently or inwardly, so as to not fall upon the wrong ears. When it is transmitted orally, as in the case of the initiation of a disciple by his teacher, this process is called “ear-to-ear” transmission (karāt karopadeśena). There exists, however, a massive textual corpus (called mantraśāstra) devoted to the discussion of secret mantras, which, in order to maintain the secrecy of these powerful, sect-specific utterances, are only given in code.135 In these sources, mantric encoding and decoding can take a number of forms, including the embedding and “extraction” (uddhti) of a mantra136 from its concealment in the midst of a mass of mundane phonemes, through one or another sort of cryptogram,137 or through more simple strategies of writing the mantra in reverse order, interchanging the syllables of a line, substituting an occult term for a phoneme, et cetera. However, we find in the texts of mantraśāstra, as well as in commentaries on texts in which mantras are given in code, “skeleton keys” that explain how to construct the mantric cryptograms, sets of equivalents for decoding occult terms, and so on. Here again, we find that a strategy of secrecy—implied in the encrypting of mantras—is undermined, in this case, by written instructions for their decryption.

It is nonetheless essential to note here that in high Hindu Tantra the knowledge and manipulation of extremely complex mantras are, by simple virtue of the fact that they are utterances in the Sanskrit tongue, the privileged prerogative of the Indian literati, who are, nearly by definition, comprised of the brahmin caste.138 For this reason, the likelihood of their being decrypted and used by non-brahmins is minimal—and high Hindu Tantra has been, from the outset, a mainly brahmanic prerogative. Now, Paul Muller-Ortega has argued, quite cogently, that the concealment of mantras through encoding/encryption, followed by their “revelation” through decoding/decryption, is of a piece with the theology of high Hindu Tantra, which maintains that these are the two modes of being that characterize the godhead in its expansion and contraction, into and out of manifest creation.139 That is, the decrypting of the mantra is, in and of itself, a mystic experience, a powerful communication of the Tantric gnosis to the initiate.

In high Hindu Tantra, the acoustic practice of the mantra flows directly into, or is simultaneous with, the visual practice of the mandala. This we have already seen in the context of the kāmakalā diagram: the bindus are simultaneously mantric utterances and photic graphemes. The Goddess is said to have a “body composed of letters” (lipitanu), which renders the act of reading them an audiovisual voyage of sorts through her body. Another grapheme will aid us in moving from this discussion of mantric encoding and encryption to an earlier time in the history of Hindu Tantra, when secrecy seemed not to have been such a vital and vexing issue. This is the phoneme E, whose grapheme, in the Sanskrit alphabet, more or less has the form of a downturned triangle. Because of its form, E is considered to be the privileged grapheme of the Goddess, the site of creation and joy, because it is identified as the “mouth of the Yoginī.” As before, the term “mouth” here refers to the Goddess’s or Yoginī’s vulva, which is called a site of creation and joy and “beautiful with the fragrance of emission”140 because, in early Hindu and Buddhist Tantra, one was reborn, re-created through initiation, and was assured the joy of liberation through the nether mouth of the Yoginī.141

Now, it is true that the Goddess, as the source of all mantras, is described in the high Hindu Tantric sources as bhinnayoni,142 “she whose vulva is spread”—but the question then arises as to how a woman embodying the Goddess would have been able to transmit mantras, sound formulas, through her vulva. This depiction of the Goddess is in fact found in a discussion, by Abhinavagupta, of Mālinī, the goddess identified with the energy of intermediate speech (madhyamā vāc) in the form of the “Garland of Phonemes”: “And this [Little Mother], by banging together with the Mass of Sounds, becomes the Garland of Phonemes, whose vulva is spread.”143

The fluid dynamic of this complex is made explicit in Kubjikā traditions, which locate the Goddess’s yoni at the level of the End of the Twelve of the subtle body, impaled there upon a subtle Śiva liga that rises out of the cranial vault.144 This yoni is simultaneously a “womb of mantras” and the nexus of the energy of transmission of gnosis, in the form of the Goddess’s “command” (āā). As the source of mantras, the triangle is subdivided into fifty smaller triangles, nested inside of it, each of which contains a Sanskrit phoneme. “Each letter is worshiped as a Bhairava or a Siddha. Each one of them lives in his own compartment, which is itself “a yoni said to be ‘wet’ with the divine Command (āā) of the energy of the transmission that takes place through the union they enjoy with their female counterparts.”145

