Early Evidence for Tantric Religion

David N. Lorenzen

The history of early Tantric religion is not easy to write. Although manuscript libraries contain hundreds, even thousands of different Tantric texts, both Hindu and Buddhist, no manuscript bearing a date before the mid-ninth century has been found, a date long after the initial rise of this movement. Relevant contemporary inscriptions, a key element in any chronological and geographical reconstruction of the early stages of Tantric religion, are unfortunately very few in number.

Another problem is that the range of phenomena covered by the term “Tantric religion” has been subject to different interpretations. In spite of all this, much headway has already been made in overcoming these problems and, today, scholars can speak with some assurance about at least the broad outlines of the early history of the movement.

The present essay attempts to give an overview of the conclusions historians have so far reached in this field.1 The first problem is that of definition. Does the term “Tantric religion” cover only those cults directly associated with the Sanskrit texts known as Tantras, Samhitas and Agamas, or does it also include a wide range of “popular” religious ¯ phenomena that can be broadly classified as being “magical” in character? Are the texts and followers of Hatha Yoga tradition, especially the N¯atha or Kanaphata yogıs, to be considered as Tantric? Are all, or nearly all, of the Hindu and Buddhist religious traditions dedicated to female deities Tantric?2

Differences of opinion about these questions exist for the simple reason that two different definitions of Tantric religion are possible and indeed both are used.

A narrow definition considers as Tantric only religious phenomena directly associated with the Tantras, Samhit¯as and ¸ Agamas. Since these texts are almost all ¯ written in Sanskrit, it can be assumed that the social base of Tantric religion narrowly defined in this way has been mostly literate, upper caste, and resident in or near towns and cities.

A wide definition of Tantric religion adds to the religion based on these Sanskrit texts an ample range of popular “magical” beliefs and practices including much of S¯´akta and Hatha Yoga traditions. To the extent that these popular ¸ religions are literate, many of their texts are written in vernacular languages. The main social base of this more widely defined Tantric religion can be assumed to have been less well-educated, lower caste, and generally more rural than its more Sanskritic counterpart.

In this essay I will accept a wide definition of Tantric religion, but this dual character of the movement remains a significant problem. Stated somewhat differently, there is a clear sense that the more elitist and Sanskritized manifestations of Tantric religion are more Tantric than those that are more popular and magical in character.

Even if we use a wide definition of Tantric religion, however, the epigraphic evidence for its existence is quite limited. This makes a close determination of its geographic spread and its historical chronology quite difficult. As for geography, we know that Tantric religion was primarily a northern phenomena, although it also had some following in parts of the South. Its chief centers of influence have been eastern north India (Bihar, Bengal, and Assam), Kashmir, Nepal, and Tibet, and perhaps the Punjab and Rajasthan (depending in part on whether one counts the Nath tradition as Tantric).

As for chronology, the earliest clear and datable evidence of full-blown Tantric religion appears in four literary texts written in Sanskrit of the seventh century .: Banabhatta’s ¸ Kadambar¯ı and Harsacartita, Mahendravarman’s Mattavilasa, and Dan¸din’s ¸ Dasakum¯aracarita. The surviving Tantric texts themselves seem to nearly all date from a slightly or considerably later period, from approximately the eighth to the eighteenth centuries.

These two facts—the northern and medieval provenance of Tantric

tradition—make the recovery of its history particularly difficult since the northern region was under the direct control of the Muslim rulers from about the beginning of the twelfth century. With the curious exception of the patronage given by several of the Mughal emperors, including both Akbar and Aurangzeb, to the N¯ath yogis of Jakhbar in the Punjab,3 none of the Muslim rulers of India is known to have been a supporter of Tantric religious cults.

An unknown number of Tantric centers, most notably the Buddhist monastaries at Nalanda and Vikramasila, were most probably destroyed by Muslim armies. In any case, royal patronage for all non-Muslim religions, except at the level of minor vassals and zamindars (land owners), evidently mostly dried up in the regions dominated by Muslim overlords.

For most of the period from 1200 to 1800 ., this included most of the Indian subcontinent.

A third major problem concerns the nature of the social institutions of Tantric religion. Surviving early epigraphs relating to religious institutions almost all register donations of land and/or money and other goods and services to temples, monasteries, and Brahmin agrahara (landgrant) villages. It is known that many Tantric ascetics organized themselves into “sects,” “orders,” or “preceptorial lines” such as those of the Kaulas, Kapalikas, and Naths.

