Jump to ContentJump to Main Navigation
Tantric TreasuresThree Collections of Mystical Verse from Buddhist India$

Print publication date: 2004

Print ISBN-13: 9780195166415

Published to Oxford Scholarship Online: May 2006

DOI: 10.1093/0195166418.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 26 August 2019



(p.3) Introduction
Tantric Treasures

Roger R. Jackson (Contributor Webpage)

Oxford University Press

Abstract and Keywords

This introductory chapter begins with a brief discussion of the possible origins of the verses translated in this book: the three surviving Apabhramśa-language collections of rhyming couplets, dohākosa, literally “treasuries of dohā” attributed to three Buddhist tantric masters who probably lived in northern India sometime around 1000 CE: Saraha, Kānha, and Tilopa. Conflicting accounts about the lives and uncertainties about their relation to the written works are described. Known facts about these three Treasuries and the men who are believed to have composed them are also presented.

Keywords:   Saraha, Kanha, Tipola, India, Nepal, Tibet, poems, siddhas, dohā

A Possible Scenario

A thousand years ago or more, a solitary yogin walks out of the Bengali jungle just after sundown and sits cross-legged under the canopy of a village banyan tree. He is dressed in little more than a loincloth. His beard and mustache are unkempt, and his long, matted hair is tied up in a bun. He carries a mendicant's staff and a double-headed hand drum. His eyes shine in the torchlight. His reputation has preceded him, and an audience quickly gathers at his feet, mostly young village men but some women, too. They've heard that he mocks the elders, teaches a way to live freely in the world, and sometimes will perform a miracle, like turning base metals into gold or flying through the sky. Older men cast suspicious glances from the edge of the crowd. They've heard that he's a dropout from the monastic university, lives near a cremation ground with a low-caste woman, participates in debauched rites, works at a low-class occupation if he works at all, and is out to subvert the social and religious order. The silence of evening is broken by the barking of dogs, the lowing of cattle, and the screeching of birds; the scent and haze of home fires fills the air.

When his audience has settled down, the man starts slowly to beat out a rhythm on his drum, and then he begins to sing. His voice is untrained and his melodies rough, but his lyrics are sharp and aphoristic. In rhyming verses, using words from the common tongue, he celebrates the ecstasy of enlightened awareness and the free-roaming life, while mocking the pretensions of ritualists, scholars, contemplatives, ascetics, and anyone who claims that realization can be found anywhere but within oneself. His words are simple, but his meanings complex and full of paradox. He sings of the sky and stars and sea, of animals and plants, of husbands and wives and kings and commoners, but in ways that seem to point below the surface. He says the mind is pure but that we have to do without it; he suggests we can live sensuously in the world but warns against the traps of pleasure; he damns obsession with religious rites but hints at mystical practices of his own; he rails against experts of every sort but venerates his guru without reserve. When he is finished, he gets up, turns his back to the crowd, and walks back alone into the jungle.

(p.4) The next morning, the village work resumes as it always does, but now some of the young people, and the old men, too, find that they've got the yogin's songs stuck in their heads, a phrase here, a rhyme there, which they try to puzzle out. At odd moments during the day, and even more so at night, they find their thoughts turning to the jungle, to truths that might be discovered beyond the village clearing, to the sound of that strange troubadour's voice, the rhythm of his drum, the look in his eyes.

Uncertainties about Siddhas and Dohās

Siddhas in General

This is at least one way of imagining the origins of the verses translated here, the three surviving Apabhraṃśa-language collections of rhyming couplets, dohākoṣa, literally “treasuries of dohā” attributed to three Buddhist tantric masters who probably lived in northern India sometime around 1000 C.E.: Saraha, Kāṇha, and Tilopa. These men, and other extraordinary men and women like them, are known collectively as mahāsiddhas (great adepts, or great perfected ones, “siddhas” for short).1 Through the songs they composed, the instructions they left, and the stories that have been told about them they have deeply influenced the shape of religious and literary culture in a number of Asian countries, especially India, Nepal, and Tibet. In India, even after the virtual disappearance of Buddhism in the thirteenth century, their criticism of the status quo and their celebration of a mystical ecstasy attainable through the human body and the grace of a guru helped to set the tone for a variety of later religious movements, including the sant tradition of Kabīr and Nānak, certain strains of bhakti devotionalism, and aspects of Sufi Islamic mysticism; while in literature they helped to hasten the eclipse of Sanskrit and the rise of various north Indian vernacular languages, whose poetic traditions still carry echoes of their rhymes, rhythm, and imagery. In Nepal, they served as models, and sometimes as deities, for the Buddhist vajrācaryas (“tantra experts”) among the Newars of the Kathmandu valley, who to this day perform rituals and sing songs that tradition traces to them. In Tibet (and culturally related areas such as Mongolia, Ladakh, Sikkim, and Bhutan), they were seen as charismatic, powerful, wise, and compassionate exemplars of the tantric Buddhist approach to life and as the crucial sources for many important lineages of spiritual practice; at the same time, their songs became models for genres of oral and written poetry that have been immensely popular and influential, whether produced (p.5) by great hermit yogins like Milarepa (Mi la ras pa, 1040–1123) or powerful clerics like the First Panchen Lama, Lozang Chokyi Gyeltsen (Blo bzang chos kyi rgyal myshan, 1567–1662).

Despite their importance and influence, the siddhas in general, and the three that concern me here in particular, remain profoundly elusive, especially to the historian. We don't know exactly who they were, what religious allegiance they claimed, where or when—or even if—they lived, or how many of the works attributed to them really are theirs.

The most widely disseminated tradition, reflected in a twelfth-century hagiographic collection by the Indian scholar Abhayadattaśrī, tells of eighty-four great siddhas, most of them adepts of the esoteric and controversial Yoginī tantras, an interrelated set of sexually and soteriologically charged texts that flourished among north Indian Buddhists starting around the eighth century, and would become especially important in the “later” (post-1000) orders of Tibetan Buddhism: the Kagyu (bKa' brgyud), Sakya (Sa skya), and Gelug (dGe lugs).2 Saraha, Kāṇha, and Tilopa all are counted among the eighty-four, as are such equally famous figures as Śavaripa, Virūpa, and Nāropa, and a Nāgārjuna who may or may not be the same as the great Mādhyamika philosopher. There are, however, other treatments of siddhas, with different enumerations and often with different names, such as a partly differing list of eighty-five attributed to Abhayadatta's contemporary Abhayākaragupta; a thirteenth-century Nepalese guru lineage text that mentions nearly two dozen; an eighteenth-century Tibetan account of lineages that relates the stories of fifty-nine—as well as texts that count as siddhas various Indian yoginīs (some of whom appear on other lists, some of whom do not) or the Indian progenitors of the great perfection (rdzogs chen) practice tradition popular in the Nyingma (rNying ma) tradition of Tibetan Buddhism (e.g., Mañjuśrīmitra, Garab Dorje [dGa' rab rdo rje], Śrī Siṃha, Vimalamitra, and above all Padmasambhava).3

The figures found in these lists are generally acknowledged to be “Buddhists.” Certainly, the legends surrounding them and the words attributed to them have influenced countless Buddhists in India, Nepal, and Tibet for a thousand years; but in their original setting, it is not always easy to separate them out—whether in terms of terminology, rhetoric, or practice—from similar figures in non-Buddhist, especially “Hindu” traditions. They seem quite closely related to Śaivite ascetics like the Paśupatas and Kāpālikas; tāntrikas like the Kashmiri Śaivas and Bengali Śaktas; or the wonder-working Nāth siddhas and Rasa siddhas. More broadly, there are general similarities between ideas and practices found in Buddhist siddha writings and those of other Indian yogic and ascetic communities—from such “textualized” movements as those (p.6) reflected in the Yoga Sūtra of Patañjali and the Saṃnyāsa Upaniṣads to such seemingly timeless and “unwritten” groups as the Nāgas, Kaṇphaṭas, and Aghoras.4 Nor can their possible connections with similar sorts of groups in, for instance, Persia, central Asia, or China be overlooked; the resonance, and possible historical connections, between Indian siddhas and Chinese Chan masters or Taoist immortals suggest an especially intriguing, if uncertain, path for further research. What is more, it is entirely possible that, as suggested long ago by Agehanada Bharati, most of the siddhas actually were pre- or nonsectarian wandering yogins, who appropriated various religious terms without intending to promote a particular religion—yet willy-nilly were appropriated by those very sectarian traditions that they resisted or ignored.5

The figures mentioned in the siddha lists often are related explicitly to one another, for instance as guru and disciple, and often are situated in a specific place and/or during the reign of a particular king—many of them, for instance, in north or northeast India during the Pāla and Sena dynasties (c. 750–1250 C.E.). Unfortunately, however, the minimal historical information supplied in one account often contradicts claims made in other sources, or simply is too vague to be interpreted clearly, so that it is very hard to specify that siddha X lived in such and such a place and time and was the disciple of siddha Y and the teacher of siddha Z. Indeed, such discussion begs the question whether many of the siddhas may not simply be literary inventions, no more reliably “historical” than stock characters in epics and folktales the world around. The hagiographies of the siddhas show unmistakable links to Indian narrative traditions dealing with wizards (vidyādhara), zombies (veṭala), and ghosts (bhūta), epic and folk treatments of powerful, capricious r̥ṣis, and Mahāyāna sūtra celebrations of the heroic bodhisattva. They also beg comparison with the traditions surrounding such universal figures as the saint, the trickster, and the superhero. These mitigating factors—which frustrate so many efforts to understand the religious and cultural history of pre-Muslim India—make it very unlikely that we ever will be able to discover the “historical siddhas.”

Hundreds of works of literature are attributed to the siddhas revered in Buddhist traditions; sometimes they have been preserved in Indic languages (usually Sanskrit or Apabhraṃśa), but more commonly they are found in their Tibetan versions, either in the Tangyur (Bstan 'gyur) collection of “new school” (post-1000) translations or the Nyingma Gyubum (Rnying ma rgyud 'bum), with its “old school” (pre-850) translations.6 The siddhas are credited with a tremendous range of types of texts, including initiation ceremonies, maṇḍala rituals, fire offerings, hymns of praise, meditation textbooks (sādhanas), and tantric treatises and commentaries, as well as works that deal with such originally nontantric philosophical topics as meditation theory, and Madhyamaka and (p.7) Yogācāra ontology and epistemology, not to mention various worldly sciences. The texts that have drawn the most attention, in part because they are the most “personalized,” in part because of their impressive literary qualities, are the collections of song-poems in such genres as the dohā (aphoristic couplet), caryāgīti (performance song), and vajragīti (diamond song). Unfortunately, there is very little way of knowing whether a particular text attributed to a particular siddha—even if that siddha was a historical figure—actually was written by that siddha, so the notion of a “corpus” of texts unambiguously belonging to a specific figure must be regarded with considerable suspicion. This may be even more true in the case of song-poems, where the “texts” most likely were originally oral and were only written down and redacted into collections years, or even centuries, after their composition.

Saraha, Kāṇha, and Tilopa

All of the problems besetting the study of the siddhas in general apply to our attempts to understand Saraha, Kāṇha, and Tilopa. We have multiple, and often utterly conflicting, accounts of their lives and great uncertainties about their relation to the written works attributed to them.

Saraha, the “arrow-maker” disciple of a female tantric practitioner (and also known as the Great Brahmin, or Rāhulabhadra the Younger), is perhaps the greatest single individual in the history of Indian tantric Buddhism, famed as its most eloquent poet; as the fountainhead for lineages of practice related to the Yoginī tantras and to meditation on the “great seal” of reality, mahāmudrā; and as a guru to the immortal Nāgārjuna. Yet we cannot locate him with any precision at all in time or place, probably confound him at times with a disciple called Saraha the Younger, and cannot be certain that two of his most notable poetic works, the “King” and “Queen” Treasuries, were written by him or by a Nepalese master of the eleventh century. Though the Treasury translated here is one of the most famous documents of late Indian Buddhism, it exists in multiple, only partly overlapping forms in both Apabhraṃśa and Tibetan. The “standard” Apabhraṃśa version, discovered in a Nepalese royal library in 1907 and published in 1916 by Haraprasād Śāstri, then worked and reworked by Muhammad Shahidullah and Prabodh Chandra Bagchi, never has been found as an independent manuscript but rather has been extracted from a later (eleventh-century?) commentary, in Sanskrit, the Dohākoṣa-Pañjikā of Advayavajra—who may be the same as the great Indian tantric theorist Maitripa. In 1929, Bagchi found in Nepal a fragment of still another Apabhraṃśa version that coincides with other editions not at all. Yet another version of the Apabhraṃśa of Saraha's text was discovered by Rahula Saṃkr̥tyāyana at Sakya (p.8) monastery in Tibet in 1934 and published in 1957; it only replicates about half the verses in the “standard” edition. Furthermore, the Tibetan translation contains both common and unique verses, adding further evidence, as if it were needed, of the complexity and fragmentation of the textual tradition surrounding Saraha's signal work.7

Kāṇha (also known as Kr̥ṣṇācārya, the “dark master”) is reputed to be an important figure in the transmission lineages of Cakrasaṃvara, a deity whose practice is the focus of a major Yoginī tantra cycle; the author of a brilliant commentary on another Yoginī tantra, the Hevajra; a disciple of the great siddha Virūpa; and the skull-bearing (kāpālika) composer of a series of controversial performance songs that speak frankly, though also in profoundly symbolic terms, of his relationship with a low-caste woman (ḍombī). Yet he, too, is very difficult to locate precisely in time or place, is easily confused with others who bear the common names of Kāṇha or Kr̥ṣṇa, and may or may not be the author of both the Treasury translated here and the performance songs that have earned “Kāṇha” so much notoriety. Kāṇha's Treasury, like Saraha's, is not attested by an independent Apabhraṃśa manuscript but rather has been extracted from a later Sanskrit commentary, an anonymous work known as the Dohākoṣa-Mekhalā-Ṭīkā.8

Tilopa, the “sesame-pounder” (also known as Tillipa, Telopa, or Tailopa), is believed to have received four great tantric teaching streams. Some of the teachings stemmed originally from Saraha, and some of them were transmitted to Tilopa by actual or visionary female figures. He is said to have distilled those into twelve profound instructions that were transmitted, amid great trials, to his disciple Nāropa, who in turn taught his own “six topics” (Tib. chos drug) to the Tibetan translator Marpa (1012–1097).9 From Marpa the teachings passed to Milarepa (1040–1123), thence to Gampopa (Sgam po pa, 1079–1153), from whom nearly all later Kagyu lineages descend. Thus Tilopa is regarded by the Kagyu traditions of Tibet as the direct human source of many of their important practice lineages, including those connected with the tantras they considered the most advanced and effective of all, the Unsurpassed Yoga tantras (Tib. bla med rnal 'byor rgyud, Skt. yoganiruttara or yogānuttara tantras) and with the radical meditative techniques of the great seal (Tib. phya rgya chen po, Skt. mahāmudrā). Yet Tilopa, too, is a mysterious character who, just as we think we may approach, melts, like so many other siddhas, into the thicket of historical and textual ambiguity, where doubts remain about his dating, his authorship, and even his historicity. His Treasury, too, is unattested as an independent Apabhraṃśa text and also has been extracted from a later Sanskrit commentary, the anonymous Dohākoṣa-Pañjikā-Sārārtha-Pañjikā.10

(p.9) What We Do Know

This raft of uncertainties notwithstanding, there are some things we can assert with modest confidence about these three Treasuries and the men who are said to have composed them.