The acoustic kāmakalā (or kāmatattva), whose practice Abhinavagupta also describes is, once again, the visarga, comprised of two bindus, as found in the Śrīvidyā kāmakalā.146 “Therefore, the venerable Kulagahvara [‘Cave of the Clan’] states that ‘this visarga, which consists of the unvoiced [avyakta] ha particle [kalā],147 is known as the Essence of Desire [kāmatattva].” Still quoting from this lost source, he continues: “[It is] the unvoiced syllable which, lodged in the throat of a beautiful woman, [arises] in the form of an unintentional sound, without forethought or concentration [on her part]: entirely directing his mind there [to that sound, the practitioner] brings the entire world under his control.”148 Here, Abhinavagupta’s bridge, between external ritual (if not sexual) practice and internalized speech acts, is the sound a woman makes while enjoying sexual intercourse—a barely articulated “ha, ha, ha.”149 It is this particle of speech (kalā) that is the essence of desire or love: in other words, the “ha” sound of the visarga is the semanticization of sex in Abhinavagupta’s system. However, as in the case of Śrīvidyā, the “practice” of the kāmakalā is reduced to meditative concentration, this time upon a syllable. Ritual doing has been reduced, once again, to a nondiscursive form of knowing. However, the presence of a sexual signifier again orients us back in the direction of a Kaula substratum that involved ritual practices of a sexual order.

7. The Masculinization of Tantric Initiation

In chapters 3 and 4, I presented a wealth of data to argue that the “insanguination” of the male initiate by a Yoginī lay at the heart, if not the source, of Kaula initiation and ritual. At the same time, many of these rituals also brought a male actor into play in the person of the teacher or master (guru or ācārya),150 with the combined sexual emissions of the pair transforming the initiand from an undetermined biologically given paśu into a kulaputra, a son of the clan. As one moves forward in time, and out of the Kaula context and into more conventional forms of Tantra, the role of the Yoginī becomes increasingly eclipsed by that of the male master. In fact, this shift toward “guru-ism” is one of the most fundamental dynamics in the development of later Tantra. The male guru gives birth to a new member of the Tantric order by inseminating his novice with male sexual fluid, which is nothing other than the seed of the male Śiva himself.

This transmission is termed initiation by penetration (vedha[mayī] dīkā) in a number of contexts,151 with the next move—the total sublimation of the sexual drop (bindu) or seed (bīja) into a seed mantra (bīja-mantra)—occurring in nearly every high Hindu Tantric tradition. Śiva, the divine revealer of the Liga Purāa, states that “initially my eternal command (āā) arose out of my mouth.”152 Mark Dyczkowski links this statement to a description found in an early Kaula work, the Kularatnoddyota, in his discussion of the term āā, which is reproduced in chapter 3. In this latter text, the guru initiates the disciple by literally transmitting the “command” to him through the recitation of mantras, at the level of the āā cakra, the “Circle of Command.”153 In his account of vedhadīkā, Abhinavagupta states that the disciple should press himself against the master, who, in order to effect a perfect fusion (samarasībhavet), should be mouth to mouth and body to body with him.154

In fact, rituals of male-to-male transmission or initiation predate Śaivism and Kaula traditions by at least two millennia. They constitute the Vedic norm, as it were, as evidenced in the Atharva Veda (AV) statement that “the teacher, when he initiates his pupil, places him, like a fetus, inside of his body. And during the three nights [of the initiation], he carries him in his belly. . . .”155 The B hadārayaka Upaniad (1.5.17) describes the transmission (sampratti) of breath from a dying father to his son in similar terms: “When he dies to this world, he penetrates his son with his breaths. Through his son, he maintains a support in this world, and the divine and immortal breaths penetrate him.” Finally the Kauśītaki Upaniad (2.15) anticipates Abhinavagupta’s instructions for vedhadīkā by at least twelve hundred years: “When the father is at the point of dying, he calls his son [to him]. . . . [The] father lies down, dressed in new clothing. Once he has arrived [there], the son lies down upon him, touching [his father’s] sense organs with his [own] sense organs.”156

