Vajrayoginī - Selections
CHAPTER 1: VAJRAYOGINĪ AND BUDDHIST TANTRAS
The cult of tantric goddess, Vajrayoginī, flowered in India between the tenth and twelfth centuries C.E. at a mature phase of the Buddhist tantras. One of the most important sources for her practice in India is a collection of sādhanas. A sādhana is a meditation and ritual text—literally, a “means of attainment” (sādhanam)—that centers upon a chosen deity, in this case, upon Vajrayoginī or one of her various manifestations. This particular collection was written and preserved in Sanskrit and drawn together under the late, collective title, the Guhyasamayasādhanamālā (GSS). It is one of these sādhanas that is edited and translated in this book, and that serves as the basis for our exploration of the goddess, particularly in her form as Vajravārāhī.
Who is Vajrayoginī? The texts refer to her reverentially as a “blessed one” (bhagavatī), as a “deity” (devatā) or “goddess” (devī). She is divine in the sense that she embodies enlightenment; and as she is worshiped at the center of a maṇḍala of other enlightened beings, the supreme focus of devotion, she has the status of a buddha. In the opening verse to the Vajravārāhī Sādhana, the author salutes her as a vajradevī, that is, as a Vajrayāna or tantric Buddhist (vajra) goddess, and in the final verse prays that all beings may become enlightened like her, that is, that they may attain “the state of the glorious vajra goddess” (śrīvajradevīpadavī).
The Buddhist Tantric Systems
Tantric Buddhism is the wing of the Mahāyāna that revolves around mantra as a path or “way,” and that is known therefore as the Mantrayāna or Mantranaya, or as the Vajrayāna after one of its primary symbols, the vajra. A pithy definition of tantra is elusive. Vaiṣṇavism, Śaivism, and other Indian religions including Buddhism all developed rich tantric traditions, and the term broadly denotes particular types of ritual employed within their various deity cults. “Tantra” also refers to the various bodies of literature within these traditions: scriptural and exegetical texts that provide instructions for attainments, both spiritual and mundane. One gains an idea of the size of the Buddhist tantric tradition alone when one considers that it evolved in India for a thousand years (from about the second century C.E.), and that this process has continued in Tibet and beyond for another thousand. The main production of tantric texts occurred in India between about the third and twelfth centuries. Some indication of the numbers involved can be gleaned from the sheer quantity of works translated from Sanskrit into Tibetan from the end of this period. The tantric portion of the Tibetan canon contains almost five hundred tantric scriptures and over three thousand commentarial texts; Isaacson (2001: personal communication) suggests there may exist as many as three thousand Buddhist tantric texts in Sanskrit, of which over a quarter—perhaps many more—have not been translated into Tibetan or any other language. In order to locate Vajrayoginī and her cult within this vast spiritual corpus, it is worth beginning with a brief summary of Buddhist tantric literature. But with so many texts to consider, and with such an array of practices and methods revealed within them, where is one to begin? The problem of how to classify and codify the material has occupied scholars from at least the eighth century and does so even today as contemporary scholars continue to propose new ways of approaching and organizing the materials (e.g., Linrothe 1999). The result is that there are various systems for categorizing the Buddhist tantras that are by no means standard, and how these different classes of texts arose, or came to be known, is something of a mystery.
It seems that one of the earliest classifications of the Buddhist tantras occurred in the eighth century by Buddhaguhya, who recognized only two classes, kriyātantras and yogatantras (Mimaki 1994: 122, n. 17). The subject-matter of some tantras, however, was neither principally kriyā (kriyāpradhāna), nor principally yoga (yogapradhāna), but seemed to combine “both” (ubhaya); these were termed ubhayatantras, and later, caryātantras (Isaacson 1998). It is this threefold classification—kriyā-, caryā-, and yogatantras—to which an eighth-century scholar/practitioner, Vilāsavajra, confidently refers. Of these classes, the earliest tantric texts are found within the kriyātantras (“action tantras”), which appeared between at least the third century, when they are known to have been translated into Chinese (Hodge 1994: 74–76), and at least the sixth century. The so-called caryātantras (“performance tantras”) were current from at least the mid seventh with the emergence of its root text, the Vairocanābhisaṃbodhi (ibid.: 65ff.) Despite their status as “tantras,” religious teachings supposedly revealed by the historical Buddha, these classes hold essentially ritual manuals and dhāraṇīs concerned with supernatural, desiderative attainments (siddhis), such as locating treasure, alchemy, flying, invisibility, forcing access to heavenly realms, warding off evils, and so on; they make little reference to soteriological goals. Sanderson (1994b: 97 n. 1) comments on the enduring popularity of the kriyāand caryātantras, even among translators of later soteriological tantras (such as Amoghavajra, d. 774), as well as their continuing importance in apotropaic rituals in Newar, Tibetan, and Japanese Buddhism. The fascination with siddhis of various types remains in later tantric literature, as the study of Vajrayoginī will show.
By distinguishing the kriyātantras (or the kriyā- and caryātantras) from the yogatantras, the eighth-century scholars were in fact pointing to the emergence of a new kind of tantra that had entered the Buddhist arena, probably from the late seventh century (Hodge op.cit.: 65–66, 58). The root text of the yogatantra is the Sarvatathāgatatattvasaṃgraha (STTS), and like the caryātantras, it centers on the supreme buddha, Vairocana. However, it reveals an important shift in emphasis. This is the first work in which tantric methodologies, such as rites of consecration, mantras, and maṇḍalas, were directly aligned to soteriological as well as to desiderative goals. The significance of bringing a liberationist slant to bear on tantric methods was not lost upon commentators, who were clearly aware of the need to bring traditional Buddhist values into the tantric field. Vilāsavajra, for example, wrote a commentary based on the Vajradhātumaṇḍala of the STTS, in which he set out “to encode and interpret tantric ritual in Mahāyānist doctrinal terms” (Tribe 1994: 4). Portions of yogatantra text are probably the oldest incorporated into the literature of Vajrayoginī.
Even within Vilāsavajra’s exegesis, however, there was other liberationist material that did not fit easily into the yogatantra category, a fact he seems to have recognized by designating his root text, the Nāmasaṃgīti, a “mahāyoga” or “great tantra” (Tribe 1997: 128, nn. 11, 18, and 20). Indeed, new kinds of texts with marked differences in subject matter were beginning to emerge, and these were soon to be contrasted with the yogatantras and given the new designation “yoginītantras.” Within the soteriological tantric realm these two terms—yogatantra and yoginītantra—seem to refer to the two main divisions of Buddhist tantras, and commentators frequently pair them together as the “yoga- and yoginītantras.” Thus, the commonest classification of tantric texts in India was probably fourfold: kriyā-, caryā-, yoga-, and yoginītantras (Isaacson 1998).
The yoginītantra class is characterized by the appearance of a new Buddha at the center of its maṇḍalas, namely Akṣobhya and his manifestations, supreme enlightened beings who belong to the vajra (“diamond” or “thunderbolt”) family of deities. These deities are wrathful in appearance with a startling affinity for places of death and impurity, the cremation grounds; they also manifest a vivid sexual symbolism. One of the key cults within this class is based on the tantric deity Hevajra and was probably emerging around or after the tenth century. In the Hevajratantra, Hevajra is seen to be a heruka form, that is, a type of wild enlightened being who dwells in cremation grounds with a retinue of cremation-ground deities and spirits. Other yoginītantra systems, probably roughly contemporary with the Hevajratantra, also center on this type of heruka deity: Cakrasaṃvara, Caṇḍamahāroṣaṇa, Buddhakapāla, Mahāmāyāhva, and Kālacakra are all heruka forms who appear as lords of their own maṇḍalas. Their appearance, accoutrements, and behavior all relate to practices that ascetics undertook while dwelling in cremation grounds. These are the kāpālika observances, or observances based on the skull (kapālaḥ, kapālam), chief tool and symbol for yogins of this kind. The heruka lord is also worshiped in embrace with his consort, while the retinue of male and female deities in his maṇḍala may also be in sexual union.