Language and Form

After some initial uncertainties, the language of the Dohākoṣas has been identified as an eastern dialect of Apabhraṃśa (sometimes called Avahaṭṭha), which seems to have been employed in Bihar and Bengal for several centuries before, and just after, the turn of the first millennium C.E. Apabhraṃśas—and they are plural, having earlier and later as well as western and eastern versions—were a group of Middle Indo-Aryan languages used in north India from approximately 300 to 1200 C.E. In the sequence of Indian languages, Apabhraṃśas generally fall after the Prākrits (e.g., Pāli, Māgadhī, and Śaurasenī)—which were themselves modifications of classical Sanskrit—and before the rise of the modern north Indian vernaculars, of which the later Apabhraṃśas are the immediate ancestors. The Apabhraṃśas were influenced both by popular speech and the late classical forms of Sanskrit, which, like Latin in the West, persisted as an elite literary language long after it ceased to be widely spoken. The late eastern Apabhraṃśa in which the Treasuries have been transmitted is related most closely to modern Bengali, though its echoes are evident in Assamese, Oriya, Mathili, Bihari, and certain forms of Hindi, too. It is important to note that while eastern Apabhraṃśa prefigures modern north Indian vernaculars in significant ways, and may have been related to the vernaculars of its time, it is—like all Apabhraṃśas—not a vernacular per se but a literary language. Thus, while eastern Apabhraṃśa is the language of the Treasuries, it probably is not the language of the original dohās, close as it may be.11

The particular poetic form that is most common in these texts is the dohā, which may refer, depending on context, to a meter or to a type of rhyming couplet dominated by that meter. The dohā gained great importance in the vernacular languages and religious traditions of late medieval north India (where it sometimes was called a sākhī), but its earliest exemplifications are found in the collections translated here, and it may derive from still earlier Sanskrit forms, such as the dodhaka. As a meter, the dohā is typified by, among other things, a strong end rhyme, a caesura midline, subdivision into smaller rhythmic units, and a close correlation between rhythm and meaning. Perhaps because of its rhythm and rhyme, the dohā frequently has been used as the vehicle for self- (p.10) contained aphorisms, and above all spiritual advice, whether by the likes of Saraha, Kāṇha, and Tilopa or later figures such as Kabīr, Nānak, or Dādū; it sometimes contains an authorial interpolation, a baṇa line, such as “Saraha says.”12 In this sense, the dohā is both a form and genre. As a form, it is also found in two other poetic styles attested in Apabhraṃśa literature, performance songs (caryāgīti) and diamond songs (vajragīti), which differ generically from dohās because of their different context and function: in performance songs, the couplets are linked together to form a larger unit of meaning, which we might reasonably compare to a Western “song” sung by a singer for an audience (though they might, too, be sung in ritual settings); diamond songs are “songs” in the same sense but differ from performance songs in their purpose: they cannot be understood except within the context of a tantric ritual feast, a “family circle” (gaṇacakra).13

If a dohā is a self-contained, often aphoristic, rhyming couplet, quite probably oral in its initial transmission, then the very idea of a collection or “treasury” of dohās (a dohākoṣa) is fraught with difficulties. Even if we did not possess multiple and partially incompatible versions of the Treasuries of Saraha, Kāṇha, and Tilopa—which (especially in the case of Saraha) we do—we might reasonably imagine that a text claiming to represent the “authorial intention” of a dohā singer was no such thing, for it is almost certain that various dohās (or groups of dohās; some clearly are related to each other) were uttered in various interactive and public settings, perhaps as the capstone to a religious discourse. Thus it is extremely unlikely that any of the Treasuries translated here—or other collections that are found only in Tibetan translation—actually has been redacted as it was performed.14 Indeed, we only can assert with confidence that when we examine the Treasury of Saraha, Kāṇha, or Tilopa, what we have before us is a later compilation by an editor who, for purposes of his own, brought together dohās or groups of dohās that had come to be associated with one or another of those names, names that might or might not once have denoted an actual person. In this sense, there is probably a considerable amount of arbitrariness built in to the compilation of any single Treasury, and though commentators on the texts find order and meaning in their arrangement (sometimes, in fact, it is they who have arranged them!), it is quite imaginable that the texts could have been ordered in many different ways and still been found meaningful by readers.

Content: The Yoginī Tantras as Background

In terms of content, the most historically significant feature of the three dohākoṣas is that each presents clear evidence that its author was familiar with, (p.11) and probably a practitioner of, the Yoginī tantras. These tantras, which include the Hevajra, Saṃvarodaya, Caṇḍamahāroṣaṇa, Mahāmudrāṭīlaka, Vajrakīlaya, CatuḤpīṭha, Buddhakapāla, and Kālacakra tantras and texts related to them, are one of the very last literary developments in Indian Buddhism, probably appearing no earlier than the eighth century and gaining importance only in the ninth and tenth centuries.15 They are related to, but distinguishable from, other late Indian Buddhist tantric textual traditions, including those of the Guhyasamāja, Yamāri, and Vajrapāṇi tantras, which frequently were designated Mahāyoga tantras. Taxonomists of tantra sometimes assigned the Mahāyoga and Yoginī tantras to separate classes and sometimes placed them together in a single, “highest” class of tantra, the Unsurpassed Yoga tantra.16 The Yoginī and other “higher” tantras are part of the broader class of Buddhist tantric texts, which began to appear in about the seventh century. All tantras purport to have been spoken by the Buddha in one or another form (most often the form for which the tantra is named), and tantras—along with their voluminous subsidiary literature—quickly began to form an important subset of the Mahāyāna Buddhist canon. By the end of the first millennium, tantra17 increasingly had come to dominate Indian Buddhist life and practice and, for that matter, to affect life and practice in nearly all Indian religious communities.

Thus, when Buddhism began to make significant inroads in Tibet in the eighth century, tantra already was an inescapable part of Indian Buddhism, and Tibetans regarded the Adamantine Vehicle of tantric practice, the Vajrayāna, as the most advanced of all the Buddha's teachings. All Tibetan Buddhists consider the Yoginī tantras to be at or near the pinnacle of the tantric path. The “old” translation school, the Nyingma, focuses primarily on tantras translated into Tibetan before the Yoginī tantras first arrived in Tibet in the tenth and eleventh centuries, but when it does incorporate aspects of the Yoginī tantras into its characteristic nine-vehicle taxonomy of Buddhist texts and practices, it generally places them in the penultimate category, Anuyoga—beyond the practices of the Disciple and Solitary Buddha paths, standard Mahāyāna, and various lower tantras but still below the apogee of the scheme, Atiyoga, whose texts and practices are the basis of the quintessential Nyingma system, the great perfection. While secondary in the Nyingma, the Yoginī tantras and their commentarial traditions are crucial to the tantric systems of latter-day (post-1000) Tibetan translation schools like the Kagyu, Sakya, and Gelug. Of Yoginī tantra texts and practices, the Kagyu tends to emphasize those related to Cakrasaṃvara (including the female Buddha Vajrayoginī), the Sakya focuses above all on Hevajra, and the Gelug concentrates on Cakrasaṃvara and Kālacakra. Whatever their preferences, all three traditions classify the Yoginī (or “Mother”) tantras—along with such “Father” (i.e., Mahāyoga) tantras as the Guhyasa (p.12) māja—among the Unsurpassed Yoga tantras that are the highest of the four sets of tantras recognized in their taxonomies (the others being Action, Performance, and Yoga tantras).18

Whatever their classification, and despite considerable differences among them with respect to both the outlines and details of theory and practice, the Yoginī tantras are characterized by a number of distinctive features, some of which are unique to them but many of which they share with other highly advanced tantras; with “lower” tantras; with Mahāyāna sūtras, commentaries, and treatises; with Buddhism in general; and with non-Buddhist Indian yogic traditions, including certain schools of Hindu tantra.

The Yoginī tantras have a number of ideas that are distinctive to them:

1. In practices related to reenvisioning the cosmos as a sacred realm, or maṇḍala, an important—and sometimes central—place is given to wrathful, naked, cremation ground–inhabiting female deities, most often called yoginīs or ḍākinīs. These figures are associated both with profound gnostic wisdom and extraordinary bliss—and the fusion of the two in reality, the mind, and tantric practice. The importance of these figures and their symbolism helps to explain why the Yoginī tantras are named as they are, and also are known as Ḍākinī tantras or Prajñā (i.e., wisdom) tantras.

2. In advanced yogas within the “subtle body” made of channels, winds, and drops, a number of different “seals” (mudrā) are identified that help to confirm and deepen one's practice. There are various listings of seals relating to various phases of practice; the most significant set, perhaps, refers to three major procedures: sexual yoga with a flesh-and-blood partner (karmamudrā), engagement with a visualized partner (jñānamudrā), and nondual contemplation of the nature of reality (the great seal, or mahāmudrā). In some Yoginī tantra systems, all are required, while in others only some or one may be applied.

3. A major function of subtle body yogas is to induce a sequence of up to four ecstasies (ānanda), which are linked to a variety of other fourfold patterns, for example, four tantric initiations, four moments on the path, four Buddha bodies, and so on. These ecstasies culminate in an enlightened awareness or gnosis—often referred to as “the innate” (sahaja) or the great seal (mahāmudrā)—in which the mind's natural purity and luminosity, its nondual realization of emptiness, and an experience of great bliss or ecstasy are indissolubly interfused.

The Yoginī tantras share a number of ideas with other highly advanced tantras:

1. Enlightenment only is possible through confronting and transforming such basic human emotions and events as passion, anger, death, and rebirth, and one must begin to do so during four highly demanding, profoundly sym (p.13) bolic, and in some cases sexually charged initiations received from one's guru, to whom one pledges absolute obedience.

2. The locus for real spiritual work is the human body, and more specifically the subtle body (sukṣma-śarira) that interpenetrates and is the basis of our coarse physical bodies. This body consists of seventy-two thousand channels (nāḍī), five to ten major breath-related energies (prāṇa), and a number of hormonal “drops” (bindu) inherited at conception from one's parents.

3. The work of transformation requires overcoming dualistic aversion to notions of pure and impure, a willingness on occasion to transgress conventional moral norms, and skillful manipulation of one's mind and energies so as to bring them to a standstill within certain nodes or centers (cakra) within the central channel of the subtle body. In order to harness one's energies, one must be willing on occasion to ingest “impure” substances such as alcohol, semen, and blood, and engage in sexual yoga practices, sometimes within the context of a tantric feast or “family circle,” a gaṇacakra. The result of controlling one's energies is the production—or revelation—within the central channel of a blissful, enlightened gnosis.

The Yoginī tantras have certain ideas in common with other Buddhist tantras:

1. One must practice here and now as if one were the Buddha-deity one someday will become, “making the goal the path” by reconstituting oneself out of one's fundamental emptiness as a sacred syllable, which becomes an enlightened being at the center of one's divine, deity-filled abode, the maṇḍala. Through this procedure, one lays the basis for overcoming the three “existential events” at the heart of saṃsāra—death, intermediate existence, and rebirth—and for achieving the dharma, enjoyment, and transformation “bodies” (dharmakāya, saṃbhogakāya, and nirmāṇakāya) of a Buddha.

2. One must employ a wide variety of ritual and meditative methods, including mantras, supplication prayers, material and immaterial offerings, and the practice of the extraordinary degree of one-pointed concentration required for seeing oneself and the cosmos in an entirely different way.

3. Tantric practice only is possible after initiation and instruction from a qualified guru, who is the key to one's access to tantric Buddha-deities, and to the lineage of gurus who have taught the practices that lead to persons' actually becoming those Buddha-deities.

The Yoginī tantras have a number of ideas in common with a broad range of Mahāyāna Buddhist movements:

1. The purpose of human (and all sentient) existence is to attain fully enlightened Buddhahood—consisting of dharma, enjoyment, and emanation bodies—at the culmination of the path of the “enlightenment hero,” the (p.14) bodhisattva, who sets out on an arduous spiritual journey motivated by compassion for the sufferings of all sentient beings, and the aspiration to free them all (bodhicitta).

2. In progressing toward enlightenment, one must employ a wide range of religious methods (upāya) for developing compassion toward all living beings, assisting beings with their worldly and spiritual problems, pleasing a multitude of powerful Buddhas and bodhisattvas, and cultivating one's own visionary experiences of those beings within their Buddha fields, or “pure lands.”

3. One must cultivate an approach to wisdom that emphasizes that the nature of all entities and concepts is nonduality, sameness, or, above all, emptiness (śūnyatā). This crucial term may variously (and sometimes simultaneously) be understood to mean that things simply lack any substantial, permanent, independent nature (as taught in the Madhyamaka school); that perceived phenomena cannot be differentiated from the mind that perceives them (as taught in the Yogācāra school); and/or that the naturally stainless and radiant mind behind all things is devoid of any of the defilements that appear to blemish it (as taught in literature on the Tathāgatagarbha, or “Buddha nature”).

The Yoginī tantras share a number of ideas with Buddhist traditions in general:

1. The essential human problem is repeated rebirth (saṃsāra), which is prompted by ignorance of the nature of things, a selfish craving for pleasure, and unskillful actions bearing inevitable effects (karma)—but which is capable of elimination (in a condition of peace, knowledge, and bliss called nirvāṇa) through proper conduct, the mastery of techniques of meditation, and direct insight into reality.

2. Those who seek spiritual freedom must begin by going for refuge to the Buddha, the doctrine he taught (dharma), and the spiritual community he founded (saṅgha). Buddhist life usually involves both active participation in a community of like-minded seekers and avoidance of the religious ideas and practices followed by outsiders to the community.

3. It is axiomatic that, in the opening words of the universally revered verse collection the Dhammapada, “all that we are is a result of what we have thought: it is founded on our thoughts, it is made up of our thoughts.” The implication of this is that as we see, so shall we be.

Finally, the Yoginī tantras share a number of perspectives with non-Buddhist Indian yogic traditions:

1. The saṃsāra-nirvāṇa cosmology is an accurate picture of the lot of all beings everywhere in the universe.

2. Through a combination of personal discipline, virtuous conduct, proper (p.15) meditation, and correct insight one may transcend the vicious circle of saṃsāra and attain the transcendent, gnostic bliss of nirvāṇa (or, in Hindu terms, mokṣa).

3. The best context in which to pursue spiritual liberation is a supportive community, ideally one built around an experienced and skillful guru.

More specifically, and quite importantly, the Yoginī tantras share with a number of more or less contemporaneous Hindu tantric traditions, whether Kashmiri or Bengali, a considerable number of technical terms, deities, yogic procedures, and social perspectives, to the point where it sometimes is quite difficult to know who has influenced whom, or, as suggested earlier, who belongs to which tradition.19


What all this, then, allows us to assert generally of the Treasuries and their authors is the following. Given their eastern Apabhraṃśa language, the dohās of Saraha, Kāṅha, and Tilopa probably were composed around the end of the first millennium C.E. in northeast India, while the “Treasuries” of dohās that have come down to us are probably somewhat later, and of less certain provenance. Given the dohās' frequent references to matters germane to Mahāyāna Buddhism in general and the Yoginī tantras in particular, their authors are more likely to have been members of a tantric Buddhist comminuty than any other. Knowing that these three dohā composers were familar with, and probably practiced, the Yoginī tantras, we may accept as poetically apt, if not historically demonstrable, the traditional stories that associate them with rejection of conventional social norms, and with resort to forbidden places and practices, including association with female companions. It is at least conceivable that there was a siddha (let us call him Saraha) who learned tantra from a an arrow-making yoginī, that there was another siddha (call him Kāṇha) who kept a low-caste mistress, and that there was still another siddha (call him Tilopa) who received visionary instructions from a ḍākinī. We also may deduce that our authors' seemingly syncretic view of the world—incorporating elements of various tantras, Mahāyāna Buddhism, basic Buddhism, and even Hinduism—simply reflects the creative blend of ideas and practices that make up the weave of the Yoginī tantras. Indeed, it was that creative blend, and their poetic and aphoristic manner of expressing it, that permitted Saraha, Kāṅha, and Tilopa (and other siddhas) to popularize the Yoginī tantras in India (where their echoes still are heard), and their successors to pass them on to Buddhists in Nepal and Tibet, who remember the siddhas to this day as the great progenitors of their traditions of spiritual practice—to the point where their status as flesh-and- (p.16) blood humans or mythological constructs hardly matters. Besides, somebody wrote the dohās, and it's as probable that those somebodies were real individuals named Saraha, Kāṅha, and Tilopa as that they were not.