As Paul Mus argued over sixty years ago, the guiding principle of these ancient sources was “not that one inherits from one’s [father]; rather, one inherits one’s father.”157 This was not, however, the implicit or explicit model of initiation in the later Tantric traditions. Rather than being the extension of a preexisting brahmanic mode of male self-reproduction, this was rather a reversal and masculinization of the Kaula model of heterosexual reproduction. That is, the Tantric vedha[mayī] dīkā and other initiations and consecrations self-consciously removed the feminine from the reproductive process, usually by internalizing and semanticizing her as the guru’s śakti, the “mother of the phonemes” and “fire stick of the mantras” passively transmitted from the guru’s mouth into that of his disciple.158 So, for example, in his general introduction to Tantric initiation in the TĀ, Abhinavagupta quotes the Ratnamālā Tantra in stating that when the master places the mālinī (mantra) on the disciple’s head, it’s effect is so powerful that it makes him fall to the ground.159 Here as well, the Yoginī—however instrumentalized she may have been in the Kaula rites in which the silent discharge from her nether mouth transformed an initiate into a member of the clan—has now been semanticized out of existence. As I argued in chapter 3, advances in Indian medical knowledge were such that a woman’s contribution, in the form of her “female discharge,” to the conception of a fetus, was well known by the time of the emergence of the Kaula rites. This understanding of the biology of reproduction, so important to the development of the Yoginī Kaula, was therefore consciously censored and sublimated in the initiation rites of later high Hindu Tantra. This paradigm nonetheless persists in the Bengali traditions of the dīkā guru (master of initiation) and the śīkā guru (teaching master). In both Sahajiyā Vaiava and Bāul traditions, the former, who is a male transmitter of mantras, plays a secondary role to the latter, who is female and whose “teaching” is received through her sexual emissions.160

In these rituals and their mythological representation, the guru inseminates his disciple by spitting into his mouth,161 a masculinized alloform of transmitting a chew of betel between mouths.162 Curiously, this type of transmission also becomes transposed into Indian Sufi hagiographical literature from the time of the Delhi Sultanate:

Now there was one man who had that very day become a disciple of the Shaykh; he was called Jamāl-al-dīn Rāvat. The Shaykh told him to go forth and give an answer to the [unnamed] Jogī’s display of powers. When Jamāl-al-dīn hesitated to do so, the Shaykh called him up close to him and took some pān out of his own mouth and placed it with his own hand in Jamāl-al-dīn’s mouth. As Jamāl-al-dīn ate the pān he was overcome by a strange exaltation and he bravely set out for the battle. He went to the Jogī. . . . When the Jogī had exhausted all his tricks he then said, “Take me to the Shaykh! I will become a believer.” . . . At the same time all the disciples of the Jogī became Muslims and made a bonfire of their religious books.163

Another tradition that blends Sufi and Tantric imagery, if not practice, is that of the Nizarpanthis of western India, whose “way of the basin” ritual was described in chapter 3.164 Here, it will be recalled that the term pāyal is used by Nizarpanthis for the mixture of sperm and chūrma that all participants consume at the end of this ritual. This terminology appears to have been inspired by Sufi traditions, in which “taking the cup” (piyālā lenā), that is, sharing a drink of milk with the master, is a transformative moment in initiation rites. However, here as well, the “milk” in the cup may have originally been the semen of the initiating pīr, diluted in water.165

8. Prescriptive Dreams and Visions

Gananath Obeyesekere and, more recently, Isabelle Nabokov166 have provided compelling analyses of the relationship between individual dreams and visions, on the one hand, and the cultural norms for the interpretation of the same, on the other. Most often dreams and visions involve possession by a demonic being, which can only be expelled by a narrativized interpretation (within the cultural idiom, in these cases, Tamil or Singalese) followed by a dramatic ritual exorcism either improvised by the dreamer or choreographed by an independent ritual specialist, a cāmi. These cāmis, generally low-caste individuals, do not choose to practice their charismatic calling. Rather, they have been “recruited” by a deity, usually a goddess; that is, they too have been the objects of repeated possessions, invasions of their person, by a foreign being. It is their prior and ongoing possession experience that empowers them to diagnose and cure other similarly possessed persons.167 In this role, cāmis are similar—if not identical—to tāntrikas. As Michel Strickmann has observed, the Tantric mantra master168 is a person who is enveloped by his dreams and visions, with the goal of Tantric ritual being to generate a sustained state of (wakeful) dreaming, such that a Tantric ritual, when properly performed, constitutes a “perfect dream.”169 Here, we may speak of what Sigmund Freud termed the “dream work,”170 the tāntrika’s ritual processing of the terrible nightmare demonesses that possess his own mind and body that permits him to see in them the one nurturing Mother goddess (or her male consort) that grounds their being as well as his own. This dream work, which is part and parcel of Kaula and Tantric initiation, continues throughout the practitioner’s career.