The principle of śakti begins to emerge in these texts as a potency manifesting in powerful female deities. It comes to the fore through the figure of the female consorts and the many types of goddesses, witches, or female spirits—yoginīs and ḍākinīs—who haunt the wilds and live in the cremation grounds. As śakti is increasingly emphasized, texts tend to redefine traditional Mahāyāna soteriology in the language of erotico-yogic techniques and mahāmudrā (p. 91). Thus, as one tantra explains: “The Mahāyāna is mahāmudrā, and yoginīs bring magical power.” It is these texts that form the direct basis for the cult of Vajrayoginī. Within the yoginītantras we see a growing preoccupation with the yoginī, or enlightened female deity. In some maṇḍalas she is worshiped as the chief deity within a predominantly female maṇḍala, even though she is still in embrace with a male partner (e.g., see ch. 2). Eventually, cults emerged in which the male consorts disappeared entirely from view, leaving the female deity to be worshiped alone at the center of a new maṇḍala. Often the form of the maṇḍala is preserved exactly as it was before, except that the male deities have simply been removed. This is typical of the maṇḍala s described in the sādhanas of the Guhyasamayasādhanamālā. Our study of the Vajravārāhī maṇḍala in Umāpatideva’s Vajravārāhī Sādhana will show that it is modeled exactly upon that of Cakrasaṃvara, except that in Vajravārāhī’s maṇḍala all the male gods of Cakrasaṃvara’s maṇḍala have disappeared, leaving the goddesses without consorts, and supreme.
Our summary so far of the tantric systems has shown the cult of Vajrayoginī to be firmly grounded within the yoginītantra class. But this classification is more complex than I have made out. On the one hand, there were already texts akin to the yoginītantras well before the maturing of the Heruka cults in the ninth and tenth centuries; the Sarvabuddhasamāyoga-ḍākinījālaśaṃvara is one such “proto-yoginītantra” that is known to have been in existence in the mid-eighth century (Sanderson 1995). Here, the lords of the maṇḍalas are heruka-type, esoteric deities, in sexual union with consorts and surrounded by retinues of female ḍākinīs. This tantra was still in use in Tibet in the eleventh century, “no doubt because of its evident kinship with the later yoginītantras” (ibid.). On the other hand, there were texts that sat uncomfortably within the yogatantra system, but that were not so markedly different that they fell naturally into the yoginītantra classification. This gave rise to another tantra class known as the “yogottara,” literally that which is “higher than the yoga [class].”
Isaacson (op.cit.) suggests the term “yogottaratantra” was a later designation. Certainly when Vilāsavajra refers to the Guhyasamājatantra, and to other texts that were later named as “yogottara,” such as the Vajrabhairavatantra and the Māyājālatantra, he seems to be unaware of any such class (Tribe 1994: 5). This stratum of tantric literature arose about a century after the yogatantras, and its root text, the Guhyasamājatantra, was codified and translated into Tibetan in the eighth century (Matsunaga 1972; Snellgrove 1987: 183). The introduction of this extra “yogottara” classification seems to reflect the fact that in the course of its evolution, the Guhyasamāja system (including its exegetical literature) came to be seen as sufficiently different from the older yogatantras—and certainly superior to it—to require a different label (Isaacson op.cit.). As in the yoginītantras, the maṇḍalas of the Guhyasamāja (or Samāja) tradition are presided over by Akṣobhya and by vajra-family deities, who are often both wrathful and erotic in character. Since the tantras of the yoginī class were deemed superior even to those of the yogottara, Isaacson suggests that they probably received the additional designation “yoganiruttaratantras,” literally: “tantras of the highest (niruttara) [division] of the yoga [class]” (translation by Sanderson 1994b: 98 n. 1).
Even this fivefold classification of kriyā-, caryā-, yoga-, yogottara-, and yoginītantras (the system almost ubiquitously expounded in our secondary literature) was not necessarily a widely accepted solution by scholars/practitioners of the day. Mimaki (1994) lists seven different classifications from various Indian exegetes and tantras, without even touching on the fourfold schema described above as possibly the most common (i.e., kriyā-, caryā-, yoga-, and yoginītantras). Atiśa, for example, writing in the early mid-eleventh century, sought to clarify works that strayed between the yoga and yogottara camps by inserting between them two more tantra classes—upāya- (“means”), and ubhaya- (“dual”)—thus presenting a new sevenfold classification of tantras.
In Tibet, the classification of texts likewise presents a complex picture (Mimaki 1994: 121). Among the gSar ma pa schools, there is the famous system of Bu ston (1290–1346), which preserves the divisions of the kriyā- (bya ba’i rgyud), caryā- (spyod pa’i rgyud), and yoga -(rnal ’byor gyi rgyud), but which classes those of the yogottara- and yoginītantras together as the anuttaratantra, or “ultimate tantra” (rnal ’byor bla na med pa’i rgyud). This fourth class is itself subdivided into father (phar gyud), mother (mar gyud), and nondual tantras (gnyis med rgyud). Mother tantras, or wisdom tantras (yeshes rgyud) are further analyzed into seven groups, one of which (itself with five subdivisions) comprises tantras connected with Heruka (Tsuda 1974: 28). The classification of the rNying ma tantric canon is based on a ninefold system of classification, in which such categories as mahāyoga (noted above) re-emerge as a distinct group (Germano 1994: 241–51 with n. 114, Williams and Tribe 2000: 203).
Complicated as the divisions and subdivisions of the tantric corpus are, they have been made more so by mistranslations in use in the West. Sanderson (1993) has pointed out that the term anuttarayogatantra found in some secondary sources does not occur in Sanskrit enumerations of the different classes of tantras and is likely to derive from an incorrect backformation from the Tibetan rnal ’byor bla med kyi rgyud or “yoganiruttaratantras.” (This refers to the class of Sanskrit works whose translations in the Tohoku catalogue are nos. 360–441, also termed rnal ’byor ma’i rgyud or “yoginītantra”; Sanderson 1994b: 98 n. 1). The term “yogānuttaratantras” (sometimes applied by secondary authors to yoganiruttaratantras) is also not attested in Sanskrit sources (Isaacson 2001: personal communication).
Within this vast and complex body of tantric literature, the practices of Vajrayoginī belong to the most developed phase of the yoginītantras. Vajrayoginī literature is unlike other systems within that class, however, in that it generally lacks its own tantras. It draws instead upon the scriptural texts of the Cakrasaṃvara cult: the Saṃvara-, or Śaṃvaratantras. Sanderson (1995) summarizes the Saṃvara corpus as follows:
The root text (mūlatantram) is the Laghuśaṃvaratantra, also called Herukābhidhāna- or Cakrasaṃvaratantra (BBK: 251). The text does not survive in its entirety; lost portions are accessible only through the early eleventh-century Tibetan translation, lemmata in tenth-century Sanskrit commentaries, and in secondary texts such as the Abhidhānottaratantra.
The Abhidhānottaratantra (BBK: 254). Its relationship with the Cakrasaṃvaratantra is that of explanatory tantra (*vyākhyātantram) to root text (mūlatantram), according to Buddhaguhya’s terminology.
Vajraḍākatantra (BBK: 255).
Saṃvarodayatantra (BBK: 256).
Ḍākārṇavatantra (BBK: 255).
Yoginīsaṃcāratantra (BBK: 258).
Herukābhyudaya (not surviving in Sanskrit).
Caturyoginīsaṃpuṭa (BBK: 259).
It is scriptures such as these—in particular, the Yoginīsaṃcāratantra, Saṃvarodayatantra, and Abhidhānottaratantra—that inform the sādhanas of the Guhyasamayasādhanamālā. One sādhana in the collection (GSS70) is based upon a unique Vajravārāhī scriptural source, the Vārāhyabhyudayatantra, itself apparently extracted from the Abhidhānottaratantra (Sanderson 2001a). In another, there is even a reference to the Lakṣābhidhāna (sometimes identified with the Khasamatantra), which is a mythical work, supposedly vast and authoritative in ten thousand verses, and allegedly the source from which the Cakrasaṃvaratantra itself was extracted (Tsuda 1974: 33). The same legendary authority is claimed in the Yoginīsaṃcāratantra following its description of the body maṇḍala, a core Cakrasaṃvara practice taken over with very little adaptation in Umāpatideva’s Vajravārāhī Sādhana.
The Vajrayoginī tradition does not simply graft itself onto the scriptural rootstock of Cakrasaṃvara; it borrows equally freely from the Cakrasaṃvara tradition of commentary and exegesis. We will see how the authors of the Guhyasamayasādhanamālā rely on the liturgical and commentarial texts at their disposal, and how they are able to adapt them for the worship of Vajrayoginī. This is most evident in the ritual portion of the sādhana, as described in chapter 3.