Common Themes

I hope that the dohās translated here will to some degree speak for themselves, but our appreciation and understanding may be enhanced if I touch at least briefly on six important themes that run through all three collections: (1) a rhetoric of paradox, (2) cultural critique, (3) focus on the innate, (4) affirmation of the body, the senses, and sexuality, (5) promotion of certain yogic techniques, and (6) celebration of the guru.20

A Rhetoric of Paradox

Saraha claims that he is declaring everything plainly and holding nothing in secret (S92–93)—yet, like so many texts that belong to what we might call “Asian wisdom literature” (including many Upaniṣads, the Buddhist Perfection of Wisdom sūtras, the classics of “philosophical” Taoism, and the great textbooks of Chan Buddhism), the Treasuries of Saraha, Kāṇha, and Tilopa are difficult to understand without a prior recognition of the rhetoric of paradox that their authors often employ. The hallmark of this rhetoric is the simultaneous or successive assertion of two facts that appear to be mutually contradictory, as in the Keṇa Upaniṣad's claim (2.3), regarding brahman, that “it is conceived of by whom it is not conceived of. … It is understood by those who … understand it not,” the Heart Sūtra's famous affirmation that “form is emptiness, emptiness is form,” the Tao Te-ching's insistence (chap. 4) that the “Tao is empty … but its capacity is never exhausted,” or the Chan Gateless Gate's statement at one time that “mind is Buddha” and at another that “there is no mind, there is no Buddha.”21 Such statements are juxtaposed in part in an attempt to force the reader or listener to push beyond the limits of conceptual thought, which is a bugaboo for almost all the traditions involved. There is more, though: one must understand that apparently contradictory statements often are asserted on two different levels of discourse, what Buddhists usually refer to as the “conventional” and “ultimate.” What is true of conventional reality may not be so from an ultimate standpoint, and vice versa, yet each has its particular claim to being “true.” Thus, to say that brahman is understood by those who understand it not is not simply to indulge in absurdity: it is to assert that “real” or “ultimate” understanding of brahman only is possible for (p.17) those who go beyond conventional, conceptual understandings of it; to maintain the identity of form and emptiness is to stress the inseparability, and mutual dependence, of apparently polar ideas; to claim that the empty Tao has inexhaustible capacity is to underline the fact that that which has no fixed nature of its own may become anything at all; while the successive insistence that mind is and is not Buddha shows that, from a certain perspective, the mind is the basis of our enlightenment but from another, even the mind, even Buddha, has no substantial existence, and the recognition of this fact will allow the mind to become Buddha. These cursory glosses by no means fully explain the paradoxes that have been offered—it is not that simple—but they do, I hope, demonstrate that interpretation is possible.

As heirs to the discursive style of Mahāyāna Buddhism (and perhaps Upaniṣadic Hinduism as well), Saraha, Kāṇha, and Tilopa frequently express themselves paradoxically, either within a single dohā or between one dohā and another, and the key to understanding at least some of what they mean is to be found precisely by recourse to an analysis in which “levels of truth” are parsed. Thus, when Saraha asserts of “the real” (or “that”) that “there is no seeing / that doesn't perceive it—/ but it's witnessed solely / at the precious guru's feet” (S17), he may be trying to prod our minds beyond conceptual categories, but he also means to say that although ultimate reality is visible everywhere, the truth evident in every moment and situation, we do not really see things that way unless we are instructed on how to do so by a qualified spiritual master. When he says that “your innate nature / neither exists nor doesn't” (S20), he certainly is indicating the inability of conventional thought to capture the ultimate, but he also is pointing out that, from the perspective of one who sees things as they are, one cannot say that the innate (or anything else) has true, substantial existence, while to assert that conventionally we do not possess such an innate, pure nature is to misunderstand our real spiritual potential. When Kāṇha instructs us to “go outside, look around, / enter the empty / and the nonempty” (K11), he certainly is presenting an absurd image of “going outside” to find emptiness and nonemptiness yet can be read, too, as suggesting that we must thoroughly explore the world around us in all its possible permutations if we are to understand it, and ourselves, aright. And when Tilopa asserts that “the precious tree / of nondual mind … bears compassion flower and fruit, / though there is no other / or doing good” (T12), he may be suggesting the difficulty of imagining compassion in a mind freed from dualities, but he also is noting that it is precisely because of the nondual mind's nondistinction between self and other (or even good and bad) that true, nondiscriminating compassion may occur, for compassion that involves discrimination falls short of what compassion can be.

(p.18) Similarly, in the course of their collections of dohās, Saraha, Kāṇha, and Tilopa seem to express contradictory views of many important concepts or practices. Thus, Saraha may instruct his listeners in one place to bow down to mind, which is “the single seed of everything” (S41), and in another to “abandon mind” (S57). Again, it is the context in which a term is being used that determines the meaning, and assures coherence: conventional, conceptual mind must be overcome if freedom is to be attained, but the mind that is clear of illusion is the experience of enlightenment, and is in some sense the foundation of all there is, hence a worthy object of worship. Or Saraha may ask in one place “if enjoying things intently / doesn't free you, / … how can consciousness be free?” (S19) yet warn in another “don't bind yourself / to sensuous things” (S71). He may have some ambivalence about the sensuous world (that would be a typically modern interpretation), but it is likelier that he is insisting, in the spirit of many Mahāyāna and tantric masters, that one must, at one and the same time, both involve oneself in the senses and maintain detachment from the traps they lay for beings who are in their thrall—that is be in the world but not of it. Or Kāṇha may instruct in one place; “the maṇḍala circle: / knowing how things are, / understand it” (K9), while claiming later: “free from the maṇḍala circle, / I live in / the innate moment” (K18). He appears simultaneously to celebrate and denounce a crucial aspect of the tantric path, the visualization of the maṇḍala, with oneself at the center as a Buddha-deity, yet if the tantric path is admitted to have different stages, for example, the steps of generation and completion, then what is appropriate at an earlier phase (maṇḍala circles) may be largely irrelevant later on. Or, finally, Tilopa may assert in one dohā “I am the cosmos, I am the Buddha, / I am the unadorned” (T16) and then claim in another “I am empty, the cosmos empty, / the triple world empty, too” (T34). The statements appear to be irreconcilable, yet if one understands that from an ultimate point of view emptiness is the nature of all things, then I am all there is in terms of their real nature, if not their conventional particulars; alternatively, if all there is is a result or a function of a pure, empty, blissful innate gnosis, and that is what I am in the most fundamental sense, then I am indeed all things.

These sorts of contradictory statements—now asserting, now denying—are made about many other important ideas, too, ranging from meditation, to emptiness, to nirvāṇa, to tantric practice itself, but in nearly every case the paradox is comprehensible when we think in terms of the different levels of discourse that are being applied in particular contexts. Again, I do not mean to limit interpretation solely to the strategy I have suggested here, for such a strategy is only one among a number that are possible, but I do think that (p.19) Saraha, Kāṇha, and Tilopa make more sense in light of such an approach—soundly based in Mahāyāna Buddhist tradition—than they do without it.

Cultural Critique

If there is a type of negation in Saraha, Kāṇha, and Tilopa that seems rather less mitigated by paradox or multiple levels of discourse, it is that related to the three siddhas' attitudes toward social and religious convention, which are aptly summarized by Groucho Marx's famous claim, in Horse Feathers, “Whatever it is, I'm against it.” In all three collections of dohās, but especially those of Saraha, there hardly is an Indian personage or practice of social or religious importance that is not subjected to mockery or critique. In the first fifteen verses of his Treasury, Saraha takes on, in succession, brahmin ritualists (S1–2), who are denounced for their pointless recitations and sacrifices; mendicant ascetics (S3–5), who are mocked for their deceit, hypocrisy, and greed; Jain renouncers (S6–9), who are ridiculed for their obsession with physical austerities; Buddhist monastics (S10), who are chastised for their dress-up games, their intellectualism, and their attempt to desiccate the mind through meditation; Mahāyānists (S11), who are described as sophists and verbal gymnasts; tāntrikas (S11, 14), who are said to be obsessed with mantras, maṇḍalas, and mystic initiations; and practitioners of all kinds (S14–15) who are lampooned for believing that offering lamps or food, going on pilgrimage, or immersing themselves in sacred rivers can purify them of defilement. In other places, Saraha criticizes the self-deception of the alchemists (S51), the absurdities of devoted meditators (S19–20, 22–23, 33, etc.) and the pretensions of scholars (S68, 76, 93) and makes evident his distaste for distinguishing pure and impure on the basis of caste (S46, 56b). Saraha also repeatedly addresses his listener as a “fool” and refers disparagingly to “bestial” or “childish” people who simply don't understand what is right in front of them. In a similar vein, but less extensively, Kāṇha and Tilopa, at various points in their dohās, criticize intellectuals (T8), scholars of sacred literature (K1, 29), people who make offerings to deities and visit pilgrimage spots (T19–21), and tantric ritualists obsessed with “chants, oblations, / and maṇḍala rites” (K29). The social and religious outlook of all three is perhaps best captured in Saraha's injunction: “Throw off / conventional nonsense” (S55).

There are a number of ways in which we might regard these thoroughgoing social and religious critiques, which do not seem to conform with the ecumenical spirit typified by so many Buddhists today. We must recall that, as irenic as it often has been, Buddhism did begin as a movement motivated at (p.20) least in part by a powerful sense of the inadequacies of the social and religious systems of the mid–first millennium B.C.E., and that satire, caricature, and scorn, along with pointed philosophical criticism, have been part of the Buddhist rhetorical arsenal for a very long time; in this sense, Saraha, Kāṇha, and Tilopa simply are upholding a long-standing Buddhist tradition of pointed inquiry into the ideas and practices of others.22

This would help to explain the criticisms of non-Buddhists, and even perhaps of various nontantric Buddhists (for intra-Buddhist polemics have been as vigorous as those directed against “outsiders”), but it does not fully explain the extension of the critique to those who practice the very tradition the siddhas are expounding, the way of the tantras. Here, perhaps, the intent is, in a manner consonant with that of the Mahāyāna Perfection of Wisdom sūtras, to undermine any basis for clinging to any idea or practice that might be taken as valuable in itself, rather than as merely instrumental to true awakening, to root out the complacency that Chögyam Trungpa Rinpoche so aptly labeled “spiritual materialism.” Put another way, any concept or technique that is said to lead to freedom through externals of any sort is doomed to failure, for it will have fallen short of touching on the true locus of the perpetuation of saṃsāra or the attainment of nirvāṇa, namely, the innate, nonconceptual, blissful, empty, originally pure mind that is our inmost nature. This innate mind is so internal, so basic, that even meditative practices that seem to point within us are inadequate for its attainment—let alone more obviously outward acts such as offering, chanting, mortifying the flesh, or going on pilgrimage.

In the final analysis, anything that leads to the experience of the innate may be celebrated, even if it appears to contravene accepted ideas and practices; and anything that hinders one from reaching it must be rejected, even if it is a hallowed idea, a well-trodden religious path, an unquestioned class distinction, or an immemorial social taboo. And, as we might expect, a practice that is conducive to experience of the innate in one context may be deleterious to it in another, so we should not be surprised that Saraha, Kāṇha, and Tilopa may, in fact, have done many of things that they criticize others for doing, including performing rituals, going on pilgrimage, engaging in meditation, and, of course, entering into tantric practice—but if they did these things in order to gain the experience of the innate, or from the standpoint of a nondualistic achievement of the innate, then there is no blame attached to their actions. As earlier, the context of critique and affirmation are crucial, and since most people are easily deceived by concepts, external practices, and the weight of tradition, it is understandable that the siddhas criticize these more often than they affirm them.

(p.21) Focus on the Innate

Whether or not Saraha, Kāṇha, and Tilopa really equivocate in their criticism of social and cultural traditions, there can be little doubt as to the great theme they do affirm in their songs: it is the innate, sahaja. In all three, it is the most common term for the ultimate, appearing a total of thirty-one times, far more often than such well-known designations as nirvāṇa, Buddha, emptiness, or the real. Indeed, the term is common enough in various Apabhraṃśa and Sanskrit Buddhist texts of the late first millennium (though especially in the Yoginī tantras and the songs of the siddhas) that early twentieth-century scholars designated the siddhas and others who sang of it “Sahajiyās” and the movement they were believed to represent the “Sahajayāna.” The term sahaja has received a variety of translations over the years, including “the Innate,” “the Together-born,” “the Simultaneously-arisen,” “the Spontaneous,” “Coemergence,” “Connate,” “Complementarity-in-Spontaneity,” and “Being.”23 In its various usages in Indic languages and in Tibetan, it probably denotes or connotes all of these, and more. In common parlance in modern north Indian languages, it refers to what is easy or natural, what one does spontaneously. In more religious or philosophical settings, it may refer to opposing qualities that emerge together, or are held in conjunction, such as existence and nonexistence, or the realization of emptiness and the experience of great bliss. And, in its most technical sense, it designates the most sublime of (usually) four “ecstasies” (Skt. ānanda) attained through the practice of the Yoginī tantras, that in which one's inmost nature is revealed, and enlightenment approached or attained. I translate sahaja as “the innate” because I think that this captures best the sense it seems to have for Saraha, Kāṇha, and Tilopa, namely, of what is most natural to us, what is most fundamentally our own, what is at the root of our experience of the world—and perhaps of the world itself.

However general or technical the siddhas' usages of “the innate” may be (and, in a given instance, it may be one or the other, or both), it clearly is a major way of describing simply what is most important and most basic—that is, what is ultimate—for those with spiritual aspirations. It is not, however, the only way of describing the ultimate: Saraha, Kāṇha, and Tilopa appear to use it more or less synonymously with a number of other important terms. These include: “great bliss,” “stainless mind,” “inmost nature,” “the real,” “great delight,” “tasting the same,” “that,” “knowledge,” “self-awareness,” “the great seal,” “rapture unceasing,” “the yoginī,” “the union of wisdom and method,” “the real nature of thought,” “nondual mind,” “emptiness,” “nirvāṇa,” “awakening,” “Buddha,” “the Thus Gone,” “the thought jewel,” “the profound,” (p.22) “the singular,” “the bodiless,” “the self,” “the unadorned,” “the utmost power,” “the single god,” “nonthought,” “perfection”—and, of course, “the ultimate.” There also is a wide range of metaphors that are applied to it, including the sky, the ocean, a mountain, a cave, an elephant, a jewel, a tree, a lotus, a king, home, a lamp, a seed, and a ripened fruit. Furthermore, a multitude of characteristics are attributed to the innate: it is single, motionless, ever the same, pure, stable, pacified—and beyond both virtue and vice, good and bad, self and other, mind and nonmind, empty and nonempty, existence and nothingness, saṃsāra and nirvāṇa. So many terms, so many images, so many predications for a concept reputed to be ineffable! For, as Tilopa remarks, “the real / can't be shown / by the guru's words” (T9a), and the others echo him in various ways, referring to the innate or its synonyms as “something that can't be described” (S52), “without syllable” (S58), “hard to approach” (K15), and most decidedly not “what falls within the range of thought” (T9).

Keeping this caveat in mind, we still may make a number of conventional observations about the ways in which Saraha, Kāṇha, and Tilopa think of the innate and its cognates. The first is that, in line with longstanding Indian ways of conceiving the ultimate, it is described both negatively and positively. While Buddhism has a reputation for characterizing reality primarily in negative terms (e.g., as nonself, or emptiness), in fact, from the earliest scriptures on, it has provided many positive terms for the ultimate, too, from specifications in the Pāli nikāyas that nirvāṇa is not just extinction but also in some sense involves knowledge and bliss; to claims in Mahāyāna that the mind, whether in potential form as Buddha nature (the matrix of the Thus Gone: Skt. tathāgatagarbha) or in fruition as the Buddha's dharma body (dharmakāya), is, though empty by nature, nevertheless replete with all possible virtues; to descriptions in tantric texts of both the nature of reality and the outcome of the path as the adamantine (vajra) body, speech, and mind, or the gnosis in which realization of emptiness and experience of great bliss are inseparable. Buddhists will disagree among themselves as to how literally this or that characterization of the real is to be interpreted, and it is worth underscoring the fact that for every thinker who sees emptiness as a simple negation of any metaphysical self or substance anywhere, there usually will be another who understands it as negating self or substance in worldly entities, while at the same time implying the enlightened mind's ultimate emptiness of any delusion or stain (i.e., anything other than it) and its maximal possession of all possible virtues. I would not claim with certainty that either of these perspectives—which in Tibet were termed the “intrinsic emptiness” (rang stong) and “extrinsic emptiness” (gzhan stong) views—is necessarily that of Saraha, Kāṇha, and Tilopa, but there is little doubt that tantric literature in general is very amenable to the “extrinsic emptiness” ap (p.23) proach, while still being interpretable from the “intrinsic emptiness” standpoint.24

The second observation is that the innate seems to possess characteristics that are both ontological and epistemological. In other words, it is “the real” (Skt. tattva), that which is most “objective,” true, or fundamental about what there is, but it also involves what, for want of a better term, we might call “experience,” a “subjective” apprehension of things that is associated with a variety of terms connoting mentality, including “mind” (Skt. citta), “thought” (manas), “comprehension” (buddhi), and knowledge, or gnosis (jñāna). One of its most common descriptions in tantric literature, especially in the traditions of the Yoginī tantras, is as a gnosis in which the realization of emptiness and an experience of great bliss are combined—and it is not just “knowledge” and “feeling” that are combined but the perception and what is perceived, so that the mind that sees emptiness also is emptiness, and the bliss that is felt by the subject also is a quality of objects that are apprehended. In a Western philosophical context, this conflation of ontological and epistemological might be seen as a category mistake of the first order; in India, however, at least from the time of the Upaniṣads on, it has been common to predicate of the ultimate (to the degree that one could predicate anything of it) that it is the most of whatever is positive, including reality, knowledge, and happiness, or bliss. Hindus maintained that brahman's nature was reality-knowledge-bliss (sat-cit-ānanda), while Buddhists insisted that nirvāṇa and/or Buddhahood involved a “permanence” lacking in worldly events, a supremacy of knowing (perhaps even omniscience), and the greatest possible pleasure. Hence, in an Indian setting (as perhaps in a theological context in the West), it made perfect sense to say that the ultimate was both objective and subjective, ontological and epistemological, for part of what made it ultimate was its transcendence—or combination—of dualities that are irreconcilable on a conventional level.