But dreams—or rather what one makes of dreams once one has “awakened,” as dictated by one’s cultural idioms—vary widely in their content and psychological impact. Here, a comparison between the prescriptive and transformative dreams and visions of two Tantric traditions—one Kaula and the other Śaivasiddhānta—will offer us a window onto the ways in which high Hindu Tantra sublimated the transformative sexuality of Kaula rites that were, as Obeyesekere’s and Nabokov’s research and analysis show, commensurate with real-life experiences of many South Asians. Here, our proof texts will be the circa ninth- to tenth-century Brahmayāmala (BY)171 and the 1095 C.E. Somaśambhupaddhati (SŚP).172

The BY passage describes a series of ritually induced dreams or visions that noncelibate practitioners are to undertake to succeed in their practice and attain a number of supernatural enjoyments, including the power of flight. Three rituals are presented, whose increasing complexity is accompanied by assurances of enhanced results; here I will only present the third and most elaborate of the series.173 As in the two envisioning rituals that precede this one, the practitioner seeks to know his past lives—in order to discern whether acts committed in those lives may be creating impediments to the success of his Kaula practice in his present life—by meditating on the vulvas of a circle of ritual consorts. These rites therefore involve the “viewing of his own yoni,” in which the yoni in question is at once the vulva that is his meditation support and the womb of the prior existence he is seeking to know.

“Concerning the ‘viewing of his own yoni,’ listen now to the following [practice on the part] of the practitioner. . . . It is the bestower of fruits: pure and impure nectar, [and] the [eight supernatural powers] beginning with atomicity.” Here, the practitioner is instructed to take a group of eight women, “the Śaktis, etc., those [women] who are devoted to the [one] Śakti, who have authority over the pure stream, and who are likewise without shame and without aversion.” Next, he is to prepare an underground chamber, equipped with a water circulation system, and a store of worship items, food (including 100,000 sweetmeats), drinking water, and a couch—but no fire: only an oil lamp is to be used to see in the dark.174 He then arranges the eight women around himself, at each of the cardinal and secondary directions. Then, he sexually arouses his Śaktis in succession,175 “effecting as many ‘rebirths’ [i.e., serial acts of sexual intercourse?] as his energy allows.” This day sets the pattern for the “sequential method” that the practitioner will observe for up to six months. Having aroused the eight Śaktis by day, the practitioner eats together with them by night. In this way, he “sees the vision of his own yoni, going back over eight births.”

At this point the text evokes a practice of the “Five Women of the Seal,” in the context of which the most vivid visions occur. The male practitioner, who is “sealed in” here by a circle of four or five women, now experiences the arising of

the great obstructors (mahāvighnāni), all of them very terrifying. . . . He should not be frightened either by these creatures . . . or when he sees a fearsome serpent that seems to be devouring [him]. He sees a she-cat with sharp teeth and a deformed body. Even seeing her, he ought not to be frightened, nor should he halt the ritual. He sees a very terrifying she-rat, with the body of an obstructor. Drawing toward (ākarya) herself the person who abandons his worship [out of fear], that Śakti . . . kills [him]. [A demoness] will say the words: “Stand up! I devour [you]!” . . . He is not to be frightened. . . . Voices will come from outside [the underground chamber]. [He will hear] the words “Kill! kill! Throw out food! Draw in [this] sinner!” . . . [and] “Get up, get up, you witless one! You are taken by the order of the king!” Seeing [these demonesses], he is not to fear, and his mind should not depart from its meditation. The practitioner [will see] dreadful gape-mouthed forms. [There will be] buck-toothed Śaktis licking [him] with their tongues. Seeing them, he should not fear. . . . Without a doubt, they lick the essence [that is] inside the practitioner for the sake of knowledge. While this is taking place, supernatural experiences [will] arise, one at a time. . . . His supernatural power, pervading the triple-world, will manifest itself. When the full six months [have passed], there is the visible manifestation of the [great] Goddess. . . . [Even when she appears] with her gape-mouthed form, she should not be feared by the possessor of mantras. . . . [The] completion [of the practice] is to be carried out by the practitioner in [the midst of the circle of] the eight [women]. In the [circle of] seven, nothing more than the viewing of the yoni occurs. In the group of eight, there is, without a doubt, the daily arising of [supermundane] wisdom. Having attracted the bodies of every one of these beings, he thereby obtains that [wisdom]. He becomes a Virile Hero, surrounded by yonis.