The Guhyasamayasādhanamālā and Its Authors
The most direct sources for our study of Vajrayoginī are the sādhanas of the Guhyasamayasādhanamālā (GSS). This is a group of some forty-six Sanskrit works drawn together as a collection centering upon Vajrayoginī and her manifestations. Fifteen of its works claim the authorship of named individuals, and it is to them that we now turn in order to gain some insight into the date of the compositions and the context in which they were written. Table 1 shows a list of our authors and the works attributed to them. Since in some cases an author’s influence upon an unattributed work may be inferred, authorship of almost half the texts in the collection can be firmly or loosely established (details are supplied in the appendix).
Establishing the dates of these authors is a thorny subject. I tentatively summarize the dates discussed here on the time chart (table 2). Various life histories survive, chiefly in Tibetan, although informed by a hagiographical and sometimes sectarian agenda (Tatz 1987: 696). Among key sources on this subject is the famous Legends of the Eighty-Four Mahāsiddhas (Grubthob brgyadcu rtsabzhi’i lorgyus, hereafter Legends), which supplies accounts of the lives of Indrabhūti, Lakṣmīṅkarā, Lūyīpāda, Śabara, and the slightly younger author Virūpa. More information on their lineages, and episodes from their lives, can be gleaned from the Blue Annals (Debther sNgonpo), written by ’Gos Lotsawa (Locchāwa) (1392–1481), and the History of Buddhism in India by Tāranātha (1575–?), but neither of these works can be relied upon for accurate dating. Scholars have often attempted to date authors according to the testimony of transmission lineages, a risky enterprise that Kvaerne describes as “methodological error” (1977: 6). Illustrative of the problem is Dowman’s attempt to date the mahāsiddhas using traditional Buddhist scholarship, according to which there are no fewer than three kings of Oḍḍiyāna called Indrabhūti (1985: 232ff; cf. Dudjom 1991: 441, 458–59, 485–87): Indrabhūti the Great, who may be as early as the seventh century (642 C.E. according to the Chinese T’ang Annals), an intermediate Indrabhūti, possibly of the eighth century (although apparently not recognized by Tāranātha, Dowman ibid.: n.233), and Indrabhūti the Younger, of the late ninth century. Davidson (2002), however, comments that even pinpointing three Indrabhūtis is “surely an underestimate” and points to “the tendency for traditional apologists and modern scholars to amalgamate the various personalities into one grand persona.” Dowman (op. cit.) also puts forward three possible candidates for Indrabhūti’s sister, Lakṣmīṅkarā, including a nun of similar name; however, even if we agree that this same Lakṣmīṅkarā is the author of our Lakṣmīsādhana (GSS24), the only certainty we can have is that she was no later than the Tibetan translator of the text, who was known to have lived 1059–1109. Virūpa, traditionally the pupil of Lakṣmīṅkarā (Blue Annals: 390), is just as elusive, and may have lived as early as the eighth century (Tāranātha History: 197) or as late as the eleventh century, when he supposedly taught Maitrīpāda (also called Advayavajra) and Mar pa the translator (Blue Annals: 390). Similar problems beset the dating of the Mahāsiddha Lūyīpāda. Kvaerne (1977: 5–6), for example, hesitantly cites Tāranātha (History: 311), according to whom “Lui” was a contemporary of Maitrī (Advayavajra) in the eleventh century, and notes that in one tradition, Lūyīpāda’s guru was Saraha, who may have flourished in the eleventh century or earlier (see also Dasgupta 1946: 6). Davidson (1991: n. 24) notes that Lūyīpāda’s Śrī-Bhagavadabhisamaya was translated into Tibetan in the first part of the eleventh century, “apparently the earliest attested practice of the Cakrasaṃvara” in the Tibetan canon. However, Sa skya legends assert that Lūyīpāda was a scribe at the court of Dharmapāla in the late eighth century (Dowman 1985: 37). The dating of Śabara is even more problematic. He appears as an early teacher in several genealogical traditions (Dowman ibid.: 65; Kvaerne 1977: 6), but also as a teacher to later authors such as Vanaratna in the fifteenth century. Dowman therefore posits a line of teachers called Śabara, the only merit of which is that it echoes the legend of Śabara’s immortality, according to which he would still be teaching today. Another of Śabara’s pupils is said to be Advayavajra, whose dates have been discussed at length by Tatz (1987: 697) and shown to be tied to the reign of King Neyapāla in the eleventh century (1007–85). Śabara also apparently initiated Vibhūticandra into the sixfold yoga system (ṣaḍaṅgayogaḥ) (Blue Annals: 727). Stearns (1996: 127–71) places Vibhūticandra in the later twelfth to early thirteenth centuries at the time of the Moslem invasions. Vibhūticandra would thus be the youngest author in our collection.
Some of the younger contributors to the Guhyasamayasādhanamālā are slightly easier to place because they admit their debt to earlier authors. One such is Śākyarakṣita, whose Flower Cluster of Clear Understanding (Abhisamayamañjarī GSS5) draws heavily on the Clear Understanding of Heruka (Herukābhisamaya) by Lūyīpāda. Apart from the similarity of his title, Śākyarakṣita refers twice to Lūyīpāda’s work, commenting on Lūyīpāda’s method for establishing the vajra ground (vajrabhūmi, K20v2) and knowledge circle (jñānacakra, K21v6), and referring to it for an in-depth treatment of Vajravārāhī’s thirty-seven-deity maṇḍala. Śākyarakṣita adds that this was taught “by my teacher in the Vajrāvalī,” which reveals that his guru was Abhayākaragupta, abbot of the monastic university Vikramaśīla during the reign of King Rāmapāla (c. 1084–1126/1130). If Śākyarakṣita was a younger contemporary of Abhayākaragupta, he would probably have flourished in the mid-twelfth century.
Our study of Umāpatideva’s Vajravārāhī Sādhana (GSS11) will show that it shares much in common with Śākyarakṣita’s work, in both its subject matter and use of sources. Fortunately, Umāpatideva’s lineage and dates are on slightly firmer ground, and these place him in the same generation as Śākyarakṣita, perhaps as an older contemporary. The colophon to the Tibetan translation describes him as “one who has the lineage of the instructions of Virūpa, śrī Umāpatidatta” (Tib 49.7), and the dates of the translators link him fairly securely to the same period as Abhayākaragupta. The translators of Umāpatideva’s two known texts in the bsTan-’gyur are Vāgīśvaragupta and Rwa Chos rab. Rwa Chos rab was active in India and Nepal in at least the first quarter of the twelfth century, and was a pupil of the Nepalese paṇḍit Samantaśrī; Samantaśrī himself flourished in the early to mid–twelfth century and received the Kālacakra teachings from Abhayākaragupta (Blue Annals: 760–61; cf. ibid.: 756, 789). Thus, the translation of Umāpatideva’s works would seem to belong to the early to mid-twelfth century, and may even have been contemporary with the author. If Umāpatideva was of the same generation as Samantaśrī (whom he is unlikely to have postdated, since his translator was a pupil of the latter), he may also have been a pupil of Abhayākaragupta’s.
In the absence of much reliable evidence for dating the authors of the Guhyasamayasādhanamālā, we must look for other clues as to their origins. First, it seems that several authors in the Guhyasamayasādhanamālā are associated with the early dissemination of tantric lineages. Indrabhūti, for example, is traditionally known as “the first tāntrika ” and was credited with initiating several tantric lineages, including the yogottara, Hevajra, and Cakrasaṃvara traditions (Blue Annals: 869; Dudjom 1991: 485, 462; Dowman 1985: 233; SM vol. 2: xxxi). Lūyīpāda is particularly associated with the Cakrasaṃvara system, on the basis of which he is traditionally known as the “original guru” (ādiguru) of the mahāmudrā (Dowman 1985: 37). According to the Tibetan tradition, he is one of three main transmitters of the Cakrasaṃvara system along with Ghaṇṭāpāda and Kṛṣṇapāda (Blue Annals: 389; Dawa-Samdup 1919: 9; Jackson 1994: 125). Śabara is also associated with the spread of mahāmudrā, according to the evidence of the Guhyasamayasādhanamālā (see chapter 2), and is an important transmitter of the Ṣaḍaṅgayoga discipline (Padma gar dbang, cited Stearns 1996: 140). One tradition putatively connects Śabara with the origins of the Trikāya-vajrayoginī tradition through his lineal descendant Kṛṣṇācārya (Dowman 1985: 320; 7.19), although Benard (1994: 12–13) prefers to credit Lakṣmīṅkarā. In Tibet, Virūpa was regarded as the “first lama or ādiguru” of the Sa skya sect (Dowman 1985: 52; Dudjom 1991: 853). The fact that the Guhyasamayasādhanamālā opens with two texts, one attributed to Indrabhūti (or elsewhere to Śabara; see GSS1 in the appendix) and an almost identical work to Lūyīpāda, is significant. It asserts the antiquity of the collection, and hence its authenticity. Similarly, Indrabhūti’s authorship implies that the geographical source of the teachings is Oḍḍiyāna, the very homeland of esoteric spiritual revelation, as many tantric colophons testify. The text itself (GSS1GSS2) reveals an East Indian influence, with its clear exchange of the consonants v for b in its mantroddhāra.