It also seems to be the case that only a few of these sects and orders established large temples or monasteries. There has always been something secretive, individualistic, and countercultural about Tantric religion, rather like Gnostic Christianity in Europe and North Africa, and this has tended to discourage the creation of Tantric temples and monasteries, although Buddhist monasteries under Tantric influence such as those at Nalanda and Vikramasila, not to mention many in Nepal and Tibet, represent an obvious exception to this rule.

Most of the sources that document the early stages of Tantric religion are reasonably well known. The discussion that follows will represent a didactic review of these sources, treating separately each of the different constituent components of the wide and loosely organized complex that comes to be known as “Tantric religion.”

This procedure should clearly illustrate that while some components are quite ancient, the complex as a whole cannot be documented before the fifth or sixth centuries . The existence of a specific Hindu Tantric sect, that of the Kapalikas, is also first documented about that time. By about the seventh century, Tantric Buddhism seems to have been flourishing in several monasteries of Bihar.

The basic categories of documentation, each relating to a major component of broadly defined Tantric religion, can be conveniently arranged as follows: (1) sources relating to shamanic and yogic beliefs and practices; (2) those relating to Sakta worship, especially worship of the Matrk¯ ¸ as and demon-killing forms of Hindu and Buddhist goddesses; (3) those relating to specific schools of Tantric religion such as the Kapalikas and Kaulas; (4) the Tantric texts themselves.

Whenever possible, emphasis will be given to epigraphic documentation.

The earliest sources relating to shamanic and yogic beliefs and practices in India are mostly literary and are ancient, abundant, and widespread. This is hardly surprising since such beliefs and practices—those that aim at control over the mind, the body, and the physical world—are a virtual universal of human behavior. The most striking early evidence for shamanic-yogic practices in India is found in the famous “wild muni” (seer) hymn of the Rig Veda ¸ (10.136), probably dating from about the beginning of the first millennium before the Common Era.

In this hymn, the munis are described as having ecstatic, altered states of consciousness and also the magical ability to fly on the wind.

What is perhaps more surprising than the evidence of this Vedic hymn,

however, is the quite early development of a systematized set of yogic beliefs and practices that eventually became codified in the classical Yoga-sutras of Patanjalı and in later Hatha Yoga texts such as the ¸ Hatha Yoga Pradıpika ¸ of Svatmarama.

These beliefs and practices are already clearly in evidence in the Chandogya Upani¸sad (8.6.6) and the Svetasvatara Upani¸s¯ad (2.8–13), texts dating respectively from about the early and middle first millennium before the Common Era. The Chandogya refers to the mystical anatomy of n¯ad¯¸ıs (veins or nerves), while the Svetasvatara describes the basic meditative posture and techniques of sense and breath control. These beliefs and practices were expounded in more systematic form in Patanjalıs Yoga-sutras, possibly about the beginning of the Common Era.

Although Patanjalıs text is not usually considered to be Tantric in character, the transition to the more Tantric Hatha Yoga involves more a shift in emphasis ¸ than a basic change in the nature of yogic beliefs and practices.

Specifically, Hatha ¸Yoga emphasizes the development of the psychic control over the natural processes of aging and death (already a significant aim of Yoga in the Yoga-sutras and the Bhagavadgıta), control over the sexual organs through such practices as the vajrolı mudra (retention of bodily fluids), and the interior visualization of and control over the mystical anatomy of n¯ad¯¸ıs and cakras.

This control over the mystical anatomy is also thought to lead to knowledge of and control over the

microcosmic-macrocosmic links between this anatomy and the external world of nature. This in turn leads to the acquisition of the supernatural powers known (siddhis). Hatha Yoga adepts also invoke the supposed magical power of sacred oral ¸ formulas (mantras) and sacred diagrams (yantras and man¸dalas ¸ ). Even these formulas and diagrams, however, have a history going back to Vedic times.

This shamanic-yogic component of Tantrism first appears in a more clearly Tantric form in the seventh century texts of Banabha ¸ t¸ta and Dandin. In Bana’s ¸ Harsacarita, a “great Saiva” ( ´ mahasaiva) ascetic from the southern Deccan named Bhairavacarya is said to have befriended Harsa’s ancestor Puspabhuti.

Puspabhuti ¸ assists Bhairav¯acarya in the realization of a powerful spell (mahamantra) called the Mahakalahrdaya. The object of the spell is to subdue a zombie ( ¸ vetala). Bhairav¯acarya is said to reside near an old temple of the Mothers (matrs¸ ).