At the conclusion of this heroic practice, the Tantric practitioner of this rite becomes a Virile Hero, standing alone, like the supreme male godhead itself, at the center of a mandala of feminine entities he now controls. This he has succeeded in doing through his “dream work,” through his ability to maintain his sangfroid when assaulted by hordes of demonesses, howling beasts with sharp teeth and long tongues seeking to drain him of his vital essence both from without and within. This is the modus operandi of the modern-day tāntrika, a visionary who induces possession by—or the vision of—a divinity, through a series of “spiritual exercises” by means of which lower beings, usually demonesses, are driven out or brought under control by the higher god, with whom he identifies. In fact, the configuration of the male practitioner at the center of a circle of eight females exactly reproduces a Kaula representation of the “clan transmission” (kulāmnāya) as described in the Netra Tantra and other sources, in which the male deity Bhairava is enthroned in the heart of a lotus, on the eight petals of which are seated eight Mother goddesses.176 As with the popular demonological traditions of South Asia, it is the family that is at the center of one’s dream-or trance-induced experience; what has changed here is that the family or clan is now comprised of superhuman families of goddesses whose powers sustain and energize the entire universe.

The specificity of this Kaula tradition further lies in its privileging of the power (and the male conquest) of female sexuality, represented by eight Śaktis in the Brahmayāmala passage. Eight is of course the “clan number”: through the kulayāga and other initiation rituals, Kaula practitioners were reborn into families of the eight Mothers, which proliferated into the broader clans of the sixty-four Yoginīs. In both the early Buddhist and Hindu Tantras, one finds ritual instructions for entering into the cosmic body of the divine clan via the powers of one of the eight Mothers, in initiation rites that involved possession (āveśa). In the controlled environment of the ritual, the danger of possession by these demonic female entities was both reduced and voluntarily induced by the heroic male practitioner: rather than being their passive victim, he actively transformed them into his instrument for the attainment of supernatural enjoyments.177 Through these rites, the eight Mothers became internalized, making the (male) body a temple in which to worship these powerful female entities. Now linked to the eight parts of the practitioner’s subtle body, the Mothers were seen as grounded in and projections of this new center, from which they were emitted and into which they were reabsorbed.178 The notion of “family” is at once retained and expanded here: the threatening female nightmare horde, now identified with the vivifying powers of a cosmic family (kula), is sublimated into the mind-body complex of the male practitioner. Yet the initiatory role of the Yoginī never fully disappears in the Clan traditions, as in the case of the “Great Feast” (mahotsava) of the Siddhas and Yoginīs, at which “only men and women initiated by a Yoginī . . . in a dream are invited. . . .”179

At the opposite end of the spectrum from the Kaula are those high Hindu Tantric traditions in which the popular goddess-based demonological traditions have been fully occulted by a direct male (god)–to–male (practitioner) initiation process and transmission of the Tantric gnosis. A case in point is the metaphysician Vāsugupta (circa 825–875 C.E.), to whom the supreme male deity Śiva revealed himself, in a dream, in the form of a text that became known as the Śivasūtra, the “Aphorisms of Śiva.”180 This sublimation of the feminine becomes the rule in elite brah-manic forms of Hindu sectarian practice, in which a technique of “divine envisioning” (divyadi) is employed to identify directly with the supreme male godhead.181

The Somaśambhupaddhati’s description of the highest Śaivasiddhānta initiation, called nirvāa-dīkā,182 epitomizes this censoring of the female from high Hindu Tantric practice. The passage opens with an instruction to the Tantric guru to “install . . . his disciples [for the night] . . . the renouncers lying on a bed of pure ashes, their heads to the south, their topknots knotted with the śikha [mantra], [and] protected by the astra [mantra]. After having prayed to the God of Dreams, he leaves them.” Here, 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 or the consciousness of the initiands. This is made clear in a number of Āgamas, which, in their discussions of this initiation rite, place great emphasis on the means of protection (rakitān) of the sleeping initiands, comparing the protective mantras to a citadel with ramparts. Two such sources mention blood offerings (bali) and blood offerings to the “beings” (bhūtabali),183 clear references to the demonological underpinnings of this ritual.