The fame of our later authors rests upon their scholarly transmission of the tantric teachings. The Blue Annals (pp. 841, 866, 976) refers to Advayavajra’s transmission of mahāmudrā, and associates him particularly with compositions of the amanasikāra class (e.g., ibid.: 842); it is in a text of this class, the *Siddha-Āmnāya, that Advayavajra’s quest for a vision of Vajrayoginī is described, and in which he is requested by his guru, Śabara, to return to academic life to commit his new understanding to writing. Advayavajra is one of the younger adepts who were working in the environment of the monastic universities in northeast India. These were centers of Mahāyāna and tantric learning established under the Pāla dynasties of Bihar and Bengal (760–1142 C.E.), which flourished until their destruction by the Moslem invaders between 1197 and 1207 (Dutt 1962: 380). The five outstanding foundations were: Vikramaśīla, founded—according to Tibetan sources—under Dharmapāla (770–810 C.E., ibid.: 359); Odantapurā and Somapura (also “Somapurī’) under Devapāla (c. 810–59 C.E., ibid.: 373–74); Jagaddala in Vārendra (north Bengal), which probably flourished under Rāmapāla (1077–1120); and the oldest establishment, Nālandā, which had been sponsored at the end of Gupta rule by Harṣa (606–47 C.E.). Although less cultivated by Pāla kings, Nālandā remained a prestigious seat of Mahāyāna philosophy, and at its peak, Chinese sources state that it catered to several thousand students, offering as many as one hundred lectures, tutorials, or debates a day on topics both brahmanical and Buddhist (ibid.: 333; Misra 1998 I: 241ff.). Vikramaśīla was the most renowned monastic universitiy in the Pāla period, with Abhayākaragupta at its head, and its various “schools” (saṃsthāḥ) conferring various posts, honors, and “degrees,” such as dvārapāla (gatekeeper), paṇḍita, or mahāpaṇḍita (ibid.: 360–63 following Tibetan accounts).
The reference to the academic milieu in the *Siddha-Āmnāya is interesting because it illustrates the contrast between the life of Advayavajra, the yogin-paṇḍit working within the monastic universities, and the supposed source of his learning, the illiterate adept and mountain-dwelling huntsman Śabara. Although our younger authors may have lived and worked in the intellectually charged milieu of the monastic universities, their sādhana texts reflect the culture of the earliest proponents of the systems. They lay down prescriptions to practice in wild, solitary places void of people, and it is this aspect of their own practice that is most attractive to legend. In many accounts, historical narrative breaks into mythic motif precisely at the point when the monk rejects formal academia in favor of tantric yogic practice. For example, the story of Advayavajra in the *Siddha-Āmnāya (p. 11) first describes his formal training in grammar and orthodox (nontantric) Buddhist disciplines at monastic universities such as Vikramaśīla; it then recounts his tantric studies (possibly under Nāropa) at Nālandā, but only finally launches him on his higher tantric career when he leaves the monastic life and sets out on his magical journey to seek Vajrayoginī, prompted by a voice in a dream. In Tibetan accounts, Advayavajra was expelled from the monastery for keeping liquor and a woman in his cell (Tatz 1987: 700–701). The same motif of expulsion is found in the account of Virūpa’s life. According to the Legends (Dowman 1985: 43–52), this mahāsiddha first became a monk of Somapura monastery, but despite his initiation into the practice of sow-faced Vajravārāhī, he failed to see her even in a dream until, after twelve years, in a depressed state, he threw his rosary into the toilet. He attained mahāmudrā after another twelve years. Virūpa’s subsequent expulsion from Somapura (for eating pigeon pie) was accompanied by various miracles, such as walking on water and holding back the sun in a ploy to avoid settling his tab at the local tavern.
The distinction between the two lifestyles—formal academic versus wandering yogic—may not have been so marked in practice. The wandering life was an integral part of the monastic experience. Practitioners would move between universities in pursuit of various teachers, and periods of retreat and prior service (pūrvasevā) were also an essential part of formal training. The perceived dichotomy may have been a natural advertising ploy for the techniques to be espoused, and a crystallization of the ideal of the solitary tantric yogin. This is an ideal firmly embedded in the Indian traditions. The Buddha’s going forth is an important role model for any would-be Buddhist saint, and the inflation of this motif to actual expulsion from a monastery is one that provides a useful exegetical comment upon tantric praxis; it is precisely from their antinomian propensities that the practices of the highest tantras draw their power. The texts themselves seem to envisage both lifestyles. At their most extreme, they advocate a type of yogic existence that transcends ritual observances, such as rites of the maṇḍala, or oblations with mantras (see the first upadeśaḥ in GSS32, appendix), but at the same time, they envisage a ritual specialist capable of performing numbers of such rites, not just for his own sake, but on behalf of others (see, for example, ch. 3 §39).
In pursuit of either lifestyle, it seems it was not altogether necessary for the practitioner to be an ordained member of the Buddhist sangha. The higher tantric initiations (ch. 3), which include the empowerments for sexual praxis, were also open to householders. This is implicit in one of the erotico-yogic texts in the Guhyasamayasādhanamālā that refers to the “[householder’s] own house” (svagṛhe) as one possible meditation site (GSS34, appendix). Umāpatideva himself, author of the Vajravārāhī Sādhana, may have been a layman, as well as an initiate into the lineage of Virūpa. His name means “lord of Umā,” that is, Śiva, while ordination into Buddhist orders would ordinarily have endowed him with a Buddhist name. If he were a lay scholar, it would seem that Umāpatideva still had access to the scriptural and exegetical sources available to those working in the monastic environment, judging by the extent of the redaction in the Vajravārāhī Sādhana.
This situation did not lie comfortably with some members of the monastic community, however. In her study of tantric antinomianism, Onians (2002: 292–93) comments that, “The tension between tantric monks and householders must reflect a time when tantric practitioners were found both in and outside monasteries, and the Sangha was compelled to reassert its primacy….” Thus, the Kriyāsamuccaya (f.3.2ff; Gellner 1992: 295) cites many tantric references to support the claim that a tantric teacher (vajrācāryaḥ) should be a monk, although the fact that it opens with a lengthy discussion on the matter raises the possibility of his not being so. Indeed, in his Vajrācāryalakṣaṇavidhi, Jagaddarpaṇa states that a tantric teacher may be of three types: monk, novice, or householder (following the Saṃvarārṇavatantra), but he asserts the superiority of the teacher who is ordained by adding that, should all three be found together, the householder should not be worshiped, for this would be disrespectful to the Three Jewels. Another tantric exegete prescribes certain “beginners” rites and observances (ādikarma) for the householder practitioner (gṛhapatibodhisattvaḥ), suggesting, perhaps, that the qualifications of the lay practitioner were inferior to those of a monk. However, Isaacson (1999: personal communication) points out that the qualifications of the lay practitioner were not necessarily inferior to those of a monk, and that Jagaddarpaṇa’s opening discussion does not reveal his final position on the matter. Indeed, it may even have been the case that practitioners who had taken the bhikṣusaṃvara were sometimes forbidden or discouraged from the actual performance of transgressive practices.