The ceremony itself takes place at “an empty building near a great cremation ground on the fourteenth night of the dark fortnight” and involves the celebration of a fire rite in the mouth of a corpse. Banas portrait of a Tantric ascetic from southern India in ¸ Kadambarı is more comic in tone but similar in content.

Dan¸din’s ¸ Dasakum¯aracarita, on the other hand, describes its Tantric ascetic as an evil siddha (one with supranormal powers). Another seventh century text, Mahendravarman’s Mattavilasa features a Tantric Kapalika ascetic, but he is portrayed more as a hedonistic clown than as a shamanic yogi. After the seventh century, Tantric ascetics are frequently mentioned in Sanskrit literature.

A second major component of Tantric religion is the worship of female deities, particularly those who manifest a fierce character. Like the shamanic-yogic component of Tantrism, the worship of female deities has a long history in India and may be regarded as a near universal characteristic of human societies.

The Vedic antecedents of goddess worship appear in a series of hymns dedicated to the goddess of the dawn, Usas, and a number of hymns dedicated in whole or part to ¸ river goddesses, to the goddess of speech, V¯ac, or to other minor female deities.

None of these hymns, however, negates the obvious fact that Vedic religion is decidedly patriarchal in character. Early hymns to the Great Goddess, the Goddess of whom all individual goddesses are merely forms or aspects, are found in the Mahabharata and Harivamsa, the Devı Mahatmya section of the Markan¸deya ¸ Purana, ¸ the Candisataka ¸ attributed (probably falsely) to Banabha ¸ t¸ta, and the ¸ Gaudavaho ¸ of Vakpati.

All these texts refer to the fierce, demon-killing forms of the Goddess, most prominently the form named Mahisamardinı, the destroyer of the buffalo demon Mahisa. The battle between this Goddess, often identified as a ¸ form of Durga-Parvati, and Mahisa is mentioned in all these sources (except perhaps the Harivamsa) and is recounted in detail in the Devı Mahatmya and the Candisataka. ¸

Sculptural representations of Mahisamardinı have been found that date to the Gupta period and the earlier Kushan period.8

The earliest epigraphic mention of this goddess is probably that found in a late sixth century . Nagarjuni Hill (Gaya District) cave inscription of Anatavarman of the Maukhari Dynasty.

May the Devi’s foot, its gleaming nails emitting a mass of rays, point the way to the abode of riches. Her foot challenges with its splendor the full beauty of a blossomed lotus. With its twinkling anklet it contemptuously rests on the head of Mahis¯¸asura. It rewards your condition as petitioner that suits the expression of firm devotion.9

The same inscription mentions Katyayani and Bhavanı as the alternate names of this same Great Goddess. This inscription and the hymns to the Great Goddess in the above-mentioned texts illustrate the relatively early development of mature Sakta religion and its increasing association with fierce, demon-killing forms of the Goddess, forms that can said to be Tantric-flavored, if not necessarily fully Tantric.

Goddess worship seems to have become more definitely Tantric in character in connection with the rise of a group of seven (or more) goddesses known as mothers or M¯atrk¯ ¸ as. They are mentioned in the Mahabharata as well as early Puranic literature, Brhatsa ¸ mhita ¸ and other relatively ancient texts. In Banabhatta’s ¸ Harsacarita, the Tantric ascetic Bhairavacarya is said to stay near an old temple dedicated to the Matrkas. Bhasa’s Carudatta, Sudraka’s Mrcchakatika, ¸ and Banabhattas ¸ Kadambarı also refer to these goddesses especially in connection with offerings made at crossroads.

In the present text, however, more important is a reference to these goddesses in the stone inscription of Vi´svavarman, found at Gangadhar in Rajasthan and dated in 423 .11 This is often identified as the earliest epigraphic evidence for Tantric religion.12

Two other important early epigraphic references to these goddesses appear in the Bihar stone pillar inscription of Skanda Gupta or Puru Gupta (fifth century ) 13 and the rock inscription of Svamibhata (sixth century ¸ ?) from Deogarh in Jhansi District.14 The Matrk¯ as are also regularly invoked in the preambles of the inscriptions of the Kadambas and Early Calukyas from the mid-fifth century onward.