The SŚP continues its exposition by stating that the aim of this initiation rite is to stimulate dreams in the initiands through the teaching of the mantra of the “Little Dream Man” (svapnamāavaka), the Indian equivalent of the Sandman, who is here identified with Śiva.184 This mantra, given in the verses that follow, concludes with this prayer to Śiva: “O God of Gods! I beg of you to reveal to me, here in my sleep, all the acts [hidden] in my heart!”185 As Hélène Brunner notes, this prayer is made to the God of Dreams to know whether there remain any karmic traces that may bar their initiation on the following day. The god’s reply will come in the form of the intervening night’s dreams, which must of course be interpreted. The Siddhāntaśekhara states the matter aphoristically: “The guru invokes the dream mantra in order that they [the disciples] have visions in their sleep.” This is the essence of the ritual: the disciples must dream!186

The extremely tame list of dreams that follows in the SŚP, dreams whose karmic content the guru is called upon to interpret and to remedy through ritual expiations (prāyaścittas), indicates that we are, in this Śaivasiddhānta initiation, worlds away from the nightmare scenarios of the Kaula envisionings of “one’s own yoni.” 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.187—precisely because the sleeping initiands have been protected from demonic invasions of their bodies and minds by a “citadel of mantras.” And such is precisely the situation in these elite traditions: no fearsome female entity troubles the sleep and dreams of the male practitioners, whose ritual practice is quite entirely directed toward the supreme male deity Śiva and his male entourage. The “dream work” has been elided here, with the total sublimation, or exclusion, of the terrifying feminine, in favor of an unthreatening male “Sandman.” Yet these troublesome beings retain an occult presence in these elite rituals, the elaborate protective measures being taken against them constituting ample evidence of their invasive power.

Despite their differences in emphasis, we can see that these two Tantric traditions share three common approaches to dreams and visions. First, both require that practitioners undergo a dream experience as part of their spiritual itinerary: for a number of reasons, they must dream. Second, both traditions work from the same worldview that places multitudes of dead and demonic beings at the periphery of a hierarchized mandala governed by a supreme deity, either male or female. Third, both give full value to the possession experience: for in high Hindu Tantra traditions, the ultimate end of the vision practice is samāveśa, the “co-penetration” or “mutual possession” of two beings—the human practitioner and the divine—until they are realized to be one. However, whereas in the earlier Kaula traditions, success is only attained through transactions with terrifying female beings, the later Śaivasiddhānta source examined here literally walls these out from the practitioner’s consciousness while denying the importance of the subconscious to his mind-body composite. The Yoginīs have no place in this all-male universe.

9. Sublimation of the Five M-words

We have shown that high Hindu Tantra defeminized Kaula practice, either internalizing the feminine or simply removing women from its sublimated ritual program. If indeed, maithuna, the fifth makāra, has been excised from these sources, what of the referents of the other four M-words, the proscribed (for the purity-conscious upper castes) substances that also gave Kaula practice its specificity? As will be shown, these substances too, and the pollution and power they represent, are dispensed with through ritual substitutes and casuistic argumentation. In the end, there is nothing left of the Kaula legacy in high Hindu Tantra, which becomes indistinguishable from other forms of orthodox householder religious practice.

We begin by looking at the transgressive language of the Kaula sources themselves. When the KJñN enjoins the Kaula practitioner to eat “the flesh, clarified butter, blood, milk, and yogurt of a cow,” in what appears to be a conflation of an orthodox abomination (bovicide) with an orthodox requirement (three of the pañcagavya, the five pure products of the cow), we find ourselves in the presence of a deliberate subversion of orthodox canons and categories of purity. A few verses later, the same source recommends the ritual consumption of the flesh of the jackal, dog, and other impure animals.188 The Vīrāvalī Tantra, quoted by Jayaratha in his commentary to TĀ 4.243, undermines orthodox categories through a historical argument, stating that “the ancient is ate both beef and human flesh”;189 and Jayaratha quotes an unnamed source in his commentary on TĀ 29.10 by declaring, “Tantric adepts who are Virile Heroes should eat that which the common man detests and, being revolting, is censorable and prohibited by the scriptures.” The Bhat Sanyāsa Upaniad, not a Tantric text, simply reverses the poles of licit and illicit when it maintains that for the renouncer, “clarified butter is like dog’s urine, and honey is equal to liquor. Oil is pig’s urine and broth is equal to garlic. Bean cake and the like are cow’s meat, and milk is equal to urine. With every effort, therefore, an ascetic should avoid clarified butter and the like. . . .”190