With their emphasis on solitary practice, the sādhanas themselves give no indication as to how they would be practiced within a monastic routine. This is particularly pertinent where the sādhana involves sexual practices that would infringe the monastic vow of celibacy (brahmacaryam). In tackling this issue, exegetes tended to argue that the tantric observances incorporate and surpass, rather than negate or contradict, earlier vows of celibacy: “[In taking tantric initiation] will he not then be guilty of abandoning his earlier vows [of celibacy]? No, for each subsequent observance transcends the preceding, just as the lay devotee becomes a novice and the novice a monk. When a person has become a monk is there the absence of the vows he took as a lay devotee, etc.? [Of course not.]” Jagaddarpaṇa (Onians op. cit.) actually redefines brahmacarya, so that for a nontantric monk it still refers to celibacy; but for a monk who has taken highest initiation (and whom he therefore understands to be spiritually superior), it refers to the retention of semen in the course of yogic sexual practices. However, the attitudes of tantric authors on this matter are complex, as Onians makes clear (op. cit.: 268–71): Atiśa, for example, has—with justification— been interpreted as insisting that for those who held full monastic ordination, the language of sexual yoga was open only to symbolic interpretation and was otherwise incompatible with monastic rule; and yet his conclusions are far more subtle than this and clearly depend upon the context in which celibates may perform the higher initiations and upon a rigorous application of the qualifications that would permit a monk to bypass or transcend his monastic precepts—crucially, the degree of insight with which sexual praxis is imbued. Such sophisticated apologetics are a reflection of the difficulty that must have arisen in bringing tantric practices within the monastic fold. Indeed, there are accounts of iconoclasm among Sthaviravādins unable to tolerate deities such as Cakrasaṃvara at Vajrāsana (Bodhgayā), which Taranātha himself recorded (1990: 279):
In a temple of Vajrāsana there was then a large silver image of Heruka and many treatises on tantra. Some of the Śrāvaka Sendhavas [“Siddhas”] of Singa island (Ceylon) and other places said that they were composed by Māra. So they burnt these and smashed the image into pieces and used the pieces as ordinary money.
But on these issues, the new tantric orthodoxy was clear, as the hagiography of Abhayākaragupta testifies (Blue Annals: 1046; Willson 2000: 397–98). Painting the picture of an exemplary abbot-scholar of traditional Buddhist hue, the lifestory of Abhayākaragupta describes his initial reluctance to embrace the new teachings, as he declines to welcome a woman into his monastic cell. When the woman turns out to be none other than Vajrayoginī in disguise, the monk sees the error of his ways, but finds that he has lost the opportunity ever to gain union with her in his lifetime. He is compensated with the promise that if he composed a “great number of commentaries on profound tantras and many rites of maṇḍalas,” he would soon become “a fortunate one”—a challenge he appears to have accepted.
Having examined the Indian milieu in which Umāpatideva’s Vajravārāhī Sādhana was written, it is time to look more closely at the compilation of the Guhyasamayasādhanamālā itself. According to the approximate dating of their authors, some texts in the collection are possibly as old as the ninth century, but perhaps only date from the eleventh century, while others are later still, dating from the twelfth century. The collection closes toward the end of the twelfth century with the work of an author who was probably a living contemporary, Vibhūticandra (GSS43). Its upper date is fixed by the oldest surviving manuscript (K), which Sanderson (1995: personal communication) suggests is from the twelfth to thirteenth centuries. This date would be roughly contemporary with the earliest manuscript of another sādhana collection, the Sādhanaśatapañcāśikā, which dates from 1165 C.E. (Cambridge add. 1686). The Guhyasamayasādhanamālā receives its title only later; the name is found in the Devanāgarī manuscript (D) alone, in which the title of the last work in the collection (Ḍākinī-guhyasamaya-sādhanamālātantrarāja) seems to serve as the basis for the collective title Śrī-Guhyasamayatantra.
The processes by which sādhanas were compiled into recognizable collections has been studied by Bühnemann (1994), who suggests that scholars were engaged in collecting such works from the eleventh century on. Bühnemann discusses four sādhana collections in all, basing her work on the four collections that Bu ston (1290–1364) drew into his catalog of the bsTan ’gyur (summarized in table 3):
1. The One Hundred and Fifty Sādhanas (*Sādhanaśatapañcāśikā, sGrub thabs brgya dang lnga bcu), consisting of about this number of sādhanas.
2. The Hundred Sādhanas (*Sādhanaśataka, sGrub thabs brgya rtsa), which contains about ninety-three sādhanas.
3. The Ocean of Sādhanas (*Sādhanasāgara in Bu ston’s catalog), also called the Collection of Sādhanas (Sādhanasamuccaya in the Peking edition P4221–4466), and the Garland of Sādhanas (Sādhanamālā in the colophon of some Sanskrit manuscripts), consisting of a large collection of 242 sādhanas.
4. The *Devāntaraviśvasādhana collection, which appears in the Peking edition as an appendix to the second collection, the *Sādhanaśataka.
It is from these collections that Bhattacharyya (1925/28) produced his edition of the so-called Sādhanamālā, accidentally conflating the largest collection of 242 sādhanas (*Sādhanasāgara) with the collection of 150 sādhanas (*Sādhanaśatapañcāśikā).
What does Bühnemann’s survey of the sādhana collections reveal about the manner and date of their compilation? Bühnemann shows that there are problems in fixing the contents of these collections since the Sanskrit manuscripts do not agree between themselves, either in the sequence in which sādhanas appear or in the number of sādhanas they contain, and the Tibetan translations do not seem to accord with the Sanskrit “originals.” The compilation of substantial numbers of sādhanas, or the addition of other collections to them, seems to coincide with the appearance of a title for the collection as a whole. This may have encouraged closure, as in the case of the *Sādhanaśatapañcāśikā, which received its title only once it had collected its one hundred and fifty works (ibid. 1994: 11). Similarly, Bühnemann hints that Bu ston’s third collection may have received its title *Sādhanasāgara in the later recensions preserved in Tibetan from its final portion of texts, entitled Devāntarasādhanasāgara (ibid. 1994: 12). In some collections, the colophon to each individual sādhana also gives the collective title, but again this practice is not standard (ibid. 1994: 11–12). Such irregularities in a title’s appearance in related recensions, and in the title itself, suggest that collective titles were a later feature of the sādhana compilations. Their introduction (possibly coupled with efforts to “round up” the collections to grandiose figures that then serve as collective titles) gives the impression that the sādhana collection was emerging as a genre in its own right. The datings given by Bühnemann indicate that the earliest translations into Tibetan of whole collections were made in the later eleventh century and around the turn of the twelfth century and continued into the thirteenth century (and beyond), that is, in the period when the monastic universities under the Pāla dynasties were at their height. Records of the Sanskrit manuscripts confirm this picture. Comparing the evidence of the manuscript collections with the dates of likely authors, it is clear that the time between the composition of a sādhana and its subsequent inclusion in a collection was often brief and that translation into Tibetan was also a rapid process.
These conclusions confirm what has been gathered of the Guhyasamayasādhanamālā collection. There are, however, notable differences. The Guhyasamayasādhanamālā is far smaller, containing only forty-six works. Moreover, it seems to have been relatively stable. Only one sādhana (GSS8) is omitted in the later recension of the collection represented by the devanāgarī manuscript, a sādhana that is anyway repeated identically later in the collection (GSS39). The Guhyasamayasādhanamālā was not translated into Tibetan, although some of its sādhanas appear in the bsTan ’gyur as part of other collections (details are given in notes to the appendix). Remarkable is that all forty-six sādhanas of the Guhyasamayasādhanamālā focus upon Vajrayoginī/Vajravārāhī. Other sādhana collections, apart from being much larger, are more diverse. They include sādhanas relating to various deities, sometimes arranged accordingly in groups inside the compilation. There are, for example, groups of sādhanas within the so-called Sādhanamālā that focus on other female deities (ch. 2), but not one of these has been preserved as a separate collection in its own right.
The reason the Guhyasamayasādhanamālā remained a discrete collection and was not absorbed into another collection is unknown. Perhaps as a grouping it was too large to be placed inside another collection, or perhaps it had its own pretensions to reach a desirable “fifty.” Another suggestion is that this collection—with its single-minded concern with Vajrayoginī and its “contemporary” nature—may have been the initiative of a single scholar. This impression is heightened by its internal organization. The collection begins with traditional-style maṇḍalas of the Cakrasaṃvara tradition adapted to the female deity Vajravārāhī. There follows a gradual shift toward maṇ˜alas exhibiting a more fully kāpālika character, a trend that is further developed in the “skeleton arch” (karaṅkatoraṇa) sādhanas, which reject the temple-palace structure of the maṇ˜ala altogether. Within this overall structure, the works seem to have been carefully, if approximately, grouped according to particular manifestations of Vajrayoginī, and to the type of work in question. These groupings may be roughly broken down as follows, with some sādhanas appearing in this list more than once where different groupings overlap (the various forms of Vajrayoginī are discussed in chapter 2, and the sādhanas are described individually in the appendix):
- GSS1≈GSS2 – The first two sādhanas in the collection deal primarily with the hogheaded ardhaparyaṅka-pose Vajravārāhī, and are attributed to the prestigious figures Indrabhūti and Lūyīpāda.