The description of the Matrkas found in the Gangadhar inscription merits some discussion. The passage that refers to them in this record has been given a somewhat different interpretation by J. F. Fleet and by A. L. Basham. Verse twenty-three states:

For the sake of religious merit, the king’s minister had them construct this terrifying home of the Mothers, filled full of female demons (d¯¸ akin¯ı) . . . these Mothers impel the great booming of the rain clouds and rouse the ocean with the mighty wind that arises from the Tantras.

In this passage from the Gangadhar inscription, the words d¯¸akin¯ı and tantra both clearly suggest an association with Tantric religion. According to Monier Williams, the dakinıs are said to feed on human flesh. It is, I think, quite probable that the word tantra here refers to the Tantras themselves, but, as Basham points out,19 the word has several other meanings including “a drug” and “a spell (mantra).” One must reluctantly agree with Basham that here “we must leave the question [of the meaning of the word tantra] open, recognizing that this inscription gives no proof of the existence of a developed literature of Tantrism in the fifth century .”

The classic description of the Matrk¯ ¸ as is found in the Dev Mahatmya, a text traditionally included as a part of the Markandeya Purana. ¸

It is generally accepted as the earliest and most important text of Sakta religion. Most portions of this text can be said with some confidence to have been written “before the close of the sixth century .”

The text describes the Matrk ¸ as as being created from the

“energies” (´saktis) of the gods Brahma, Siva, Skanda, Visnu (the ´ ´saktis Vaisnavı, ¸ Varahı, and Narasimhı) Indra, and Candika in order to help the Goddess destroy ¸ the armies of the demons Sumbha and Nisumba.

An interesting Kalacuri inscription from Pujaripali, near Sarangarh, Chhattisgarh, praises several of these and other demon-killing goddesses in verses that are evidently directly inspired by the Devı Mahatmya. The inscription is dated either in about 1150 . or in 1088 . It clearly shows that, by this time, the Devı Mahatmya was accepted as a basic source of Sakta religion.

The early evidence for the existence of specific sects and vows of Tantric religion pertains mostly to the Kapalikas, sometimes identified as Somasiddhatins or Mahavratins. They are first mentioned in several literary sources including dubious references in the Maitrayanıya Upani¸sad and the Yajnavalkya smrti ¸ and a more credible reference in Hala’s Gatha-saptasatı (third to fifth century .) and in two texts of the astronomer-mathematician Varahamihira (c. 500–575 .).

Starting with Mahendravaman’s early seventh century farce, the Mattavilasa, literary references to Kapalikas become quite common.24 As far as epigraphs are concerned, there are in fact only three or four that have been clearly identified as registering donations to or from Kapalika ascetics.

These are the following: (1) an Igatpuri (Nasik district) copper plate inscription of the early Calukya king Nagavardhana (seventh century .) that registers a donation to a Kapalesvara temple and the Mahavratin ascetics residing in it; (2) a Tilakwada (Baroda District) copper plate inscription (1047 .) of a subordinate of the Paramara king Bhoja that registers a donation to “the muni Dinkara, a Mahavratadhara who was like the Kapalin Sankara in bodily form”; (3) the Kalanupaka (Nalgonda District, A.P.) inscription 1050 .. that registers a land grant made by a Kapalika ascetic named Somibhattaraka to an individual named Can¸damayya; and (4) a sixth century ¸

inscription from Bangalore District that registers a land grant by a king Durvinıta to a Brahman named Kapalisarman (who may or may not be Kapalika). In addition, a clear reference to a Somasiddhantin ascetic named Vagisa Bhatta is found in a 1171 ¸ inscription from Tiruvorriyur (Chingleput District, Tamilnadu). The 1050 .. inscription of the Kapalika Somibhat¸taraka is particularly ¸ important since it includes a physical description of this ascetic and his vestments that agrees remarkably well with the descriptions of the Kapalika vestments in texts by the Vaisnava theologians Yamunacarya and Ramanuja, even to common ¸ use of the term ¸sanmudra ¸ (six insignia) to identify the key items.26 The same two theologians also identify the Kalamukhas as being Tantric ascetics, but this attribution of a Tantric character to them was probably willfully mistaken.