A third and final observation about the innate is that it seems to be conceived in a manner that is at once psychological and cosmological. Clearly, if it is to be identified with “the mind,” at least in its purified condition, it is generally what we might call a “psychological” concept; yet it is not merely a psychological notion, for Saraha, Kāṇha, and Tilopa all clearly believe that the mind is not just a subjective capacity that this or that individual possesses, through which the world might be “known,” with or without the help of concepts. No, the awakened mind is far greater than that: it is the “single seed of everything” (S41), from which “you'll emit / the whole triple world, / then draw it back again” (K17), where “things rise and … set” (S88a), and “all is forever fixed” (S103), and which, though “drained of all color, / lacking a (p.24) shape,” is nevertheless “fulfilled in every appearance” (T32). In short, the mind, hence the innate, seems to function as a cosmological—even cosmogonic—principle, serving as the source and substance of things as surely as it is a capacity for the cognition of things. In this respect, again, the tantric authors of the dohās belong to a well-established tradition in Indian religious thought, wherein there is a close relation between the microcosm of human mental activity and the macrocosm of the creative processes of the cosmos. Whether they claim, as do many Hindus, that our individual psychology is an aspect of an ultimate reality that in some sense is or has “mind,” or maintain with various Buddhists that the cosmos is affected or even effected by our thoughts, Indian philosophers repeatedly have recognized the relation between ontogeny and cosmogony, between psychology and cosmology—and Saraha, Kāṇha, and Tilopa appear to accept the relation, too. How “strongly” they intend the relation is another question. Just as there are interpreters of the positive images of the innate in the dohās who will insist that all such imagery must be taken as merely metaphoric (since ultimately, reality only can be described negatively, as empty), so will there be interpreters of the dohās' cosmopsychological language who will take it symbolically, as, for instance, merely indicating the importance of mind in the world rather than mind's substantive, magical creation of it; or indicating that the mind's identity with the world is merely in terms of the emptiness that is the common nature of all things and minds. As earlier, multiple readings are possible—but also as earlier, the dohās probably lend themselves more naturally to the “literal” reading, even while leaving the figurative readings open, as well.

However we regard the innate—whether we accept its ineffability, try to resolve its complexities, or simply see it as a classic Indian attempt to articulate the ultimate in paradoxical terms—it remains the crucial term in the lexicon of Saraha, Kāṇha, and Tilopa, for it is what, if we are to be free, we must attain, realize, experience, become. In a certain sense, of course, everything is it, and everything is already awake—yet most of us maintain at least the illusion of error and pain, and until we are rid of that, until we understand who we really are and always have been, we will remain in bondage. Thus, though the innate already is perfected in all of us, we still require methods of one sort or another to help us realize this fact, and recover what never has been lost. The two most important methods prescribed by Saraha, Kāṇha, and Tilopa are (1) an affirmation of the body, the senses, and sexuality, and (2) the promotion of certain yogic techniques—though, as I will show, they do not prescribe either method without noting some important contraindications.

(p.25) Affirmation of the Body, the Senses, and Sexuality

India is famous for having developed some of the world's most extreme forms of asceticism. As early as the time of the Buddha, and probably before, various wanderers and forest dwellers sought to overcome mortal afflictions by conquering, and ultimately transcending, the body. To do so, they deprived themselves of food, shelter, and clothing, stopped their minds and senses, and in some cases went as far as standing in a single spot for years on end or fasting unto death. We are told in canonical texts that the Buddha, after rejecting the hedonism of royal palace life, pursued such self-mortification with great zeal but in the end rejected that way of life as well, arriving at a “middle way” between the two extremes, which found its expression primarily in the rules of conduct for Buddhist monastic orders. While moderate by ancient Indian standards, however, the rules of Buddhist monastic life, and the attitudes informing them, clearly reflect a considerable suspicion of the body, the senses, and sexuality, all of which are said to feed desire—and desire, or craving, or attachment, was believed by the Buddha to be a primary cause for the continuing rebirth in saṃsāra that he saw as the basic problem of sentient existence.

Such an attitude is predominant in the canonical and postcanonical literature of non-Mahāyāna schools and to this day affects the normative Theravāda outlook that is the “official” Buddhism of Sri Lanka and most of southeast Asia. And although a few Mahāyāna texts suggest that a bodhisattva's compassionate action in the world may require engagement with the body, the senses, and even sexuality, or that a bodhisattva with perfect wisdom will not distinguish between pure and impure, or worldly and transcendent, in most Mahāyāna sources the traditional suspicions remain firmly in place. It is only with the rise of tantric movements in the late first millennium C.E. that a real shift in Buddhist values becomes evident. As I have shown, tantric movements in general believe that the human body is a crucial locus for spiritual work and that the various energies and emotions housed within that body may, if properly reenvisioned or transformed, serve as causes of enlightenment rather than hindrances to it. I also have shown that the more “advanced” tantras, including the Yoginī tantras practiced by many of the siddhas, may incorporate sex, and other types of religiously unconventional behavior, into the Buddhist path. Tantric attitudes and actions are justified by their theorists and practitioners as a logical extension of the nondual outlook and commitment to skillful religious methods that are so much a part of the Mahāyāna—but that only makes them seem slightly less at odds with a puritanical code that held sway in India for so long.

Certainly, Saraha, Kāṇha, and Tilopa are quite explicit in their insistence (p.26) that the body, the senses, and even, in certain contexts, sexuality are to be embraced rather than shunned by those who seek freedom. One of Saraha's major critiques of the Jains is that they seek to gain freedom by discarding the body (S9), when, in his own experience, “I've seen / no place of pilgrimage / more blissful than the body” (S48). He criticizes the scholar who does not know that “Buddha / dwells within his body” (S68), and instructs, as well, that “what's bodiless / is hidden in the body: / know this, and you're freed there” (S89). Kāṇha, similarly, sings of gaining “awakening / in this body” (K29) and specifies that “natural bliss is seen / … in your inmost body” (K3)—probably a reference to the subtle body that is so crucial to success in practicing the Yoginī tantras.

It is not just that the body is to be accepted and understood; the senses that are part of it must be used as well. Saraha instructs his listeners to “look to the senses—/ there's nothing / I'd exclude from them” (S43a); and, addressing one who is “beyond all renouncing,” he asks, “if enjoying things intently / doesn't free you, / … how can consciousness be free?” (S19). In the same spirit, Kāṇha invites us to “indulge in play” and asks why we should not “entertain / the five senses” (K28), while Tilopa points out that “aggregates, elements, / fields, and the senses—/ all are bound up / in your innate nature” (T1).

Not only may the senses be used and embraced but sexuality itself, often seen as the most destructive manifestation of sense desire, is acceptable, too. Saraha refers casually to “living at home with your wife” (S19); includes among the practices that “perfect the transmundane” “eating and drinking, / enjoying bliss” (S24); devotes a number of verses to celebration of a yoginī who may well be his female companion (S83–87); and asks pointedly (referring to tantric symbols for the vagina and penis): “delighting in / the rapture between / lotus and vajra / … who in the triple world / could not have their hopes fulfilled?” (S94). Similarly, Kāṇha asks a “maiden” of his acquaintance, “without your ceaseless passion, / how will I gain awakening?” (K29) and Tilopa instructs his listeners: “don't disparage the physical woman” (T25)—the action seal (karmamudrā) through sexual practices with whom enlightenment may be hastened—and adds that sublime experience is possible in “a passionate woman's embrace” (T27a).25

All this might seem to add up to clear proof that the siddhas, as both ancient and modern critics have claimed, are guilty of hedonism and antinomianism, their quest for the innate fueled by indulgence in pleasures that more sober and upstanding Indian spiritual traditions have found to be the path to ruin—an indulgence that they cannot or will not counteract because it is part of their conception of the ultimate that “in it, there is / no vice or virtue at (p.27) all” (K10). It is not nearly so simple, though, for scattered throughout the Treasuries are enough echoes of traditional Buddhist cautionary rhetoric that we cannot assert that their attitudes toward the body, the senses, and sexuality are entirely unambiguous. Saraha, for instance, speaks of a time when “the body's bonds are broken” (S46), “senses subside” (S29) “desire indeed is destroyed” (S50), or “all appearance / collapses before your eyes” (S74a); and, citing various desire-entranced animals as negative examples, instructs his listeners: “Don't bind yourself / to sensuous things” (S71). He also cautions against indulgence in sex—especially in a tantric context—without proper insight and discipline, asking, “If you don't grasp / everything as it is, / how, in the midst of sex, / will you perfect great bliss?” (S91). Similarly, Tilopa speaks of a point on the path when “senses and their objects / no longer appear” (T5). More broadly, all three siddhas make it clear that ethical distinctions do matter, as when they criticize pseudosages who “don't know right / any more than wrong” (S3) or exalt such classic Buddhist virtues as compassion and generosity (S15ab, 107–109, 112; T2, 12).

Are we then to believe that Saraha, Kāṇha, and Tilopa really are just good old-fashioned Buddhists, maintaining traditional suspicions of the body, the senses, and sexuality, promoting an ascetic ideal and a transcendent ultimate? The answer, unsurprisingly, seems to be yes and no. There certainly are enough cautionary indications in their dohās to suggest that the siddhas did not seek to obliterate all ethical distinctions and, furthermore, understood very well that the body, the senses, and sexuality really could lead one astray and needed to be controlled if one was to attain enlightenment. Unlike their more traditional predecessors, however, they did not seek control through rejection or avoidance but rather through active engagement guided by wisdom, skill in various methods, and—crucially—the advice of an experienced guru. Indeed, there are a number of songs where the ideal of being “in the world but not of it” is made fairly explicit, as when Saraha describes a yogin who is “enjoying things, / unstained by things. … / untroubled by things, enjoying things” (S64), or another who “seized / by the elephant trunk of senses,/ … seems to die, / but … / like a skillful trainer, / escapes and goes away” (S101); and when Tilopa instructs his audience: “Like a poison expert / partaking of poison, / delight in existence / but don't get hooked on existence” (T24).

There is, thus, a certain balance that the practitioner must strike. Virtue, discipline, detachment: all are required, and if they seem to be denigrated, that is because, when one adopts the perspective of the ultimate, they do not really exist; conventionally, however, they are a part of what one must practice if one is to experience the ultimate, for as Saraha reminds his listeners, “not comprehending the innate, / you're in the grip of vice” (S63). Once one has compre (p.28) hended the innate, the ultimate, it turns out that it is a condition that itself transcends the classical distinction between turbulent saṃsāric existence and a quiescent nirvāṇa, what is referred to in Mahāyāna as “nonabiding nirvāṇa” (apratiṣṭhanirvāṇa): Saraha describes the nondual mind as “free from existence—/ and from nirvāṇa” (S110), Kāṇha tells of a state where “delighting in existence, / you'll still perfect nirvāṇa” (K22), and Tilopa urges his listeners to “honor Buddha / by nonconceptual mind; / Hey! Don't get stuck in existence—/ or nirvāṇa” (T22).

In short, the body is the locus of spiritual work, and one should not seek freedom anywhere but within it (or the subtle body that is its “inmost” aspect)—yet one should not simply indulge it; the senses are to be entertained, for they are no less (or more) real and important than less tangible aspects of the world—yet they cannot be allowed to master one; and sexuality may be used in the pursuit of the spiritual, for it is an intensive expression of many of our (and the cosmos's) most fundamental energies—yet one must be sufficiently disciplined in practicing it that it will not lead to a deterioration of compassion or wisdom. And, as I now will show, the way one maintains the balance required for living “naturally” in the world, sensuous and spiritual at the same time, is through certain yogic practices.

Promotion of Certain Yogic Techniques

As is evident from the commentaries written on them, the Treasuries of Saraha, Kāṇha, and Tilopa have been regarded by Buddhist practitioners in India, Nepal, and especially Tibet not just as works of poetic or philosophical genius or incisive social commentary but as systematic guides to practicing the Buddhist path, and in particular to describing and promoting certain yogic techniques that are conducive to enlightenment.26 Before we concede the truth of this assumption, however, we must acknowledge that in none of the three texts are verses of advice on yogic practice laid out in a completely straightforward manner; rather, they come up here and there, interspersed among other verses with other concerns, which makes it difficult to ascertain what the priorities of the author must have been—especially when we recall that the Treasuries reflect choices most likely made by editors rather than authors, and the connection between a collection of dohās and an individual author is tenuous at best.

Furthermore, all three collections contain what we might regard as “antiyogic” rhetoric, which may leave us further confused as to what the siddhas were promoting and what they were denigrating. There are, for instance, numerous passages in which Saraha warns his listeners that meditation (Skt. (p.29) dhyāna) and contemplation (bhavanā), which are quintessentially yogic exercises, are actually worthless: “You're deceived by meditation, / so why meditate?” (S22); “mind is unstained—/ don't taint it with meditation” (S23); “meditation: / why look for freedom in a lie?” (S33); “the whole cosmos / is deluded by meditation” (S35). Saraha also questions the value of the breath control that is essential to so many yogic practices, wondering in one place what a yogin who has succeeded in stopping the breath will do when death time comes (S66) and in another instructing a “wretched yogin”: “Don't hold your breath / and think on yourself; / … don't focus in / on the tip of your nose” (S44). Rather more surprisingly, given my earlier claim that the siddhas were almost certainly practitioners of the Yoginī tantras, there are as well a number of passages that disparage the yogic techniques most associated with tantra, such as repetition of mantras (S14, 23, 39), visualization of the maṇḍala circle of deities (S11, K18, 29), and, as I have shown, sexual yoga (S94).

These passages do pose prima facie problems for an interpreter who wishes to see the Treasuries as guides to meditation or other yogic techniques, but the problems are not insurmountable. We must recall that in any Buddhist wisdom text a negation, whatever the object, may at times be conveyed solely at an ultimate level of discourse, such that, for instance, when Saraha criticizes meditation or maṇḍala practice, he does so in order to “empty” them of any substantial existence we might attribute to them or any attachment we might develop toward them. Such an ultimate level of negation does not mean that, in a conventional sense, one does not meditate or visualize a maṇḍala. Quite the contrary, it most likely is through yogic practices of one sort or another that one will truly learn how to overcome attachment to entities instinctively taken to be substantially existent. We also must recall that, in the complex sequence that is any individual's Buddhist path, practices that are necessary at one point may be counterproductive at another: there are disciples for whom meditation or maṇḍala practice may be useful at one time but for whom at other times the same practices may turn into dead ends that must be transcended. Many of the dohās collected in the Treasuries may first have been given as a particular piece of advice for a particular audience in a particular context, but we no longer know that context, so we simply juxtapose statements that seem, at times, to contradict each other.

In fact, the preponderance of the evidence from the three Treasuries indicates that Saraha, Kāṇha, and Tilopa all did promote specific yogic practices, which were believed crucial to the listener's ability to attain the ultimate. There are several yogic themes that appear repeatedly, not all of them consistent at first glance, including exhortations to: (1) control, or even stop, the mind and/or breath; (2) relax the mind, or simply see it as it actually is; (3) meditate with (p.30) the aid of mantras and maṇḍalas; (4) understand the subtle body and its component parts; and (5) induce—perhaps with the aid of sexual yoga—a series of four momentary ecstasies, of which the “innate” is the highest.