- GSS2, GSS3, GSS4, GSS5 – The next manifestation is of Vajravārāhī in her classic warrior-stance form. She appears by herself (GSS2, GSS4), in her fivefold maṇ˜ala (GSS3), and finally in the full thirty-seven-fold maṇḍala (GSS5).
- GSS3, GSS4, GSS5 (GSS11, GSS16) – The third sādhana (GSS3) is by another eminent figure, Advayavajra. It is the first in a group of essentially Cakrasaṃvara-based works, all similar in their exposition of the warrior-stance Vajravārāhī within a maṇ˜ala based on the temple palace. All sādhanas in this group salute Vajravārāhī in their opening reverence. Umāpatideva’s Vajravārāhī Sādhana (GSS11) is also of this type. An interesting sādhana that belongs in part to the Advayavajra group and in part to the Śabara-related texts, is the sādhana of the thirteenfold Vajraḍākinī Vajravārāhī (GSS16).
- GSS6, GSS7 – The next group is of two sādhanas redacted from the Abhidhānottaratantra, the first presenting a six-armed, seated manifestation of Vajravārāhī in embrace with her consort within a thirteenfold maṇ˜ala (GSS6), the second a twelve-armed ardhaparyaṅka-pose Vajravārāhī in a forty-one-fold maṇḍala (GSS7).
- GSS8≈GSS39, GSS13, GSS14, GSS41 – The oblation ritual (homavidhi¯) that follows is one of a more dispersed group of oblation rituals in the collection.
- GSS10, GSS43 – There follow some distinctive, erotic practices of Vajrayoginī, notably Vajravilāsinī (GSS10), who is also the subject of a stotra (stotram) or praise work (GSS43).
- GSS12, GSS17≈GSS45 – Similarly amorous are the “raised-foot” (ūrdhvapāda-) pose deities, first the red Vajravārāhī (GSS12), and then the white Vajrayoginī (GSS17≈GSS45).
- GSS15, GSS18, GSS38 – Next comes the red hog-headed “Vajraghoṇā” manifestation of Vajravārāhī (GSS15, GSS18), possibly related to a white manifestation of the same deity (GSS5, GSS38).
- GSS19 – The next section of the Guhyasamayasādhanamālā deals primarily with magical erotic forms of Vajrayoginī, such as a two-armed Vajrayoginī at the center of a fivefold maṇḍala (GSS19).
- GSS20, GSS24, GSS25, GSS26, GSS27, GSS30≈GSS9 – Another magico-yogic manifestation is the striking, self-decapitated Trikāyavajrayoginī (“Chinnamastā”) in sādhanas GSS20, GSS24, and GSS25, and in verse works related to Virūpa, GSS26 and GSS27. This form is related to the deity to be visualized in GSS9≈GSS30.
- GSS21, GSS22, GSS23 – Another such group is that of the flying Vidyādharī Vajrayoginī forms of the Śabara school.
- *GSS28?, GSS29, GSS30, GSS31, GSS39 – Next, the Guhyasamayasādhanamālā contains a collection of repeated works that are almost identical to those transcribed earlier in the GSS, but with minor differences. This section includes: *GSS28 (≈GSS19?), GSS29 (⊕GSS4), GSS30 (⊕GSS9), GSS31 (⊕GSS3), and GSS39 (=GSS8).
- GSS32, GSS33, GSS34, GSS35 – The collection then provides three svādhiṣ˛hāna-method sādhanas (GSS32, GSS33, and GSS34), the internalized nature of which is also reflected in a rare four-armed form of warrior-stance Vajrayoginī (GSS35).
- GSS36, GSS37, GSS38 – Some unusual Vajrayoginī forms follow, such as the yellow Vajrayoginī in falling-turtle pose (GSS36), and two white warrior-stance Vajrayoginī forms, GSS37 and GSS38.
- GSS42, GSS43 – There are two Vajrayoginī stotras in the collection grouped together.
- GSS40, GSS46 – Finally, there are two commentarial works.
While these groupings are not entirely even, they are marked enough to suggest a conscious arrangement of the materials. What is even more striking is that this arrangement is complemented by the internal structure of the Abhisamayamañjarī by Śākyarakṣita (GSS5). ⁄ākyarakṣita’s work begins with classic sādhana meditations on Vajravārāhī’s thirty-seven-fold maṇḍala, after which it becomes a compendium of alternative visualizations of the deity in her different manifestations (see appendix). The catalog of visualizations supplied in the Abhisamayamañjarī mirrors the sequence of the Guhyasamayasādhanamālā collection as a whole, so that the classic warrior-stance Vajravārāhī of the first part of the work is followed by the ūrdhvapāda-pose Vajrayoginī, Vajraghoṇā, the two-armed Vajrayoginī, and the Trikāyavajrayoginī forms. Thus, it looks as if the Abhisamayamañjarī may have been used as a blueprint for the arrangement of sādhanas by the compiler of the Guhyasamayasādhanamālā.
The importance of the Guhyasamayasādhanamālā collection to the Vajrayoginī tradition, and its uniqueness as a collection, have now been established. However, the decision to edit and translate the Vajravārāhī Sādhana by Umāpatideva (GSS11) still requires some explanation. Not only are there many sādhanas in the Guhyasamayasādhanamālā deserving of attention, but tantric literature in general is vast, and sādhana itself forms only one genre within it. What, then, is the significance of the sādhana within the tantras? And what is of particular interest in Umāpatideva’s Vajravārāhī Sādhana?
A sādhana is a progressive sequence of meditative and ritual procedures that focus upon a particular deity or set of deities. It is a relatively late addition to the tantric corpus. The first embryonic sādhanas appeared only in the eighth century with the yogatantras, and their form was still not standardized by the time of the Hevajratantra. Perhaps one of the earliest references to sādhana practices appears in the yogatantra scripture, the Sarvadurgatipariśodhanatantra, which recounts “a sādhana taught by
Śākyanātha” (p. 132 9b). The meditation is to be done “with the method of generation” (p. 130 8a: utpattikrameṇa), and it is described as “the highest deity yoga” (devatāyogaṃ…uttamaṃ). Despite the inclusion of material not generally found in later sādhanas, it clearly sets out the key features of a mature sādhana, all of which will be seen as distinct stages in the Vajravārāhī Sādhana (GSS11) studied here in chapter 3. Thus, it includes the usual preliminaries, the construction of a circle of protection, the accumulations of merit and wisdom, pūjā, and the merging of the maṇḍala in space with the maṇḍala in the heart. The eighth-century commentator Buddhaguhya recognized the sādhana material in the Sarvadurgatipariśodhanatantra as distinct from the rest of the tantra, describing it as an “introduction” (gleng gzhi = nidāna) to “the actual text” (mdo bshad) dealing with maṇ˜alas (Skorupski 1983: xxvii). Another yogatantra commentator, Vilāsavajra, also approaches the topic in his Nāmamantrārthāvalokinī. In adhikāra IV, Vilāsavajra produces his own proto-sādhana, which includes key preliminary meditations and the generation of deities within a maṇḍala, but which lacks other established features of the later sādhana, such as developed stages of generating oneself as the deity, or the merging of the pledge and knowledge forms of deities (Tribe 1994; 1997: 115–17, 123–25).
The eighth century also saw the emergence of the Guhyasamājatantra (GST), and the beginnings of the exegetical schools based upon it. This tantra begins to systematize the components of deity practice. It distinguishes a fourfold sequence of meditations as a prelude to ritual undertakings (e.g., GST ch. 12, vv. 60–65) that it refers to as: (1) service (sevā); (2) auxiliary attainment (upasādhanam); (3) attainment (sādhanam); and (4) great attainment (mahāsādhanam). These cover introductory and preliminary meditations (in the first and second stages), with the “urging” (codanam) and summoning of the deity, and its final visualization (in the third and fourth stages). A related schema in the Guhyasamājatantra, also in four stages, focuses just upon the generation of the deity. This is the “[set of] four vajras” (vajracatuṣka), which corresponds in yoginītantra texts to the sequence of five awakenings. The Guhyasamājatantra also distinguishes a stage of “generation” (utpatti), from a stage of “completion” (utpanna/niṣpanna) (e.g., GST ch. 18, v. 84; see Wayman 1977: 23), an important classification that we will see in the mature sādhanas of the yoginītantra. The two stages or methods (kramaḥ), the generation stage (utpattikramaḥ) and the completion stage (utpannakramaḥ, niṣpannakramaḥ), were elaborated upon in the two schools of Guhyasamāja exegesis, each of which produced its own texts based on the classification.