The numerous inscriptions registering donations to K¯al¯amukha ascetics and temples clearly show them to belong to a non-Tantric South Indian sect descended from the Pasupatas. Whatever the case may be, the earliest epigraphs refer to the Kalamukhas, dated 806 and 810 ., were found at Nandi Hill in Kolar District at Karnataka.28 Another early record, the undated Tandikonda grant of the eastern Calukya king Ammaraja II (946–970 .), registers a donation to a group of Kalamukhas located at a temple at Vijayawada (Bezwada) about sixty miles from the mouth of the Krishna River. Also worth mentioning is a short inscription found at Anaji in Dharwar District that records a gift of land to a temple connected with the Kalamukha Saktiparisad. ¸

Epigraphic references to other Tantric sects such as the Kaulas are apparently quite rare, but systematic research on this question remains to be done. Mark Dyczkowski has, however, made considerable progress in sorting out the Kaula affiliation of many Tantric texts.

Also relevant in this context is a Cambodian inscription of about 1052 . that tells how “king Jayavarman II’s court priest Sivalakaivalya at the beginning of the ninth century ( ´ . 802?) installed a royal cult based upon the four Tantric books brought from elsewhere . . . The texts in question are the Sirascheda, Nayottara, Sammohana, and Vınasikha. ” Although these texts may not be specifically Kaula, they belong to the tradition of vama (left) Tantras, some of which are associated directly with the Kaulas.

When we turn to the earliest evidence for the existence of Tantric Buddhism, we find that this consists primarily of sculptures of fierce deities such as Trailokyavijaya, Cunda, and Samvara and of sexually engaged ( ¸ yuganaddha) male and female deities.32 The principal monasteries where such Tantric sculptures have been found are those at Nalanda in Patna District, at Antichak in Bhagaipur District (often identified with ancient Vikramasila), at Paharpur in Rajshahi District, and in other sites in the northeastern region.Nalanda seems to be the oldest of these monastic sites. Its foundation has been dated to the mid-fifth century . It is unclear, however, whether the Tantric images found at this site belong to the earliest stages of its development.

Finally, we come to the question of the dates of the earliest specifically Tantric texts, especially the Tantras, Samhitas, and ¸ Agamas belonging to different ¯ Tantric sects or schools. Apart from the somewhat dubious reference to tantra in the Gangadhar inscription of 423 ., the earliest clear reference to Tantric texts seems to occur in Banabhattas ¸ Kadambarı. In his description of a South Indian Tantric ascetic, Banabhtta says that “he had made a collection of manuscripts of ¸ jugglery, Tantras, and mantras (which were written) in letters of red lac on palm leaves (tinged with) smoke.”

According to D. C. Sircar,36 Buddhist tradition claims that “Padmavajra, author of the Hevajra Tantra, was the preceptor of Anagavajra, a son of king Gopala who founded the Pala dynasty in Bengal about the middle of the eighth century ” Sircar also notes that some scholars date the composition of this text as early as “shortly before 693 ..” D. L. Snellgrove similarly estimates that “the Hevajra-Tantra [was] existing in its present form towards the end of the eight century.”37 On the other hand, Alex Wayman has ascribed another early Buddhist Tantra, the Guhyasamajatantra, “on a purely tentative basis, . . . to the fourth century .”

38 His reason for suggesting this early date does not bear scrutiny.

Hindu Tantrism is probably slightly older than its Buddhist counterpart, but early Hindu Tantras cannot be dated with any precision. Some earlier Pancaratra texts, insofar as these are Tantric in character, may date from the fifth century ., but these dates are highly speculative.

Much the same comments can be made about the Saiva´Agamas preserved mostly in southern India. ¯ 40 Goudriaan claims that the oldest surviving Tantric manuscript known, a copy of the Paramesvaramata, bears a ninth century date equivalent to 858 or 859 .41 He also notes, however, that the mention of many other Tantric texts “as venerable authorities” in Abhinavagupta’s great Tantraloka, written sometime around 1000 ., “renders it at least probable that Tantric literature existed already two or more centuries before. . . .” 42

In terms of its philosophical sophistication, Kashmiri Saiva tradition represents the richest development of Tantric literature. In recent years there has been a veritable flood of scholarly publications in this field.43 Its greatest traditional scholar, Abhinavagupta, wrote such Tantric works as Tantraloka, Tantrasara and Paratrimsikavivarana¸ in about the early middle part of the eleventh century. A fair amount about the earlier history of Kashmiri Saivism is known, above all from ´ the discussions of Abhinavagupta’s Tantraloka. Nonetheless, few if any of the earlier sources, except for some of the Saiva Agamas themselves can be dated before ¯ the eighth or ninth centuries.