Stopping Mind

Saraha advises his audience several times to bring thought and/or breath to a standstill, urging, for instance, “where thought and breath / no longer roam, / and sun and moon don't shine—/ there, fool, repose / your mind!” (S25) and describing a point when “winds enter mind … [and] utmost bliss / can't be contained” (S42c). Kāṇha sings in symbols of how “thought / can't possibly escape / when motionless breath, / the mistress, remains at home” (K13). And Tilopa describes as “the self-aware fruit of the real” the place where “mind has died,” hence “breath / is completely dissolved” (T7), and asserts that “where mind is becalmed, / the triple world dissolves” (T11a). There are a number of passages, too, in which the mind (or thought) is singled out for denigration and identified as something to be transcended on the path. Saraha sings: “When thought's been changed / to nonthought, / you'll gain thereby / unsurpassed awakening” (S42a) and asserts that “the world / is bound by mind, / and no one at all can grasp / the nonmind” (S78); he also celebrates the yoginī who appears before him “mind destroyed” (S85). Tilopa, similarly, exclaims: “Mind must be killed! / Destroy it with nirvāṇa” (T3), and advises: “Don't disparage unthinking” (T4), since “what falls within / the range of thought / is not the ultimate” (T9).

These passages in which mind is in one way or another negated echo longstanding attitudes in yogic communities, where, as early as the time of the Yoga Sūtra, Patañjali could define yoga itself as “stopping the fluctuations of mind” (citta-vr̥tti-nirodha) and where language and conceptual thought always have been regarded as hindrances to a direct experience of reality as it actually is—however that reality might be defined.27 More specifically, it is assumed in nearly all yogic systems that mind and breath are intimately related, so control of the one will enhance control of the other. In the specific yoga system of the Yoginī tantras, the experiences of ecstasy and enlightenment that lie at the culmination of the path are impossible without becalming the conceptual mind and the breath, so that one may work with clarity and care in the subtle body.

Relaxing Mind, Seeing Mind

In still other passages, the siddhas seem to suggest that, far from controlling or trying to “kill” mind (and/or breath), one ought simply to let it go, or just see it as it is. Thus, Saraha recommends: “Releasing thought / and breath, / like unsteady horses, / dwell in your innate nature” (S45) and notes that, like a (p.31) camel, “bound, it runs / in all directions, / freed, it stands there motionless” (S43). He sings of the yoginī, a symbol for enlightened awareness, that “she eats, she drinks, / she doesn't care / what appears” (S86), and also likens the mind to a massive elephant that must be freed, and allowed to “drink the river's water, / and dwell on the shore as it pleases” (S100). In a similar fashion, Tilopa instructs: “Let thought go / where it wishes—/ it can't go wrong there” (T35). There also are a number of important verses in which seeing mind just as it is, as empty and/or pure, is required. Saraha advises us to “grasp the mind / as being like space; / as naturally spacious / grasp the mind to be” (S42a); asserts that “if mind is indicated / by mind, / concepts are held in suspension … / [and] mind dissolves / into its nature” (S78b); and instructs: “Witness mind with mind, fool, / and be free / from every base view” (S99). He also urges his listener to “just recognize mind wherever you are / [since] all is unceasingly fixed / in awakening” (S103), and to “bring forth / the stainless nature of mind, / quite beyond concepts” (S104). The mind's innate flawlessness and luminosity is celebrated repeatedly by all three siddhas, from Saraha's assertion that it is “is quite naturally pure” (S106), to Kāṇha's description of it as “beyond defilement” (K10), to Tilopa's claim that it “is stainless, beyond existence and nothingness” (T11).

All of these instructions to release mind, or see it as it is, or let its purity shine seem quite different in spirit from the denigrations of mind and thought I examined in the previous subsection, yet they need not be seen as contradictory if we acknowledge that a word like “mind” or “thought” may be used in more than one way and with more than one sort of valuation. This double usage is nowhere more evident than in Kāṇha's description of nirvāṇa as the place “where thought / has nothing to do / with thinking” (K20) or in Tilopa's advice to “kill the thought / that is not rooted in mind” (T33). Thus, conceptual mind or thought may be an obstacle to enlightenment, but mind or thought released into its essential nature (as pure, empty, blissful, etc.) is enlightenment.28 The latter is a characterization of the goal (though also, at times, a means to the goal) while the former is a concern solely on the path—as long as one recognizes which stage of practice is being described, one need not feel (pardon the expression) any cognitive dissonance.

Using Mantras and Maṇḍalas

There are several passages that make it clear that, whatever their occasional criticisms of reciting mantras or visualizing maṇḍalas, these traditionally tantric practices are a part of the siddhas' spiritual repertoire. Saraha sings cryptically of the “single syllable” he knows, though he admits, “friend, / I don't know its name” (S90), and he describes a syllable, amid the “uncorrupted three” (p.32) (perhaps the purity of body, speech, and mind), that is synonymous with “the god” (S90a). Saraha also discusses the maṇḍala circle, asserting that it is created by Śukra—a multivalent term that refers to the planet Venus but also to the drop of semen whose manipulation within the subtle body is crucial to the attainment of tantric ecstasies (S98). Kāṇha refers in a number of places to the mantric syllables or phrases, including haṃ (which is, in fact, associated with Śukra) (K4); evaṃ, which, in the Yoginī tantras and other advanced tantric systems, refers to a complex set of polarities, including emptiness and compassion, female and male, and breath and mind (K6, 21); and an unidentified “syllable of truth” at which an accomplished master stops his mind (K23) and where “zenith and nadir / both are unseen” (K24). He also instructs his listeners to understand the maṇḍala circle (K9) and speaks of emitting “the whole triple world” and drawing it back again (K17) and attaining the ultimate “in the company of wrathful goddesses” (K18), presumably those of a Yoginī tantra maṇḍala.

These references do not add up to a clear picture of what sort of mantra and/or maṇḍala practice the siddhas might have performed, but it does indicate that such practice was part of the religious world in which they moved, and may have been a part of what they advised their listeners to practice, too. Certainly, the notion of sound and syllables as creative realities has been part of Indian thought for millennia, and the notion of a yogin's participation in a divine world just as old. They form an important aspect of virtually all tantric practices, including those of the advanced tantras with which Saraha, Kāṇha, and Tilopa were familiar. On the basis of what we know about the practice of the Yoginī tantras from such texts as the Hevajra Tantra and Saṃvarodaya Tantra, we might speculate that the siddhas would engage in intensive visualization practices, in which they would reduce themselves and the world to emptiness, then reimagine both their world and themselves as divine. First, out of emptiness, they would generate certain Sanskrit “seed” syllables, corresponding to the great elements, which would transform into the basic maṇḍala mansion that would serve to house divine inhabitants. Next, from the emptiness where they themselves had been, they would generate another seed syllable, which in turn became themselves in the form of a male, female, or sexually joined male/female Buddha-deity, surrounded by other deities generated from their own seed syllables, often wrathful goddesses; these were the maṇḍala's inhabitants. Maintaining a vision of themselves as possessing a divine body, uttering mantras as their deity speech, and maintaining as best they could a nondual, blissful mind like that of a Buddha, they would practice emitting and withdrawing from themselves the transformed world of the maṇḍala, along with its deities, perfecting thereby their powers of concentration, and strengthening (p.33) their recognition of their own ultimate divinity. These practices, complex and important as they are, were in the Yoginī tantras only preliminary to the more subtle and truly transformative practices performed within the subtle body, and it may be because of their preparatory nature that the siddhas seem at times to disparage elements of maṇḍala-centered yogas.

Working Within the Subtle Body

That we possess a subtle body (sukṣma-śarira) is, as already noted, a common assumption in many yogic systems and, though never named as such, an implicit context for many instructions in the dohās of Saraha, Kāṇha, and Tilopa. The subtle body is probably the referent when Kāṇha sings of the “inmost body” (K3) and is likened by him to both a lotus (K4–5) and a mountain (K14–15, 25–26). References to “sun and moon” (e.g., S25, 47, K4) or the Ganges and Yamuna rivers (S47) are in a tantric context commonly understood to denote the left and right “outer” channels of the subtle body, and a term like “the in between” (S30) or “the middle” (K24) may easily be interpreted to refer to the central channel, which is where the mind and the breath-related physical energies (prāṇa) that are at the basis of ourselves and the world originate, and to which they must return, if real spiritual work is to be performed. The nodes where the “outer” channels intersect the central channel are called “wheels” (cakras), each of which (there usually are four or five) is a focal point for various substances and attributes, as well as for various meditative practices. Saraha sings of “ever filling the cakras / again and again” (S24), and Kāṇha makes numerous references to the “lotus” that, among other things, connotes the cakras (e.g., K3–7). Within the subtle body, the yogin must manipulate mind, breath, and various drops so as to induce the ultimate awareness that is the goal of completion-stage practices. Therefore, the many references (mentioned earlier) to the “place” where mind and breath are stopped must be understood in most cases to connote the central channel, and in particular either the crown cakra or the heart cakra.

The crown cakra (symbolized by Kāṇha as the summit of a mountain) is asserted in many Yoginī tantras to be the residence of the blissful white semen drop we each inherit from our father, and it is the manipulation, retention, and refinement of this drop—also referred to as the awakening mind (bodhicitta) (K3), the thought jewel (K16, K31), or the utmost gem (T27a)—that is the basis of the ecstatic experiences so important to practitioners of the Yoginī tantras. The heart cakra is mentioned somewhat less frequently in the Treasuries—Saraha does note that “what rises in the heart / settles in the heart” (S73)—but it is an important locus in the context of many advanced tantric practices, where it is in the “indestructible drop” at the heart that the subtlest mentality and mate (p.34) riality, the basis of life, death, and rebirth, is to be found. The subtlest mentality-materiality is therefore our inmost nature, or innate awareness, our “Buddha nature,” which must be purified if we are to attain enlightenment. In any case, whatever their specific cakra references, the siddhas do believe that it is only through drawing mind and breath-related energies into the central channel, then to the crown and/or heart cakras, that conceptuality and physical restlessness will be stilled and real progress toward freedom made possible.

Ecstasy and Sexual Yoga

Within the context of the subtle body, the practice of greatest importance to Saraha, Kāṇha, and Tilopa would seem to be the production of a series of four momentary ecstasies of increasing intensity and ultimacy: ecstasy (ānanda), utmost ecstasy (paramānanda), ecstasy of cessation (viramānanda), and innate ecstasy (sahajānanda). Though they do not use the term “ecstasy” in more than a handful of verses (e.g., S96, T25, T27, T28), it is probable that most, if not all, of their many references to “the innate” are a sort of shorthand for “innate ecstasy,” which is the fourth and highest in the series, perhaps identical to enlightenment itself.

There are a variety of descriptions of how the ecstasies are induced. They first are encountered in the course of the four initiations—Vase, Secret, Wisdom-Gnosis, and Fourth—that are required for undertaking Yoginī tantra practice.29 In most accounts of postinitiation yogic practice, the ecstasies presuppose the “dissolution” of conceptual mind and coarse energies within the central channel, followed by the generation of heat energy from the red female drop at the navel cakra, which rises up the central channel and melts the blissful white male semen drop that resides at the crown cakra. In some accounts, then, that drop descends immediately to the navel cakra, where it is held, and ecstasy experienced; when the drop is raised to the heart cakra, utmost ecstasy is experienced; when it reaches the throat cakra, the ecstasy of cessation is experienced; and finally, when it returns to the crown cakra, innate ecstasy—which utterly transcends the other three and is tantamount to enlightenment—is experienced. In other accounts, the four ecstasies are said to proceed in descending and ascending phases. In the descending phase, when the white drop at the crown reaches the throat cakra, ecstasy is attained; when it reaches the heart, utmost ecstasy is attained; when it reaches the navel, ecstasy of cessation is attained; and when it reaches the tip of the sexual organ, innate ecstasy is attained. These ecstasies are “ordinary,” but when ecstasy is conjoined with the realization of emptiness, one experiences extraordinary ecstasies: one draws the drop back up the central channel, through ecstasy at the navel, utmost ecstasy at the heart, ecstasy of cessation at the throat, and, finally, innate ecstasy at the crown. In any (p.35) scenario, the culmination of the process is the attainment of an innate ecstasy that involves the simultaneous experience of bliss and realization of emptiness.30

I have noted already that Saraha, Kāṇha, and Tilopa all make references to female partners with whom they appear to have sexual relationships, and it is probably within the context of these subtle body practices that these references should be understood. Again, less from their own dohās than from other Yoginī tantra contexts, we may suggest that the peerless coquettes, passionate maidens, and incomparable yoginīs of whom the siddhas sing were their partners in highly specialized and difficult practices whose primary purpose was to “force” energies into or within the subtle body so that they could be employed there for the purpose of hastening enlightenment. One of the reasons why sexuality may be used yogically is that, more than any other human activity, sexual intercourse, even in an “ordinary” context, has the effect of bringing thought and energy into the central channel, stilling conceptuality, inducing pleasure, and melting the white drop at the crown cakra, which then is “emitted” at the time of orgasm. In general, however, the ecstasies of tantra only are possible if, rather than being emitted, the white drop is retained, and one's bliss combined with a realization of the empty nature of phenomena, which may form the basis of one's transformation into an enlightened Buddha-deity. Thus, tantric sexual practices require immense discipline, great mental and physical self-control. Ironically, though they use “desire” to hasten the achievement of enlightenment, they cannot be practiced successfully by people in whom desire is uncontrolled and reality misunderstood. This, no doubt, is one of the reasons why Saraha cautions those who, failing to grasp things as they are, think that “in the midst of sex” they will “perfect great bliss” (S91). If, however, one is able to utilize sexual activity properly, that is, as a controlled form of yogic sublimation, one may harness one's mental and physical energies toward the attainment of Buddhahood, a state that, in its Yoginī tantra context, involves both absolute gnosis and an experience of rapture, bliss, and ecstasy that is analogous (and even related) to the pleasure of “ordinary” sex yet infintely beyond it in its intensity, duration, and soteriological import.31

Is There a System Here?

The Yoginī and other advanced tantras and their commentaries often refer to the procedures involved in seeing oneself as a Buddha-deity at the center of a divine mansion, or maṇḍala, as the “generation stage” (utpattikrama) and the procedures for manipulating forces within the subtle body so as to completely transform oneself into a Buddha as the “completion stage” (utpannakrama or saṃpannakrama). It was understood that in both stages, stilling the mind and seeing the nature of mind as emptiness, that is, developing concentration and (p.36) insight, were necessary for progress in uprooting defilements and that the final, perfected, Buddha state attained at the conclusion of the completion stage was one in which the innate gnosis of empty, blissful awareness was revealed. The completion stage, in turn, sometimes was subdivided into practices with and without signs, or into a “path of means” and a “path of liberation”; in each case, the former involved manipulation of energies within the subtle body and the latter the realization of emptiness. Because realization of emptiness—by an innate, blissful gnosis—was central to tantric definitions of enlightenment, some theoreticians made it virtually definitive of the completion stage, to the point where any formal practice (whether involving maṇḍalas or the subtle body) might be regarded as the stage of generation, and all “emptying” of those forms as the stage of completion.32 None of the three siddhas under consideration ever refers to either stage in his dohās, yet it is quite evident that the procedures involved were part of these siddhas' ritual and contemplative world, and it is not surprising that later, systematically inclined commentators would read their verses as an instruction in the various, sequenced stages of tantric practice.

Is there, then, a “system” of yoga that is promoted by Saraha, Kāṇha, and Tilopa? On the basis of the foregoing, we might be tempted to say that if there is, it requires a focus on inward practice rather than outward ritual, and a classically Mahāyāna balance between extremes of worldliness and transcendence and of compassion and emptiness. It requires that one receive tantric initiation from a guru and develop skill in reenvisioning oneself in the bodily form of a Buddha-deity at the center of one's maṇḍala, while uttering one's speech as mantra, and keeping in mind the nondual wisdom consciousness of a Buddha. One must not, however, be content with these generation-stage meditations and must progress to completion-stage practices within the subtle body. First, one controls and pacifies conceptual thought and the breath. Then (perhaps with the help of a sexual partner) one introduces one's mental and physical energies into the central channel. Within that channel and its cakras (and again, perhaps with the assistance of a sexual partner) one manipulates mental and physical energies, and certain inherited drops, so as to induce a series of four ecstasies, the last of which, innate ecstasy, is a blissful realization of reality that is tantamount to full enlightenment. Finally, when the innate has been fully realized, then one no longer needs to control or suppress thought, but simply sees mind and the world exactly as they are and always have been: empty, blissful, pure, awake—and one goes about living in the world, compassionately and wisely.