The period of yogottara systematization took place in the ninth to tenth centuries in the setting of the great monastic universities (Mimaki and Tomabechi 1994: ix), a period that coincided with the emergence of the new yoginītantras. The highest tantra scriptures develop the deity meditations into sādhana-type practices that bear much the same form as the mature sādhana (e.g., Hevajratantra, devatāpaṭala 1.3 and Saṃvarodayatantra, śrīherukodayanirdeśapaṭala ch. 13). The four stages of the yogottara system (sevā, etc.) are still current—both implicitly in a fourfold structure of the sādhana-type passages, and explicitly through direct reference (e.g., HT1.1.25; ADUT ch. 14: 317ff). It is also notable that the internal structure of these tantras may demonstrate the same sequence of meditative and ritual events as those we will see in our study of a mature sādhana. The Saṃvarodayatantra, for example, begins with the methods of generating the deity and his wider maṇḍala, followed by the ritual practices grounded in that self-generation. The structure of the Hevajratantra is similar and also mirrors the composition of a sādhana. The scriptural sources of the yoginītantras therefore draw closely on the methods of the sādhana, and may be seen as products of existing praxis that cultivated sādhana or sādhana-type techniques. Without an understanding of these stages within the sādhana practice, the intended meaning of the tantras is lost.
At the same time, this period saw important developments in the form and structure of the sādhana itself. Such developments were doubtless stimulated by the new trends of the highest tantras and perhaps also reflected the need to clarify the practices outlined in the scriptures. Thus, features of the sādhana already evident in the yogatantra corpus underwent gradual definition. The process is detectable in certain sets of sādhanas in the Sādhanamālā collection, such as the sizable collections of sādhanas grouped around manifestations of Avalokiteśvara (SM6 to SM42) and Mañjuśrī (SM44 to SM84). Here one sees how the peaceful cults of princely cakravartin-style bodhisattvas are increasingly permeated by tantric elements, such as the preeminence of the guru, the use of transgressive substances, erotic and wrathful
Śaiva-based iconography, erotico-yogic praxis, and cremation-ground motifs. The method of generating the deity is also refined, and evolves into the series of five awakenings found in the mature sādhanas, to be followed by the merging of its pledge and knowledge forms. By the time of the yoginītantra sādhanas of the Guhyasamayasādhanamālā, the form of the tantric sādhana was well established, and yet our study of the collection will reveal that the genre was still developing. In response to developments in the yoginītantra scriptures, some sādhanas will be seen to introduce cremation-ground (kāpālika) features within the standard format of the sādhana, and others to reject mainstream formats altogether.
The sādhana is significant within tantric literature as a whole in that it mirrors and clarifies developments in content and method. As a genre it is particularly flexible, because its form may be easily adapted to cater to changing currents and trends in praxis. In this way, the sādhana is able to elaborate and develop tantric practices that are lacking or marginal in the scriptural material itself. This is particularly pertinent in the cult of Vajrayoginī/Vajravārāhī, which has no scriptural corpus of its own, but borrows from the scriptural tradition of Cakrasaṃvara. One reason Umāpatideva’s Vajravārāhī Sādhana is a useful subject for analysis is that it highlights the processes of redaction by which new tantric techniques were adapted from existing ones—that is, how the author borrows from scriptural and exegetical sources concerning the Cakrasaṃvara maṇ˜ala and its rituals, and alters them to describe the Vajravārāhī maṇḍala and its rituals. Since sādhanas are not tied to a particular scriptural source, we will see that different works in the Guhyasamayasādhanamālā draw on different parts of the Buddhist (and Śaiva) traditions and produce a range of forms and practices of Vajrayoginī.
Above all, sādhanas are manuals of practice; they are the “means of attainment” (sādhanam) whereby the goals of the highest tantras may be realized. Their prescriptions encompass a range of meditation techniques and ritual procedures, the length and complexity of which suggest a fulltime commitment to the practices. As shown earlier, little in the sādhana suggests the practitioner’s broader lifestyle. His daily routine is indicated only by general injunctions that are embedded into the sādhana itself, namely, to rise early, to wash, to perform the sādhana in a solitary place preceded by certain preliminary rites, to repeat it three or four times a day, and to perform various external rites on the basis of this meditation. Sādhana texts also say little of the previous spiritual practice that has prepared the practitioner for taking up the sādhana or of the initiations that have qualified him to do so. Such preliminaries are so fundamental to the tantric system that they are usually taken for granted by the author of a sādhana, whose audience is understood to be made up exclusively of initiates into the cult. As one sādhana in the Guhyasamayasādhanamālā puts it, the practitioner should be someone “who has an undivided attitude of devotion toward his teacher and the Buddha, who has firmly seized the will to enlightenment, [and] who has correctly obtained initiation.” The topic of initiation or consecration is a vast and complex one; it is discussed briefly in our study of the sādhana at the point when the meditator visualizes his own consecration by celestial deities, a process that mirrors the types of consecrations employed by tantric teachers in their initiatory empowerment of pupils. It is only after such initiations have taken place that certain practices may be undertaken, indeed, that the sādhaka becomes obliged to fulfil his vows to practice.
The role of the guru in this process is, of course, central. It is upon his authority alone that the tantric systems depend. It is the teacher who transmits teachings, authorizes praxis, and performs the initiations that qualify pupils to identify themselves with their chosen deity in the practice of deity yoga. The importance of understanding the guru to “be” the Buddha (that is, the central deity of the particular tantric cult), the benefits of worshiping him, and the evils of transgressing his instructions, are therefore favorite themes in tantric literature and often appear in frame verses to sādhana texts, for example:
The guru is the Buddha, the guru is the Dharma, and the guru is the Sangha. The guru is the glorious Vajradhara; in this life only the guru is the means [to awakening]. Therefore, someone wishing to attain the state of buddhahood should please the guru.
The post-initiatory observances are known as the observances of the pledge or samaya (samayācāraḥ). Their supreme importance to the newly consecrated yogin is often emphasized by the texts with the insistence that the samaya be “protected.” The yogin does this by practicing it faithfully, and by maintaining a strict code of secrecy. Reminders that the practices are secret (guhya) and solemn injunctions to secrecy are therefore common, especially when the texts invert traditional ethical norms by prescribing transgressive disciplines, such as sexual yoga. This leads us back to the centrality of the guru, who is the source of teachings that may well remain purely oral. The first sādhana in the Guhyasamayasādhanamālā is itself described as the “Oral Teaching of Vajrayoginī” (Vajrayoginīmukhāgama), and its secret practices are said to go from “ear to ear, mouth to mouth.” The sādhana includes a description of a parvapūjā in which the yogin is to worship a young virgin and, while naked, make transgressive offerings of sexual fluids to the deity; it then enjoins secrecy, and reminds the yogin that if he wants to obtain siddhi, he must preserve the samaya. In a Vajraghoṇā pūjā (GSS18), midnight offerings include incense made of powdered human flesh, following which the text states that “this is not to be told to anyone.” Secrecy preserved and enhanced the efficacy of the teachings, and was understood to be the crucial context for their practice, the essential ingredient that gave the antinomian cults of esoteric Buddhism their power. It was (and still is) of crucial importance to the practices of Vajrayoginī/Vajravārāhī, who is described in one text as “mother of the guhyakas [lit: ‘those with secrets,’ i.e., initiates]” (n. 75).
Following his initiation into a sādhana practice, the pupil would next undertake a prolonged period of mantra recitation, which forms a preparatory foundation for undertaking the sādhana itself. This is referred to as “prior service” (pūrvasevā). Without this bedrock it is said that the sādhaka “would stun, damage, and harm himself ” (mKhas grub rje: 275–79). The number of mantra recitations required varies according to different systems. For example, in the Hevajra system (HT1.10.25ab) there are to be one hundred thousand for the lord of the maṇḍala and ten thousand for the maṇḍala retinue; similar numbers are given in the Vajrāvalī for the Kālacakra texts, while in the Saṃvara system, there are said to be both medium and brief periods of service (mKhas grub rje ibid.).