One interesting Buddhist Tantric school is represented by the Buddhist Siddha authors of the Caryagıtikosa, a collection of religious songs written in a language most scholars regard as an early form of Bengali. These songs, in fact represent the oldest examples of Tantric literature written in an early form of a modern vernacular language. D. L. Snellgrove and Per Kvaerne place most of these songs in about the eleventh century. It is possible that a few of the Siddhas to whom some of the songs are attributed may have lived a century or two earlier. In particular, Saraha may date from the ninth century and Lui (perhaps the same as Matsyendra) from the late ninth or early tenth century.

Another Tantric (or Tantra-influenced) sect that has had an impressive literary output, in both Sanskrit and vernacular languages, is that of the Naths or Kanphata Yogıs. ¸ Most of its texts deal with aspects of Hatha Yoga. None can be ¸safely dated before the tenth century .., however, the legendary founders of the sect, Gorakhnath and his teacher Matsyendra, probably did not live much earlier than a century or two before this date. It is notable, however, that this tradition spread throughout India in the medieval period, including even in the South where it is represented by the Tamil Siddhas. Basing himself on Zvelebil,5 Goudriaan claims that: “the oldest Tamil Siddha, Tirumular, perhaps flourished in the seventh century .; the apogee of the Tamil Siddha literature, however, lasted from the tenth to the fifteenth century.”

The Nath tradition seems to have historical connections with the earlier Kapalikas,52 but this is too poorly documented to be of much help in dating the beginnings of Nath tradition. The main influence of this tradition was in northern India during the later medieval period. During the early nineteenth century it became a virtual state religion in the kingdom of M¯an Singh in Jodhpur, Rajasthan.5 It also had a strong historical influence on the devotional Varakarı tradition of Maharashtra (through the preceptorial line of Jnane´svar), and later Kabır, Raidas, Guru, Nanak, Dadu and others.

In summary, it can be said that Tantric religion as a recognizable complex of beliefs and practices is first documented, in very sketchy fashion, in the fifth century .. and relatively rapidly increased its influence in succeeding centuries within both Hinduism and Buddhism. It became particularly strong in North India (excepting perhaps the state of Uttar Pradesh, but including Bangladesh and parts of Pakistan), in Nepal and Tibet, and in parts of southern India.

By the ninth or tenth centuries, Tantric religion, both Hindu and Buddhist, had become extremely influential, perhaps even dominant, in many of these areas.

Buddhist Tantrism together with other forms of Buddhism, died out in India by the late twelfth and thirteenth centuries. It has survived in Nepal and Tibet but has lost influence in Tibet since Chinese occupation and the introduction of modern secular education. Hindu Tantrism remained popular during all the medieval period, but it seems to have lost most of its popular and intellectual support during the nineteenth century, largely as a result of the efforts of Indian reformers, both liberals and conservatives, to “purify” Hindu tradition.

Nonethelss, Tantrism continued to have an active presence in at least the Benares region until the early decades of the twentieth century. Today it no longer exists as a significant organized force in India or other countries (with the possible exceptions of Bali, Bhutan, Tibet, and Nepal). Nonetheless, many of its beliefs and practices are now well-integrated within more mainstream Hinduism and Buddhism. In at least this assimilated form, Tantric religion remains alive and well.

NOTES

1. The best surveys are those of Teun Goudriaan, “Introduction, History and Philosophy,” Hindu Tantrism, ed. Teun Goudriaan, Sanjukta Gupta, and Dirk Jan Hoens (Leiden: E.J. Brill, 1979); and “Hindu Tantric Literature in Sanskrit,” Hindu Tantric and Sakta Literature, ed., Teun Goudriaan and Sanjukta Gupta, eds., (Weisbaden: Otto Harrassowitz, 1981). See also the discussions of Andre Padoux, Recherches sur la symbolique et l’energie de sa parole, 2d ed. (Paris: Edicions E. de Boccard, 1975), and, “Tantrism,” in Encyclopedia of Religion, vol. 14, ed. Mircea Eliade (New York: Macmillan, 1987), 273–75; P. V. Kane, History of Dharma´s¯astra, vol. 5, pt. 2

2. Goudriaan and Padoux have written intelligent attempts to define Tantrism. See Goudrian, “Introduction, History and Philosophy,” 7–91; and Padoux, “Tantrism,” 14:273–75. 3. B. N. Goswamy and J. S. Grewel, The Mughals and the Jogis of Jakhbar (Simla: Indian Institute of Advanced Study, 1967).