Lest we complacently assume, though, that we now have settled what Saraha, Kāṇha, and Tilopa “really were up to” in their yogic practice, let us remember how scattered and unsystematic their references generally are, and (p.37) how much guesswork is involved in trying to reestablish either the context in which their dohās were uttered or the meanings they were intended to convey. Let us also recall the often confusing mix of negative and positive discourse, the assertions counterposed with denials, that inform their discussion of nearly every topic—and their insistence that the ultimate, in any case, cannot be captured in words, and cannot be confined to any single technique or terminology. Thus, while it certainly is possible to understand their various references to mind, body, sex, and enlightenment in the way I have just suggested, we must admit that this is only one possible explanation. It is, in fact, quite possible that, in the face of achieving the ultimate, the entire raft of tantric practices I have just systematized may be a hindrance rather than an aid, and that when the siddhas sing of going beyond such things, they mean it. Similarly, it is possible that when they sing of releasing the mind to go where it will, or simply seeing it as it is, wherever one is, they mean this to be a part of the path and not just an expression of what it is like for one who has attained the goal.

At the very least, we must understand that, by their rhetoric and their indirection, Saraha, Kāṇha, and Tilopa may be warning their listeners (and us) that the crucial thing for anyone intent on freedom is simply to see and reveal the innate mind as it is. This is the value that trumps all others, and the process of attaining it has no necessary connection to thought or thoughtlessness; controlling or releasing mind; mantras and maṇḍalas or their absence; coarse body, subtle body, or no body; ecstasy or indifference; sex or celibacy. This is why both Indian and Tibetan Buddhists were able to discover in the dohās justifications for a “natural,” even profane, way of life in an everyday laicized setting, or for the practice of sexual yoga and other complex tantric rites and meditations inside a tightly controlled esoteric community, or for tantric meditation within the context of a monastery full of celibate scholar-monks, or for an approach to spiritual life that—whatever one's social milieu—emphasized direct, unmediated realization of the innate mind, without any recourse at all to complex techniques. Which did the siddhas intend? We do not know. As in the writings of so many mystics, it is precisely this ambiguity in the verses of the siddhas that has made them so fruitful over the years, allowing each generation to find in them meanings appropriate to its own situation—and it is this ambiguity that we would do well to keep in mind, if we would not reduce them to the “conventional nonsense” they so vociferously rejected.

Celebration of the Guru

There is, however, one topic on which the siddhas seem to entertain no ambivalence and permit no equivocation, and that is the centrality of the guru (p.38) to whatever the spiritual path might be. Saraha sings of how the ultimate is “witnessed solely / at the precious guru's feet” (S17) and adds that “if the guru's words / but enter your heart, / it's as if you've been handed / assurance” (S18). If you seek the innate, he says, you must know that “it's got from the guru” (S29)—indeed, “apart from the guru's teaching, / it's never seen” (S38). When you've managed to purify your mind, he adds, then “the guru's virtues / enter your heart” (S39). The guru's teaching is “ambrosial,” and those who fail to drink it up “die of thirst / in the desert / of variant texts and meanings” (S56), whereas those who are “firmly devout / in the guru's word” will experience the wave of the innate (S57) and know that “through [the guru's teaching] / awareness is purified” (S69). However, Saraha notes, those who are willing and able to find the “virtue / in grace at the guru's feet / … are rare” (S95). Kāṇha celebrates the guru rather less openly but does note that the spiritual master “makes thought motionless / by the syllable of truth” (K23). Tilopa proclaims that “at the precious guru's feet / is the nondual declared” (T6; see 28a) and asks how for one “blessed at the guru's feet” the innate mind can seem unapproachable (T8), for in fact “with the guru's teaching / it enters the heart” (T31).

The only verses that seem in the slightest to impugn to power of the guru are two in which the guru's capacity really to teach is questioned: Saraha poses the dilemma as follows: “If the guru / doesn't explain the teaching, / the pupil won't understand—/ but the ambrosial taste / of the innate, / who can teach what that's like?” (S56a). Tilopa poses a similar quandary, in very similar words: “The real / can't be shown / by the guru's words, / so the disciple / cannot comprehend. / The fruit of the innate / tastes ambrosial; / who teaches the real to whom?” (T9a). Clearly, the “limitation” on the guru's power expressed here is only that of language itself, which never can fully encompass the ultimate. Yet to the degree that conventional speech can at least point us in the direction of the ineffable experience of the innate, it is most certainly the guru—and only the guru—who will be able to orient us properly.

The term guru literally means “heavy,” “weighty,” “serious,” and in Indian tradition it has been the guru's task to inform us of the weightier truths of life, though a guru may be a teacher of any skill or subject, including mundane arts and sciences. I deliberately leave the word untranslated here because none of the possible English equivalents carries sufficient nuance, especially for the tantric context in which the siddhas use the term. A “teacher” in the West is someone who informs us, without necessarily forming us, as a guru does. A “mentor” gives sage advice and serves as an example but does not usually bear as much authority as a guru. A “master” clearly exercises control in some way but is not necessarily either a role model or teacher. The guru, especially the (p.39) tantric guru, is exceptional in the senses conveyed by all three English alternatives: as a teacher, as a role model, and as an authoritative guide. This is so primarily because the tantras—and especially advanced systems like the Yoginī tantras—provide an extraordinarily difficult and dangerous path to follow. In the tantras, as I have shown, one deals with complex meaning systems and attempts to confront and control the most basic energies and events to which we are heir: desire and anger, life, death, and rebirth. Neither the understanding of complex symbols nor, especially, the manipulation of basic human forces can be undertaken without instruction from someone who is experienced and knowledgeable about what is involved. And, in fact, because tantric systems are technically esoteric, limited to a circle of initiates, one cannot undertake their practice without initiation from a guru. If we try to do it on our own or, once initiated, ignore our guru's instructions, we almost certainly will be destroyed by the forces we seek to control—and those thus destroyed have a special circle of Buddhist hell reserved for them, the Adamantine Hell (varjranaraka) that is lower even than the deepest “conventional” hell, the dreaded “Unceasing Hell” (avīci).

Therefore, anyone who is serious about practicing tantra, especially the more advanced varieties, must find a guru who will give initiation and instruction. This is an important enough step on the spiritual path that a large body of literature exists in India and Tibet that is devoted either to analyzing the qualities necessary from both sides for a proper guru-disciple relationship or to telling stories about how gurus and disciples met and interacted. These stories, which are among the most appealing and provocative narratives in Buddhist literature, tend to emphasize above all the absolute submission that a tantric disciple owes to his or her guru. One of the best known from the Tibetan tradition is that of the reformed sorcerer Milarepa (a great-grand-disciple of Tilopa), who underwent extraordinary and dispiriting trials at the hands of his chosen guru, Marpa—including building, demolishing, and rebuilding the same tower numerous times—before the latter would grant him teachings; Marpa's “cruelty,” of course, is revealed as an exercise in compassionate, skillful methods, which serve both to test and to ripen Milarepa.

This story, in turn, echoes many that are told in Indian tradition, including a number involving the authors of the three Treasuries presented here. Thus, Saraha is reported in a number of Tibetan sources to have apprenticed himself to an arrow-making yoginī, though Abhayadatta's influential collection of hagiographies tells a different, more amusing story, in which Saraha, living in the mountains with his young wife, goes into a twelve-year trance just as she is offering him some radishes. When he rises from his trance, he immediately asks for his radishes—but she points out to him that they are no longer in season, (p.40) and that, besides, if he still clings to the idea of radishes, his labors have accomplished little; at this, Saraha abandons conceptual thought. Abhayadatta's account of Kāṇha similarly subjects him to gentle mockery, telling of his study under several gurus, both male and female, none of whom, because of his own jealousy and pride, he can fully obey; as a result, his spiritual progress is stunted until late in his life. Tilopa is known for having apprenticed himself to the daughter of a sesame-pounder, but is more renowned, perhaps, for his own actions as guru to Nāropa: the trials to which he subjects Nāropa before he will grant him teachings make Marpa's testing of Milarepa seem mild by comparison; they include jumping off a multistoried building, building a bridge over leech-infested waters, and abducting the queen of the realm, only to be beaten by her retainers.33

These legendary exemplifications of guru devotion, and cautionary tales of what befalls those with insufficient dedication, are simply ornamentation to the dohās in which, as I have shown, Saraha, Kāṇha, and Tilopa celebrate the guru as the sine qua non for practicing tantra, seeing the real, and experiencing the innate. In a modern context, the authority and devotion granted traditional gurus may seem out of place, if not downright dangerous, especially to those raised in societies supposedly guided by egalitarian assumptions. It is tempting, perhaps, to displace the outer, human guru with an incorporeal substitute, a kind of Protestantized “conscience” or “inner voice.” Certainly, the siddhas would admit that the guru is more than just a flesh-and-blood human being and that in any case, the guru's instruction must be internalized and assimilated if it is to be of any value. It is dubious, however, that in the context of complex tantric practices, the human guru—and the respect one owes to him or her—ever could be eliminated, and if the unquestioning devotion displayed toward their gurus by Saraha, Kāṇha, and Tilopa leaves us uneasy, then that, perhaps, is yet another of the paradoxes that they pose for us, as surely as they posed them to their audiences in the villages of Bengal, a thousand years ago or more.

Some Differences

In the previous sections I have focused on points of style and substance on which the authors of the three Treasuries seem to be in general agreement. There are, however, some differences among them worth mentioning.

The most obvious difference is that Saraha's Treasury (if we include, as I do, verses attested in Tibetan but undiscovered in Apabhraṃśa) is approximately four times larger than either Kāṇha's or Tilopa's collection. Furthermore, it is by far the most ambiguous and problematic of the three texts. It contains a (p.41) greater variety of verse styles, a larger range of themes, more seeming paradoxes and contradictions, and a less self-evident overall plan. There are certain “clusters” of verses, such as those critical of other religions (S1–15), those celebrating the yoginī (S84–87), or those exploring the metaphor of the “precious tree of emptiness” (S107–110) that naturally fit together, but such clusters are widely scattered, and not always clearly related to verses before or after them. These problems may be endemic to Saraha's Treasury because, more than the others, it is the product of a long and rather unsystematic editing process. These issues of coherence and redaction may undermine the historian's certainties about the integrity of the personage called Saraha or the text attributed to him, but for the Buddhist practitioner or for the religion scholar, it is precisely such ambiguity that gives Saraha's collection its extraordinary richness and suggestiveness. It is not surprising that both commentators and translators have been drawn to Saraha's Treasury above all others and that his reputation is greater, too—for within it one may find the seeds of seeing the world, and living life, in any number of different ways, from hedonistic to ascetic, from ritualistic to spontaneous, from lay to monastic, from arcane and complex to direct and natural, from tantric to transtantric.

Kāṇha's and Tilopa's Treasuries, on the other hand, are shorter, and somewhat more uniform in terms of style and content, indicating, perhaps, a less problematic process of redaction, if not absolute certainty that their current state reflects the intention of their purported authors. Kāṇha's collection contains several large and relatively systematic clusters of verses, including a series that refers to the subtle body and the various elements within it, all seen as rooted in the syllable evaṃ (K3–9); two series that describe the subtle body, and meditation within it, in terms of the metaphor of a mountain (K14–15, 25–27); and yet another series that criticizes conventional religious practices while exalting play with one's “inmost mistress” (K28–30). In terms of general emphasis, we might regard Kāṇha's as the most heavily “tantric” of the three collections: very little within it cannot be interpreted in terms of Yoginī tantra practices on the “completion stage,” that is, within the context of the subtle body. Kāṇha details the subtle body somewhat differently from Saraha or Tilopa, concentrating on the metaphors of lotus and mountain and saying little about the four ecstasies.

Like Kāṇha's, Tilopa's collection is stylistically fairly consistent and contains a number of key verse clusters, which include treatments of the nonconceptual nature of the innate (T3–9), the inadequacies of various traditional religious practices in comparison to Buddhist approaches (T19–22), and the details of producing the four ecstasies within the subtle body (T25–28). Thematically, Tilopa ranges a bit more widely than Kāṇha, including more verses that touch (p.42) on the metaphysics and ontology of the innate, and he also introduces technical terms that are infrequent or not present at all in the other two collections, most notably “unthinking,” (Skt. amanasikāra) a meditative procedure in which one does not allow anything to stay fixed in the mind, and the “great seal” (mahāmudrā), which designates, among other things, the final achievement of the tantric path, the nature of reality that must be comprehended on that path, and a set of meditation techniques built around seeing the nature of mind just as it is and being liberated into that nature. Mediated through figures like Maitrīpa, Nāropa, and Marpa, these concepts—especially the great seal—would come to have great importance in the Tibetan traditions stemming from the siddhas.

Influences and Echoes

In General

The dohās of Saraha, Kāṇha, and Tilopa—and others whose collections have not survived in Apabhraṃśa—have, as I suggested near the outset, been profoundly influential in Indic and Asian culture. They are the direct (though not the only) ancestors of the vernacular-shaping songs of north Indian devotional poets of the fourteenth through seventeenth centuries, from Kabīr and Ravidās in the Hindi world, to Jñānadeva and Tukaram in Marathi, to Nānak and Sultan Bahu in Punjabi, to Lallā in Kashmiri, to Chandidās, Vidyāpati, and Caitanya in Bengali. To this day in north India, their images, their message—and perhaps even some of their melodies—still are conveyed by wandering yogins, including, most notably, the widely discussed and recorded Bāuls of Bengal, and their less studied cousins the Kartābhajās.34 In Nepal, the songs of the siddhas (though diamond songs and performance songs more often than dohās) have served as a part of the liturgy of Newar Buddhist tantric priests (vajrācāryas) for a millennium and continue to shape both elite and popular conceptions of religion and of life.35 In Tibet, the dohās inspired the development of the most personal and spiritually profound of poetic forms, the “song of experience” (nyams mgur), which reached an early apogee in the works attributed to Milarepa. Just like the Indian Treasuries, Mila's verse is replete with a mix of paradoxical utterances, caustic social criticism, celebrations of enlightened experience, and profound practical advice; and others followed in Mila's wake, from other “mad” yogins like Tsangpa Gyarepa (gTsang pa rgya ras pa, 1161–1211) and Drukpa Kunlek ('Brug pa kun legs, 1455–1529), to great scholastics like Longchen Rabjampa (Klong chen rab 'byams pa, 1308–1363), Tsong Khapa (Tsong kha pa, 1357–1419), and Pema Karpo (Padma dkar po, 1527–1592), to important political (p.43) figures like the first Panchen Lama and the seventh Dalai Lama (1708–1757), to modern figures like Geshe Rabten and Chögyam Trungpa Rinpoche.36

Outside the three cultures in which the dohās have had a detectable influence, we may still discern their echoes—though they are not the sort of echoes whose causal lines may be traced but are, rather, the sort described by Rilke: “Echo answers echo: all is reverberation.” There may or may not, for instance, be any but the most tenuous historical connections between the Indian siddhas and, to the east, a great and mysterious figure like the Buddhist mountain poet Han Shan or, to the west, the supreme voice of lyrical and didactic Sufism, Rumi—yet the mix of paradoxical rhetoric, social critique, and celebration of a radiant, blissful, yet ineffable ultimate are common elements in all their works. There is no direct connection at all between the siddhas and Western poet-mystics like St. John of the Cross, William Blake, or Walt Whitman, yet again, we can identify in the rhymes and in the reasoning of them all a common passion for the absolute and for attaining it in and through a human body not easily separable from the soul. More recently (and in an interesting historical development) a number of American poets in whom we also can hear echoes of the siddhas actually can be shown to have been aware of them, and in some cases specifically influenced by them: some of the earlier poetry of Jack Kerouac (e.g., in Mexico City Blues) was self-consciously Mahāyāna Buddhist in its orientation, with a special fondness for paradox; and some of the later work of Allen Ginsberg (by then a disciple of the Tibetan teacher and poet Chögyam Trungpa Rinpoche) looked back to Indian or Tibetan Buddhist themes of a thousand years ago and tried self-consciously to replicate their concern with expressing spiritual experience.37 Finally, who can claim that, wherever in the world there still is troubadour trying to shake people into awareness—a young Bob Dylan or Bob Marley, perhaps—we don't still hear something of the songs of the siddhas, by an incarnation, in a new millennium, of Saraha, Kāṇha, or Tilopa?