Since authors of sādhanas say little about the preparation and training undergone by a sādhaka prior to his undertaking the practice, they assume that he has already acquired the necessary meditative, ritual, and conceptual skills. This most important meditative tool is the technique of visualization meditation. This demands that the practitioner be able to visualize the object of meditation, located either in space in front of him or at the center of his own body. Texts state that he should “see” (paśyet, avalokayet, īkṣeta) the object of meditation “very clearly” (GSS11 v.17: vispaṣṭataram) and unwaveringly (SM123: 254); he should “contemplate” (vi-cintayet), “imagine” (vi-bhāvayet), “meditate upon” (dhyāyāt), or “be convinced of” (adhimuñcet) it. The manner of producing the visualized object in the mind is described as the arising or generation (utpattiḥ) of the object and usually begins with a mantra syllable representing the essence or source of the object to be visualized. This is known as its seed (bījam) or seed-syllable (bījākṣaram), and it has both an aural dimension, such as the sound of the mantra syllable hūṃ, and a visual dimension as the written form of that syllable, seen with the mind. The seed-syllable then undergoes an imaginative transformation into the object for which it is the more essential symbol, which is expressed in Sanskrit as the object being “produced” or “born” (-ja, -bhūta, -niṣpanna), or—where there is a whole sequence of such visualizations—by their “evolution,” “development,” or “transformation” (pariṇāmena) into the final object. The visualized forms are understood to be made of light; they are vibrant, incandescent, pellucid, and yet as insubstantial as any other simile for emptiness. They scintillate with the emission and retraction of light rays that function as powerful agents of the meditation, acting to remove ignorance and impurities, destroy obstacles, give succor to beings, or praise or coerce deities.
The locus of the visualization is significant because it differs according to different rites, and plays an important part in the classification of the sādhana. The deity may be visualized “in space” (khadhātau) in front of the meditator, as in the preliminary pūjā, or be generated within an external ritual object, such as a maṇ˜ala diagram drawn upon the ground or upon the meditator’s own hand. The process of generating objects of meditation is at its most elaborate in the section that deals with the yogin’s generation of himself as the deity. Here the generation is located at the center of the yogin’s own body, inducing in him the conviction that he “is” the deity. The sādhana is therefore a “means of attainment” because it is a tool for the transformation of the mundane into the the transcendental.
The application of the self-generation method at this stage generally classifies the sādhana as a generation-stage practice (utpattikramaḥ, see ch. 3). In a self-generation sādhana, the subject of prescription changes in midcourse. The mundane personage of the practitioner who begins the practice is designated variously as the sādhaka, the yogin, the mantrin (literally, “the possessor of mantra”), or by some traditional laudatory epithet acknowledging that he is “a skilful one” (vicakṣaṇaḥ), wise (budhaḥ), or learned in mantric lore (mantravit). In the course of the self-generation, the meditator acquires the transcendental identity of the chosen deity. The new agent is described as “one conjoined with the deity” (devatāyuktavān), the practitioner of “deity yoga” (devatāyogaḥ). He is the “yogin-as-deity” or, as in the context of our Vajravārāhī visualization, the “yogin-as-goddess.”
Another means of transforming a mundane object into a transcendental one is by symbolically equating one with the other. This is termed, literally, a “purification” (viśuddhiḥ). The correspondence is made on the firm understanding or conviction (adhimokṣaḥ, niścayaḥ) of the mundane object “as” the supramundane counterpart. The yogin understands that the true essence or inherent nature (svabhāvaḥ) of the mundane element is ontologically equivalent to that of the supramundane, because both are empty (śūnya). The mundane is “purified” through the practitioner’s realization that emptiness pervades both sides of the equation. For example, a practice well attested in yogottara and yoginītantra sources is the purification of the yogin’s entire pyschophysical being as a preliminary to undertaking the sādhana. Here, each of his five skandhas, his sense organs and the five elements in the body, are correlated imaginatively with a particular buddha, bodhisattva, or buddha-consort. The viśuddhi is more than a means of imbuing an object with a symbolic value to an object, although a complex web of symbolic relationships may be implied, connecting together different levels of reality. It is rather a “purifying correspondence” that associates the mundane with the supramundane on the basis of emptiness, and thereby purifies the former. As Sferra (1999) notes, in his discussion of the topic, the term viśuddhi indicates on one hand “pureness,” Buddha nature itself, “the ever shining and pure condition that is always present in all things…. On the other hand, the term indicates purification and therefore a process or a means.”
In addition to the sādhaka’s skill in visualization and meditation techniques, he is expected to be a ritual specialist. Tantric ritual in general revolves around the methods of the sādhana, which provide the means and the rationale for rites, both on private and public levels. In the sphere of private practice, rituals of worship and propitiation are generally prescribed following the main body of the sādhana, and we will see how, according to the highest tantric systems, they can be undertaken only on the basis of deity yoga. The transformation of the sādhaka into the deity during the course of the sādhana is therefore the necessary preliminary to all other ritual acts whatsoever, and it is really the transcendental deity itself—in this case, Vajrayoginī—who performs the rituals, and not the (unenlightened) practitioner. In the public arena, the transformative tools of sādhana meditation are just as crucial. For example, rites of consecration (pratiṣṭhā) play a key part in communal practice, as all objects for religious use must be consecrated, from buildings such as monasteries and stūpas, to objects such as statues of deities, painted images, the cloth on which those images are drawn, religious texts and manuals, initiation vases, ritual implements, and so on. In order to undertake the rites of consecration, the tantric officiant must first have generated himself as the deity by means of the sādhana, and then, in his transcendental persona, must set about transforming the mundane object into a receptacle for the deity to enter, recreating it as the locus in which the deity becomes present and established (pratiṣṭhā). In this process, the tools of sādhana meditation are employed to generate the form of the deity within the object, to infuse it with supramundane wisdom, and then to initiate it according to the tantric system of initiations. In her detailed study of the consecration of images and stūpas in tantric Buddhism, Bentor traces the elements of these complex public rites, and shows how they are in themselves a “special application” of the “basic transformative ritual” that is the sādhana (1996: especially 1–13; Tanemura 2002).
The transformative influence of the sādhana is intended to permeate the sādhaka’s entire life. Rites are sometimes distinguished according to whether they are “outer” (*bāhyakriyā) or “inner” (adhyātmayogaḥ) (e.g., mKhas grub rje: 219), and it is clear that the different elements of the sādhana cover both planes. On an outer level, sādhana prescriptions govern bodily actions and speech, as when the yogin performs his morning ablutions or prepares a suitable site for the meditation through mantra recitation. On an internal level, we have seen how mental, imaginal, and experiential faculties all come into play in visualization meditation to create the conviction of new transcendental reality. But the rites and meditations of the sādhana cannot really be so clearly divided. External ritual actions also play an important part in the yogin’s internal world, as the visualization meditations themselves also include bodily movements such as hand gesture (mudrā), verbal utterance (mantraḥ), or the complex mental activity of preparing and visualizing offerings to deities. In some meditations, the inner world the yogin has conjured up in the course of the sādhana is itself treated as if it were an “external” object and subjected to meditative practices that seek to internalize it even further, integrating it within his experience on less and less conceptual levels. Note, for example, the increasingly subtle meditations prescribed within the context of yogic meditations, practices such as the contemplation of iconic and aniconic forms of deities and “drops” that are perceived within the yogin’s own “veins” (nāḍīs) and “body centers” (cakras) (ch. 3). Looked at another way, the internal world that is created through the practice of deity yoga must also be externalized and made to imbue all the yogin’s outer actions in his daily life. This happens at the end of the sādhana, when the sādhaka is instructed to keep the internal convictions produced through his visualization meditation and to maintain an awareness of himself with the form and nature of Vajravārāhī while he goes about his everyday business. In this way, his whole life becomes a meditative ritual. The inner and outer levels are thoroughly interwoven and interconnected, and come together to forge the practitioner’s conviction that he is the deity on all levels of his being: on the external planes of his bodily and verbal action, on the internal planes of thought process and existential conviction, and on the subtle experiential dimensions beyond conceptualization. The method is thus perfectly allied to the goal of unification with the deity, or “deity yoga.”
The same methodology is reflected in the structure of the sādhana. It begins with a series of preparations that allow the sādhaka to assimilate himself to the outer and inner character of his chosen deity, and intensifies as he imagines himself reborn as Vajravārāhī and infused with her wisdom. Since the sādhana is to be performed at least once daily, it results in a spiraling circularity. It establishes and reestablishes the yogin in a form that he already believes himself to possess. The significance of the sādhana within tantric literature therefore lies in the fact that it is the basic tool of all tantric praxis; it supplies the means with which the practitioner is to recreate ordinary reality as transcendental reality, and thus to achieve his—or her—ultimate aim.
How to cite this document:
© Elizabeth English, Vajrayoginī (Wisdom Publications, 2002)
Vajrayoginī by Elizabeth English is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Based on a work at http://www.wisdompubs.org/book/vajrayogini.
Permissions beyond the scope of this license may be available at http://www.wisdompubs.org/terms-use.