4. See Sheldon Pollock for a partly opposing argument that patronage for nonMuslim culture, at least in the case of Sanskrit literature, did not dry up to the extent claimed by Alberuni and most modern historians; “R¯am¯ayana and the Political Imaginary in Medieval India,” The Journal of Asian Studies 52, no. 2 (1993): 261–97. 5. The descriptions of these and other Tantric ascetics and rites found in such literary texts are quoted in translation and discussed in more detail in David Lorenzen, The K¯ap¯alikas and K¯al¯amukhas: Two Lost Saivite Sects, ´ 2d rev. ed. (Delhi: Motilal Banarsidass, 1991), 16–23, 54–55.

6. J. N. Tiwari, Goddess Cults in Ancient India: With Special Reference to the First Seven Centuries A.D. (Delhi: Sundeep Prakashan, 1985), 61–94. 7. Thomas B. Coburn, Dev¯ı M¯ah¯atmya: the Crystallization of the Goddesss Tradition (Delhi: Motilal Banarsidass, 1984), 92–93; and J.N. Banerjea, The Development of Hindu Iconography, 3rd ed. (New Delhi: Munshiram Manoharlal, 1974), 497–500.

8. Gritli Von Mitterwallner, “The K¯us¯¸ana Type of the Goddess Mahis¯¸asuramardin¯ı as Compared to the Gupta and Medieval Types,” German Scholars on India (Bombay: 1976), 2: 196–213.

9. My translation. See John Faithful Fleet, Inscriptions of the Early Gupta Kings and Their Successors, Corpus Inscriptionum Indicarum 3 (Varanasi: Indological Book House, 1963r), 226–28. The goddess Bhadr¯ary¯a or Bhadr¯ayak¯a, possibly a form of P¯arvat¯ı-Durg¯a, is mentioned in the Bihar pillar inscription of Skanda Gupta or P¯uru Gupta; see D. C.Sircar, Select Inscriptions Bearing on Indian History and Civilization, 2nd ed. (Calcutta: University of Calcutta, 1966), 1:325–28.

10. Katherine Anne Harper, Seven Hindu Goddesses of Spiritual Transformation: The Iconography of the Saptamatrikas (Lewiston, N.Y.: The Edwin Mellen Press, 1989); and Tiwari, Goddess Cults, 95–99.

11. Fleet, Inscriptions, 72–78; Sircar, Select Inscriptions, 399–405. 12. A. L. Basham, “Notes on the Origins of S¯´aktism and Tantrism,” Religion and Society in Ancient India: Sudhakar Chattopadhyaya Commemoration Volume (Calcutta: Roy & Early Evidence for Tantric Religion 35

Chowdhury, 1984), 148–54; and M. C. Joshi, “S¯´akta-Tantrism in the Gupta Age,” Aruna¸ Bh¯arati: Prof. A. N. Jani Felicitation Volume (Baroda: Oriental Institute, 1983), 77–81. 13. Fleet, Inscriptions, 47–52; Sircar, Select Inscriptions, 325–28.

14. Daya Ram Sahni, “Deogarh Rock Inscription of Sv¯amibhata,” ¸ Epigraphica Indica 18 (1925–1926): 125–27. 15. See: J. N. Tiwari’s brilliant historical study of goddess cults in ancient India, 94– 181; and N. N. Bhattacharyya, The Indian Mother Goddess, 2d ed. (Columbia, Mo.: South Asia Books, 1977).

16. Fleet, Inscriptions, 72–78. 17. Basham, “Notes,” 148–50.

18. I have substituted the reconstruction “[pracu]dita-” for Fleet’s and Basham’s [pramu]dita-.” Neither Fleet’s nor Basham’s translation is completely satisfactory. In particular, I am not convinced by Basham’s renderings of ambhonidhi as “cloud’ rather than “ocean” and of ghana as “cymbal” rather than “dense,” “thick,” “multitude,” “cloud,” or “darkness.” My rendering supports Basham’s suggestion of a connection with rain-making better than his own translation.

19. Basham, “Notes, 149–50. 20. Ibid., 150. 21. See Thomas B. Coburn, Dev¯ı-M¯ah¯atmya: the Crystallization of the Goddess Tradition (Delhi: Motilal Banarsidass, 1984); and Encountering the Goddess: A Translation of the Dev¯ıM¯ah¯atmya and a Study of Its Interpretation (Albany: State University of New York Press, 1991).