I do not have the space here to provide examples of the various influences and echoes I have suggested (let alone those I have not). I do, though, want briefly to detail one influence and one echo. The influence is Kabīr, the echo is Blake.

An Influence: Kabīr

Kabīr lived in and around Varanasi, in north India, probably from 1398 to 1448. Born to Muslim weaver caste parents, he studied with a famous guru of the era, Ramānanda, and himself became one of the most important and beloved of the transsectarian sants (“saints”) who were the great religious charismatics, (p.44) and founding poets, of modern vernacular language religious culture in north India. Like the other sants, Kabīr often was harshly critical of the cultural status quo, and he devoted himself to interior worship of a formless God, referred to variously as Ram, the Word, the Name, Hari—or even the Innate. Like the Buddhist siddhas a half millennium earlier, he left a legacy of hundreds of songs in a variety of styles, which still are known and recited today. The form he used that most closely approximated the dohā was the aphoristic rhyming couplet called the sākhī (“witness”), of which 353 are collected in the Kabīr anthology, the Bījak. In those verses, we encounter many, though not of course all, of the attitudes and themes expressed by the Buddhist siddhas. Here is one in which the centrality of the guru and a critique of scholasticism and Brahmanism are combined:

The guru's word is one,

ideas about it endless.

Sages and pandits exhaust themselves,

the Vedas can't touch its limit.38

In the following, using a metaphor much like one of Saraha's, he expresses the importance of finding the truth within oneself:

Why run around offering water?

There's a sea in every house.

If anyone is thirsty,

by hook or crook, he'll drink.39

The ultimate, as “the word,” is celebrated, but the inadequacies of language are mentioned as well:

Hit by the word, one fell down,

another dropped his kingdom.

Whoever can discern the word,

his work is done.

. …

Everyone says words, words.

That word is bodiless.

It won't come

on the tongue.

See it, test it, take it.40

As in the siddhas, mind is either criticized or exalted, depending on the con-text:


The mind is a nervous thief.

The mind is a pure cheat.

The ruin of sages, men and gods,

the mind has a thousand gates.

. …

Moving within limits: man.

Moving without limits: saint.

Dropping both limits and no-limits—

unfathomable thought.

. …

Make the guru your burnisher,

polish, polish your mind.

Scour, scour with the word.

Make consciousness a mirror.41

There even are suggestions, now and again, of a yogic or meditative technique that may distantly derive from that of the siddhas, and achieve the same ecstasy as they did:

Meditated in the sky,

opened the thunderbolt door,

saw his own reflection.

The three [worlds] filled with joy.42

These are only samples of verses in which the influence of the likes of Saraha, Kāṇha, and Tilopa are evident. We must, of course, keep in mind as well the differences in the setting, poets, and poems: the sants are devotees of a monotheistic God in a cultural milieu shaped mostly by Hinduism and Islam, the siddhas tantric yogins in an environment dominated by Buddhist and Hindu meditative and ritual traditions. Nevertheless, the siddhas are there in the sants, and to the degree that the sants still live in the religious movements of contemporary north India, including the Kabīr panth, the siddhas live today, too—even though they, and the Buddhism they practiced, disappeared from India almost a millennium ago.

An Echo: Blake

William Blake (1757–1827), the English engraver, painter, printer, poet, and visionary, knew of India but almost certainly had never heard of Saraha, Kāṇha, or Tilopa, whose names were barely known in the West before the twentieth century. As a seer, he wove a private mythology out of elements of Judaism (p.46) and Christianity, the ideas of Emmanuel Swedenborg, and his own idiosyncratic reflections and experiences, expounding his ideas in long, visionary, prophetic poems. Blake's longer poems are difficult for the modern reader and not widely read today, but his shorter poetic works and epigrams have taken their place in the canon of English literature—and of Western mysticism. In these works, Blake decries the joy-suppressing conventions—the “mind-forg'd manacles”—of rationalism and institutional religion and exalts the imagination, the body, the senses, and sexuality. Thus, in criticism of rationalists, he writes:

Mock on, Mock on Voltaire, Rousseau:

Mock on, Mock on: 'tis all in vain!

You throw the sand against the wind,

And the wind blows it back again.

And every sand becomes a Gem

Reflected in the beams divine.43

Similarly, he decries three causes of error promoted by “all Bibles or sacred codes”:

  1. 1. That Man has two real existing principles: Viz: a Body & a Soul.

  2. 2. That Energy, called Evil, is alone from the body; and that Reason, call'd Good, is alone from the soul.

  3. 3. That God will torment Man in Eternity for following his Energies.

Against these, Blake counterposes three true “contraries”:

  1. 1. Man has no Body distinct from his Soul; for that call'd Body is a portion of Soul discerned by the five Senses, the chief inlets of Soul in this age.

  2. 2. Energy is the only life, and is from the Body; and Reason is the bound or outward circumference of energy.

  3. 3. Energy is eternal delight.44

He asserts, further, that both “Attraction and Repulsion, Reason and Energy, Love and Hate, are necessary to Human existence.”45 As a result, it sometimes is the case that “the tygers of wrath are wiser than the horses of instruction,”46 and that

Love to faults is always blind,

Always is to joy inclined,

Lawless, wing'd, and unconfin'd,

And breaks all chains from every mind.47

Blake also is famed for his frank celebrations of sexuality, as in the following:


What is it men in women do require?

The lineaments of Gratified Desire.

What is it women do in men require?

The lineaments of Gratified Desire.48

Further, it is clear that, to Blake, human capacity is infinite, for he comments:

  1. I. Man's perceptions are not bounded by organs of perception; he perceives more than sense (tho' ever so acute) can discover …

  2. VII. The Desire of Man being Infinite, the possession is Infinite & Himself Infinite.49

And, in one of his most oft-quoted observations, he insists:

If the doors of perception were cleansed every thing would appear to man as it is, infinite.50

There are, of course, many more differences between Blake and the siddhas than similarities, but the echoes of certain themes from one to the other is undeniable, and undeniable, too, is a striking formal similarity: just like Saraha, Kāṇha, and Tilopa, Blake expresses himself much of the time in aphoristic rhyming couplets. Indeed, I would argue that, if there is an English language Dohākoṣa, it is Blake's “Auguries of Innocence,” which begins with the instruction “to see a World in a Grain of Sand / And a Heaven in a Wild Flower, / Hold Infinity in the palm of your hand. / And eternity in an hour”51 and then goes on to present a series of didactic couplets, some in evident sequence, some appearing rather random. It includes many observations about the interrelations among things, along the lines of: “The Game Cock clipp'd & arm'd for fight / Does the rising sun affright” or “The Catterpiller on the Leaf / Repeats to thee thy mother's grief.”52 There also are implicit or explicit reflections on human attitudes and behavior, such as: “If the Sun & Moon should doubt, / They'd immediately Go out,” and “To be in a Passion you Good may do, / But no Good if a Passion is in you.”53 There are theological observations, too, such as the verses with which the poem ends: “God Appears & God is Light / To those poor Souls who dwell in Night, / But does a Human Form Display / To those who Dwell in Realms of day.”54 I will leave the matter with a set of Blakean “dohās” that recapitulate a number of his favorite themes, and at least some of those of the siddhas, too:

Man was made for Joy and Woe;

And when this we rightly know

Thro' the World we safely go,

Joy & Woe are woven fine,


A Clothing for the Soul divine;

Under every grief & pine

Runs a joy with silken twine.55

About the Translation

The edition of the Dohākoṣas of Saraha, Kāṇha, and Tilopa on which I base my translation is that of the Apabhraṃśa version published by Prabodh Chandra Bagchi in Calcutta, in 1938. I have drawn supplementary material from the recent editions of the Tibetan of Saraha prepared by Kurtis Schaeffer, and of Tilopa by Fabio Torricelli. As noted earlier, it is quite obvious from reading the various available Apabhraṃśa and Tibetan versions of the Treasuries that there were a number of different recensions in circulation in South Asia early in the second millennium C.E., which (especially in the case of Saraha) only partly overlapped.56 No editor yet has resolved all the difficulties and discrepancies among these various versions, nor is anyone likely to do so soon. Like all versions, Bagchi's edition (again, especially in the case of Saraha) is simply one arrangement of material found in a number of different forms. Like other recensions of the Treasuries, it reflects to a considerable degree editorial decisions not of the redactors of the “original” collections (let alone their purported authors) but of later Indian commentators (e.g., Advayavajra) from whose works the Treasuries have to a large extent been extracted.57 Furthermore, though it is not unique in this, either, it contains its share of misprints. Nevertheless, Bagchi's edition does represent the one available source in which versions of all three Treasuries surviving in Apabhraṃśa have been brought together, and it contains much supplementary material that is useful, including Bagchi's running Sanskrit translation of the Apabhraṃśa texts, and Sanskrit commentaries on each from medieval writers. Schaeffer's and Torricelli's editions of the Tibetan of Saraha's and Tilopa's Treasuries are the result of a careful comparison among the various Tibetan translations, and are likely to be definitive for some time to come. I have employed them not in order to produce a “complete” or “definitive” version of any of the three Treasuries but to provide a text that reflects the dohās most commonly encountered in the “standard” Apabhraṃśa and Tibetan editions.

My basic plan is to present the three Treasuries as Indic works, so I have everywhere translated the Apabhraṃśa (usually that of Bagchi's edition but in a few instances that of Shahidullah's or Saṃkr̥tyāyana's) in preference to the Tibetan, except where (as sometimes in Saraha or Tilopa) only the Tibetan is available.

(p.49) Each dohā presented includes, from top to bottom: (1) the number of the verse according to Bagchi's edition, except for verses he omits; (2) a transliterated Apabhraṃśa (or Tibetan) version of the verse; (3) my own translation of the verse; (4) information (if applicable) on alternative numberings of the verse in other editions and translations; and, in a number of cases, (5) supplemental information that will assist the reader in understanding the verse.

Further explanation of each element follows.

1. To reflect the ordering of Bagchi's edition of each of the texts, I have maintained his system of numbering the verses, with supplementary Apabhraṃśa or Tibetan verses or portions of verses placed after the verse they most closely succeed in Schaeffer's or Torricelli's Tibetan version, for example, as 42a. In instances where I am not translating from Bagchi, I have italicized the edition from which I am drawing. It must be noted that in any number of cases (especially in Saraha) what is numbered as a single dohā actually is a combination of two dohās with different end rhymes and, sometimes, different themes. Nevertheless, because they have been combined as a single unit at some point in the process of redaction or commentary, I have chosen not to break them apart into their constituent units but to remain faithful in general to the numbering and grouping systems adopted by previous editors and translators.

2. The Apabhraṃśa is presented not as a definitive edition (though occasionally I have corrected Bagchi) but so the reader may have some idea of (1) the text on which I base my translation and (2) what the original might have sounded like. With regard to the latter, it must be admitted that, at our historical remove, we have only a faint sense of the original pronunciation and rhythm (let alone such melody as may have been attached), but there still is a certain value in giving the reader access even to an echo of an echo of what Saraha, Kāṇha, and Tilopa might have sung.58 In the Apabhraṃśa, a roman-type bracket indicates an interpolation made by Bagchi himself; an italicized bracket indicates my own change. In two instances in Saraha (42d and 56a), I have included the Apabhraṃśa from Saṃkr̥tyāyana's edition for verses found in Tibetan but missing from Bagchi. Where no Apabhraṃśa is available, I have included the Tibetan, transliterated in the Wylie system, from the editions of Schaeffer (Saraha) or Torricelli (Tilopa); my occasional emendations also are indicated by an italicized bracket.

3. In preparing my translations, I have tried to remain faithful to the meaning of the original, and reasonably consistent in my use of translation terms, such that, for instance, citta always is translated by “mind,” jhāna by “meditation,” and so on. (On the other hand, dhamma/dharma is translated variously, according to context.) I have broken up the poetic line in the (p.50) translation, but in a manner that reflects the length of the line in the original. Thus, for instance, a longer line (whether of Apabhraṃśa or Tibetan) generally will merit three lines of translation, a shorter line, two. I have not attempted to replicate the end rhyme of the original—if Blake had known Apabhraṃśa, he might have been able to do so! Because Saraha, Kāṇha, and Tilopa—whatever the complexity of their meanings—express themselves in language that is direct and colloquial, I have tried to translate in an equally straightforward manner. One convention I have adopted to make the presentation more direct is to employ the second person more than is warranted by the original. Where the original speaks impersonally, or in the third singular, I have generally inserted a “you” to emphasize the point that these are instructions from a guru to an audience of disciples. Similarly, Indic languages often frame aphorisms by complex relative-correlative constructions, of the style: “Who cannot stand the heat, let that one get out of the kitchen”; I have almost invariably replaced these with conditionals that use the second person, for example, “If you can't stand the heat, get out of the kitchen.”

4. My references to alternative enumerations of the verses are keyed to the abbreviations found at the beginning of the book. Where applicable, the alternative enumerations of Saraha's Treasury are drawn from the Apabhraṃśa edition and French translation of Shahidullah, the Tibetan edition and English translation of Schaeffer, the English translation (from Apabhraṃśa) of David Snellgrove, the English translation (from Tibetan) of Herbert Guenther, and the Apabhraṃśa edition of Saṃkr̥tyāyana. There is no alternative enumeration of Kāṇha's Treasury. In the case of Tilopa's Treasury, the alternative enumeration is drawn solely from Torricelli's Tibetan edition and English translation. When one of those sources is italicized, I have drawn my original from it, rather than from Bagchi.

5. In the notes to particular dohās, I have tried to only provide information that seems clearly implied by, or necessary to the understanding of, the original verse. Later commentators have read into individual dohās a great many details (e.g., of Buddhist tantric practice or meditation theory) that are not evident from the words before us on the page,59 and, while aware of and interested in these glosses, I have usually chosen to ignore them, so as not to constrict the siddhas within a particular interpretive grid. I could not, however, resist noting a number of cases where there were striking parallels between a topic treated in the dohā and material found in the two Yoginī tantras that have been substantially translated into English, the Saṃvarodaya (in Tsuda 1974) and the Hevajra (Snellgrove 1959, Farrow and Menon 1992)—since, as noted near the outset, I do believe that the Yoginī tantras are the most promising single interpretive key to understanding the Treasuries.

(p.51) Saraha, Kāṇha, and Tilopa never will be entirely transparent to us, as they probably were not to their original audiences. If the reader can catch in these versions, though, something of their freshness and some of their wisdom, and sense a bit better the context in which they may have lived and sung, then I will have succeeded in minimizing my “traduttorial” sense of being yet another tradittore to poets and sages who probably were misunderstood the first time around, but kept coming out of the jungle anyway, into the village clearing, to sing for the few with ears to hear, a curious heart, and a pure passion to taste the innate. (p.52)


(1.) On the mahāsiddhas in general, see, e.g., Dasgupta 1976, Dowman 1985, Ray 1986, Davidson 2002a: chaps. 5–7, White 2003: chap. 6.

(2.) For translations of Abhayadattaśrī's text, see Robinson 1979, Dowman 1985. For material relating to songs sung by, and liturgies related to, the eighty-four, see Egyed 1984, Kapstein 2000.

(3.) On the eighty-five, see Schmid 1958; for the Nepalese guru list, see Tucci 1930; on the fifty-nine, see Templeman 1983; for Nyingma versions, see Dudjom 1991; on yoginīs, see Shaw 1994.

(4.) On Paśupatas and Kāpālikas, see Lorenzen 1972, Davidson 2002a: chap. 5; on Kashmiri Śaivites, see Mishra 1993, Sanderson 1988; on Bengali Śaktas, see Avalon 1974, McDaniel 1989; on the Nāth and Rasa siddhas, see White 1996; for Patañjali, see, for example, Radhakrishnan and Moore 1957: 453–485; on Saṃnyāsa Upaniṣads, see Olivelle 1992; on Nāgas, etc., see Bedi 1991.

(5.) Bharati 1970: chap. 1, especially 28–30.

(6.) For representative lists, see, for example, Robinson 1979: 289–307, Dudjom 1991: 2: 199–290.