Of course, the Kaula traditions were less concerned with shocking the conventional sensibilities of the wider South Asian society than they were with the transformative effects, for the uninitiated, of eating the final member of the series, the clan fluid. We should also recall here that five M-words do not belong to the original Kaula traditions. One instead finds other sets of five—the Five Essences (pañcatattva),191 Five Nectars, Five Lamps (in Kālacakra Buddhism),192 or Five Jewels (urine, semen, menstrual blood, feces, and phlegm). A set comprised of these five, combined with seven other prohibited foods, is termed the “Set of Twelve Fluids” (dravyadvādaśaka).193 Alternately, it is an earlier set of three M-words, comprised of liquor, meat, and sexual fluids, that one finds in the TĀ and other Trika sources.194 When, however, one examines treatments of the five M-words in works belonging to the later Tantric revivals, they are most often presented in order to be equated with ritual substitutes.

Even when the language is subversive in these later sources, the intention is casuistic. A prime example is a rhetorical sally found in the fifteenth-century Hahayogapradīpikā of Svātmarāman: “[The practitioner] should always eat the flesh of a cow (gomāsa) and drink strong liquor (amara-vāruī). Him I consider to be well-born, whereas those who do otherwise are the ruin of their families.” Svātmarāman, however, provides his own gloss on this verse: “By the word go (cow), the tongue is intended. Its entry into [a cavity in] the palate is gomāsa-bhakaa (‘eating the flesh of a cow’). This indeed destroys the [five] great sins.”195 In the later Tantric works, one finds for each of the five M-words a ritual substitute (pratinidhi). The Āgamasāra196 declares that madya, liquor, refers to the nectar internally drunk in the cranial vault at the culmination of yogic practice; that māsa, flesh, refers to the practitioner’s tongue in the yogic technique of khecarī mudrā; that matsya, fish, refers to the absorption into the medial channel of the breaths moving in the right and left channels of the subtle body (these breaths, styled as two fish swimming in the Ganges and Yamuna Rivers, are to be swallowed into the central suumā channel, the Sarasvati River); that mudrā refers to the dawning of inner knowledge in the sahasrāra cakra, located in the cranial vault; and that maithuna, sexual union, refers to the supreme essence (paramatattva), from which siddhis and the knowledge of the absolute arise.197 A similar yogic interpretation is found in the Kulārava Tantra,198 which, in spite of the rhetorical glorification of the Kaula in its opening chapter,199 shows itself to be an altogether conventionalist work in its fourth chapter when it provides long lists of such equivalents and condemns the consumption of the referents of the original five M-words.

Casuistry gives rise to countercasuistry, as in the case of the ānārava,200 a pre-sixteenth-century work that quite nearly reverses all of the Āgamasāra’s “soft” readings of the five M-words, no doubt to make a point concerning the doctrine of radical nonduality (advaya)—that is, that there is no difference between the absolute and the lowest forms of manifest being—as held by many of the Kaula schools:201

From the perspective of one who has a consummate knowledge of dharma and adharma, there is purity even in the things of this world. . . . The eating of cow dung and urine . . . is prescribed as an expiation for such sins as the murder of a brahmin, so what stain can there be in [human] excrement and urine . . . ? It is by means of menstrual blood that a body is in fact generated. How then can that by means of which liberation is realized be polluting (duaa)? . . . Semen, as the root cause of the body, is assuredly pure. How is it that [men] revile it?202

But countercasuistry can, in turn, give rise to counter-countercasuistry. The KĀN repeats the verse quoted above from the ānārava on the subject of menstrual blood and semen.203 Yet, after having effected a reductio ad absurdum of every sort of brahmanic purity regulation, this same source then goes on to catalog the numerous ritual substitutes that orthodox brahmin practitioners were permitted to resort to in their practice.204