22. Tiwari, Goddess Cults, 63–64 and 74–75. 23. See especially chapters seven and eight of the Dev¯ı M¯ah¯atmya. The numbers and names of these M¯atrk¯ ¸ as vary considerably in different texts; see Tiwari, Goddess Cults, 94– 181.

24. Lorenzen, K¯ap¯alikas, 13–71. 25. Ibid., 24–31, 219–22. 26. Ibid., 219–20. 27. Ibid., 4–6, 107–110. 28. Ibid., 160–61. 29. Ibid., 160–61, 141–42, 232.

30. Mark Dyczkowski, The Canon of the Saiv¯agama and the Kubjik¯a Tantras of the ´ Western Kaula Tradition (Albany: State University of New York Press, 1988). 31. Goudriaan, “Hindu Tantric Literature in Sanskrit,” 21. See also ibid., 36–38.

32. Susan L.Huntington, The Pala-Sena Schools of Sculpture (Leiden: E.J. Brill, 1984) 7, 17–18n. 33. Ibid., 88–131, 153–54, 160–64. 34. Ibid., 96, 108–116. 35. Lorenzen K¯ap¯alikas, (1991), 181. 36. The S¯´akta Pithas (Delhi: Motilal Banarsidass, n.d.[first published in 1948]), 12. 37. The Hevajra Tantra: Critical Study, Part 1 (London: Oxford University Press, 1959), 14.

36 David N. Lorenzen 38. Yoga of the Guhyasam¯ajatantra: The Arcane Lore of Forty Verses. A Buddhist Tantra Commentary (Delhi: Motilal Barnarsidass, 1977), 99. 39. See the discussions by Goudriaan, “Introduction, History and Philosophy,” 9–11, 20–21; and Sanjukta Gupta, “The Changing Pattern of Pancaratra Initiation: A Case Study in the Reinterpretation of Ritual,” Selected Studies on Ritual in the Indian Religions: Essays to D. H. Hoens, ed. Ria Kloppenborg (Leiden: E.J. Brill, 1983): 69–71.

40. Jan Gonda, Medieval Religious Literature in Sanskrit (Wiesbaden: Otto Harrassowitz, 1977), 163–215; and Alexis Sanderson, “Review of N. R. Bhatt’s editions of Mata˙ngap¯arame´svar¯agama and Rauravottar¯agama” in Bulletin of the School of Oriental and African Studies 48 (1985): 564–68. 41. Goudriaan, “Hindu Tantric Literature,” 21.

42. Ibid, 20–21.

43. For an up-to-date bibliography on this subject, see Paul Eduardo Muller-Ortega, The Triadic Heart of Siva ´ (Albany: State University of New York Press, 1989). 44. Ibid, 45–47. 45. See David L. Snellgrove’s contribution in Buddhist Texts Through the Ages, ed. Edward Conze in collaboration with I. B. Homer, David Snellgrove and Arthur Waley (Boston: Shambala, 1990), I:13–14n. 46. Per Kvaerne, An Anthology of Buddhist Tantric Songs: A Study of the Cary¯ag¯ıti, 2nd ed. (Bangkok: White Orchid Press, 1986), 5–7. 47. Ibid.

48. On the sect’s early history, see the works of George W. Briggs, Gorakhn¯ath and the K¯anphata Yog¯ıs (Calcutta: YMCA Publishing House, 1938); Shashibhusan Dasgupta, Obscure Religious Cults, 2nd ed. (Calcutta: Firma K.L. Mudhopadhyay, 1962); K. V. Zvelibel, The Poets of the Powers (London: Rider, 1973); R. Venkataraman, History of the Tamil Siddha Cult (Ennes: 1990); and Hajariprasad Dvivedi, Nath-sampraday (in Hindi) (Varanasi: Naivedya Niketan, 1966).

49. David Lorenzen, “Gorakhnath,” Encyclopedia of Religions, ed. Mircea Eliade (New York: Macmillan, 1987), 6:77–78. 50. Zvelebil, Poets, 18, 73. 51. Goudriaan, “Introduction, History and Philosophy,” 23. 52. Lorenzen, K¯ap¯alikas, 35–38. 53. Daniel Gold, “Ascenso y ca´ıda del poder de los yogu¯ıs: Jodhpur, 1803–1842,” Estudio de Asia y Africa 27, no.1 (1992): 9–27. 54. Goudriaan and Gupta, Hindu Tantric, 1991