(7.) On issues related to Saraha, see Shahidullah 1928: chap. 2, Y. Nara 1966, Guenther 1969: 3–20, Dowman 1985: 69–72, Guenther 1993: 3–15, R. Jackson 1994, 1996a, Schaeffer 2000 (especially chap. 8). For traditional biographies, see Robinson 1979: 41–43, Dowman 1985: 66–69, Templeman 1983: 2–3, Guenther 1993: 3–7. For translations of texts attributed to Saraha, see, e.g., Shahidullah 1928: 169–181, 233–234, Snellgrove 1954, Kvaerne 1977: 168–171, 199–202, 222–231, Guenther 1969: 63–71, Guenther 1993: 89–157, Cleary 1998: 111–112, 137–139, 161–169, Schaeffer 2000: 271–348. The Dohākoṣa translated here is, in the Tibetan setting, referred to as the “People” Treasury of Couplets, in contradistinction to the “Queen” and “King” collections; Guenther 1969 is translation and commentary on the King Treasury of Couplets, and Guenther 1993 includes translations of all three. Given his “foundational” role in various tantric lineages, especially that of the great seal, Saraha—if he lived at all—may have been the earliest of the three authors, but his dates cannot reasonably be narrowed beyond the general assertion that he probably lived some time between the eighth and eleventh centuries.

(8.) On issues related to Kāṇha, see Shahidullah 1928: 3–4, and chap. 2, Dowman 1985: 128–131, Templeman 1989: 107, R. Jackson 1992. For traditional biographies, see Robinson 1979: 81–85, Dowman 1985: 123–127, Templeman 1983: 43–44, Templeman 1989. For translations of texts attributed to Kāṇha under one or another of his names, see Shahidullah 1928: 85–88, 117–122, Beyer 1974: 258–261, Kvaerne 1977: 100–104, 109–135, 150–158, 214–218, 231–234, 238–241, 248–250, Farrow and Menon 1992, Cleary 1998: 41–43, 49–72, 97–103, 171–173, 179–181, 191–193. Like Saraha, Kāṇha is extremely difficult to date; (p.144) we have a bit more evidence to go on in his case, but much of it is conflicting, and the range of dates we might supply for him is, in the end, rather similar to Saraha's (sometime between the eighth and eleventh centuries); for a discussion, see Snellgrove 1959: 1:13, n. 4, Templeman 1989: 107, n. 3.

(9.) The six are the practices of inner heat, illusory body, clear light, dream, transference of consciousness, and the intermediate state; for discussions, see, for example, Chang 1963, Mullin 1996, Mullin 1997.

(10.) On issues related to Tilopa, see Dowman 1985: 151–155, Nālandā Translation Committee 1982: xxxi–xxxii, Torricelli 1997. For traditional biographies, see Robinson 1979: 98–99, Dowman 1985: 151, Templeman 1983: 45–46, Gyaltsen 1990: 33–54, Thrangu 1993: 5–42, Nālandā Translation Committee 1997, as well as Guenther 1963, where he is discussed in the context of his role as guru to Nāropa. For translations of texts attributed to Tilopa, see, for example, Chang 1963: 25–30, Bhattacharyya 1982: 289–291, Thaye 1990: 75–76, Bercholz and Kohn 1993: 266–272, Mullin 1997: 27–29. Tilopa is the only one of the three authors herein to whom precise dates sometimes are assigned: the most commonly given are 988–1069, but these may be too late, if we accept 1012–1097 as Marpa's dates and assume that Marpa was a disciple of Nāropa, and Nāropa of Tilopa.

(11.) This presumes—and most scholars do presume—that the dohās were, in fact, originally oral compositions; if they were not, they might well simply have been composed in Apabhraṃśa. On the language of the Dohākoṣas, see, for example, Shahidullah 1928: 53–55, Bagchi 1934, T. Nara 1961, T. Nara 1962, De 1993; on Apabhraṃśa, see, for example, Tagare 1987, Ghosal 1956, S. Sen 1973, Dimock et al. 1974: 12–13, Bubenik 1996.

(12.) On the dohā, see, e.g., Shahidullah 1928: chap. 4, N. Sen 1973, Schomer 1987, Templeman 1994: 17–26. The various metric patterns of the dohā are too complex to analyze here. For the beginnings of a discussion, see Shahidullah 1928: 60–62.

(13.) On performance songs and diamond songs, both in and of themselves and vis-à-vis dohās, see Templeman 1994. For examples of performance songs in translation, see Kvaerne 1977, Cleary 1998. For diamond songs, see, for example, HT 2: 4, 6–10, Templeman 1994: 26–29.

(14.) For examples of English translations of texts preserved only in Tibetan, see, e.g., Thaye 1990: 75–86, Kunga and Cutillo 1995: 26–27; for a discussion of related issues, see Schaeffer 2000: chap. 7.

(15.) For editions and translations of the Hevajra, see Snellgrove 1959, Farrow and Menon 1992; for the Saṃvarodaya (which is part of the Cakrasaṃvara tantra cycle), see Tsuda 1974; for the Caṇḍamahārosana, see George 1974; for the Vajrakīlaya, see Mayer 1996; for a discussion of aspects of the Buddhakapāla (on which Saraha apparently wrote a commentary), see Davidson 2002a: 247–252; for portions of the Kālackra, see Newman 1987; see also Wallace 2000.

(16.) Alternative names for Yoginī tantras include Prajñā (or wisdom) tantras and Ḍākinī tantras. Alternative names for Mahāyoga tantras include Upāya (method) tantras and Ḍāka tantras; see, for example, Lessing and Wayman 1978: 251.

(17.) As a number of scholars have pointed out, the usage of “tantra” as a generic term for a style of religious thought and praxis in India and beyond is to some degree an innovation of Western scholarship, which all too often has projected onto Asian traditions monolithic concepts that are not recognized by Asians themselves. As David White notes, an examination of the phenomena that scholars have called “tantra” reveals “a complex array of ritual, theoretical, and narrative strategies that are specific to their various religious, cultural, (p.145) sociopolitical, geographical and historical contexts.” He adds, however, that “there nonetheless exists a grouping of common denominators that should permit us to classify these as so many varieties of a single tradition … of Tantra” (White 2000: 5). White then goes on to supply a definition, which (fortunately) he admits is subject to modification according to particular contexts: “Tantra is an Asian body of beliefs and practices which, working from the principle that the universe we experience is nothing other than the concrete manifestation of the divine energy of the godhead that creates and maintains the universe, seeks to ritually appropriate and channel that energy, within the human microscosm, in creative and emancipatory ways” (9).

(18.) On taxonomies of tantra, see, for example, Wayman 1973: 233–239, Tsong-ka-pa 1977, Thondup 1982, Snellgrove 1987: vol. 1, Karmay 1988: 146–151, Thondup 1989: 29–35, Dudjom 1991, K. Gyatso 1994.

(19.) On idiosyncracies of the Yoginī tantras, see, for example, Snellgrove 1987: 1:243–266, Williams 2000: 213–217, English 2002, Davidson 2002b; on Unsurpassed Yoga tantras, see K. Gyatso 1982, Cozort 1986, Snellgrove 1987: vol. 1, Newman, 2000; on Buddhist tantras in general, see Lessing and Wayman 1978, Snellgrove 1987: vol. 1, Dudjom 1991, Samuel 1993, K. Gyatso 1994, Sanderson 1994, Williams 2000: chap. 7; on Mahāyāna, see Williams 1989; on Buddhism in general, see Rahula 1974, Gethin 1998; on yoga as a pan-Indian movement, see Eliade 1969; on the relation between Buddhist and Hindu tantras, see Sanderson 1988, Sanderson 1994; on tantra in general, see Bharati 1970, White 2000, White 2003: chap. 1.

(20.) For an excellent summary of major themes in the dohās and caryāgīti, arranged somewhat differently, see Dasgupta 1976.

(21.) For the Keṇa, see, for example, Radhakrishnan and Moore 1957: 41; for the Heart Sūtra, see, for example, Conze 1958; for the Lao Tzu, see, for example, Chan 1963: 141; and for the Gateless Gate, see, for example, Shibayama 1974: cases 30 and 33.

(22.) Davidson (2002a) points out that “agonistic” rhetoric is especially prominent in Buddhist tantric literature, which was created against the background of competition by Buddhists and others for scarce sources of patronage in a fragmented medieval polity.

(23.) For translations as “the Innate,” see, for example, Shahidullah 1928, Dasgupta 1976, and Snellgrove 1954; for “Together-born,” see Lessing and Wayman 1978; for “Simultaneously-arisen,” see Kvaerne 1975; for “the Spontaneous,” see Beyer 1974; for “Coemergence” see Guenther 1969, Namgyal 1986; for “Connate,” see Newman 2000; for “Complementarity-in-Spontaneity,” see Guenther 1993, for “Being,” see Torricelli 1997. Davidson (2002b) argues that the term must be translated according to its context. Still others (e.g., Urban 2001) simply leave the term untranslated. Snellgrove 1987: 1:245, n. 208, argues that “the innate” remains the best translation for sahaja, and I agree. Kvaerne 1975 and Davidson 2002b are the most cogent overall discussions of the term and its usages, though there also is much useful information in Shendge 1967 and Dasgupta 1976 (e.g., 77–86).

(24.) On the contrast between these views, see, for example, Ruegg 1963, Ruegg 1969, Williams 1989, Hookham 1991, Stearns 1999.

(25.) The siddhas' references to females (and their use of gender polarity symbolism) are significant and complex. The “actual” women to whom they refer in an apparently sexual manner may also, in other (e.g., monastic) contexts, merely be visualized partners, used to effect similar if not identical results. Women also may be taken as symbols of, for example, wisdom, or emptiness, or the blissful, empty gnosis of the innate, with whom the yogin “joins” (physically or not) to complete the classic Mahāyāna union of wisdom and method, (p.146) which is a prerequisite for Buddhahood. In other contexts, for example, as a house mistress, a woman may symbolize the breath, which must be “kept at home,” or the unquiet mind, which mistakenly seeks its “husband,” the true mind of the innate, in externals, rather than within. On the question the female as symbol and social being in Buddhist tantra, see, for example, Willis 1989, Shaw 1994, J. Gyatso 1998, Simmer-Brown 2001; on the question in Hindu tantra, see Kinsley 1997, White 2003.

(26.) For examples of this with relation to Saraha, see Guenther 1993 and Schaeffer 2000.

(27.) Yoga Sūtra 1.2; see, for example, Radhakrishnan and Moore 1957: 454, where the phrase is rendered: “Yoga is the restraint of mental modifications.”

(28.) This sort of distinction is very important in Tibetan meditative systems influenced by the siddhas, such as the great perfection (Tib. rdzogs chen) of the Nyingma, where a clear line is drawn between ordinary mind (sems) and exalted, primordial mind (rig pa), and the great seal of the Kagyu, where a distinction sometimes is drawn between mind (sems) and mind itself (sems nyid). On the great perfection, see, for example, Karmay 1988, Thondup 1989, Norbu and Clemente 1999; on the great seal, see, for example, Gyaltsen 1983, Namgyal 1986, Martin 1992, D. Jackson 1994.

(29.) On these, see, for example, Snellgrove 1987: vol. 1.: 213–277.

(30.) See the accounts of Yoginī tantra ecstasy practice found in, for example, Chan 1963, Shendge 1967, K. Gyatso 1982, Snellgrove 1987: vol. 1, K. Gyatso 1991, Mullin 1996, and Mullin 1997, and the descriptions and analyses in Kvaerne 1975, Dasgupta 1976, and Davidson 2002b.

(31.) On this question, see, for example, Snellgrove 1987: vol. 1, R. Jackson 1992.

(32.) Two classic, early sources for understanding the two stages are ST chaps. 2–3, HT 1:8, 23; see also, for example, Beyer 1973, Gyaltsen 1983, Cozort 1986, Thurman 1995: chaps. 7–8. Beyer 1973 remains the most thorough and stimulating discussion of these two stages, which he designates the “Process of Generation” and the “Process of Perfection.”

(33.) For Milarepa's story, see Lhalungpa 1986; for Saraha, Kāṇha, and Tilopa, see notes 7–8, 10.

(34.) For Kabīr, Ravidās, and Nānak, see Hawley and Jurgensmeyer 1988; for Jñānadeva, see Pradhan 1967; for Tukaram, see Chitre 1991; for Sultan Bahu, see Elias 1998; for Lallā, see Barks 1992; for Chandidās and Vidyāpati, see Dimock and Levertov 1967; on the Bāuls, see McDaniel 1989: chap. 4; on the Kartābhajās, see Urban 2001.

(35.) On Newar religion, see, for example, Gellner 1992, Lewis 2000.

(36.) On songs of experience, see R. Jackson 1996b; for Milarepa, see, for example, Chang 1989, Kunga and Cutillo 1995; for translations of a range of songs of experience from the premodern era, including those of several figures listed here, see Nālandā Translation Committee 1980, Jinpa and Elsner 2000; for modern figures, see, for example, Rabten 1983, Trungpa 1983.

(37.) For examples of Beat poetry linked to Buddhist themes, see Tonkinson 1995.

(38.) No. 124, Hess and Singh 1983: 104.

(39.) No. 12, Hess and Singh 1983: 91.

(40.) Nos. 6, 35, Hess and Singh 1983: 90, 93.

(41.) Nos. 96, 189, 160, Hess and Singh 1983: 100, 112, 107–108.

(42.) No. 348, Hess and Singh 1983: 130.

(43.) From “Poems and Fragments,” Hillyer 1941: 586–587.

(44.) From “The Marriage of Heaven and Hell” (“The Voice of the Devil”), Hillyer 1941: 651–652.


(45.) Ibid. (“The Argument”), Hillyer 1941: 651.

(46.) Ibid. (“Proverbs of Hell”), Hillyer 1941: 654.

(47.) From “Poems and Fragments,” Hillyer 1941: 575.

(48.) “The Question Answer'd,” Hillyer 1941: 579.

(49.) From “There is no Natural Religion,” Hillyer 1941: 619–620.

(50.) From “The Marriage of Heaven and Hell,” Hillyer 1941: 657. The phrase “doors of perception” was taken by Aldous Huxley as the title of his book on mescaline and mysticism, and “the doors” was adopted by Jim Morrison for the name of his late-1960s rock group. In the Doors' song “End of the Night,” Morrison directly quotes Blake's “Auguries of Innocence”: “Some are Born to Sweet Delight, / Some are Born to Endless Night.”

(51.) Hillyer 1941: 597.

(52.) Hillyer 1941: 597, 598.

(53.) Hillyer 1941: 599.

(54.) Hillyer 1941: 600.

(55.) Hillyer 1941: 598.

(56.) For different Apabhraṃśa versions of Saraha, see Śāstri 1916, Shahidullah 1928, Bagchi 1935, Bagchi 1938, Saṃkr̥tyāyayana 1957; for the Tibetan, see Shahidullah 1928, Schaeffer 2000, and the recensions in, for example, the Derge (Tohoku no. 2224, Barber 1991: vol. 28) and Peking (Peking no. 3068, Suzuki 1957: vol. 68) editions of the Tibetan Tripiṭaka. For a concordance of verses among the editions of Śāstri, Bagchi, and Saṃkr̥tyāyayana, see Saṃkr̥tyāyayana 1957: 459–467. For the Apabhraṃśa of Kāṇha, see Shahidullah 1928, Bagchi 1935, Bagchi 1938; for the Tibetan, see Shahidullah 1928, as well as Barber 1991: vol. 28 (Tohoku no. 2301), Suzuki 1957: vol. 69 (Peking no. 3150). For the Apabhraṃśa of Tilopa, see Bagchi 1935, Bagchi 1938; for the Tibetan, see Torricelli 1997, as well as Barber 1991: vol. 28 (Tohoku no. 2281), Suzuki 1957: vol. 69 (Peking no. 3128).

(57.) See Schaeffer 2000: 220–230 for a clear discussion of some of these issues.

(58.) Some points worth noting with regard to pronunciation: (1) long vowels—indicated by a macron—receive greater stress than short ones, (2) in words of three or more syllables where no vowel is long, the stress is often on the third-to-last syllable (e.g., Niccala), (3) compound vowels are probably best pronounced if each component of the compound is separately articulated (e.g., paḍhi-a-u), and (4) many of the consonants transliterated as a “v” (e.g., in vuddha) may have been pronounced as a “b”—indeed, in Shahidullah's edition, b is generally used in preference to v, perhaps reflecting tendencies in Bengali, Apabhraṃśa's modern descendant.

(59.) Readers interested in such commentarial expansions may consult the Sanskrit commentaries to each provided by Bagchi. For Saraha, see also Guenther 1993 (especially the notes, in which he draws on a range of Indian and Tibetan commentaries) and Schaeffer 2000: 264–348 (which is a translation the commentary by the Tibetan scholar bCom ldan ral gri); for Tilopa, see Torricelli 1997.