This spirit of scrupulousness and attention to purity regulations, inspired no doubt by the real fear that contact with dangerous fluids would destroy one’s very being, flies in the face of the fundamental Kaula and Tantric ideologies. More than this, they epitomize the “contracted consciousness” that is the antitype of the expanded god-consciousness of the Tantric practitioner, and they bar the path to the powers and supernatural enjoyments that give Tantra its specificity. The inhibitions, or sorts of dread or fear (śakā) by which the consciousness of the orthodox brahmin becomes contracted, have been enumerated by Jayaratha, in his TĀ commentary on the kulayāga. They are dread of loss of identity, dread of participation in non-Vedic rites, dread of impure Tantric incantations, dread of fluids (dravyaśakā), dread of contamination by untouchables in caste-promiscuous sexual rites, dread of entering the cremation grounds and the other impure sites in which the Tantric rites are observed, dread of assault or possession by the demonic beings that inhabit these sites (bhūtaśakā) and dread of the human body (śarīraśakā), and, finally, dread of [non-brahmanic] categories (tattvaśakā).205 Finally, in a total reversal of standard notions of demonic possession, Abhinavagupta, citing the Kulagahvara Tantra and the Niśisacāra Tantra, dismisses such orthodox obsessions with conventionalist categories as so many Seizers (Grahas), “because they conceal the true self (autonomous, unitary consciousness) beneath a phantasmagoric pseudo-identity, contaminating and impoverishing it with categories unrelated to its essence.” These “Eight Seizers,” which combine to create the limited persona of the orthodox householder practitioner, are conformist obsessions with birth, traditional knowledge, family (kula), discipline, the body, one’s country, conventional virtues, and wealth.206

In this exegetical synthesis, the ultimate referent of purity regulations, like that of sexual practice, becomes an exalted state of consciousness. Once again, doing becomes subsumed under knowing. Thus, while there remains a place in the secret initiations of the Abhinavan corpus for the consumption of prohibited foods and sexual fluids, the goal of such practice has now become a breakthrough of consciousness rather than the transformative effects of these substances themselves. Kaula practice and its effects are now explained in terms of the value of transcending the “psychosis” of conformity to the exoteric religion,207 as epitomized in the use of the Five Jewels, mentioned above.

The hesitation which prevents the majority from accepting the validity of the Kaula and Tantric revelation becomes identical in this perspective with the contraction which consciousness takes on when it projects itself as bound individuals and their world. . . . Such is the power attributed to this contact with impurity that it is believed that it may take the place of the conventional process of initiation (dīkā) into the Kaula cult. Instead of that ritual the Kaula officiant may simply present the candidate with a skull-cup containing wine and the [Five] Jewels or other such substances. If he swallows the contents without hesitation (śakā) he is considered to have attained direct realization of consciousness in its essential nature uncontaminated by conceptual or ethical dualities (vikalpa). Termed the “consumption of the oblation” (caruprāśanam) this act is listed accordingly among the contexts in which enlightenment may occur without recourse to meditation, ritual or any other means of liberation.208

Even as, however, Abhinavagupta allows for the ingestion of the dravyams of the original Kaula rituals in his high Hindu Tantric synthesis, the referent of the practice has been entirely displaced, from the power inherent in the clan fluids themselves—to transform a biologically given “inert being” (paśu) into a Virile Hero or Perfected Being—to the transformative psychological effect of overcoming conventional notions of propriety through the consumption of polluting substances. This emphasis on aesthetic experience and gnoseological transformation, coupled with a system of equivalencies between Vedic and Tantric ritual, could not help but have a leveling effect on all later forms of Tantra, whether of the more Kaula or Śākta “left” or the more Śaiva “right.”209 In the end and regardless of Tantric theory, the impetus behind the Tantric rituals became one of achieving parity with the Vedic rituals, with the “need to match these orthodox rituals . . . strong enough to compromise the very beliefs which justified the separate existence of the Tantric system.”210 This is precisely what has happened over the centuries. Heterodox Kaula or Tantric ritual has so shaded into orthodox Vedic or Śaivasiddhānta ritual as to become indistinguishable from it, as in the case of South Indian Smārta communities. The various waves of “Tantric revival” have only further clouded the picture. Everything becomes Tantra, because nothing is Tantra. In the late twentieth century, New Age Tantra has rushed in to fill the vacuum.

But this is not where the story ends, nor is it where it began.